Negative feedback loop squealing?

[Yoda]

Hello all,

So I’ve been tinkering with my recent home brew 5F1 build and I added a negative feedback loop with a 10k resistor on a switch but when I turn the NFB on amplifier squeals extremely loudly even if the volume is all the way down and I immediately hit the NFB switch to turn it off so hopefully nothing bad happens to the amp.

The NFB resistor is wired between the 8 ohm speaker output and the V1B cathode, at the junction where the cathode resistor and socket solder tab meet.

Any ideas what would cause this squealing when engaging the NFB?

Thanks in advance.

[roberto]

Because you have PFB: reverse the wires going to the OT.

[martin manning]

Probably no permanent damage done. To fix it you need to invert the feedback signal by reversing either the primary or secondary output transformer leads. If your OT has more than one speaker tap, it will have to be the primary leads.

[CraigGa]

In my Cornford Roadhouse the OT primary colours didn't match the schematic so I reversed them.

Howl!!!
lesson learned.

Craig

[Yoda]

I reversed the OT primary wires and all is well. It’s such a dramatic difference now when using the NFB switch I’m even thinking of doing a 3-way switch for 2 different levels of NFB instead of just On or Off.

Thanks again for the help you guys rock!

[roberto]

Increase the diameter of the hole in the chassis and install a potentiometer instead.
It is a very useful control.

[Yoda]

Increase the diameter of the hole in the chassis and install a potentiometer instead.
It is a very useful control.

What’s a good way to wire up a variable NFB? What value pot?

[martin manning]

What’s a good way to wire up a variable NFB? What value pot?

Wire the pot as a variable resistor, and insert it between the 22k fixed resistor and the speaker, or or between the 22k and the 1k5 cathode resistor. You might reduce the 22k to 10 or 15k so you can get a little more NFB than standard. I'd try a 100k pot, which would reduce the FB to ~20% of the original level at its max resistance.

[roberto]

Reduce the 22k NFB resistor to 10k/12k then add a 100k LOG potentiometer in series:
with the pot at noon you'll have the same amount of NFB, then you can change it as you like.

[Yoda]

I’m already using a 10k resistor instead of the 22k called for on the original 5F1 schematic. Is this considered a lot of negative feedback already? Because I see that the blackface circuits are often 820 ohms which seems even more extreme.

[martin manning]

You need to consider the voltage divider made up of the feedback resistor (originally 22k) and the 1k5 cathode resistor going to ground (which is also setting the DC bias on the 12AX7). Blackface Deluxe (e.g.) has an 820Ω feedback resistor, but the lower leg of its feedback divider is the PI tail resistor which is only 47Ω.

[roberto]

What aspect do you like of the increased feedback?

[Yoda]

Well I never tried the 22k in this build actually but with the NFB on it’s completely clean and mild-mannered. A totally different amp. So I can see how this is a versatile control.

[Yoda]

I just experimented with a 10k + 25k pot for 10-35k of sweep on the NFB as well as with 22k and 27k resistor. I found that I actually prefer the fixed 10k on a switch at least for this amp so I think I will leave it and incorporate an adjustable NFB on the next build.

[Yoda]

So I have another question; The V1B triode on this amp has a switchable cathode bypass cap with values of .68uF or 10uF, as well as no bypass. The negative feedback switch has an affect when the cathode bypass switch is either unbypassed or on the .68uF position but no effect at all on the 10uF postition. What is the reason for this?

On another note I inserted the NFB signal where the cathode resistor and bypass cap of the 6V6 meet the tube socket. I noticed however that on the blackface Champ schematic the negative feedback is inserted on the ground side of the same two components. What difference does this make?

Thanks.