Loud Noise due to NFB

[sonata149]

I have just built the attached 6L6-GC SE circuit by JC Morrison with EF86 driver. When I first fired it there was a loud noise, so I checked the coupling caps. They were OK. What silenced it was by removing the NFB connection from the +ve side of the OPT secondary.

It is to be noted that in the published circuit the nfb is connected to a higher impedance than 8 ohm, presumably 16 ohm. My OPT secondary is only 0-8 ohm. What value of nfb resistor do you suggest I use instead of 8.2K? Do I go up or down in resistor value?

Thanks for your help.
Joe A

[Svein_B]

First you have to check if your global feedback is POSITIVE or NEGATIVE.
If the amp is unstable with NFB, you may try to swap the output wires from the OPT.
When you connect the feedback the volume should become lower.

The 8 instead of 16 ohm OPT means that you have a little less feedback (about 70% of the original with 16 ohm tap. You could then reduce the feedback resistor a little for the same feedback ratio. The optimal ratio should be determined by listening.

Svein.

[kavermei]

This might be caused by having the feedback wired in the wrong way. In this case you wind up with feed-forward instead of feedback, resulting in oscillations.

Try to swap the connections to the output transformer secondary winding.

Kenneth

[kstagger]

I've built something very similar to this circuit (EF86 -> UL connected EL156) with no problems at all - one of my favorite amps. Definitely try swapping your transformer leads.

[sonata149]

Thanks, friends. I wired the output from the transformer exactly as specified by the manufacturer (Transcendar). I will swap the connections and see how it fares.

I have hooked the amp (without feedback) to my main (expensive) system and while the loud noise is not there, there is a noticeable hum with no program playing. I wired my grounds in star fashion and I'm not sure if this is also part of the incorrectly wired output transformer.

Anyway, we'll see how it goes with the swapped connections.

Thanks for your help.
JA

[Svein_B]

The hum, it is important to find out if this is 100Hz PS ripple, or 50Hz hum.

The first can be reduced with larger PS caps after the choke.

50Hz hum may be magnetic coupling between PT and OPT, or grounding issues.


Svein

[sonata149]

Thanks Svein and the other posters. For the sake of those who find themselves in the same situation I'm posting how I solved my problem/s. Well, basically it's all based on what you have said in your posts. Swapped the secondary OPT connections, increased the capacitance after the choke by 100uF and added a 100R resistor || with .22uF between the -ve PS wires and chassis ground. The amp is now very quiet with no noticeable hum. As to NFB I soldered a 39K || with the 8.2K resistor for a target 6K8. To my ears hi-freqs are a bit pronounced, but not obtrusive.

Svein wrote:

The 8 instead of 16 ohm OPT means that you have a little less feedback (about 70% of the original with 16 ohm tap. You could then reduce the feedback resistor a little for the same feedback ratio. The optimal ratio should be determined by listening.

OK I determined the value without any mathematical reasoning. I thought that reducing it to about 6.8K I would be more or less there. What I can't understand is why we reduce the feedback resistor to increase the feedback. Can someone provide a not-too-complex explanation, possibly with a rule-of-thumb formula.

Thanks, and best regards.
Joe A

[Cale06]

Originally Posted by sonata149

Can someone provide a not-too-complex explanation

The feedback resistor forms a voltage divider with the unbypassed cathode resistor.

[artosalo]

I have hooked the amp (without feedback) to my main (expensive) system and while the loud noise is not there, there is a noticeable hum with no program playing.

The amp is now very quiet with no noticeable hum. As to NFB I soldered a 39K || with the 8.2K resistor for a target 6K8.

One and maybe most important reason that the hum disappeared is that the NFB is now "on". Asuuming the NFB is, say 18 dB, the hum level is also attenuated 18 dB compared to situation when NFB was "off".

To my ears hi-freqs are a bit pronounced, but not obtrusive.

Try to add a 470pF...1 nF capacitor in parallel to 6800 ohms NFB-resistor.
This will attenuate the top discants.

[artosalo]

Quote:
Originally Posted by sonata149
Can someone provide a not-too-complex explanation

The feedback resistor forms a voltage divider with the unbypassed cathode resistor...

...and when the upper resistor of a voltage divider is decreased, the voltage at the junction of the cathode resistor and NFB-resistor is increased.
When this negative feedback voltage is increased at the cathode of the 1st. tube, the gain of the whole amplifier is decreased.