I removed the tone controls, balance pot and all the front panel switches except the tape monitor. I also installed my capacitor board, and replaced every cap and resistor in the thing. It's a different amp and quite capable now. I also replaced thevolume pot with a noble conductive plastic pot.
Sheldon
Sheldon
Re: Dynaco SCA-35
Hi, I also forgot to mention that a very knowledgeable audio engineer by the name of Dave Gillespie has developed a patented design of a EFB
enhanced fixed bias circuit that makes the SCA-35
effortless at 29% gain in wattage per channel and 43% less distortion. It has been said that by adding higher voltage premium coupling caps improve the sonics as well. Sonicap's brand I heard is oustanding as well as Mundorf's silver/gold's. using metal film resistors I would go with caddocks and polystyrene in the pF region. volume control with a log/taper Audio Note or TAD would also be good.
I also know that audioregenesis sells a module to replace the can caps as well. Get creative use the large Blackgate electrolytic caps to replace the can caps, and wow would be the sound you will get if you can do all of these you would have a great sounding vintage amplifier without possibly having to bypass all the controls and I have seen photos of a re-work and it would be a job to do that right! the person wanted to use the filter switch to take out the tone controls, but to switch the other way to add them back in.... go wonder. anyway my 2cents worth. I love talking with dave he sure knows all about dynaco amplifiers.
Lydia
Hi, I also forgot to mention that a very knowledgeable audio engineer by the name of Dave Gillespie has developed a patented design of a EFB
enhanced fixed bias circuit that makes the SCA-35
effortless at 29% gain in wattage per channel and 43% less distortion. It has been said that by adding higher voltage premium coupling caps improve the sonics as well. Sonicap's brand I heard is oustanding as well as Mundorf's silver/gold's. using metal film resistors I would go with caddocks and polystyrene in the pF region. volume control with a log/taper Audio Note or TAD would also be good.
I also know that audioregenesis sells a module to replace the can caps as well. Get creative use the large Blackgate electrolytic caps to replace the can caps, and wow would be the sound you will get if you can do all of these you would have a great sounding vintage amplifier without possibly having to bypass all the controls and I have seen photos of a re-work and it would be a job to do that right! the person wanted to use the filter switch to take out the tone controls, but to switch the other way to add them back in.... go wonder. anyway my 2cents worth. I love talking with dave he sure knows all about dynaco amplifiers.
Lydia
Last edited:
Do you have a number for that patent? Looking at the circuit (a bypassed CCS to provide cathode bias), it seems like there's a lot of prior art and I'm unable to find any patents issued to David Gillespie relevant to this. Thanks!
edit: On careful reading, he does have a novel feature (at least novel to me) of varying the idle current with B+ sag. I still don't see any patent, though, or any mention of one in his writeup. Dave was good enough to at least provide some back-up measurements, so kudos to him for that. He clearly doesn't like my LED bias scheme, but missed a few of the essential elements of it.
enhanced fixed bias
SY, I don't think he has claimed a patent. He uses a negative three terminal regulator to establish a fixed voltage (not constant current) relative to the B+ by using the B+ rail to feed the adjusting terminal. I used this approach in a triode strapped el84 amp and it does work very well, and is a compact, easy to implement improvement over a simple resistor/cap combination. My decision, when I was comparing it to an array of leds was based on ease of implementation. I still bypass it with a cap (some bad habits die hard), but he (Dave G) says cap isn't needed. I am going to try this with a 7591 based amp somewhere in the future (have tubes and trannies laying around) as they seem to fit the bill as suitable candidates as well (low bias voltage).
SY, I don't think he has claimed a patent. He uses a negative three terminal regulator to establish a fixed voltage (not constant current) relative to the B+ by using the B+ rail to feed the adjusting terminal. I used this approach in a triode strapped el84 amp and it does work very well, and is a compact, easy to implement improvement over a simple resistor/cap combination. My decision, when I was comparing it to an array of leds was based on ease of implementation. I still bypass it with a cap (some bad habits die hard), but he (Dave G) says cap isn't needed. I am going to try this with a 7591 based amp somewhere in the future (have tubes and trannies laying around) as they seem to fit the bill as suitable candidates as well (low bias voltage).
The output impedance of an LM317 (the text above says negative regulator so it would actually be an LM337) rises sharply with frequency. So if you want a neutral use (or lack there of) of local cathode feedback, you won't get it with just a LM317 alone. At low frequencies, the LM317 acts like a short and thus provides no local feedback on the output tubes, but at high frequencies, it can be several ohms. So I'd bypass it with a healthy sized cap to counteract the rising impedance with the falling impedance of the cap. As always, I'd also bypass the electrolytic cap with a film cap.
BTW, this thread is pretty old, I guess they never die...
Sheldon
BTW, this thread is pretty old, I guess they never die...
Sheldon
Hi Sy -- Regarding a patent on my EFB(tm) modification, that was an effort I chose not to pursue. Rather, I simply chose to introduce the concept into the public domain through an article I wrote and published on the Tronola website, so that all diyers could benefit from the concept if they wished to.
I do have a trademark claim on "EFB" and "Enhanced Fixed Bias", supported (in part) by the use of that notation with the circuit I published, as offered by the Audio Regenesis group in their offering of replacement boards for Dynaco 6BQ5 products. In particular, the power supply boards that Audio Regenesis offers to replace the can caps in the SCA-35 and ST-35, are both available with the EFB(tm) modification included on the board.
Regarding the bypass cap, as mentioned, the characteristic of a rising impedance with respect to frequency at the output of a typical 3T regulator is well established. The LM337 is certainly no different in this regard. However, in developing the modification, I found that by simply using a bypass cap of equal value to the original (to keep the modification simple), the output impedance of the regulator circuit remained extremely low throughout the audio spectrum (20 Hz - 20 kHz). Typically, this was well under .10 ohm through out this bandwidth, so as to be largely insignificant. The cap can certainly be enhanced with larger values and cap bundles for maximum effect. The bypass cap also plays a part in maintaining the stability of the regulator as would be expected.
Regarding your LED approach to biasing, I will certainly go back and review that approach to determine the essential elements I missed. My comments regarding that approach were largely aimed at the concern of a changing B+ level, without a corresponding change in output stage bias level. The comments are largely theory based, but also backed by work with other designs where the output stage bias was regulated, while the B+ was not. Against this backdrop, with the EFB biasing method, the output stage bias is automatically adjusted appropriately (under any operating condition) for changing B+ conditions only (due to any reason), while otherwise remaining firmly fixed in the face of changing current draw produced by the output stage itself. In short, it offers the best of both cathode and fixed biasing methods.
I hope this helps clear up any confusion regarding my EFB modification.
Dave
I do have a trademark claim on "EFB" and "Enhanced Fixed Bias", supported (in part) by the use of that notation with the circuit I published, as offered by the Audio Regenesis group in their offering of replacement boards for Dynaco 6BQ5 products. In particular, the power supply boards that Audio Regenesis offers to replace the can caps in the SCA-35 and ST-35, are both available with the EFB(tm) modification included on the board.
Regarding the bypass cap, as mentioned, the characteristic of a rising impedance with respect to frequency at the output of a typical 3T regulator is well established. The LM337 is certainly no different in this regard. However, in developing the modification, I found that by simply using a bypass cap of equal value to the original (to keep the modification simple), the output impedance of the regulator circuit remained extremely low throughout the audio spectrum (20 Hz - 20 kHz). Typically, this was well under .10 ohm through out this bandwidth, so as to be largely insignificant. The cap can certainly be enhanced with larger values and cap bundles for maximum effect. The bypass cap also plays a part in maintaining the stability of the regulator as would be expected.
Regarding your LED approach to biasing, I will certainly go back and review that approach to determine the essential elements I missed. My comments regarding that approach were largely aimed at the concern of a changing B+ level, without a corresponding change in output stage bias level. The comments are largely theory based, but also backed by work with other designs where the output stage bias was regulated, while the B+ was not. Against this backdrop, with the EFB biasing method, the output stage bias is automatically adjusted appropriately (under any operating condition) for changing B+ conditions only (due to any reason), while otherwise remaining firmly fixed in the face of changing current draw produced by the output stage itself. In short, it offers the best of both cathode and fixed biasing methods.
I hope this helps clear up any confusion regarding my EFB modification.
Dave
Thanks for clearing that up, Dave!
The thing you missed on the LED bias was running the tubes in pentode and regulating the screens to adjust idle. So it's really no different than any fixed bias amp in that respect. It's not a suitable choice for UL or triode, yours is a better approach for those topologies.
The thing you missed on the LED bias was running the tubes in pentode and regulating the screens to adjust idle. So it's really no different than any fixed bias amp in that respect. It's not a suitable choice for UL or triode, yours is a better approach for those topologies.
Hi Sy -- Regulating the screens in combination with your LED approach is certainly an "essential element" I missed! That would clearly maximize the LED biasing approach when used, and obviously eliminate the concern I was mentioning. So implemented for pentode operation, LED biasing is certainly a very worthwhile option then.
Thank-you for clearing that up for me!
Dave
Thank-you for clearing that up for me!
Dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.