Kofi,
Your case is sounding pretty complicated, but I will say that I've gotten the "Squeeeeeeeeee!" when I bypassed the feedback resistor, that is, shorted rather than opened the NFB loop when the leads were reversed. 100% *positive* feedback.
You might check to see if you have inadvertantly created a shorted loop (i.e. no resistance) among the adjacent components. Also, if you're running ultralinear, be sure you're not shorting to ground or screen to anode among connections at the OPT end.
Also, if you've changed leads on one side of the trafos, you may need to check to be sure you're pulling the feedback off the correct secondary taps. Try switching the taps see of those were also reversed.
Just swatting at gnats, I realize. But I hope you at least don't feel too alone. 🙂
--Jeff
Your case is sounding pretty complicated, but I will say that I've gotten the "Squeeeeeeeeee!" when I bypassed the feedback resistor, that is, shorted rather than opened the NFB loop when the leads were reversed. 100% *positive* feedback.
You might check to see if you have inadvertantly created a shorted loop (i.e. no resistance) among the adjacent components. Also, if you're running ultralinear, be sure you're not shorting to ground or screen to anode among connections at the OPT end.
Also, if you've changed leads on one side of the trafos, you may need to check to be sure you're pulling the feedback off the correct secondary taps. Try switching the taps see of those were also reversed.
Just swatting at gnats, I realize. But I hope you at least don't feel too alone. 🙂
--Jeff
OK, you've ruled out a reversed connection and positive feedbacck. It's time for an internal photograph or two.
I used a choke (20H, I think) followed by a 100uF electrloytic. That was suggested here, so I went with it.
Being pretty stupid, I'm not exactly sure what this means. Can you tell me which resistances I should check here?
Let me present this issue, which I encountered when I was initially looking for the source of the squeeeeeeeeeee. Since one channel was initially working fine with feedback and the other was not, I started to check resistances on both sides to see what didn't match.
When I checked resistance from the positive speaker terminal to the chassis on the left (working) channel, I got 330R. On the right one I got .68, I think. Feeling that the 330R resistance was correct, I started to check the right channel and stumbled upon the concept that the NFB was causing the problem.
I re-soldered a few components and replaced the right channel's NFB resistor / cap combo with a spare one that I bought for just such an emergency. After that, the left channel was squeeeeeeeeeeeing too. I fidgeted for a bit more and found that when I disconnected the NFB, everything was OK.
Also, I remember getting the exact same squeeeeeeeee when I was working on that cheapie K12 or S12 or S12-K amp (you know the one-- like, $120 mail order) about a year ago. I abandoned that amp as I got busy with other things at the time, so I never found the cause.
So, maybe I'm just screwing up something basic here? Probably.
I was afraid of that. OK, I'll promise to do that (tonight if I'm not at the UN too late), but be aware that I know I'm no Johan Potgieter. More like Jackson Pollock.
Kofi
Kofi,
All I meant was to make sure you aren't bypassing the FB resistor via shorts, such as by having component leads touching. Sounds like you've been checking that.
You mentioned a difference in speaker resistance positive to chassis. Check the feedback tap to ground resistance too. One other thing to try, although it can be a real pain, is to switch OPTs from left to right and see if the problem moves.
--Jeff
All I meant was to make sure you aren't bypassing the FB resistor via shorts, such as by having component leads touching. Sounds like you've been checking that.
You mentioned a difference in speaker resistance positive to chassis. Check the feedback tap to ground resistance too. One other thing to try, although it can be a real pain, is to switch OPTs from left to right and see if the problem moves.
--Jeff
Kofi Annan said:
Kofi,
Shut up.
You only read about the good advice of the rest of us; you do not know of the buggerups we (? Speak for yourself, Potgieter .... I do!) made. As you found I have just made one by honestly trying to sort out a circuit diagram via the manufacturer's data. (I would still like to know what the Hammond lead phases are then; I do think I got my circuit analysis right. Anybody on this?)
I also must now strongly supprt the idea of an under-chassis colour photo (including the chassis' sides). Hope you can get that arranged one way or another.
Regarding the first paragraph, I think a little further philosophy is in order: After the first great film success of US actor Paul Newman, he was asked a typical (stupid) journalist's question: "How does it feel to be an overnight success, Mr Newman?" He shot back with a smashing reply, to be heeded as a lesson to all.
"Overnight?? Big deal! You do not know about the previous ten years when I often did not have enough money for the bus fare home."
You do not know about all the circuits I built up so perfectly. Only they did not work.
Regards.
Johan Potgieter said:
Absolutely. How do you think we are able to suggest so many possible causes for your problem?
Hi Kofi,
I haven't been following this thread and have in fact just read the last few messages. However, you wrote....
---
>When I checked resistance from the positive speaker terminal to the chassis on the left (working) channel, I got 330R. On the right one I got .68, I think. Feeling that the 330R resistance was correct, I started to check the right channel and stumbled upon the concept that the NFB was causing the problem.
---
Respectfully, I would say that .68 ohm reading from the secondary of the output transformer to chassis is correct, as one winding of the winding should be grounded to chassis.
The reading of 330 ohms to chassis appears then to be incorrect. It would seem that the secondary winding of the one amplifier is not grounded.
Please take this in the spirit I present it, as I said, I have not been following he thread but it seems you have built the Bevois Valley schematic, and one side of the output is definitely at ground potential.
-Eric
I haven't been following this thread and have in fact just read the last few messages. However, you wrote....
---
>When I checked resistance from the positive speaker terminal to the chassis on the left (working) channel, I got 330R. On the right one I got .68, I think. Feeling that the 330R resistance was correct, I started to check the right channel and stumbled upon the concept that the NFB was causing the problem.
---
Respectfully, I would say that .68 ohm reading from the secondary of the output transformer to chassis is correct, as one winding of the winding should be grounded to chassis.
The reading of 330 ohms to chassis appears then to be incorrect. It would seem that the secondary winding of the one amplifier is not grounded.
Please take this in the spirit I present it, as I said, I have not been following he thread but it seems you have built the Bevois Valley schematic, and one side of the output is definitely at ground potential.
-Eric
Just a quick reply...
So, you're saying that the negative (black) speaker binding terminal should be independently grounded to the chassis in addition to having the black and black / yellow secondary attached?
Why do I feel like this was the stupid mistake I was looking for?
Kofi
So, you're saying that the negative (black) speaker binding terminal should be independently grounded to the chassis in addition to having the black and black / yellow secondary attached?
Why do I feel like this was the stupid mistake I was looking for?
Kofi
Are you using the Hammond 1608 OPT? Then, for a 4-ohm output the Black & Yellow/Black leads are grounded, while you take the "live" connection from the Green & Green/Yellow winding.
If you're using an 8-ohm speaker, then the same two wires as above are grounded, while you take the live side from the Yellow lead, and join the Green & Green/Yellow leads together.
These connections as per the Hammond url you provided earlier.
-Eric
If you're using an 8-ohm speaker, then the same two wires as above are grounded, while you take the live side from the Yellow lead, and join the Green & Green/Yellow leads together.
These connections as per the Hammond url you provided earlier.
-Eric
Hate to quote myself but...
I made sure not to ground the black and black / yellow secondaries, of course. I just attached them to the negative speaker binding post and let it rip.
So, that explains why when I "fixed" the one channel to ensure it was seeing 330R at from the positive terminal to ground I really "effed it up" since it was correct all along.
So, if this is the real issue (and it sounds like it is), here's my uninformed theory:
Since the negative terminal was essentially "floating" (or at least, not at ground potential), this was causing a problem at extreme ends of the pot because the signal became "confused" about what ground was.
Again, I ask for partial credit on this, but a clearer explanation would help.
Man. This could be it. I'll correct this tonight and post the results.
Yay! Learning!
Kofi
I made sure not to ground the black and black / yellow secondaries, of course. I just attached them to the negative speaker binding post and let it rip.
So, that explains why when I "fixed" the one channel to ensure it was seeing 330R at from the positive terminal to ground I really "effed it up" since it was correct all along.
So, if this is the real issue (and it sounds like it is), here's my uninformed theory:
Since the negative terminal was essentially "floating" (or at least, not at ground potential), this was causing a problem at extreme ends of the pot because the signal became "confused" about what ground was.
Again, I ask for partial credit on this, but a clearer explanation would help.
Man. This could be it. I'll correct this tonight and post the results.
Yay! Learning!
Kofi
If it were so, the you got FB from the lossy capacitance between primary and secondary of the OPT, the speaker and its wire could be a fine antenna, and the FB could be either positive or negative depending of which side of the primary had more leakage...
Don't let to report back how it goes 🙂
Gastón
Don't let to report back how it goes 🙂
Gastón
I think you got it. If, like me, you've got piles of clip leads around, it's a one minute experiment.
Ah, man!!! Definitely.
Now why did I not pick up on that when Kofi mentioned the 330 ohm, instead of philosophising forth!
Vaguely wondering where the 330 ohm came from since then the total NFB resistance of about 3,6K would have been in the circuit to ground. Not to mind ......
Now why did I not pick up on that when Kofi mentioned the 330 ohm, instead of philosophising forth!
Vaguely wondering where the 330 ohm came from since then the total NFB resistance of about 3,6K would have been in the circuit to ground. Not to mind ......
One more thing that I did forget - if you now get a squeal when you power up after grounding the Black & Black/Green wires, then use these two as your "live" connection and ground the other side instead, thus reversing the phase of your feedback.
Kofi Annan said:
Sorry, fellows - knowing that it is I who should probably "shut up" this time, something is bothering me though. We have all decided (me too) that Family Dog was right - all logical.
But why is Kofi saying that it he measured the 330R on the working channel? Do we have a further puzzle?
I think I might have gotten that backwards. I'll check when I get home, but I'm pretty sure its the other way around.
Man, thanks so much for the advice. I know this is it and I'll make sure I reply after I make the adjustments.
Kofi
Man, thanks so much for the advice. I know this is it and I'll make sure I reply after I make the adjustments.
Kofi
Hi Johan,
I was wondering about that too. I have looked at the BV circuit but I cannot see any way that Kofi can get 330 ohm, unless there is a component discrepancy. I figure about ten times that amount should be about correct if the OPT secondary is floating from ground potential.
I was wondering about that too. I have looked at the BV circuit but I cannot see any way that Kofi can get 330 ohm, unless there is a component discrepancy. I figure about ten times that amount should be about correct if the OPT secondary is floating from ground potential.
No fine Eric.
Let us hope that your solution is the right one, for Kofi's sake. (Otherwise, a photo.) I think we have all learnt from this.
Then we thought the UN had problems! It should be easy for Kofi there.
Regards.
Let us hope that your solution is the right one, for Kofi's sake. (Otherwise, a photo.) I think we have all learnt from this.
Then we thought the UN had problems! It should be easy for Kofi there.
Regards.
So, in a moment of divine intervention, right as I got home the power went out due to a bad storm. Geez. That wasn't too aggrivating.
So, about 20 minutes ago the power came on and Kofi started solderin'. And it works. It just works. With the NFB in and enverything. Despite my ignorance. Despite my disregard of basic electricity. Despite my re-re-resoldering. It just works.
So, in the words of Micheal Jackson, who is currently on the turntable:
Keep on with the force
Don't Stop
Don't Stop 'Till You Get Enough
Thanks to all! There is absolutely no frogging way I could have done this without you. It's a really nice amp. Really. Nice detail and fantastic soundstage.
Thank you so much.
Of course, however, I have another request of you. With the NFB in, I like the sound fine, but I have it pinned now and it's loud alright, but I'd like a bit more gain. Before it was really rockin' the house, albeit waaaaaaaay too much. But now its up to eleven and not really that loud... and I have pretty sensitive speakers (96dB Fostex 206Es).
Honestly, my 3.5W SETs are louder than this, so I am surprised that this is as low as it is right now.
Of course, there's a good possibility that I screwed this up again....
But really, as it is it's really, really nice and I just can't thank you all enough.
Greatest. Forum. Ever.
Kofi
So, about 20 minutes ago the power came on and Kofi started solderin'. And it works. It just works. With the NFB in and enverything. Despite my ignorance. Despite my disregard of basic electricity. Despite my re-re-resoldering. It just works.
So, in the words of Micheal Jackson, who is currently on the turntable:
Keep on with the force
Don't Stop
Don't Stop 'Till You Get Enough
Thanks to all! There is absolutely no frogging way I could have done this without you. It's a really nice amp. Really. Nice detail and fantastic soundstage.
Thank you so much.
Of course, however, I have another request of you. With the NFB in, I like the sound fine, but I have it pinned now and it's loud alright, but I'd like a bit more gain. Before it was really rockin' the house, albeit waaaaaaaay too much. But now its up to eleven and not really that loud... and I have pretty sensitive speakers (96dB Fostex 206Es).
Honestly, my 3.5W SETs are louder than this, so I am surprised that this is as low as it is right now.
Of course, there's a good possibility that I screwed this up again....
But really, as it is it's really, really nice and I just can't thank you all enough.
Greatest. Forum. Ever.
Kofi
More gain is easy. Increase the value of the feedback resistor that goes from the OPT to the input tube's cathode. Try increasing it by 1/3 to start and see if that's enough.
- Status
- Not open for further replies.