Home |
Search |
Today's Posts |
#1
|
|||
|
|||
The man wants to make a 212 SET amp.
Anyone have any experience in using the 212 triode
for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner |
#2
|
|||
|
|||
It seems its a 212E, or 4212E. I found a reference to Kronzilla at http://www.enjoythemusic.com/magazin...okronzilla.htm But no details of loads, although Kronzilla boasts 42 watts at 5% thd from a pair of paralleled tubes. This suggests class A1, and a low B+, and very conservative op conditions. I found the data on the 212E and ONE could be set up for class A2 and at 1.3kV, 140mA, Pda = 182 watts, 11.2k RL, and 54 watts is available, with 28 watts in class A1. Drive voltage in A2 would be about 56Vrms to G1 biased at -50V. Ra is 2,500 ohms, so DF isn't too bad. Another question, is WE making the WE212E again or not? Patrick Turner. Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner |
#3
|
|||
|
|||
Patrick Turner wrote:
Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner See WE data at http://frank.pocnet.net/sheets/136/2/212E.pdf JLS |
#4
|
|||
|
|||
Patrick Turner wrote:
Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Theres a yahoo group or two devoted to the 212. The message archives for the GM70 group has a lot of threads about it. A few people agreed that the 212 was powerful but still didn't sound as nice as smaller triodes like the 211 or 845. Adam |
#5
|
|||
|
|||
John Stewart wrote: Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner See WE data at http://frank.pocnet.net/sheets/136/2/212E.pdf JLS Thanks John, I found all that last night in my searches. I was wondering if anyone here had actually had real expereince with the 212E. Its not used commonly, but it looks like a good candidate for high SET power. A guy may need me to wind 4 output trannies for a batch of amps, and as usual, he hasn't the foggiest notion of what is required for the OPT, and nor has the the guy in NZ who is doing the chassis for him. You can't buy 212E OPTs off the shelf, anywhere. Patrick Turner. |
#6
|
|||
|
|||
Adam Stouffer wrote: Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Theres a yahoo group or two devoted to the 212. The message archives for the GM70 group has a lot of threads about it. A few people agreed that the 212 was powerful but still didn't sound as nice as smaller triodes like the 211 or 845. Maybe, but did they know to to use it? The curves indicate it will have excellent performance technically at least. The only bother is the A2 operation, but only if its set up for A2. A2 doesn't have to be used, although if it is, there is a lot more power than A1. With A1 a high B+ is needed, perhaps 1,600V, Ia = about 120mA and a high RL of about 14k, to get a wide linear V swing without the negative going anode swing extending past the Ra line for Eg = 0V, ie, when Ea = about 500V on the load line. For A2, the B+ can be lower, Ia higher, and RL value at 11k, since the negative going anode swing can extend down to Ea = 220V. To use such tubes optimally, you have to throw lots of voltage, lots of RL ohms, lots of turns, lots of iron at the tube. We live in low impedance times, and there is a reluctance to use high voltage, ohms, turns or iron. I have to make the OPTs able to have varying impedance connections, and with ability for use with perhaps other tubes, and maybe also set up in a tank of oil, and well sealed, so these OPTs are not child's play to get right. But we do have terrific iron now days....... All this may lead me to thouroughly spoil my springtime by having to wind OPTs for such tubes. Patrick Turner. Adam |
#7
|
|||
|
|||
See the GEC book "Audio frequency amplifier design", chapter 8. It deals
extensively with the use of the GEC V1505, aka 4212, unfortunately only in push-pull, but transforming one of these circuits in SE is as easy as "cutting" the drawing between the "push" and the "pull" part. Some data are provided too. There's a schematic at http://www10.big.or.jp/~dh/work/0402.html , but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner |
#8
|
|||
|
|||
On Mon, 25 Jul 2005 16:28:16 GMT, Patrick Turner
wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Patrick, I don't know if this is much use, but at http://audiobizarro.com/ you can see a couple of photographs of amps created with the 212E, the first is on three chassis, whilst the second, is on more reasonable one. I seem to remembe seeing diagrams of these, but the site has changed somewhat and they have dissapeared/ watch your fingers. bill |
#9
|
|||
|
|||
Patrick Turner wrote:
John Stewart wrote: Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner See WE data at http://frank.pocnet.net/sheets/136/2/212E.pdf JLS Thanks John, I found all that last night in my searches. I was wondering if anyone here had actually had real expereince with the 212E. Its not used commonly, but it looks like a good candidate for high SET power. A guy may need me to wind 4 output trannies for a batch of amps, and as usual, he hasn't the foggiest notion of what is required for the OPT, and nor has the the guy in NZ who is doing the chassis for him. You can't buy 212E OPTs off the shelf, anywhere. Patrick Turner. Should be an interesting project. But get your money for the transformers up front!! Since your prospective client seems somewhat less than experienced, he could become frustrated & may lose interest partway thru. I found the 2nd paragraph under Class A operation on the WE212E data sheet interesting. It sounds as though local spot heating is a potential problem on the anode. OK I guess at lower voltages. Cheers, John Stewart |
#10
|
|||
|
|||
Fabio Berutti wrote: See the GEC book "Audio frequency amplifier design", chapter 8. It deals extensively with the use of the GEC V1505, aka 4212, unfortunately only in push-pull, but transforming one of these circuits in SE is as easy as "cutting" the drawing between the "push" and the "pull" part. Some data are provided too. There's a schematic at http://www10.big.or.jp/~dh/work/0402.html , but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes Ciao So the 4212, or 212E is the same tube as the V1505? I have the GE book with 17 schematics from 5 to 1,100 watts, with two V1505 used for the 1,100 watt amp. Such an amp was used for all the horns in a stadium. I used to think showground PA systems sounded dreadful when i was a lad. Its not quite all that easy to cut a class B amp in 1/2 and get a class A SE amp. One still needs to work from first principles. The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. The IST input tranny is 150 ohms :100kohms, with the 150 ohm input primary fed from a 1.5 watt 100 ohm pot to set the level. so the turn ratio is 1:25.8, so for 100vrms grid to grid max drive, you need to apply 3.9vrms to the input tranny, which is doable with a small power amp. The stray C and miller C would limit the F response somewhat since the secondary impedance is 100k, getting rather high. Patrick Turner. |
#11
|
|||
|
|||
bill ramsay wrote: On Mon, 25 Jul 2005 16:28:16 GMT, Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Patrick, I don't know if this is much use, but at http://audiobizarro.com/ you can see a couple of photographs of amps created with the 212E, the first is on three chassis, whilst the second, is on more reasonable one. You mean http://audiobizarro.com/billeci5.jpg yes, but no schematics. Patrick Turner. I seem to remembe seeing diagrams of these, but the site has changed somewhat and they have dissapeared/ watch your fingers. bill |
#12
|
|||
|
|||
John Stewart wrote: Patrick Turner wrote: John Stewart wrote: Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner See WE data at http://frank.pocnet.net/sheets/136/2/212E.pdf JLS Thanks John, I found all that last night in my searches. I was wondering if anyone here had actually had real expereince with the 212E. Its not used commonly, but it looks like a good candidate for high SET power. A guy may need me to wind 4 output trannies for a batch of amps, and as usual, he hasn't the foggiest notion of what is required for the OPT, and nor has the the guy in NZ who is doing the chassis for him. You can't buy 212E OPTs off the shelf, anywhere. Patrick Turner. Should be an interesting project. But get your money for the transformers up front!! Since your prospective client seems somewhat less than experienced, he could become frustrated & may lose interest partway thru. I warn them of the delay that is normal for when anything quite exotic is ordered that isn't what is probably available froma shelf in a shop somewhere in Japan. I never start a project like this without at least 50% of the payment. If they lose interest, or die from some ailment during the 3 mths I may take to complete works in hand and do their items, and they don't complete the deal, then I am at least paid for my material and some labour. I normally give late completers of deals a long time to get their pennys together and pay me. Its normal for me to have up to six or seven items that i have repaired within two weeks of item reciept, but which sit around here for months before pick up. I wait a year, and if i don't hear from them I may sell the items, and usually I don't get the price agreed on the original contract; OPTs like these may well suit builders of other high powered SET amps using multiple low power tubes or large transmitter tubes, so hence i'd never build such an item without having several impedance ratios available, including the ability to parallel 2 halves of the primary, so multiple small tubes could be used. The design of a 10k:5 OPT can thus be made to suit 2.5k:5. I found the 2nd paragraph under Class A operation on the WE212E data sheet interesting. It sounds as though local spot heating is a potential problem on the anode. OK I guess at lower voltages. I'd never run such tubes anywhere near their rated max Pd, when hot spots are very likely. Pda = about 150watts would be about max practical, and typical class A1 efficiency is 20% with low B+, so you get only 30 watts, about 4% thd. Class A2 is better, perhaps efficiency is 40%, so 60 watts max but with a lot more thd at 60 watts, about 8%. But from the plate curves, the first 20 watts of A1 would be very good with maybe 2%. 2 watts would about 0.6%, all 2H. The thd at 2 watts could be about halved if RL is allowed to be high, so probably thd would then never exceed 0.3% in normal use. 2 watts from an EL34 or 300B would have a lot more thd. To avoid feedback, you have to throw a lotta glass, iron, copper, turns, volts, and ohms at the problem but its no worse than a rich man spending 10 million on carbon fibre for a yacht to break the Sydney to Hobart race record, or a cyclist also spending his life's savings on a carbon fibre bike frame to be able to pedal up a hill faster. We only borrow the iron and copper and glass for awhile. In 100 years maybe it has become something else, but I do not know what becomes of a carbon fibre yacht. Patrick Turner. Cheers, John Stewart |
#13
|
|||
|
|||
"Patrick Turner" ha scritto nel messaggio ... Fabio Berutti wrote: See the GEC book "Audio frequency amplifier design", chapter 8. It deals extensively with the use of the GEC V1505, aka 4212, unfortunately only in push-pull, but transforming one of these circuits in SE is as easy as "cutting" the drawing between the "push" and the "pull" part. Some data are provided too. There's a schematic at http://www10.big.or.jp/~dh/work/0402.html , but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes Ciao So the 4212, or 212E is the same tube as the V1505? AFAIK it is indeed so. I have the GE book with 17 schematics from 5 to 1,100 watts, with two V1505 used for the 1,100 watt amp. They're used to get anything in between 400 (most A1 class) and 1100W (AB2 class), varying B+ and driving circuit Such an amp was used for all the horns in a stadium. I used to think showground PA systems sounded dreadful when i was a lad. Maybe the OPTs were on the cheap side (only speech bandwidth required) and surely the HORNS were definitely NOT HiFi... would You place an Altec box costing M$ on the top of a pole and then expose it to an Aussie thunderstorm (or to the outrages of diarrhoic flying foxes)? Its not quite all that easy to cut a class B amp in 1/2 and get a class A SE amp. One still needs to work from first principles. I just meant to "take inspiration" from the circuit, using the idea of having a KT66 (ok a 6550 is easier to find) cathode-coupled thru a transformer to the monster triode in order to have a suitably low driver impedance. Changing to SE is straightforward, as far as I can understand, and surely an experienced technician as You will have no problems when re-working the most suitable operating points. I like very much that GEC book, it really seems written by true professionals. The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. The IST input tranny is 150 ohms :100kohms, with the 150 ohm input primary fed from a 1.5 watt 100 ohm pot to set the level. so the turn ratio is 1:25.8, so for 100vrms grid to grid max drive, you need to apply 3.9vrms to the input tranny, which is doable with a small power amp. The stray C and miller C would limit the F response somewhat since the secondary impedance is 100k, getting rather high. I just cited it, I don't like it either. As You know I think that each tube should be used for what it was meant for: the 4212 is a monster, there's no reason to use it unless You want 100W class A SE. A Chinese 845 or 2/3 KT88s in parallel or Your 13E1 will smoothly deliver the 20-30W that are needed to enjoy SE designs with almost any LS in almost any home (the U2 concert in S.Siro stadium with its 100'000 W is not included), if I were to spend that money and risk my butt with 2500V I'd want to squeeze all the juice out of it. Ciao Fabio Patrick Turner. |
#14
|
|||
|
|||
Patrick Turner wrote:
Fabio Berutti wrote: There's a schematic at http://www10.big.or.jp/~dh/work/0402.html but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html -- Oskari Heinonen * University of Helsinki * Department of Computer Science * http://www.cs.Helsinki.FI/Oskari.Heinonen/ |
#15
|
|||
|
|||
Oskari Heinonen wrote: Patrick Turner wrote: Fabio Berutti wrote: There's a schematic at http://www10.big.or.jp/~dh/work/0402.html but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html -- Hi Oskari, your eye for detail is as fine as ever, and no I didn't see the above, but one has to say its a spectacular amp. It uses a little iddy biddy 845 as the driver....... Patrick Turner. |
#16
|
|||
|
|||
That unit is very pretty indeed...in the picture at least. The label is
written in German, but the schematic is surely "made in Japan". From an electronic-common-sense viewpoint it simply makes no sense. An 845 used as a driver, lots of transformers with their relevant phase shifts, then a monster tube used at "only" 1000V with CATHODE BIAS!!!! OK, I can trust it sounds good, but making all that stuff for perhaps 20W is like using a Peterbuilt truck to move a box of cigarettes, and I'd bet a couple of $ that it is not better sounding than any other "sensible" project, costing like the volume knob of that insanity. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... Oskari Heinonen wrote: Patrick Turner wrote: Fabio Berutti wrote: There's a schematic at http://www10.big.or.jp/~dh/work/0402.html but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html -- Hi Oskari, your eye for detail is as fine as ever, and no I didn't see the above, but one has to say its a spectacular amp. It uses a little iddy biddy 845 as the driver....... Patrick Turner. |
#17
|
|||
|
|||
"Oskari Heinonen" wrote
Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html Hi Oskari The operating point doesn't quite square with the datasheet posted by John. Stepping up from the 845 seems perverse to me. Can the 212 be driven into A2 with a 10k winding I wonder? Looks to me like grid resistance drops to 1k pretty quick when it goes +ve. Also wondering what the cross-hatching on many of the cap symbols means. Is it permissible to use a distributed load with these things? Not much gain to play with but DF could be significantly improved all the same with perhaps 30% of the winding on the cathode. Need plenty drive volts though. All scary stuff to me. cheers, Ian |
#18
|
|||
|
|||
Fabio Berutti wrote: That unit is very pretty indeed...in the picture at least. The label is written in German, but the schematic is surely "made in Japan". From an electronic-common-sense viewpoint it simply makes no sense. An 845 used as a driver, lots of transformers with their relevant phase shifts, then a monster tube used at "only" 1000V with CATHODE BIAS!!!! OK, I can trust it sounds good, but making all that stuff for perhaps 20W is like using a Peterbuilt truck to move a box of cigarettes, and I'd bet a couple of $ that it is not better sounding than any other "sensible" project, costing like the volume knob of that insanity. Yes, well, look, ah, hmm, yeah, I know. But then 20 watts is a heck of a lotta power to some folks, including my prospective client who said he only wants 20 watts. Its the sound they are chasing, and 20 watts is like setting up a 300B and settling for 4 watts only, with a higher load value than "normal". But at the higher primary loads, the triode is most linear, and the OPT has the highest ratio, so the damping factor is much better than if we set up the triode to screw the most power out of it. The 845 on it's own could give enough power. But some folks like to use whatever is out there, and you gotta admit it, the 4212E does have a certain flair and style that not many other tubes have. It makes the 845 look like a wussy teensy weenie toobe. Now they say the Pda of the 212 is 275 watts max, and the linearity going into the grid current area looks fairly good, assuming the driver tube has a nice low Ra, so if one set up such a tube at 275 watts Pda, then an efficiency of 40% in A2 would give 110 watts, or about what a quad of 211 etc in PSE would give, or what 14 x 6CA7 in PSE tetrode with CFB would give, or what 16 x KT88 in triode would give, etc. But one would never dare have such a tube set up on 275 watts Pda at idle. It surely wouldn't last long, and not many are around, and afaik none are being made now. They are designed to be able to handle 275 watts of Pd, but in practice the duty cycle in most class AB apps would make the average Pda far lower; they were designed to make huge amounts of power in PP circuits with a low idle bias current. I think about 150 watts of Pd would be about right, so if Ea was 1,500V at 100mA Ia, it would be fine, and a class A1 Vswing of about 700Vrms into 14k is possible giving only 35 watts, as much as my SECFB amp with a quad of 6CA7. If class A2 is used with Pd = 150watts and Ea = 1,500V, then you get 1,000 Vrms into 15k, and about 65 watts. That's an efficiency of 43%, and not too bad, since any other class A tube amp won't achieve much higher efficiency, bearing in mind the max theoretical possible class A efficiency is 50%. With a 15k load, and Ra = about 2,500 ohms at Iaq = 100 mA, the DF without any NFB is quite good at 6 and as good as a 300B loaded with 6k. The linearity at 15k looks good, I reckoned about 2% thd at 50 watts from the load lines. Expect about 0.63% at 5 watts, and at 0.5 watts about 0.2%, all of which would be mainly 2H. 12 dB of NFB would make the amp measure about 0.5 % at 50 watts, 0.15% at 5 watts, and 0.05% at 0.5 watts. These figures would be actually drastically reduced in practice if a tube such as a 300B 2A3, or trioded KT66 etc was was used as a driver and loaded appropriately since the thd of the driver having to produce 63vrms, which is not much for a power tube, will be about 2%, and it will cancel the 2H in the output stage to a large extent, and since nothing is badly loaded, the amount of second order products due to distortion cancelling would have to be tiny indeed. So without any loop FB, a substantially linear 60 watt SET amp can be made using only 3 tubes, perhaps 4 if we includ a cathode follower buffer between the driver and 212 grids. People can argue that when linearity is dialled into the design of an SET amp it won't sound any good, but I've never heard any evidence to support this idea. I figured the 15k OPT would need to be about 14Kgs, and a 400VA PT would also be about 10Kgs, and with a choke in the CLC supply, the amp won't be light. I don't see any need for mercury vapour rectifiers or oiler caps when standard practice for HV supplies in recent years is to series electros, so small chokes may be used, and to series SS diodes. But a mono will be around 30 Kgs. My prospective client says he has tonnes of gear, including 30,000 tubes. He says he has 6,000 x 12AT7, all NOS. Maybe he has a few 866, and many chokes and oil caps. Patrick Turner. Ciao Fabio "Patrick Turner" ha scritto nel messaggio ... Oskari Heinonen wrote: Patrick Turner wrote: Fabio Berutti wrote: There's a schematic at http://www10.big.or.jp/~dh/work/0402.html but I'm sure I saw the project of a guy who made a monster SE using some STC 4212 tubes The url you gave has a PP amp with 1,100V applied to the CT and cathode biasing, so it must be a largely class A1 PP amp. Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html -- Hi Oskari, your eye for detail is as fine as ever, and no I didn't see the above, but one has to say its a spectacular amp. It uses a little iddy biddy 845 as the driver....... Patrick Turner. |
#19
|
|||
|
|||
I don't know if this is much use, but at http://audiobizarro.com/ you can see a couple of photographs of amps created with the 212E, the first is on three chassis, whilst the second, is on more reasonable one. By definition, there is nothing reasonable about building an amplifier with a 212 today. Might as well go the full monty! |
#20
|
|||
|
|||
"Bret Ludwig" wrote
...there is nothing reasonable about building an amplifier with a 212 today... Exactly the point. The key element of the design process is getting the styling right. Ludicrously over-engineered build quality then has to complete the impression of a thoroughgoing but civilised unreasonableness. They will need totally unreasonable speakers to complete the statement. If I had a Peterbuilt I'd go to the shop for cigs in it, even though I don't smoke. cheers, Ian |
#21
|
|||
|
|||
Ian Iveson wrote: "Oskari Heinonen" wrote Did you notice this 845/4212 SE amp? http://www10.big.or.jp/~dh/gallery/pic/4212.html http://www10.big.or.jp/~dh/gallery/pic/4212c.html Hi Oskari The operating point doesn't quite square with the datasheet posted by John. Stepping up from the 845 seems perverse to me. Can the 212 be driven into A2 with a 10k winding I wonder? Looks to me like grid resistance drops to 1k pretty quick when it goes +ve. Part of the grid swing can be +ve to drive into 10k as the above schematic proposes, but if A2 isn't addopted, then po in A1 is rather low. The lower the Ea, higher the Ia, and lower the RL, the worse things become, +ve grid drive is increased, thd is increased, even in the A1 part of the op, and Ro increases, all of which would spoil the sonic and measured technical outcome. Rgin of 1k is easily overcome by using a CF tube with Ro = 100 ohms. Even when gI isn't used, a CF is supposed to improve the sound as borne out with the Ongaku amp, which has a CF drive to the 211 grid anyway. Also wondering what the cross-hatching on many of the cap symbols means. Is it permissible to use a distributed load with these things? Not much gain to play with but DF could be significantly improved all the same with perhaps 30% of the winding on the cathode. Need plenty drive volts though. As I explained in several posts, the last of which was last night, the use of a high Ea, RL of 15k, and almost all A1 op will give a DF of 6 without CFB. 6dB of CFB will double the drive voltage from around 70vrms to 140vrms, and make matters worse, with very little reduction of Ro and thd if any. If FB is to be used, it should be global, and around a tube drive stage with no ISTs, and with an OPT with about 50kHz open loop bw to start with. Patrick Turner. All scary stuff to me. 1,600 volts isn't something you take lightly, you need to be frightened of such voltages at all times. Patrick Turner. cheers, Ian |
#22
|
|||
|
|||
Ian Iveson wrote:
"Bret Ludwig" wrote ...there is nothing reasonable about building an amplifier with a 212 today... Exactly the point. The key element of the design process is getting the styling right. Ludicrously over-engineered build quality then has to complete the impression of a thoroughgoing but civilised unreasonableness. They will need totally unreasonable speakers to complete the statement. Ian, may I suggest the Acapella Sphäron? http://www.zulum.co.kr/bbs/data/ACAPELLA/spharon.jpg Ion tweeters... If I had a Peterbuilt I'd go to the shop for cigs in it, even though I don't smoke. :-) cheers, Ian O. -- Oskari Heinonen * University of Helsinki * Department of Computer Science * http://www.cs.Helsinki.FI/Oskari.Heinonen/ |
#23
|
|||
|
|||
bill ramsay wrote: On Mon, 25 Jul 2005 16:28:16 GMT, Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Patrick, I don't know if this is much use, but at http://audiobizarro.com/ you can see a couple of photographs of amps created with the 212E, the first is on three chassis, whilst the second, is on more reasonable one. I seem to remembe seeing diagrams of these, but the site has changed somewhat and they have dissapeared/ watch your fingers. bill Here's a fine example of agricultural functionality over aesthetics. It was built in the late '60s for the Seventh Day Adventists for their big ralleys and was the "prime mover" for a stadium full of Vitavox horns/GP1 drivers. It uses a pair of 212s driven by a pair of 845s driven by a pair of KT66s. The clock on the mixer works, haven't got around to finding out if the rest works (chicken!!!). http://homepages.kcbbs.gen.nz/moby/bigamp.JPG Caution - 1.7MB M |
#24
|
|||
|
|||
"Patrick Turner" wrote
...Rgin of 1k is easily overcome by using a CF tube with Ro = 100 ohms. Of course, I was commenting on the design posted by Oskari. Have you discussed this CF idea with your customer? Even when gI isn't used, a CF is supposed to improve the sound Some say, others disagree. as borne out with the Ongaku amp, which has a CF drive to the 211 grid anyway. Also wondering what the cross-hatching on many of the cap symbols means. Is it permissible to use a distributed load with these things? Not much gain to play with but DF could be significantly improved all the same with perhaps 30% of the winding on the cathode. Need plenty drive volts though. As I explained in several posts, the last of which was last night, I don't generally read your posts. the use of a high Ea, RL of 15k, and almost all A1 op I was talking about the circuit posted by Oskari will give a DF of 6 without CFB. Par for a SET. 6dB of CFB will double the drive voltage from around 70vrms to 140vrms, Correct. And halve the effective input impedance. You have 1000V HT, what's the problem? and make matters worse, What matters? How? with very little reduction of Ro and thd if any. Rubbish. In proportion to the amount of feedback. As I said, not much gain to play with, but the improvement in DF without the use of global feedback could be worthwhile. Do you know what speakers your target final user has? My question was whether a distributed load was allowable in the tradition of the genre. As for thd, much would depend on how well the voltage amp and output stages suit each other. Feedback round the output stage will reduce the distortion produced by that stage in proportion. The problem of designing a voltage amp for best fit, distortion-wise, would be the same. If FB is to be used, it should be global Says who? and around a tube drive stage with no ISTs, and with an OPT with about 50kHz open loop bw to start with. Have you discussed this global fb idea with your customer? And why ask a question if you already know all the answers? Still wondering what the crosshatching means... cheers, Ian |
#25
|
|||
|
|||
" wrote: bill ramsay wrote: On Mon, 25 Jul 2005 16:28:16 GMT, Patrick Turner wrote: Anyone have any experience in using the 212 triode for an SET amp? I have been asked for some input to a project a man is about to embark on; he has the tubes, but needs a little help. Patrick Turner Patrick, I don't know if this is much use, but at http://audiobizarro.com/ you can see a couple of photographs of amps created with the 212E, the first is on three chassis, whilst the second, is on more reasonable one. I seem to remembe seeing diagrams of these, but the site has changed somewhat and they have dissapeared/ watch your fingers. bill Here's a fine example of agricultural functionality over aesthetics. It was built in the late '60s for the Seventh Day Adventists for their big ralleys and was the "prime mover" for a stadium full of Vitavox horns/GP1 drivers. It uses a pair of 212s driven by a pair of 845s driven by a pair of KT66s. The clock on the mixer works, haven't got around to finding out if the rest works (chicken!!!). http://homepages.kcbbs.gen.nz/moby/bigamp.JPG Caution - 1.7MB M Well now Bill, that's quite some lotta gear you got there. But yeah, quite some complexity, and obviously tailor made for the prosletizing of the Faith, and I notice that on the bottom left hand corner a meter with control knob for Demonic Control Effect, and top right hand corner God's Grace Effect, and in the middle there's a knob labelled Converted Soul Level, my my, what a fine old time they musta had back in those glory to be days. Sure it was the late sixties? more like late fifties maybe, but hey maybe the gear was purchased from Billy Graham, he was big back then, and they used to broadcast live in Oz via special heavenly transmit link from the US that could only have been arranged between Billy and Our Redeemer, after a few serious prayers. Patrick Turner. |
#26
|
|||
|
|||
Ian Iveson wrote: "Patrick Turner" wrote ...Rgin of 1k is easily overcome by using a CF tube with Ro = 100 ohms. Of course, I was commenting on the design posted by Oskari. Have you discussed this CF idea with your customer? I doubt he'd have the foggiest idea if i did. But no, not yet. Even when gI isn't used, a CF is supposed to improve the sound Some say, others disagree. as borne out with the Ongaku amp, which has a CF drive to the 211 grid anyway. Also wondering what the cross-hatching on many of the cap symbols means. Is it permissible to use a distributed load with these things? Not much gain to play with but DF could be significantly improved all the same with perhaps 30% of the winding on the cathode. Need plenty drive volts though. As I explained in several posts, the last of which was last night, I don't generally read your posts. Yes, I am aware of that, and I have no comment. the use of a high Ea, RL of 15k, and almost all A1 op I was talking about the circuit posted by Oskari will give a DF of 6 without CFB. Par for a SET. Its not, actually. Eg, many ppl use about a 3.5k RL for 300B, and go for maybe 9 watts, and since Ra is 800 ohms, and the OPT winding R is equivalent to say 350 ohms, then the DF = 3,500 / ( 800 + 350 ) = approx 3, not 6, if you are lucky. 6dB of CFB will double the drive voltage from around 70vrms to 140vrms, Correct. And halve the effective input impedance. You have 1000V HT, what's the problem? The problem is that in producing twice the drive voltage, you have twice the thd in the drive amp, which may/may not cancel the output stage's thd, maybe the thd outcome is worse with a the higher drive voltage. and make matters worse, What matters? How? with very little reduction of Ro and thd if any. Rubbish. In proportion to the amount of feedback. Its not quite that simple as you suggest. Using only 6 dB of CFB in the output stage means that there will be some of the 2H converted to 3H and higher products, and the thd of the drive stage will add to this. The output stage without the transformer complexities of a CFB winding is a fine output stage in its own right; there is just no need for the local CFB. Other tubes, 805, 833, would benefit from CFB because the gains of these other triodes is high, and the Ra also high, which needs a lot of reduction. Using CFB with the 212 is like operating a Quad II in triode, and of course one must retain the CFB since one cannot afford to waste the windings, but all that does is apply a feeble 3dB of NFB to the already lowish Ro output stage. I have done this though, and used KT88 as well, and the CFB merely compensates for the high winding R of the Quad OPT. I then applied 12 dB of global FB around the revised Quad amp, and the sound became better. We'd tried 6 dB and triode, and that wasn't so good. The spectral content in the thd is slightly better with triodes in such amps, compared to a tetrode output stage. But I digress. A big fat triode like the 212 is a fairly fine thing without local NFB. Global FB requires careful design, but its very effective if used right. As I said, not much gain to play with, but the improvement in DF without the use of global feedback could be worthwhile. Do you know what speakers your target final user has? He wants a switch to change from "with FB" to "without FB". The only time the DF will be really healthy with no global FB is where the ratio of RL to Ra is especially high, and I place that at around 15k for 212, and the plate voltage is high, and the load line is thus as flat as possible, with as wide a max swing voltage as possible. Its a challenge to get wide BW with a 15k load though; at full power voltage, to get the onset of oveload distortion at 20Hz would require ZLp = 15kohms at around 14Hz, so Lp should be 170H, an almost impossible figure to attain. My question was whether a distributed load was allowable in the tradition of the genre. By all means be my guest and use CFB if you want to with a 212. I would decline to. CFB is just series voltage FB, and as such it is frowned upon by those involved in the "tradition of the genre", presumably you mean the cult of the SET amp. Not a bad cult, they don't have a strange messiah who tries to get them all to suicide at the right moment. As for thd, much would depend on how well the voltage amp and output stages suit each other. Feedback round the output stage will reduce the distortion produced by that stage in proportion. Yes, but as i said, the drive stage then increases its distortion, and ther's no global FB to correct that. Quad II didn't adopt CFB without also using some global, since the pentode drive stage isn't the world's cleanest driver stage. The role of CFB with power tetrodes is altogether different to use with triodes. The CFB with tetrodes also means the FB is applied around the screen circuit, and the effect is somewhat unique, since the 3H of the tube is reduced more than one would think it would. based on gain-feedback equations very much like it is with UL. But the reduction in Ra with CFB around a tetrode is about twice as a effective as plain UL, which does have a big linearizing effect, but the Ra is still high with plain UL. The beauty of the CFB around tetrodes is that the lion's share of thd reduction Ra reduction, and phase shift reduction is done in the output stage without paying much of a price in raising the required drive voltage above what a triode output stage would need. Of course ppl forget they could have class A2, or AB2 PP triode amps and get oddles more power, but not have to use CFB or UL with tubes such as 6L6, 807, 6550, KT66 et all. The GE book I have with a pair of DA1505 in PP which are the same as 212 afaik simply had a normal output tranny with a 2.5kV supply. The power output was about 1,100 watts, and no need for CFB because the drive voltage was already large. A pair of KT66 in CF were used to drive a 1:1 coupling tranny to the 1505 grids. The primary was cathode to cathode, and the two 1/2 secondaries were taken to the low impedance fixed grid bias supply. Caps shunt the ends of P and S windings at the same signal to extend the HF coupling. the drive voltage to the CF buffer stage is simply a pair of 6BQ5 in PP triode and tranny coupled to the CF grids with yet another 1:1 coupling tranny. considerable signal voltage needed to be applied to the 'Q5 grids. These lovely old bangers were PA amps, they were only required to go from 75Hz to 8 kHz, any more might be a problem. Anyway, there is simply no need whatever to use an 845 to drive a 212. But since the tubes exist, there is always someone who will use them in ways that seem unecessary. The PA stadium amp didn't have global FB; that would have been quite impossible around so many stages and trannies, but thd was 7% at clip, which is a lot for a PP amp, since its all mainly odd order, but 3% at 250 watts. The problem of designing a voltage amp for best fit, distortion-wise, would be the same. If FB is to be used, it should be global Says who? I am, with regard to this particular tube and the intended use. Feel free to disagree. and around a tube drive stage with no ISTs, and with an OPT with about 50kHz open loop bw to start with. Have you discussed this global fb idea with your customer? As I said, he wants switchable FB. And why ask a question if you already know all the answers? I designed a colleagues GM70 amp with a CF driver buffer, CR coupled gain tube, normal input and global FB, and the amp ran very well and sounded good. I never assume I know all the answers. Patrick Turner. Still wondering what the crosshatching means... cheers, Ian |
#27
|
|||
|
|||
"Oskari Heinonen" wrote
Ian, may I suggest the Acapella Sphäron? http://www.zulum.co.kr/bbs/data/ACAPELLA/spharon.jpg Ion tweeters... Perfect! Er...odd place to put the wardrobe :-) cheers, Ian |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Some Recording Techniques | Pro Audio | |||
Some Mixing Techniques | Pro Audio |