Feedback

[Greg.]

Sorry, slightly off topic but yesterday I reduced global feedback on a WAD6550 amp in accordance with the directions in the FAQ. This particular amp is fitted with Sowter O/P transformers.

The owner reports a very positive outcome in terms of detail resolution and better sound stage. No doubt this is a Sowter thing.

[Richard]

Hi Greg, totally on-topic . I can't remember anyone posting about lowering fb in Kit6550 so good to hear how it went. Yes the Sowter's are very good and perhaps need less fb than other TXs due to their generous build quality. I found with Sowters on Kit88 they needed less and sounded more open and natural with less fb - but they cost as much as the kit so fb is a good tool to improve performance in the right situation - so give it a go - 5dB is a good starting point

[Greg.]

Thank you Richard. This is fast becoming a thread that maybe should be linked elsewhere.

Anyway, having done the feedback conversion on my friend’s amp as per the faq, although initially happy, he is now experiencing a sharp edge to the treble to the point of sibilance and he’s not a happy bunny

In consequence, the obvious to me is to re-increase feedback (sibilance has never been a problem before and I have worked exclusively on this amp over a long period) but maybe not so much as the original value. So, if I change from a 50% reduction to a 25% reduction, is it the simple arithmetic of halving the component value between the 50% reduction (faq) and the original values of R45/46 and C21/22 to find a 25% reduction or is it more complicated?

As you can tell, I don’t have a clue, so do you have any recommendation on what I should do? Having followed the FAQ which has resulted in sibilance, I’m keen to sort this out not only for my own satisfaction but to ensure my friend can properly enjoy his system.

[Richard]

Hi Greg, I'd just swap it back to original, as otherwise the doubt will always be in your friend's mind

[Greg.]

Thanks, Richard. It’s often difficult with these sort of things because I do the work here, listen to check all is good (measure when necessary) and then the amp on return home continues to perform well until after a few days things change. I have absolute faith in my friends listening abilities so there is no criticsm of him. Anyway, without any other pointer to a cause for this, we have taken your suggestion and are returning feedback to the original setting.

[Richard]

Morning Greg, Yes it's difficult to remember sounds across a period of days. There will also have been a volume/sensitivity change and anticipation/expectations involved.

Your friend may well be right about a change in the sibilance but listening critically for a change (after having been used to relaxed listening for many months to a system he was familiar with) may well cause him to notice things he hadn't noticed before. Going back to the original is the only way to be sure he isn't imagining it.

It's happened to me a few times, I change back and sure enough it's there but I just hadn't paid it any attention before. If he's right then it won't be there and he'll be happy again. Halfway adjustment would just confuse things and prolong any doubt

[bikerhifinut]

Originally Posted by Richard

Morning Greg, Yes it's difficult to remember sounds across a period of days. There will also have been a volume/sensitivity change and anticipation/expectations involved.

Your friend may well be right about a change in the sibilance but listening critically for a change (after having been used to relaxed listening for many months to a system he was familiar with) may well cause him to notice things he hadn't noticed before. Going back to the original is the only way to be sure he isn't imagining it.

It's happened to me a few times, I change back and sure enough it's there but I just hadn't paid it any attention before. If he's right then it won't be there and he'll be happy again. Halfway adjustment would just confuse things and prolong any doubt

Just a thought Richard/Greg, Is there space to incorporate a switch on the rear of the amp for instance, and then you can switch in a small network of feedback resistors/compensation capacitors? If nothing else it would give you a fairly instant A/B comparison.
I see that you seem to have taken into account the compensation capacitor Greg as I did wonder if that was possibly altering the HF response and giving rise to a bit of sibilance, a bit like having a bit of HF boost but very high up the audio band?
Is there a compensation network elsewhere in the amp, I do not know the circuit? Maybe in the anode circuit of the first valve or in between where there may be a bite of "pole slugging"?
Do forgive me if I stated the bleedin' obvious chaps.

A.

[Greg.]

Thanks, Richard for your observations. Yes, original values are being returned to.

Yes, Andy, it would be possible to build a switchable network, but the trouble with that is once you have the feedback value of choice, there it will stay making the rest of the network and switch redundant. Of, course, I could build a temporary external network to establish what works best and then stick with those values, but it’s kind of awkward when it’s not your amp and you are doing the work for someone else when it might take them several days to reach a conclusion on the preferred sound. Actually changing the feedback value on a 6550 is pretty simple. Two caps and two resistors mounted directly on the amplifier tag boards. A 5 minute job