Merlin High Gain SE Amp NFB question

[andrewRneumann]

My 2 cents...
I like idea 1, although I've never tried it myself. I don't like the idea of messing with the cathode bias resistor or bypass cap on the power tube. Idea 4 is too much hassle. Good luck!

 

[printer2]

OK, I've been researching my NFB options. I'll list them in order of easiest to implement to hardest.

1. Move my Power Tube Cathode ground to the positive speaker terminal. This seems a bit strange to me, but I have found a few references to the practice. I guess it works or it doesn't, there doesn't appear to be any way to vary the results.

2. Take out my power tube cathode bypass resistor and bring NFB back to the power tube cathode. Search for a value that sounds good.

3. Change the cathode resistor to a series pair, divider. Place the current bypass cap in parallel to the first resistor and bring the NFB back to the junction between them. I'm not sure where to start with values here. Obviously the series total needs to be the same as what I have now, but how do I decide on the split ratio. It seems there could be a lot of experimenting there.

4. Re-order my stages to place the tone stack and follower earlier in the signal path. Bring NFB back to the "new" last gain stage. I have sketched the wires that I would have to move for this and it's not a lot of work. That said, I would probably want to change the gain of the two stages involved.

 

[printer2]

OK, I've been researching my NFB options. I'll list them in order of easiest to implement to hardest.

1. Move my Power Tube Cathode ground to the positive speaker terminal. This seems a bit strange to me, but I have found a few references to the practice. I guess it works or it doesn't, there doesn't appear to be any way to vary the results.

2. Take out my power tube cathode bypass resistor and bring NFB back to the power tube cathode. Search for a value that sounds good.

3. Change the cathode resistor to a series pair, divider. Place the current bypass cap in parallel to the first resistor and bring the NFB back to the junction between them. I'm not sure where to start with values here. Obviously the series total needs to be the same as what I have now, but how do I decide on the split ratio. It seems there could be a lot of experimenting there.

4. Re-order my stages to place the tone stack and follower earlier in the signal path. Bring NFB back to the "new" last gain stage. I have sketched the wires that I would have to move for this and it's not a lot of work. That said, I would probably want to change the gain of the two stages involved.

Thoughts?

You know your cathode resistor and the current through it (0 mA to 2 x bias mA) so you can calculate your voltage swing across the resistor. Your output power should allow you to figure out how much voltage swing you have there. Then you figure out how much NFB you want.

 

[SerpentRuss]

My 2 cents...
I like idea 1, although I've never tried it myself. I don't like the idea of messing with the cathode bias resistor or bypass cap on the power tube. Idea 4 is too much hassle. Goo

OK,

I've tried option 1. It worked a bit too well perhaps. I might have mentioned that I have selector switches for my OT taps. I did a really nice job of soldering and heat shrinking the connects from the OT wires to the switch, so I just moved the cathode ground to my speaker jack. On first power up, I got the squealing shriek of doom. First time for me, but I think everyone needs to experience that at least once. So I checked the OT documentation and even though I had the "correct" wire connected to my B+, it must have been marked wrong.

So, I switched the OT primary leads and the amp sounded much better and much quieter. I decided that before I moved it I would send some sine waves through the amp and write down their voltages at the OT secondary, so right across the speaker.

Here are the results:

Frequency	Before NFB	After NFB
80 HZ	1.49 Volts	1.20 Volts
160 HZ	1.78 Volts	1.13 Volts
320 HZ	2.19 Volts	1.25 Volts
640 HZ	2.94 Volts	1.51 Volts
1280 HZ	3.99 Volts	1.87 Volts

As you can see, it attenuated the speaker signal quite a bit. The amp sounded much better, but the only clean sound I could achieve I could talk over.

So, I went in search of swing. My pre-amp started almost exactly like the schematic I linked to. So the first thing I did was bypass R10 which was attenuating the signal going into the 3rd gain stage. I dialed in a 200 hz test tone and measured the swing before 38 volts at the OT primary. After removing the resistor and without changing the pots I got 58 volts.

Next, I changed the load resistor for the 3rd stage from 100K to 150K. I got to 62 Volts and I could get a louder, cleaner sound.

That's where I'm at so far. I want to try another OT tap for the cathode ground to decrease the amount of feedback, but I wanted to ask which way I should go. I'm thinking it has to be the 4 ohm tap, but I'm not sure.

 

[andrewRneumann]

I'm thinking it has to be the 4 ohm tap, but I'm not sure.

Yes this is the way to go to get the least amount of NFB. The 4-ohm tap has the least amount of voltage.

You are now entering the phase of fiddling. We are here for moral support.

 

[SerpentRuss]

OK,

Here is the latest update.

It turns out that I was loosing 85% of my signal voltage across my tone stack. I triple checked the tone stack wiring and component values. Of course checking the caps required de-soldering one end of each. What a mess. Anyway, the tone stack works fine. Bass goes up and down, treble goes up and down, it seems right, just ultra lossy.

So, I re-ordered the stages. Not all that hard really, about 5 to 6 wires to re-run.

The new order is:

Gain 1 (V1B) -> volume -> Gain 2 (V2B) -> follower (V2A) -> james tone stack -> Volume -> Gain 3 (V1A) -> Master Volume -> Power Tube (V3)

It sounds much, better and is capable of getting a louder clean sound.

So, now that I have that done, there's really no reason not to do the NFB in a more normal fashion. Maybe, I get a crack at that tomorrow.

BTW, the bright cap goes around the first volume is obnoxiously bright. The value is 200 pF. I think I'll try some different values or maybe make that a NFB switch.

 

[SerpentRuss]

OK, I got the NFB now working back to the driver cathode. Yay for me, LOL. Of course I had to switch my OT Primary polarity again.

Right now I have it on 100K pot and a series resistor to set the minimum value at 5.6K, so 5.6 to 105.6 ohms. That stage currently has a 2.2K cathode resistor. I also placed a 560K resistor in the ground connection to the voicing switch that chooses which bypass cap to use. I haven't measured the setting, but I think it's a linear pot and is probably about 75K.

I feel like I've had this amp completely apart about 5 times now. That said, it's been a great education, since it's the most complex thing I've tackled so far.


Here are some pics.

 

[drew1d]

Looking good. I haven't tried anything that complex, I'm glad it's working for you. A quick question about the 6n2p-ev tubes you're using, did you ground pin 9?

I'm still trying to wrap my brain around what you did for NFB. I'll get there, but it's going to take me a minute.

 

[SerpentRuss]

Looking good. I haven't tried anything that complex, I'm glad it's working for you. A quick question about the 6n2p-ev tubes you're using, did you ground pin 9?

I'm still trying to wrap my brain around what you did for NFB. I'll get there, but it's going to take me a minute.

Thanks.

The NFB is set up like most amps now, going from the OT secondary back to the cathode of the driver gain stage. Now that the stages are re-ordered, there's no TS between that stage and the power tube, just a volume control. The resistor I placed below my stage, cathode bypass capacitor is there to make sure the NFB does not just get shunted to ground and can actually have an influence on the signal. NFB is obviously AC, so if I did not do this, it would just pass right across the bypass capacitor.

I noticed I did not put in a 6P1P-EV, just a plain 6P1P. I have both, so I grabbed the wrong box. I don't know if I'll change it out or not. I may save the few of those I have for a push-pull design down the road. As far as I know the only difference is longevity and the amount of G's that tube will take while energized. Not critical for guitar amps.

As for pin 9 of the 6N2Ps, I definitely think I'll ground them. I read two schools of thought on pin 9, ground it, and don't bother. I don't think I'll be able to hear the difference, but for my own personal experience I wanted to have the amp working before I grounded them.

As a general note: I think if you're building a tried and true design it's best to stay with the tubes that design used. If you're experimenting, if you're already planning on getting off the beaten path, it's hard to beat NOS Russian tubes. I bought a few small lots of new tubes and a lot of 20 used, tested, and documented 6P14P tubes from one Ukrainian dealer for 90 dollars to my door. It turned out to be less than 2 dollars a tube. I have purchase a few other small lots for just slightly more.

I'm going to continue to experiment and try a couple of things before I decided to put this in a case. I still have to wire a front headphone jack, make a guard for the wiring near the PT, remove the NFB pot and put in a resistor and wire a line out jack in that location.

With no instrument plugged in and with all the gains dimed, it's dead silent. When the signal chain has a guitar, there is some high end harshness that I'm going to chase.

I'll keep everyone posted.

 

[printer2]

With the volume control in the NFB loop it might cancel out some of the NFB and then run opemn loop (depending on volume position).

 

[SerpentRuss]

With the volume control in the NFB loop it might cancel out some of the NFB and then run opemn loop (depending on volume position).

If the NFB goes down with volume I think it will be ok.

 

[printer2]

If the NFB goes down with volume I think it will be ok.

Then way do you want NFB?

 

[SerpentRuss]

Then way do you want NFB?

I'm not sure what you're getting at here. It seems logical to me that the benefits that are derived from NFB are more necessary when the power stage is being driven hard. If the volume control is turned up, most of the NFB is passed by the driver to the power tube, when it is turned down less NFB is passed. It actually seems like a good idea in principle though I'm certainly not an expert.

In an amp like a Champ or a Princeton, can't you look at that grid leak resistor as a volume pot that's not adjustable?