Thank you Dave and KevinKr too.
Dave have you a suggestion for the value/wattage for the screen resistor, ie the value that Sherwood used or is that flawed and an issue? You mentioned 100 ohms before, do you believe that is still right, after looking at the schematic? By the way, the cap for the screens is a IL, TTA, 22uf at 500v, would a TVA or another brand be be a better choice? Any suggestion on stabilizing the power supply for these doubler type circuits?
And, Kevinkr thanks, I wasn't paying that much attention to the layout. Do you have a suggestion for the fl pole change? The 150K and .068 works well with the EH tubes, better than the 300k and .047 that Fisher and other used which the new manuf. tubes do not seem to like.
By the way, the help from all concerned is very much appreciated!
If you have a suggestion for decreasing the feedback, I am all ears. Seems now it must be at least 20db (not to worried about a lower output overall, it is quite loud now). One thing at a time, so I need to get this screen issue settled, and pitch the old tubes (have a second new set).
Dave have you a suggestion for the value/wattage for the screen resistor, ie the value that Sherwood used or is that flawed and an issue? You mentioned 100 ohms before, do you believe that is still right, after looking at the schematic? By the way, the cap for the screens is a IL, TTA, 22uf at 500v, would a TVA or another brand be be a better choice? Any suggestion on stabilizing the power supply for these doubler type circuits?
And, Kevinkr thanks, I wasn't paying that much attention to the layout. Do you have a suggestion for the fl pole change? The 150K and .068 works well with the EH tubes, better than the 300k and .047 that Fisher and other used which the new manuf. tubes do not seem to like.
By the way, the help from all concerned is very much appreciated!
If you have a suggestion for decreasing the feedback, I am all ears. Seems now it must be at least 20db (not to worried about a lower output overall, it is quite loud now). One thing at a time, so I need to get this screen issue settled, and pitch the old tubes (have a second new set).
When output stage components get unexpectedly fried this is often a sign of parasitic oscillation. You think you have an audio amp, but the valves don't know that. They will happily act as a high-power VHF/UHF oscillator if they see the right circuit conditions. Maybe one of the capacitors you changed has more or less inductance, or sufficiently lower ESR that an HF resonance is no longer damped. The "best" capacitors can prove disastrous for some circuits! The fact that you got the same result by "improving" another amp suggests that something is inappropriate with your new components or wiring technique.
Screen grids can sometimes do strange things, as they can be either an input or output.
Screen grids can sometimes do strange things, as they can be either an input or output.
fyi.. Reducing the feedback will increase the output.
Graphics: GIF format usually works much better for diagrams. Use JPEG for photos.
..Todd
Last edited:
I have found that it takes at least 47 ohms to keep 7591s and 7189s out of trouble in this regard. Other tubes like 6550 and the like can get by with less. I always use 100 ohms for 7591 type tubes, and it has always been completely effective. Keep the original 470 ohm dropping resistor and screen filter cap in place, or better yet, replace the 470 resistor with a suitably lower resistance choke for improved regulation. Just add the new stability resistors at each output tube as well. 1/2 watt, of any type construction is fine, as this is non-critical in these locations.
Dave
Dave
thanks Dave, I will do that tonight.
To reply to DF96: None of the Fisher power caps were bad, only the grid supply. But I have found that thebrand/type of cap does make a difference in the end.
To Todd: never fooled with feedback before. I know it reduces distortion. A friend has an amp with adjustable feedback, and when he turns it from 10 to 3 to zero, the sound decreases and the sound stage moves forward of the speakers. The later is what I was aiming at.
To reply to DF96: None of the Fisher power caps were bad, only the grid supply. But I have found that thebrand/type of cap does make a difference in the end.
To Todd: never fooled with feedback before. I know it reduces distortion. A friend has an amp with adjustable feedback, and when he turns it from 10 to 3 to zero, the sound decreases and the sound stage moves forward of the speakers. The later is what I was aiming at.
I think you would need to look at a schematic of that implementation to see what's really going on. Adjusting feedback up and down with a pot is not a straightforward matter.
..Todd
One way to experiment with adjustable feedback would be to replace the 1200 cathode resistor in the AF amp stage with a 1K linear pot and 200 ohm resistor. The top of the pot would go to the cathode, the bottom of the pot to ground through the 200 ohm resistor. Then connect the feedback connection to the wiper of the pot. When the wiper is up full so as to connect to the cathode, the circuit is as it always was. As the wiper is turned down towards the 200 ohm resistor end of the pot, feedback is ultimately reduced to 1/6 the level of the original amount. If the original amount was 20 db as you suggest, it will only be ~ 4.5 db at the low end of the pot. This will produce a significant change in gain, distortion, and damping factor.
Dave
Dave
Deleted my original post expressing an objection to this idea - from a bias perspective it should be OK. From the perspective of the switchable 2nd/3rd order Sallen-Key VCVS HP filter in the front end I am less certain. Changing the loop feedback may affect the impedance at the cathode and should reduce the voltage efficiency of the pseudo cathode follower the VCVS is relying on for its positive feedback - this may affect the Q of the filter. (With full feedback applied to this node I would perhaps naively expect the voltage at the cathode to be very nearly 100% of the value at the grid which should allow for the implementation of higher Q filters than would otherwise be possible.)
Unless an amplifier is sufficiently distortion-free that it really needs no feedback at all, 4.5dB of feedback will almost certainly sound worse than no feedback at all. This is because it is enough to create noticeable higher-order terms but insufficient to suppress them. The correct level of feedback is an engineering design decision.
Hi Dave,
I decided to give the added 100 ohm resistor some time to see what effect or affect it would have. Does seem to stablize the output tubes and the plate current too. The tube that started all this was bad, a hot spot on the screen...it is now in the trash.
Never did anything further with the feedback thought, but for those who were expressing their ideas, it was local feedback on the power tubes only not global or stage. It just seems that 20 db of feedback is a heck of a lot. Engineers decisions on the design aside, I have found many did things wrong ( like 300k ohms on a grid for a 7591 tube) and other assorted things...so the idea of 20 db feedback just seemed like another attempt to make it loud in 1962 whereas in 2011 the speakers are more efficient and don't need the forced power.
On the cathode resistors, every one of the Xicon brands have failed in every amp I had them in. Since the first post I have switched to Vishay CMF or CFF type and have had none fail.
Thanks again Dave and everone else.
Last edited:
20 dB of (negative) feedback was not to make the amp louder, it will reduce the volume. As discussed above, (negative) feedback reduces distortion and increases damping factor. If your friends amp had the volume decreasing with decreasing feedback, then the feedback was positive (good way to make an oscillator, not an audio amplifier) or the label in the switch/pot was wrong. Was the amp DIY?
I know it reduces distortion, and also reduced the volume for the same setting. Something has gotten lost in my posting, sorry it must be me.
What I was after with the notion was to get the stage more forward (right now it is 90% behind the speakers).
My friend has a Mesa Baron, and when the feedback is zero the stage is completely in front of the speakers and as you add negative feedback, the stage receeds (and yes the volume is a bit lower). I realize that amp is very different than a 1960's 7591 based amp and the Baron looks to only use one layer of global feedback rather than each stage having local.
My amp is crystal clear, dramatic detailing with a black background, but the stage is not where I think it could be with some carefull adjustments.
By the way, if any one has any doubts as to how much of an improvement OCC mono-crystal copper wire will make, I can assure you it is jaw dropping.
What I was after with the notion was to get the stage more forward (right now it is 90% behind the speakers).
My friend has a Mesa Baron, and when the feedback is zero the stage is completely in front of the speakers and as you add negative feedback, the stage receeds (and yes the volume is a bit lower). I realize that amp is very different than a 1960's 7591 based amp and the Baron looks to only use one layer of global feedback rather than each stage having local.
My amp is crystal clear, dramatic detailing with a black background, but the stage is not where I think it could be with some carefull adjustments.
By the way, if any one has any doubts as to how much of an improvement OCC mono-crystal copper wire will make, I can assure you it is jaw dropping.
To what? Your bank balance?
I use teflon coated, but only because it is easier to solder in tight spots without melting the insulation.
I believe layout, correct component values and topology are going to have more of an effect than 'magic' wire or capacitors
The biggest variable in the system is probably going to be your speakers and room acoustics. Was the effect you saw in your friends amp while it was in his system or in yours? Have you tried moving your speakers (away from wall and corners, a little toe in etc)?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.