Home |
Search |
Today's Posts |
#1
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]()
I wrote before about my old HP606A 6 band oscilator with AM
modulation. The switchable 10dB stepped output attenuator is enclosed and buried in metalwork and too hard to pull out to examine. No need, there's nothing wrong, and its quite nicely designed for constant resistance 50 ohms output R and compensated loading on the 6CL6 output tubes. Operating the concentric small knob on the attenuator switch gives you fine adjustment of any chosen level of the switch between that level and no signal. The actual adjustment to the sig level is by adjusting the Vdc bias on the 12B4 modulator with a pot so that the 6CL6 can be run and normal idle full signal with 4mA idle current, or cut off, so no signal appears. But if the internal or external modulation is set for 100%, the sigal becomes much over modded when one reduces the carrier level with this fine adjust vernier knob to alter Vdc bias. So the level must be chosen carefully, preferably without using the vernier at all, which is possible if the radio under test has a vari cap coupling to its antenna, and the signal must be monitored with CRO because the RF level meter gives different mod % reading for different ranges of F for the same % of modulation. The 606A has an interesting feature in that it has a 6AW8 differential amp which allows dc control to calibrate output levels and also to amplify the modulation signal and have a detected AF signal applied as negative feed back to keep the modulation envelope linear. The gain reduction is 1/65, so that if I had 3% AF envelope THD at 75% modulation, with NFB this almost completely dissapeared to a value of about 0.05%.I tried shunting the FB, and the modulation linearity was poor above 66%, and of course it only took a tiny amount of drive to get high modulation. Thus its possible to get full 100% mod with under 1%THD at below 1kHz. But above 1 kHz, and by about 10kHz, the modulation cannot be increased to 100% despite all that NFB, more like 50% before THD sets in bad by 20KHz. The distortion is a slew rate type and I could not find the cause. The internal detector workings looked fine. But the AF BW of a receiver can be tested using 50% mod when the detected AF to 20kHz is way under 1% and quite negligible if the incoming modulating from a low THD source is used. With carrier at 455kHz for IF testing it is a very good tool to use, and if an AM radio reproduces an clean AF wave from a 100% modded incoming carrier then the radio is usually very good sounding. The audio energy lessens above 1kHz and its rare 100% mod occurs. If the AM radio is flat to say 5kHz, with -3dB at 7kHz and fast roll off above, then usually the radio sounds extraordinarily good compared to the usual crap where its -3dB at 2 kHz, with rough sounding THD&IMD products replacing the real source HF content. The 606A will help keep the workshop warm in winter. My own SE pentode AM sig gene is 20db simpler and does most of what I want but not any short wave and the short wave gene I made with a 6BE6 and about 60% AM, very rough, is rather primitive and a bit harmoniky. The 606 would be better. But an afternoons work to change my own gene to have audio FB and a differential amp drive for the grid modulation should reduce the approx 1.3% THD at 90% mod I now get to very low levels indeed, and push the envelope to real 100% mod without quitting at 95%. As it is, my own gene gives wider BW with 95% mod than the 606. Because the 606 goes to 65MHz, its possible to make an FM gene and then double the F in an external box, and cover the FM band. But I probly won't get around to it, and my home brew stereo FM gene in a metal cigar box is quite good enough. The BA4014 runs off 1.4Vdc, no need for 15Kg and lotsa toobes and volts. My own AM gene can be set for 455kHz and I have a switched in group of 7 x 65V zener diodes used as varicap diodes set up so they don't conduct with a variable Vdc bias which changes their capacitance nicely. So the gene produces a wobbulated 455kHz signal, +/- 44kHz, and you see the IF selectivity shape while adjusting it if wanted. I guess the 606A could be made to wobble its oscillator F about +/- 100kHz quite easily at about 47MHz carrier, not much C change is involved. I think I know why aircraft were so slow to get off the ground in 1955. Their electronics were so heavy. And generally, even with their constant smoking, young blokes were fitter in 1955, and because there was so much luggin 'evvy gear round. Nowdays, nearly all don't smoke but they are mostly overweight, and fobidden to carry more than 16 Kg. Patrick Turner. |
#2
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]() "Patrick Turner" wrote in message ... Thus its possible to get full 100% mod with under 1%THD at below 1kHz. But above 1 kHz, and by about 10kHz, the modulation cannot be increased to 100% despite all that NFB, more like 50% before THD sets in bad by 20KHz. The distortion is a slew rate type and I could not find the cause. The internal detector workings looked fine. Perhaps a load of the modulator is a high-Q LC tank, and it simply can not pass sidebands at high enough AF? |
#3
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]()
On Nov 21, 6:34*am, Patrick Turner wrote:
I wrote before about my old HP606A 6 band oscilator with AM modulation. The switchable 10dB stepped output attenuator is enclosed and buried in metalwork and too hard to pull out to examine. No need, there's nothing wrong, and its quite nicely designed for constant resistance 50 ohms output R and compensated loading on the 6CL6 output tubes. Operating the concentric small knob on the attenuator switch gives you fine adjustment of any chosen level of the switch between that level and no signal. The actual adjustment to the sig level is by adjusting the Vdc bias on the 12B4 modulator with a pot so that the 6CL6 can be run and normal idle full signal with 4mA idle current, or cut off, so no signal appears. But if the internal or external modulation is set for 100%, the sigal becomes much over modded when one reduces the carrier level with this fine adjust vernier knob to alter Vdc bias. So the level must be chosen carefully, preferably without using the vernier at all, which is possible if the radio under test has a vari cap coupling to its antenna, and the signal must be monitored with CRO because the RF level meter gives different mod % reading for different ranges of F for the same % of modulation. The 606A has an interesting feature in that it has a 6AW8 differential amp which allows dc control to calibrate output levels and also to amplify the modulation signal and have a detected AF signal applied as negative feed back to keep the modulation envelope linear. The gain reduction is 1/65, so that if I had 3% AF envelope THD at 75% modulation, with NFB this almost completely dissapeared to a value of about 0.05%.I tried shunting the FB, and the modulation linearity was poor above 66%, and of course it only took a tiny amount of drive to get high modulation. Thus its possible to get full 100% mod with under 1%THD at below 1kHz. But above 1 kHz, and by about 10kHz, the modulation cannot be increased to 100% despite all that NFB, more like 50% before THD sets in bad by 20KHz. The distortion is a slew rate type and I could not find the cause. The internal detector workings looked fine. But the *AF BW of a receiver can be tested using 50% mod when the detected AF to 20kHz is way under 1% and quite negligible if the incoming modulating from a low THD source is used. With carrier at 455kHz for IF testing it is a very good tool to use, and if an AM radio reproduces an clean AF wave from a 100% modded incoming carrier then the radio is usually very good sounding. The audio energy lessens above 1kHz and its rare 100% mod occurs. If the AM radio is flat to say 5kHz, with -3dB at 7kHz and fast roll off above, then usually the radio sounds extraordinarily good compared to the usual crap where its -3dB at 2 kHz, with rough sounding THD&IMD products replacing the real source HF content. The 606A will help keep the workshop warm in winter. My own SE pentode AM sig gene is 20db simpler and does most of what I want but not any short wave and the short wave gene I made with a 6BE6 and about 60% AM, very rough, is rather primitive and a bit harmoniky. The 606 would be better. But an afternoons work to change my own gene to have audio FB and a differential amp drive for the grid modulation should reduce the approx 1.3% THD at 90% mod I now get to very low levels indeed, and push the envelope to real 100% mod without quitting at 95%. As it is, my own gene gives wider BW with 95% mod than the 606. Because the 606 goes to 65MHz, its possible to make an FM gene and then double the F in an external box, and cover the FM band. But I probly won't get around to it, and my home brew stereo FM gene in a metal cigar box is quite good enough. The BA4014 runs off 1.4Vdc, no need for 15Kg and lotsa toobes and volts. My own AM gene can be set for 455kHz and I have a switched in group of 7 x 65V zener diodes used as varicap diodes set up so they don't conduct with a variable Vdc bias which changes their capacitance nicely. So the gene produces a wobbulated 455kHz signal, +/- 44kHz, and you see the IF selectivity shape while adjusting it if wanted. I guess the 606A could be made to wobble its oscillator F about +/- 100kHz quite easily at about 47MHz carrier, not much C change is involved. I think I know why aircraft were so slow to get off the ground in 1955. Their electronics were so heavy. And generally, even with their constant smoking, young blokes were fitter in 1955, and because there was so much luggin 'evvy gear round. Nowdays, nearly all don't smoke but they are mostly overweight, and fobidden to carry more than 16 Kg. Patrick Turner. Just a coincidence, but I'm in the process of troubleshooting a 606B - very similar to the 606A but with a solid state P/S - the rest is tubes like the 606A. It won't oscillate on any band! So far the P/S checks out (300, 200 and 25 VDC), although the 25 VDC rectifier diodes seem to get a bit warm - must check that. I'm about to tackle the RF oscillator section. I have the manual, actually for both the 606Aand 606B - the latter downloaded from an HP site, the 606A a paper copy that came with the 606B unit - not the same, the "B" has a tube-based P/S. The thing is very large and heavy! It's built like a brick outhouse and takes up a huge amount of bench space. Cheers, Roger |
#4
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]()
On Nov 23, 1:41*pm, "Alex" wrote:
"Patrick Turner" wrote in message ... Thus its possible to get full 100% mod with under 1%THD at below 1kHz. But above 1 kHz, and by about 10kHz, the modulation cannot be increased to 100% despite all that NFB, more like 50% before THD sets in bad by 20KHz. The distortion is a slew rate type and I could not find the cause. The internal detector workings looked fine. Perhaps a load of the modulator is a high-Q LC tank, and it simply can not pass sidebands at high enough AF? The actual modulator does have a PP tuned RF transformer with tuning only on the primary. The anodes of a pair of 6CL6 connect to the RF tank at about 1/2 the the primary turns out from the CT, and the coil is litz wire, so maybe the Q is is high enough to cut sidebands. I don't have the full spec sheets. The 606 has 6 bands, and presumably the Q varies considerably, but there is an internal two tone oscillator which produces either 400Hz or 1,000 Hz and you can get full 100% mod at these F. Patrick Turner. |
#5
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]()
On Nov 24, 8:41*am, Engineer wrote:
On Nov 21, 6:34*am, Patrick Turner wrote: I wrote before about my old HP606A 6 band oscilator with AM modulation. The switchable 10dB stepped output attenuator is enclosed and buried in metalwork and too hard to pull out to examine. No need, there's nothing wrong, and its quite nicely designed for constant resistance 50 ohms output R and compensated loading on the 6CL6 output tubes. Operating the concentric small knob on the attenuator switch gives you fine adjustment of any chosen level of the switch between that level and no signal. The actual adjustment to the sig level is by adjusting the Vdc bias on the 12B4 modulator with a pot so that the 6CL6 can be run and normal idle full signal with 4mA idle current, or cut off, so no signal appears. But if the internal or external modulation is set for 100%, the sigal becomes much over modded when one reduces the carrier level with this fine adjust vernier knob to alter Vdc bias. So the level must be chosen carefully, preferably without using the vernier at all, which is possible if the radio under test has a vari cap coupling to its antenna, and the signal must be monitored with CRO because the RF level meter gives different mod % reading for different ranges of F for the same % of modulation. The 606A has an interesting feature in that it has a 6AW8 differential amp which allows dc control to calibrate output levels and also to amplify the modulation signal and have a detected AF signal applied as negative feed back to keep the modulation envelope linear. The gain reduction is 1/65, so that if I had 3% AF envelope THD at 75% modulation, with NFB this almost completely dissapeared to a value of about 0.05%.I tried shunting the FB, and the modulation linearity was poor above 66%, and of course it only took a tiny amount of drive to get high modulation. Thus its possible to get full 100% mod with under 1%THD at below 1kHz. But above 1 kHz, and by about 10kHz, the modulation cannot be increased to 100% despite all that NFB, more like 50% before THD sets in bad by 20KHz. The distortion is a slew rate type and I could not find the cause. The internal detector workings looked fine. But the *AF BW of a receiver can be tested using 50% mod when the detected AF to 20kHz is way under 1% and quite negligible if the incoming modulating from a low THD source is used. With carrier at 455kHz for IF testing it is a very good tool to use, and if an AM radio reproduces an clean AF wave from a 100% modded incoming carrier then the radio is usually very good sounding. The audio energy lessens above 1kHz and its rare 100% mod occurs. If the AM radio is flat to say 5kHz, with -3dB at 7kHz and fast roll off above, then usually the radio sounds extraordinarily good compared to the usual crap where its -3dB at 2 kHz, with rough sounding THD&IMD products replacing the real source HF content. The 606A will help keep the workshop warm in winter. My own SE pentode AM sig gene is 20db simpler and does most of what I want but not any short wave and the short wave gene I made with a 6BE6 and about 60% AM, very rough, is rather primitive and a bit harmoniky. The 606 would be better. But an afternoons work to change my own gene to have audio FB and a differential amp drive for the grid modulation should reduce the approx 1.3% THD at 90% mod I now get to very low levels indeed, and push the envelope to real 100% mod without quitting at 95%. As it is, my own gene gives wider BW with 95% mod than the 606. Because the 606 goes to 65MHz, its possible to make an FM gene and then double the F in an external box, and cover the FM band. But I probly won't get around to it, and my home brew stereo FM gene in a metal cigar box is quite good enough. The BA4014 runs off 1.4Vdc, no need for 15Kg and lotsa toobes and volts. My own AM gene can be set for 455kHz and I have a switched in group of 7 x 65V zener diodes used as varicap diodes set up so they don't conduct with a variable Vdc bias which changes their capacitance nicely. So the gene produces a wobbulated 455kHz signal, +/- 44kHz, and you see the IF selectivity shape while adjusting it if wanted. I guess the 606A could be made to wobble its oscillator F about +/- 100kHz quite easily at about 47MHz carrier, not much C change is involved. I think I know why aircraft were so slow to get off the ground in 1955. Their electronics were so heavy. And generally, even with their constant smoking, young blokes were fitter in 1955, and because there was so much luggin 'evvy gear round. Nowdays, nearly all don't smoke but they are mostly overweight, and fobidden to carry more than 16 Kg. Patrick Turner. Just a coincidence, but I'm in the process of troubleshooting a 606B - very similar to the 606A but with a solid state P/S - the rest is tubes like the 606A. *It won't oscillate on any band! *So far the P/S checks out (300, 200 and 25 VDC), although the 25 VDC rectifier diodes seem to get a bit warm - must check that. I'm about to tackle the RF oscillator section. *I have the manual, actually for both the 606Aand 606B - the latter downloaded from an HP site, the 606A a paper copy that came with the 606B unit - not the same, the "B" has a tube-based P/S. *The thing is very large and heavy! *It's built like a brick outhouse and takes up a huge amount of bench space. Cheers, Roger- Hide quoted text - - Show quoted text - You confuse me. Did your 606B come with SS PSU or not? Anyway, my 606A has SS diodes, but there are 6 x 12B4 regulator tubes plus a couple of 6AW8 to drive the regs. There are a couple of heater supplies, plus rails of +300Vreg, and -200V reg. There should be a fuse on the front panel which is for the dc supply to the RF amp and oscillator stages and this is 0.15A. Its beside the mains fuse which is a higher value. The RF fuse blew soon after trying to use it and it was faulty 6CL6, probably an original tubes from 1956. The rest of the tubes are OK and still original. Once I replaced the pair of 6CL6, the unit came to life and all's well. Its a heavy big thing, and maybe it will go for another 50 years. Patrick Turner. |
#6
![]()
Posted to rec.audio.tubes
|
|||
|
|||
![]()
On Nov 24, 9:00*am, flipper wrote:
On Sun, 21 Nov 2010 03:34:51 -0800 (PST), Patrick Turner wrote: snip You might want to review this article I found. http://www.hpl.hp.com/hpjournal/pdfs...Fs/1959-04.pdf According to them the 'standard' receiver test is with 30% modulation and, as you can see from the graphs (about midway down), the 606A is not capable of doing 100% modulation to 20Khz at 'rated max' (apparently 3%) distortion. Thanks for the link to the .pdf above. The two 6CL6 work with low bias currents in mainly class B. Their modulation AF signal is applied to their commoned cathodes via a 12B4 in series with both cathodes and 0V. The grid signal to the 12B4 is derived from a differential amp using a 6AW8 triode-pentode tube. The AF oscillator is applied to the triode side of the 6AW8, and the detected AF signal from the AM output signal is applied to the 6AW8 LTP side as a series voltage negative FB signal. The differerential gain of the LTP is about 60, and is reduced to about unity with NFB so any distortion in the envelope shape, detector network with diode, or the LTP itself is cleaned up by the NFB. But because the PP 6CL6 have so much distortion without NFB, the distortion can only be reduced so much by FB. The FB loop employed is very prone to parasitics being generated. To overcome the likelhood of parasitic instabilities the open loop bandwidth of the LTP diff amp has to be carefully limited and the detected signal must be fairly free of phase shift as AF mod F rise past 2kHz. Yesterday, I spent some time re-designing my own smaller 2 band AM modulated RF gene so that it includes a diff amp with incoming AF applied to one side and the detected signal to the other. After much time spend experimenting with some counter intuitive ideas which I didn't think would work, then finding that they did work, I was able to get similar performance to the 606A just using a single grid modulated 6EJ7 as the modulator, with Ea at 220V, and Ia at about 2mA, and working into a single tuned RF tranny which I'd made 13 years ago. I used a 12AT7 as the LTP. Distortion becomes high when full modulation is tried above 6kHz like the 606A. But at low mod, say 30%, then I can get mod to over 20kHz and flat which allows testing of radio pass bandwidth. The distortion at 98% mod using 1 kHz is at least as good as the 606A, and at 75% it is probably less than 0.2%. Usually the distortions in recovered AF generated in old radios is much worse than what is produced in the sig generators so if there is 5%THD at 1 kHz AF at 75% mod, you know the radio could be doing a lot better, and IF alignment and detector design could be improved perhaps. Patrick Turner. |
Reply |
Thread Tools | |
Display Modes | |
|
|
![]() |
||||
Thread | Forum | |||
I have GOIA's recent meltdowns figured out | Audio Opinions | |||
Word Clock Generator with a fine frequency adjustment? | Pro Audio | |||
What freq is impedance figured at? | Vacuum Tubes | |||
Insertion of attenuator | Vacuum Tubes | |||
HAHA I FIGURED OUT THE holy grail of distorted guitar micing | Pro Audio |