Remove NFB from 4R tap. Apply 100mV into lifted wire (820/2n2 end). Put 8R load on 8R tap. With amp on measure signal on 4R tap. Should be 18dB bigger than that on 820/2n2 end. Should be easy as you have scope and sig gen. If scope has 2 inputs just check phase is 180deg different.
Great. And thanks for your patience. The NFB wire from the PCB to the OPT is screwed into a terminal block on the PCB, so I can just lift it there and connect the sig gen instead. The scope is two channel so I can check the phase too. BTW, your picture is out of phase. Cool! 😀
It may take me a day or two, as I have just sold my A21, so I need to organise a courier, preferably without going out.
It may take me a day or two, as I have just sold my A21, so I need to organise a courier, preferably without going out.
I've not cut the NFB loop and measured (nervous of blowing it up!), but I moved it to the 4 ohm tap and it works OK into dynamic speakers, even with capacitive Litz speaker leads.
So I added the 1uH//10 ohms in series with the output on the 8 ohm tap, and 8R2 on the end of that. Connected one channel of the scope directly to the 8 ohm tap, and the other channel to the output end of the inductor. I would have expected there to be a difference between them at higher frequencies, but both show a 6dB drop at about 80kHz. Shorted across the inductor and measured again; no difference.
So what am I doing wrong? Just because I have a scope doesn't mean I know what I am doing with it!
Is the inductor too small? I can't measure inductance that low, but both the online calculators I used say 20 turns on a 9mm former gives 1uH. Is that simply the wrong value, or am I barking up the wrong tree?
So I added the 1uH//10 ohms in series with the output on the 8 ohm tap, and 8R2 on the end of that. Connected one channel of the scope directly to the 8 ohm tap, and the other channel to the output end of the inductor. I would have expected there to be a difference between them at higher frequencies, but both show a 6dB drop at about 80kHz. Shorted across the inductor and measured again; no difference.
So what am I doing wrong? Just because I have a scope doesn't mean I know what I am doing with it!
Is the inductor too small? I can't measure inductance that low, but both the online calculators I used say 20 turns on a 9mm former gives 1uH. Is that simply the wrong value, or am I barking up the wrong tree?
Reactance = 2*PI*f*L = .5R at 80KHz so is going to have virtually no effect. I think the 10R/1uH is more aimed at solid state amps which tend to go up-to 1-5MHz on the output. If your nervous about removing the NF and injecting a signal. You can just inject the signal on the normal input and measure on the 8R dummy load. Without the negative feedback it will go up by 7 fold.
If you want to keep with the 1uH/10R it will do no harm. I think you have solved the problem.
If you want to keep with the 1uH/10R it will do no harm. I think you have solved the problem.
Last edited:
Thanks. It was running the amp without the NFB connected that worried me!
If I increase the inductor until it DOES have an effect, will that cause harm? Apart from not wanting to risk the Quads until I am more confident in the HF stability, I have to get a 30kg valve amp down a ladder and the stairs to try it! And I can't ask anyone to help.
I've just used an online reactance calculator, and 10uH gives me about 5 ohms at 80kHz an 1.2 ohms at 20kHz.
If I increase the inductor until it DOES have an effect, will that cause harm? Apart from not wanting to risk the Quads until I am more confident in the HF stability, I have to get a 30kg valve amp down a ladder and the stairs to try it! And I can't ask anyone to help.
I've just used an online reactance calculator, and 10uH gives me about 5 ohms at 80kHz an 1.2 ohms at 20kHz.
I think your issue is solved. Valve amps don't tend to need the series inductor and resistor. I would just try it. I guess your working in the loft!
Yes, working in the loft. But the amp is now back to the state that had instability into capacitive leads, and the only difference from the Quad situation is the NFB. Which I guess is now less, being on the 4 ohm tap. I will test with a bigger inductor, as I would rather lose some HF than risk damage to the Quads. Anyway, this should only be for a short while, until I can travel. So maybe a long while!
Ok, I'll make up the bigger inductor set just in case, and try it without it first. Fingers crossed. But I will have to exercise a bit more to build up my strength first!
Just been reading a little about NFB (better late than never, I guess), where it is pointed out that more NFB can increase oscillation (I know, you have told me! 🙄), which makes me wonder if I can safely reduce it any more. I tried different value inductors on the output, and anything that affects the ultra-sonic level lowers the audible signal too much.
When designed correctly there is a sweet spot where the correct gain gives the best phase margin. Too low can also cause instability. Have a read about bode diagrams. I would leave it alone.
I have at last dragged it down from the loft, nearly killing myself in the process. At one point I thought it was going to fall on me sliding it down the ladder, but I made it!
And it's playing now, not warmed up properly yet, nor the Quads, but so far so good. A little hollow sounding at the moment, but hopefully that will improve. I know it's not the most refined performer, but at least it is driving the Quads with no drama. So far! 😎
So thanks for the help, as I was definitely a bit lost.
And it's playing now, not warmed up properly yet, nor the Quads, but so far so good. A little hollow sounding at the moment, but hopefully that will improve. I know it's not the most refined performer, but at least it is driving the Quads with no drama. So far! 😎
So thanks for the help, as I was definitely a bit lost.