Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Lundahl LL1623SE impressions.
I took a gander at the Lundahl 1623 SE specified
by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#2
|
|||
|
|||
The discussion is way too complicated for me... my point is only about 300B
and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). Less is more, unless the designer knows exactly what to add. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#3
|
|||
|
|||
Hello-
What Patrick says is too complicated for me too. Back when I first built a tube amp Tango transformers were still available in New York. Andre recommended them as alternatives to Lundahl. I bought Tango and used them for the first version of my amp. They sounded good to me. Then a more experienced audiophile told me thaat I should always use the transformer the designer chooses because more depends on it the transformer than on the tube. I bought my Lundahls in Belgium and even with carriage they were less costly than the Tango transformers. I can't say that the Lundahl transformers are x amount better than the Tangos but they sound as good for much less, therefore they are better value. Best, Gray Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). Less is more, unless the designer knows exactly what to add. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#4
|
|||
|
|||
In article , Patrick Turner
wrote: I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. I don't see the big discrepancy here? A loss of 0.5 dB is the same as a loss of about 10.9%, not all that far from the 16.2% figure you calculated from the winding resistances. By the way what winding resistance values did you use to arrive at your 16.2% loss figure? Regards, John Byrns Surf my web pages at, http://users.rcn.com/jbyrns/ |
#5
|
|||
|
|||
Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). How could you mention 6L6 and 2A3 in the same breath? Anyway, since its Easter, we will instruct the Bunny to be gentle with you, and this year the visiting bunny is a nice young lass, so be gentle Fabio, an don't force her to lay an easter egg. My upcoming 2A3 project involves building a remote PS, and replacing the walnut size Hammond low grade OPTs on the amp chassis. It was an Ebay "bargain", but like so much rotten junk flogged off at Ebay, it has shorted turns in one OPT. Less is more, unless the designer knows exactly what to add. Yeah, I might stick with the 6SL7 driver already in the amp. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. I have no idea if the figure of 2.5 kG is correct. There wasn't a close up picture of the tranny, and one could not work out exactly what one was buying. Its a typical tactic of OPT sellers, to not give out all the details, lest ppl copy or wind their own. I wish ppl would copy mine, and if they did a good job, I'd buy them. But only 2 guys, one from Taiwan and one from china have enquired, and they soon ran off when I said I wanted just one sample please, to see if they understood my details. I wasn't interested in a batch of 10,000. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Well, the enemy musn't be in such good condition; let's hope he has mended up ok. But yeah, the Swedes have a way with iron. But looking at the Lundahl factory, which is no bigger than my little weekend lodge, I see no smoke stacks, so no iron is made there afaik, and I'd say they buy in pre-cut rolls of GOSS and make the C-cores to the size they want and to the orders they get. They say they make the cores in-house, which is clean simple work, but I doubt they make the actual steel. Same would go for the amorphous cores. They are supposed to be a lot superior to GOSS, but I am not so sure. Put it this way, nobody has succeeded in convincing me amorphous is better. The LF performance is worse, and the HF better, but the cost is outrageous, if one can ever find a usable size. A lot is said for foil wound windings, and silver wire, teflon insulation. All manner of claims abound. And yet the thd of the tubes is an order of magnitude greater. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#6
|
|||
|
|||
John Byrns wrote: In article , Patrick Turner wrote: I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. I don't see the big discrepancy here? A loss of 0.5 dB is the same as a loss of about 10.9%, not all that far from the 16.2% figure you calculated from the winding resistances. By the way what winding resistance values did you use to arrive at your 16.2% loss figure? I like less than 6% total winding losses. If you go to the Lundahl website, you will see the data they have, and you can make up your own mind. The winding resistances are given for P and S windings, and with all S in series, the RwS is 0.4 ohms, and I forget RwP. Patrick Turner. Regards, John Byrns Surf my web pages at, http://users.rcn.com/jbyrns/ |
#7
|
|||
|
|||
"Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). How could you mention 6L6 and 2A3 in the same breath? IMHO this humble geetah workhorse is usually mistreated. Since You have lots of stuff around, take one of these beam tetrodes, connect it in UL single-ended at normal "design center" parameters thru any decent OPT, add a one-triode gain/driver stage and play it with any decently efficient LS. I bet You'll find a pleasant 4W surprise in this cheap egg. OK, not the 2A3 "finesse", but think how easy it is to make a stereo amp with 2x Russian cheap-as-dirt 6N3P (or eq.) and 1x of any double triode. I'm sure it would be the ideal entry point to the Realm of Single-Ended Magic for all beginner RATs. By the way, the KT88/KT90/6550C can do the same job as a 300B, which costs 3x as much; add a switch and You can have UL or triode at almost zero extra price to better suit taste and music type... IMO 90% of an engineer's skill is to fit into a sensible budget (sometimes using a large hammer) what any damn fool can do with mega-bucks. Happy Easter... by the way, I thought that in Oz You had Easter 'roos, not rabbits.. Ciao Fabio Anyway, since its Easter, we will instruct the Bunny to be gentle with you, and this year the visiting bunny is a nice young lass, so be gentle Fabio, an don't force her to lay an easter egg. My upcoming 2A3 project involves building a remote PS, and replacing the walnut size Hammond low grade OPTs on the amp chassis. It was an Ebay "bargain", but like so much rotten junk flogged off at Ebay, it has shorted turns in one OPT. Less is more, unless the designer knows exactly what to add. Yeah, I might stick with the 6SL7 driver already in the amp. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. I have no idea if the figure of 2.5 kG is correct. There wasn't a close up picture of the tranny, and one could not work out exactly what one was buying. Its a typical tactic of OPT sellers, to not give out all the details, lest ppl copy or wind their own. I wish ppl would copy mine, and if they did a good job, I'd buy them. But only 2 guys, one from Taiwan and one from china have enquired, and they soon ran off when I said I wanted just one sample please, to see if they understood my details. I wasn't interested in a batch of 10,000. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Well, the enemy musn't be in such good condition; let's hope he has mended up ok. But yeah, the Swedes have a way with iron. But looking at the Lundahl factory, which is no bigger than my little weekend lodge, I see no smoke stacks, so no iron is made there afaik, and I'd say they buy in pre-cut rolls of GOSS and make the C-cores to the size they want and to the orders they get. They say they make the cores in-house, which is clean simple work, but I doubt they make the actual steel. Same would go for the amorphous cores. They are supposed to be a lot superior to GOSS, but I am not so sure. Put it this way, nobody has succeeded in convincing me amorphous is better. The LF performance is worse, and the HF better, but the cost is outrageous, if one can ever find a usable size. A lot is said for foil wound windings, and silver wire, teflon insulation. All manner of claims abound. And yet the thd of the tubes is an order of magnitude greater. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#8
|
|||
|
|||
Fabio Berutti wrote: "Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). How could you mention 6L6 and 2A3 in the same breath? IMHO this humble geetah workhorse is usually mistreated. Since You have lots of stuff around, take one of these beam tetrodes, connect it in UL single-ended at normal "design center" parameters thru any decent OPT, add a one-triode gain/driver stage and play it with any decently efficient LS. I bet You'll find a pleasant 4W surprise in this cheap egg. 4W from 6L6 is certainly possible in triode, a nice egg power amount. I built a couple of SE35 amps last year using 4 6CA7 in parallel. These are russian made copies of the Sylvania 6CA7, and are beam tetrodes, not pentodes, like the EL34 which is a plug in replacement for most 6CA7, versions of which sometimes have EL34 /6CA7 marked on the bottle. Because I use CFB, the tubes work in beam tetrode with FB mode, rather similar to 50% plain UL, so I get about 9 watts max per 6CA7. Its much more linear than triode. Odd order harmonics don't seem any higher than triode. I am sure 6L6 would work similarly in my circuit. I use an EL84 in triode to drive the quad of tubes. I could also use 6L6 copies of the originals, 807, or better, the 5881, or perhaps KT66, KT88, KT90, all biased with the same low 60 mA of plate current, same Ea, but I would have to fiddle with the grid bias a little. I don't hold much against any power tube, even 6L6. In my circuit with only Ea = +380v, Eg2 = +270v, any old version of 6L6 would be nice, since I am not going for 50 watts from mainly class AB1 from a pair in PP. No need for 6L6GC. I am only after the comfortable power the tube is happy to give me with a conservative total of 24 watts of plate and screen dissipation ( with 6CA7 ), or a little less with 6L6 etc. Use 4 tubes to get around 30 watts was my aim. Use only 2 watts most days for comfortable listening with modern speakers. Sound is detailed, smooth, warm, accurate, etc. OK, not the 2A3 "finesse", but think how easy it is to make a stereo amp with 2x Russian cheap-as-dirt 6N3P (or eq.) and 1x of any double triode. I'm sure it would be the ideal entry point to the Realm of Single-Ended Magic for all beginner RATs. 4 parallel EL84 in triode might be nice; Pd total 44 watts, efficiency at 25%, po = 11 watts. By the way, the KT88/KT90/6550C can do the same job as a 300B, which costs 3x as much; add a switch and You can have UL or triode at almost zero extra price to better suit taste and music type... Yes, 13E1 in 66% UL does what 4 EL34 or 3 KT88 do in triode. IMO 90% of an engineer's skill is to fit into a sensible budget (sometimes using a large hammer) what any damn fool can do with mega-bucks. I think you got all that wrong :-). 90% of an accountant's skill is to fit into a sensible budget. Hammers have little place in tube craft processes. And good sound doesn't come from megabuck spending. So how to do? Never hire an accountant. Please put that hammer back into the carpenters tool box, and lock it. Throw away the key. Think very hard about the engineering, and have a clear total design package worked out, using medium power dissipation in large tubes, and use easy loaded driver stages. At least use enough NFB to get Ro down to a reasonable level. It means triode operation with little if any global NFB, moderate global NFB for UL/CFB in the output stage is used for the multigrids. More than low global NFB can be used if the sound is OK with increased NFB. A few listening tests should confirm what is best. And for whatever NFB is used, make sure the critical damping is appropriate for the amount of NFB applied. Triode connection works best if we choose low efficiency with a higer than usual load relative to Ra. Happy Easter... by the way, I thought that in Oz You had Easter 'roos, not rabbits.. We got bloody rabbits alright. Back in the early days, ppl brought rabbits here to allow hunting like in old europe. They bred into a national menace; there were billions of them, and they ate more than cattle, sheep, and kangarroos all combined. So we introduced some sexually transmitted pox, and that wiped a lot out, but they developed resistance to the pox, and they are still trying to find a bug that would wipe out the millions of them. We got foxes too, courtesy of the early europeans, and they attacked all the native small marsupial fauna, and many beautiful species have become extinct. They now control some of the rabbits. We have wild pigs weighing 1/2 a tonne, and which can rip a truck tire off, and rip through the door to eat a man with a gun. Then there are wild dogs that attack the wild goats, horses, and anything else edible, and wild domestic cats that have evolved to 10 Kgs of pure muscle and bone, and can slice up a man quite easily, and will see off most of the dogs. In the far west camels roam wild, not much bothers them except old age and parasites. If we had lions and tigers and so forth, they'd demolish the larger roos which grow to about 6', and can rip a man open with a large front toenail on a hind leg. Then they'd move to the horses and cattle, so nobody has allowed the real big cats to go wild yet. But if WW3 comes, maybe they will open the gates of the large zoos out west where the lions and tigers are, and it'll be every animal for itself, the rule of tooth and claw, a la Africa. We got large buffalo up north, 24' long crocodiles, that occasionally eat tourists, and lots of varieties of wild other game. We also have introduced cane toads from sth america, millions of them. They have just spread from cane farms in Qld to NT, where they have begun to infest the magnificent wetland national parks, and all the populations of large lizards have plummeted, since they are quite poisonous to eat. Water from Australia's river systems has been mainly diverted to grow stupid crops like cotton, which uses absurd amounts of water. So many wetlands never get a decent amount of water, reducing the natural bird life dramatically. Since the natural flood cycles of the Murray and Darling rivers is so restricted by man, the rivers are dying, becoming a mere polluted salty drain, and the mouth of the Murray has actually closed up. Salination results in a huge % of arable land becoming useless. We export food, but in 200 years we will barely feed ourselves, and the environment will be truly stuffed. Sure we got roos, I see plenty dead ones along the road sides. Ppl in regional areas pay most of their panel beating repairs for roo hits. If a roo comes over the bonnet, and through the windshield, and kicks madly around when inside, you is in deep donga.... They got **** for brains, and can hop very unpredictably right in front of the car. So in Oz you fit some very substantial bars to the front of the vehicle, and avoid travelling at night. The taipan snake of the central Oz area gives you about 15 minutes to die with a bite on the arse, around which it is difficult to place a torniquet. Maybe you last a little longer if you tie up a leg bite. Way out there, there is no local hospital. Ppl still occasionally die of thirst when their vehicle breaks down. Mobile networks are extending further though. Meanwhile, excellent italian restaurants are to be found here in my town, and its said we have the most restaurants per head of any Oz town. Patrick Turner. Ciao Fabio Anyway, since its Easter, we will instruct the Bunny to be gentle with you, and this year the visiting bunny is a nice young lass, so be gentle Fabio, an don't force her to lay an easter egg. My upcoming 2A3 project involves building a remote PS, and replacing the walnut size Hammond low grade OPTs on the amp chassis. It was an Ebay "bargain", but like so much rotten junk flogged off at Ebay, it has shorted turns in one OPT. Less is more, unless the designer knows exactly what to add. Yeah, I might stick with the 6SL7 driver already in the amp. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. I have no idea if the figure of 2.5 kG is correct. There wasn't a close up picture of the tranny, and one could not work out exactly what one was buying. Its a typical tactic of OPT sellers, to not give out all the details, lest ppl copy or wind their own. I wish ppl would copy mine, and if they did a good job, I'd buy them. But only 2 guys, one from Taiwan and one from china have enquired, and they soon ran off when I said I wanted just one sample please, to see if they understood my details. I wasn't interested in a batch of 10,000. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Well, the enemy musn't be in such good condition; let's hope he has mended up ok. But yeah, the Swedes have a way with iron. But looking at the Lundahl factory, which is no bigger than my little weekend lodge, I see no smoke stacks, so no iron is made there afaik, and I'd say they buy in pre-cut rolls of GOSS and make the C-cores to the size they want and to the orders they get. They say they make the cores in-house, which is clean simple work, but I doubt they make the actual steel. Same would go for the amorphous cores. They are supposed to be a lot superior to GOSS, but I am not so sure. Put it this way, nobody has succeeded in convincing me amorphous is better. The LF performance is worse, and the HF better, but the cost is outrageous, if one can ever find a usable size. A lot is said for foil wound windings, and silver wire, teflon insulation. All manner of claims abound. And yet the thd of the tubes is an order of magnitude greater. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#9
|
|||
|
|||
On Sat, 26 Mar 2005 11:52:35 +0000, Fabio Berutti wrote:
snip IMHO this humble geetah workhorse is usually mistreated. Since You have lots of stuff around, take one of these beam tetrodes, connect it in UL single-ended at normal "design center" parameters thru any decent OPT, add a one-triode gain/driver stage and play it with any decently efficient LS. I bet You'll find a pleasant 4W surprise in this cheap egg. OK, not the 2A3 "finesse", but think how easy it is to make a stereo amp with 2x Russian cheap-as-dirt 6N3P (or eq.) and 1x of any double triode. I'm sure it would snip At the moment I'm running 1/2 6SL7 d/c to 1/2 6SN7 cap/c to 6N3P in pentode with g2 stabilised at 195v by 3 series neons! Rather nice sound and lights up even more... :-) I might try 2A3s - never got into SE DHTs yet and I can't afford 300Bs. -- Mick (no M$ software on here... :-) ) Web: http://www.nascom.info Web: http://projectedsound.tk |
#10
|
|||
|
|||
If they were really swift they'd make a replacement McIntosh
transformer. If the Australians were really swift, they'd get rid of their gun laws so they could conveniently dispatch the really dangerous game, and they'd put a bounty on the crocs and some of the snakes. They would also go into the aircraft industry in a big way and get everyone to fly. That'd open up the outback, and also give Piper, Beech and Cessna a good well deserved swift one in the nuts. In America, which has a lot of stupid laws as well, we have the starling, which isn't all that pernicious, but it has driven out native songbirds. I kill starlings and other pest varmints whenever possible with a good handgun. I recommend the Thompson Center Contender. We have McDonalds, you have Vegemite, neither of which are actually edible. |
#11
|
|||
|
|||
wrote in message
ups.com... We have McDonalds, you have Vegemite, neither of which are actually edible. In that case, go and bite your arse ... (an old Australian saying, _and_ correct spelling, FYI). |
#12
|
|||
|
|||
Dear mr. Turner,
You are risking to be sued by Your national tourist board.... after all I don't believe that Aussie outback can be more dangerous than any European dark alley, in a bad quarter, at late night: after all the most dangerous beasts usually have only 2 legs. Anyway, if I'll ever manage to visit "down under" as I hope, I'll watch my back. Ciao Fabio PS: listening just now to a blues CD through the 2A3SE with 97dB TQWT speakers: the knob is only half-way but John Lee Hooker's guitar is barely sustainable for my ears. I don't have a wattmeter connected, but I think it's no more than 2W out and as You wrote is enough. Sure, the electro-statics (80dB) wouldn't make a move with such a power.. "Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: "Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). How could you mention 6L6 and 2A3 in the same breath? IMHO this humble geetah workhorse is usually mistreated. Since You have lots of stuff around, take one of these beam tetrodes, connect it in UL single-ended at normal "design center" parameters thru any decent OPT, add a one-triode gain/driver stage and play it with any decently efficient LS. I bet You'll find a pleasant 4W surprise in this cheap egg. 4W from 6L6 is certainly possible in triode, a nice egg power amount. I built a couple of SE35 amps last year using 4 6CA7 in parallel. These are russian made copies of the Sylvania 6CA7, and are beam tetrodes, not pentodes, like the EL34 which is a plug in replacement for most 6CA7, versions of which sometimes have EL34 /6CA7 marked on the bottle. Because I use CFB, the tubes work in beam tetrode with FB mode, rather similar to 50% plain UL, so I get about 9 watts max per 6CA7. Its much more linear than triode. Odd order harmonics don't seem any higher than triode. I am sure 6L6 would work similarly in my circuit. I use an EL84 in triode to drive the quad of tubes. I could also use 6L6 copies of the originals, 807, or better, the 5881, or perhaps KT66, KT88, KT90, all biased with the same low 60 mA of plate current, same Ea, but I would have to fiddle with the grid bias a little. I don't hold much against any power tube, even 6L6. In my circuit with only Ea = +380v, Eg2 = +270v, any old version of 6L6 would be nice, since I am not going for 50 watts from mainly class AB1 from a pair in PP. No need for 6L6GC. I am only after the comfortable power the tube is happy to give me with a conservative total of 24 watts of plate and screen dissipation ( with 6CA7 ), or a little less with 6L6 etc. Use 4 tubes to get around 30 watts was my aim. Use only 2 watts most days for comfortable listening with modern speakers. Sound is detailed, smooth, warm, accurate, etc. OK, not the 2A3 "finesse", but think how easy it is to make a stereo amp with 2x Russian cheap-as-dirt 6N3P (or eq.) and 1x of any double triode. I'm sure it would be the ideal entry point to the Realm of Single-Ended Magic for all beginner RATs. 4 parallel EL84 in triode might be nice; Pd total 44 watts, efficiency at 25%, po = 11 watts. By the way, the KT88/KT90/6550C can do the same job as a 300B, which costs 3x as much; add a switch and You can have UL or triode at almost zero extra price to better suit taste and music type... Yes, 13E1 in 66% UL does what 4 EL34 or 3 KT88 do in triode. IMO 90% of an engineer's skill is to fit into a sensible budget (sometimes using a large hammer) what any damn fool can do with mega-bucks. I think you got all that wrong :-). 90% of an accountant's skill is to fit into a sensible budget. Hammers have little place in tube craft processes. And good sound doesn't come from megabuck spending. So how to do? Never hire an accountant. Please put that hammer back into the carpenters tool box, and lock it. Throw away the key. Think very hard about the engineering, and have a clear total design package worked out, using medium power dissipation in large tubes, and use easy loaded driver stages. At least use enough NFB to get Ro down to a reasonable level. It means triode operation with little if any global NFB, moderate global NFB for UL/CFB in the output stage is used for the multigrids. More than low global NFB can be used if the sound is OK with increased NFB. A few listening tests should confirm what is best. And for whatever NFB is used, make sure the critical damping is appropriate for the amount of NFB applied. Triode connection works best if we choose low efficiency with a higer than usual load relative to Ra. Happy Easter... by the way, I thought that in Oz You had Easter 'roos, not rabbits.. We got bloody rabbits alright. Back in the early days, ppl brought rabbits here to allow hunting like in old europe. They bred into a national menace; there were billions of them, and they ate more than cattle, sheep, and kangarroos all combined. So we introduced some sexually transmitted pox, and that wiped a lot out, but they developed resistance to the pox, and they are still trying to find a bug that would wipe out the millions of them. We got foxes too, courtesy of the early europeans, and they attacked all the native small marsupial fauna, and many beautiful species have become extinct. They now control some of the rabbits. We have wild pigs weighing 1/2 a tonne, and which can rip a truck tire off, and rip through the door to eat a man with a gun. Then there are wild dogs that attack the wild goats, horses, and anything else edible, and wild domestic cats that have evolved to 10 Kgs of pure muscle and bone, and can slice up a man quite easily, and will see off most of the dogs. In the far west camels roam wild, not much bothers them except old age and parasites. If we had lions and tigers and so forth, they'd demolish the larger roos which grow to about 6', and can rip a man open with a large front toenail on a hind leg. Then they'd move to the horses and cattle, so nobody has allowed the real big cats to go wild yet. But if WW3 comes, maybe they will open the gates of the large zoos out west where the lions and tigers are, and it'll be every animal for itself, the rule of tooth and claw, a la Africa. We got large buffalo up north, 24' long crocodiles, that occasionally eat tourists, and lots of varieties of wild other game. We also have introduced cane toads from sth america, millions of them. They have just spread from cane farms in Qld to NT, where they have begun to infest the magnificent wetland national parks, and all the populations of large lizards have plummeted, since they are quite poisonous to eat. Water from Australia's river systems has been mainly diverted to grow stupid crops like cotton, which uses absurd amounts of water. So many wetlands never get a decent amount of water, reducing the natural bird life dramatically. Since the natural flood cycles of the Murray and Darling rivers is so restricted by man, the rivers are dying, becoming a mere polluted salty drain, and the mouth of the Murray has actually closed up. Salination results in a huge % of arable land becoming useless. We export food, but in 200 years we will barely feed ourselves, and the environment will be truly stuffed. Sure we got roos, I see plenty dead ones along the road sides. Ppl in regional areas pay most of their panel beating repairs for roo hits. If a roo comes over the bonnet, and through the windshield, and kicks madly around when inside, you is in deep donga.... They got **** for brains, and can hop very unpredictably right in front of the car. So in Oz you fit some very substantial bars to the front of the vehicle, and avoid travelling at night. The taipan snake of the central Oz area gives you about 15 minutes to die with a bite on the arse, around which it is difficult to place a torniquet. Maybe you last a little longer if you tie up a leg bite. Way out there, there is no local hospital. Ppl still occasionally die of thirst when their vehicle breaks down. Mobile networks are extending further though. Meanwhile, excellent italian restaurants are to be found here in my town, and its said we have the most restaurants per head of any Oz town. Patrick Turner. Ciao Fabio Anyway, since its Easter, we will instruct the Bunny to be gentle with you, and this year the visiting bunny is a nice young lass, so be gentle Fabio, an don't force her to lay an easter egg. My upcoming 2A3 project involves building a remote PS, and replacing the walnut size Hammond low grade OPTs on the amp chassis. It was an Ebay "bargain", but like so much rotten junk flogged off at Ebay, it has shorted turns in one OPT. Less is more, unless the designer knows exactly what to add. Yeah, I might stick with the 6SL7 driver already in the amp. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. I have no idea if the figure of 2.5 kG is correct. There wasn't a close up picture of the tranny, and one could not work out exactly what one was buying. Its a typical tactic of OPT sellers, to not give out all the details, lest ppl copy or wind their own. I wish ppl would copy mine, and if they did a good job, I'd buy them. But only 2 guys, one from Taiwan and one from china have enquired, and they soon ran off when I said I wanted just one sample please, to see if they understood my details. I wasn't interested in a batch of 10,000. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Well, the enemy musn't be in such good condition; let's hope he has mended up ok. But yeah, the Swedes have a way with iron. But looking at the Lundahl factory, which is no bigger than my little weekend lodge, I see no smoke stacks, so no iron is made there afaik, and I'd say they buy in pre-cut rolls of GOSS and make the C-cores to the size they want and to the orders they get. They say they make the cores in-house, which is clean simple work, but I doubt they make the actual steel. Same would go for the amorphous cores. They are supposed to be a lot superior to GOSS, but I am not so sure. Put it this way, nobody has succeeded in convincing me amorphous is better. The LF performance is worse, and the HF better, but the cost is outrageous, if one can ever find a usable size. A lot is said for foil wound windings, and silver wire, teflon insulation. All manner of claims abound. And yet the thd of the tubes is an order of magnitude greater. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#13
|
|||
|
|||
On Mon, 28 Mar 2005 11:17:13 GMT, "Fabio Berutti"
wrote: Dear mr. Turner, You are risking to be sued by Your national tourist board.... after all I don't believe that Aussie outback can be more dangerous than any European dark alley, in a bad quarter, at late night: after all the most dangerous beasts usually have only 2 legs. In Oz, they mostly have none! Anyway, if I'll ever manage to visit "down under" as I hope, I'll watch my back. Better to watch where you're putting your feet........ :-) -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#14
|
|||
|
|||
Fabio Berutti wrote: Dear mr. Turner, You are risking to be sued by Your national tourist board.... after all I don't believe that Aussie outback can be more dangerous than any European dark alley, in a bad quarter, at late night: after all the most dangerous beasts usually have only 2 legs. Anyway, if I'll ever manage to visit "down under" as I hope, I'll watch my back. I didn't mean to frighten you. Many Australians have died of skin cancer from too much sunshine, so if you come, wear the biggest hat you can find. All the risks I mentioned pale into insignificance to the risks you run in the city. Too legged sharks abound, and out drivers are slightly absent minded. But the size of Oz is nearly as big as the US of A, so all those risks are so spread out and so unlikely to get you, that most tourists have a great time here and go back home to europe with all their bits intact. We have large numbers of germans, swedes, danes, norwegians et all travelling around here because Oz is one of the least spoilt continents on the planet. There are places you can go which have not changed much in 50,000 years. Young ppl adore travelling a bit before they give their freedom up for marriage and grey conservative existance in europe. Oz is a kind of fantasy world for them. Ciao Fabio PS: listening just now to a blues CD through the 2A3SE with 97dB TQWT speakers: the knob is only half-way but John Lee Hooker's guitar is barely sustainable for my ears. I don't have a wattmeter connected, but I think it's no more than 2W out and as You wrote is enough. Sure, the electro-statics (80dB) wouldn't make a move with such a power.. Funny you should mention JLH, ( not the amp designer ). I heard a vinyl from a loooong time ago, and a recent CD taken from the same master, but the idiot knob jockeys added some way to make the audio level more constant, cos JLH tends to disregard the distance to the mic, and the modern CD version sounded bland with flattened dynamics. So much for some modern re-mastering. Don't they realize the old mastering was done by masters? Patrick Turner. "Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: "Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: The discussion is way too complicated for me... my point is only about 300B and 2A3 tubes. Apart from the fact that I have a 2A3, this latter tube has MUCH lower requirements than the 300B, particularly in terms of driving voltage and OPT DC/power handling capabilities. This in turn means that there are less things that can go wrong. Briefly, it's not the power tube but all the rest that surrounds it which makes (generally speaking) any decent 2A3SE a great sounding amp. Therefore, if I only wanted 3W, I'd use a 2A3 (BTW a measle trioded 6L6 does the same job and has zero filament hum...). How could you mention 6L6 and 2A3 in the same breath? IMHO this humble geetah workhorse is usually mistreated. Since You have lots of stuff around, take one of these beam tetrodes, connect it in UL single-ended at normal "design center" parameters thru any decent OPT, add a one-triode gain/driver stage and play it with any decently efficient LS. I bet You'll find a pleasant 4W surprise in this cheap egg. 4W from 6L6 is certainly possible in triode, a nice egg power amount. I built a couple of SE35 amps last year using 4 6CA7 in parallel. These are russian made copies of the Sylvania 6CA7, and are beam tetrodes, not pentodes, like the EL34 which is a plug in replacement for most 6CA7, versions of which sometimes have EL34 /6CA7 marked on the bottle. Because I use CFB, the tubes work in beam tetrode with FB mode, rather similar to 50% plain UL, so I get about 9 watts max per 6CA7. Its much more linear than triode. Odd order harmonics don't seem any higher than triode. I am sure 6L6 would work similarly in my circuit. I use an EL84 in triode to drive the quad of tubes. I could also use 6L6 copies of the originals, 807, or better, the 5881, or perhaps KT66, KT88, KT90, all biased with the same low 60 mA of plate current, same Ea, but I would have to fiddle with the grid bias a little. I don't hold much against any power tube, even 6L6. In my circuit with only Ea = +380v, Eg2 = +270v, any old version of 6L6 would be nice, since I am not going for 50 watts from mainly class AB1 from a pair in PP. No need for 6L6GC. I am only after the comfortable power the tube is happy to give me with a conservative total of 24 watts of plate and screen dissipation ( with 6CA7 ), or a little less with 6L6 etc. Use 4 tubes to get around 30 watts was my aim. Use only 2 watts most days for comfortable listening with modern speakers. Sound is detailed, smooth, warm, accurate, etc. OK, not the 2A3 "finesse", but think how easy it is to make a stereo amp with 2x Russian cheap-as-dirt 6N3P (or eq.) and 1x of any double triode. I'm sure it would be the ideal entry point to the Realm of Single-Ended Magic for all beginner RATs. 4 parallel EL84 in triode might be nice; Pd total 44 watts, efficiency at 25%, po = 11 watts. By the way, the KT88/KT90/6550C can do the same job as a 300B, which costs 3x as much; add a switch and You can have UL or triode at almost zero extra price to better suit taste and music type... Yes, 13E1 in 66% UL does what 4 EL34 or 3 KT88 do in triode. IMO 90% of an engineer's skill is to fit into a sensible budget (sometimes using a large hammer) what any damn fool can do with mega-bucks. I think you got all that wrong :-). 90% of an accountant's skill is to fit into a sensible budget. Hammers have little place in tube craft processes. And good sound doesn't come from megabuck spending. So how to do? Never hire an accountant. Please put that hammer back into the carpenters tool box, and lock it. Throw away the key. Think very hard about the engineering, and have a clear total design package worked out, using medium power dissipation in large tubes, and use easy loaded driver stages. At least use enough NFB to get Ro down to a reasonable level. It means triode operation with little if any global NFB, moderate global NFB for UL/CFB in the output stage is used for the multigrids. More than low global NFB can be used if the sound is OK with increased NFB. A few listening tests should confirm what is best. And for whatever NFB is used, make sure the critical damping is appropriate for the amount of NFB applied. Triode connection works best if we choose low efficiency with a higer than usual load relative to Ra. Happy Easter... by the way, I thought that in Oz You had Easter 'roos, not rabbits.. We got bloody rabbits alright. Back in the early days, ppl brought rabbits here to allow hunting like in old europe. They bred into a national menace; there were billions of them, and they ate more than cattle, sheep, and kangarroos all combined. So we introduced some sexually transmitted pox, and that wiped a lot out, but they developed resistance to the pox, and they are still trying to find a bug that would wipe out the millions of them. We got foxes too, courtesy of the early europeans, and they attacked all the native small marsupial fauna, and many beautiful species have become extinct. They now control some of the rabbits. We have wild pigs weighing 1/2 a tonne, and which can rip a truck tire off, and rip through the door to eat a man with a gun. Then there are wild dogs that attack the wild goats, horses, and anything else edible, and wild domestic cats that have evolved to 10 Kgs of pure muscle and bone, and can slice up a man quite easily, and will see off most of the dogs. In the far west camels roam wild, not much bothers them except old age and parasites. If we had lions and tigers and so forth, they'd demolish the larger roos which grow to about 6', and can rip a man open with a large front toenail on a hind leg. Then they'd move to the horses and cattle, so nobody has allowed the real big cats to go wild yet. But if WW3 comes, maybe they will open the gates of the large zoos out west where the lions and tigers are, and it'll be every animal for itself, the rule of tooth and claw, a la Africa. We got large buffalo up north, 24' long crocodiles, that occasionally eat tourists, and lots of varieties of wild other game. We also have introduced cane toads from sth america, millions of them. They have just spread from cane farms in Qld to NT, where they have begun to infest the magnificent wetland national parks, and all the populations of large lizards have plummeted, since they are quite poisonous to eat. Water from Australia's river systems has been mainly diverted to grow stupid crops like cotton, which uses absurd amounts of water. So many wetlands never get a decent amount of water, reducing the natural bird life dramatically. Since the natural flood cycles of the Murray and Darling rivers is so restricted by man, the rivers are dying, becoming a mere polluted salty drain, and the mouth of the Murray has actually closed up. Salination results in a huge % of arable land becoming useless. We export food, but in 200 years we will barely feed ourselves, and the environment will be truly stuffed. Sure we got roos, I see plenty dead ones along the road sides. Ppl in regional areas pay most of their panel beating repairs for roo hits. If a roo comes over the bonnet, and through the windshield, and kicks madly around when inside, you is in deep donga.... They got **** for brains, and can hop very unpredictably right in front of the car. So in Oz you fit some very substantial bars to the front of the vehicle, and avoid travelling at night. The taipan snake of the central Oz area gives you about 15 minutes to die with a bite on the arse, around which it is difficult to place a torniquet. Maybe you last a little longer if you tie up a leg bite. Way out there, there is no local hospital. Ppl still occasionally die of thirst when their vehicle breaks down. Mobile networks are extending further though. Meanwhile, excellent italian restaurants are to be found here in my town, and its said we have the most restaurants per head of any Oz town. Patrick Turner. Ciao Fabio Anyway, since its Easter, we will instruct the Bunny to be gentle with you, and this year the visiting bunny is a nice young lass, so be gentle Fabio, an don't force her to lay an easter egg. My upcoming 2A3 project involves building a remote PS, and replacing the walnut size Hammond low grade OPTs on the amp chassis. It was an Ebay "bargain", but like so much rotten junk flogged off at Ebay, it has shorted turns in one OPT. Less is more, unless the designer knows exactly what to add. Yeah, I might stick with the 6SL7 driver already in the amp. BTW the OPTs I use with my 2A3s (Ripley UK 3k to 6R) weigh about 9 pounds (4 kg). I'd say that 2.5kg is not that much for a full-power 300B OPT, but Swedish steel is really in a league of its own.. I have no idea if the figure of 2.5 kG is correct. There wasn't a close up picture of the tranny, and one could not work out exactly what one was buying. Its a typical tactic of OPT sellers, to not give out all the details, lest ppl copy or wind their own. I wish ppl would copy mine, and if they did a good job, I'd buy them. But only 2 guys, one from Taiwan and one from china have enquired, and they soon ran off when I said I wanted just one sample please, to see if they understood my details. I wasn't interested in a batch of 10,000. My Swedish Mauser's barrel is still perfect, and it surely fired many thousand rounds since it was rifled in 1916! Well, the enemy musn't be in such good condition; let's hope he has mended up ok. But yeah, the Swedes have a way with iron. But looking at the Lundahl factory, which is no bigger than my little weekend lodge, I see no smoke stacks, so no iron is made there afaik, and I'd say they buy in pre-cut rolls of GOSS and make the C-cores to the size they want and to the orders they get. They say they make the cores in-house, which is clean simple work, but I doubt they make the actual steel. Same would go for the amorphous cores. They are supposed to be a lot superior to GOSS, but I am not so sure. Put it this way, nobody has succeeded in convincing me amorphous is better. The LF performance is worse, and the HF better, but the cost is outrageous, if one can ever find a usable size. A lot is said for foil wound windings, and silver wire, teflon insulation. All manner of claims abound. And yet the thd of the tubes is an order of magnitude greater. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... I took a gander at the Lundahl 1623 SE specified by Andre for the KISS amp. The details were a little hard to understand, with the usual tongue in cheek way of giving information out. However, not a bad OPT for a 3 watt SET project. The 30H inductance seems fine, and with a load of 3k and Ra of 800 ohms in parallel, the -3 dB point in the LF response with 300B should be at 3.3 Hz at low levels where the OPT will mainly be used. I dount it will saturate at a highish F because its good for 25 watts at 30 Hz. The worst case HF pole is where the source R = zero ohms. So with 3k load, and the 4.6 mH leakage L the lowest pole is at 104 kHz, a quite respectable figure. The insertion loss is quoted at 0.5 dB. I assume this is voltage drop. So if 106v is applied to the primary, 100v is across the 3k load, with the 0.5 dB drop of 6v across the winding resistance as seen from the primary, so the winding R is 181 ohms. So losses are 5.6% total. But elsewhere when they quote winding resistances, i figured the total losses would be 5.2% on the P and 11% on the S, for a total of 16.2%, not such a good figure. I wish these ppl would be clearer when they dish out the info. There are 4 primary coils and eight secondary coils. from the tables I could discern that that when all P are in series, and all the S are in series, there is a 3kohm to 3.2 ohm load match, which converts to 5.25 : 5.6 ohms, a very healthy load match for 300B and to modern speakers, which rarely measure their 8 ohms. I only saw one weight figure of 2.5 Kg, and I have to say that is very light indeed and I would tend to use more iron and less copper. I would guess that the transformer has 2 P windings and 4 S windings on each leg of a single C core, so that an S-P-S-S-P-S arrangement of sections is used on each leg in two bobbins. There are enough interleavings. From a sample invoice, the price of a 1623 could be US$137 ex the quaint little factory in Sweeden. If one ventures to the Lundahl site one cannot help seeing references to the same trannies but with amorphous cores. Alas this material can only take about 0.7 of the voltage for the same Fsat of the GOSS cores which are standard, so only 13 watts instead of 25 watts is available. But when one reads the local discussion group reviews of the amorphous, it appears no effort is spared in heaping praise on this material, but without a single technical syllable about why this stuff must be so much better, especially with PP transformers whose performance isn't "dominated by the air gap" as reported. Nobody has checked out the harmonic distortion. In particular a big deal is made of the "burn in" qualities of the amorphous cores, and how, after 50 hrs of use, the sonics improve so much it makes the sound from the standard GOSS cores like its coming from a "broken" transformer. Well, big fat claims indeed, when we know how good the sound of GOSS is without resorting to amorphous. Basically the theme of the email responses from the local "discussion group" tend to say its costs a lot more but its worth it. Hmm, I'd rather just use bigger cores..... Meanwhile I have a 2A3 project coming up, and the client has had a bunch of Hammond iron dumped at the door for me. The OPTs are 4.5 Kg, with 2" stack of 1.5" tongue E&I. It will be interesting to compare the performance I get for these 3 watt amps. 2A3 is reputed to be even better than 300B by some audio enthusiasts. Patrick Turner. |
#15
|
|||
|
|||
On Tue, 29 Mar 2005 00:07:14 +1000, Patrick Turner
wrote: We have large numbers of germans, swedes, danes, norwegians et all travelling around here because Oz is one of the least spoilt continents on the planet. There are places you can go which have not changed much in 50,000 years. Indeed so. Now, if we could just get rid of all that dross that the Brits dumped there for a couple of centuries............ :-) I heard a vinyl from a loooong time ago, and a recent CD taken from the same master, but the idiot knob jockeys Oh dear, Churches will be after you for that one! :-) -- Stewart Pinkerton | Music is Art - Audio is Engineering |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Does anyone know of this challenge? | High End Audio | |||
GT 12AX7M (Mullard "repro") - first impressions | Vacuum Tubes | |||
1rst impressions of 5th of RAP CD | Pro Audio | |||
RIP dearly beloved Lundahl LL1635 | Vacuum Tubes | |||
PodXT Pro, First Impressions | Pro Audio |