NFB "Feedback Feedback" - stumped...

[wjdunham]

I'm having a problem with the NFB circuit on my "project" - but before I describe the problem, I should give a little history of the "project". The short version is that I'm in the process of turning my fender Prosonic into a "dumble like" amplifier. OK, put the weapons down, I can explain I ordered a d'lite kit a while back and during the interval from order to ship I have enough time to read every post on this forum three times and learn as much as I can. Armed with all that dangerous knowledge and still no kit, I sit down with my Prosonic schematics and to see if I can figure out why I can't get a tone I like out of the OD channel. Long story short, I decided that since all the right pieces are there, I could do some major reconfiguration to the preamp using a Dumble architecture and maybe come up with something I like better. The PI and power section are very similar already. It's a big project, but I figured I could do it in steps. I started with the "Early 80s grail tone non HRM" schematic, with the post OD treble bleed, and OD input filters, 100k plates all around, as a baseline that I could then tweak from to get a tone I like. Keep in mind, I'm not trying to make it into a clone, but just trying to get a better tone.

Everything has gone according to plan until last night - I've got the clean channel and OD channel up and running, tone stack in the right place but I haven't converted to the mid 80's config, still using the stock fender config. I'm not worried about tone yet, just basic functionality. OD trimmer working - sound on the clean is OK, OD channel stinks but I still don't have it up to baseline yet.

So, last night I tackle the NFB circuit since the Prosonic doesn't have any. The PI has all the right values except the tail is 18K in stead of 24K. I pull the tail resistor, add the 4.7K, 390, 24K and .01uf/2K presence wired per mid 80's schematic. I wired it to the 4OHM tap of the OT. Amp is currently set up with a 16OHM vintage 30 driven from the 16OHM OT tap. I turn the amp on and start playing, sounds good, and then after 1 or two MINUTES I get a very deep feedback. If I disconnect the OT input then everything is fine. I tried the 8OHM tap with same results. I'm completely mystified. Why after two minutes? All the component leads are as short as possible, OT wire is completely separated and away from any other wires. In general the amp is as clean or cleaner than it was originally, so I'm not inducing a lot of extra noise. What gives? Any suggestions as to what might be happening would be greatly appreciated. I don't have a picture of the PI mods, but here's a snap of the V1, V2 mods so you can get an idea of the work I've done.

Thanks for any help!
Bill

PS, this is an awesome forum, one of the most constructive and knowledgable. Hopefully I'll be able to contribute something interesting once I have this beast tamed.

[ChrisM]

Are they primaries on the OT maybe reversed?

Try swapping them around.

[Structo]

Nah, I don't think the OT wires are reversed.
This was a functioning amp before the mods.

I have gotten that type feedback before.
What I did was reduce the feedback by increasing the 4K7 resistor.
Try doubling that each time until it stops.

Another thing it could be is the way you have the presence pot grounded.
Are you taking the cap and resistor to ground? Or do you have the third lug of the pot going to ground?

Right now I have an amp that does a high pitched squeal if I open the presence all the way up.

[wjdunham]

I have the presence pot/cap wired like your second diagram, as shown in the 80's schematic I was working from. Interesting you mentioned the high pitch squeal - I had originally left off the presence pot/cap just to get things working and I got a similar high-pitch squeal after a few seconds of power. I thought for sure adding the cap would cure it. It just lowered the frequency of the feedback significantly. These's a clue there somewhere.

I can try increasing the 4.7K until it stops. I was wondering whether I needed the FB at all and how much effect that would have on the sound, I've not played with this before so my ears don't know, any guidance here? would doubling the FB resistor (decreasing the feedback) effect the sound much?

I thought the point of negative feedback was to increase the stability of the amp? Something's causing a pretty significant unexpected phase shift, I agree swapped OT primaries is a good candidate, I'll check those, but as Tom said, it was working fine before I added the FB.

Thanks for all the suggestions so far! I have some hope I can figure it out at least.
bill

[wjdunham]

Just checked some photos of the OT section, OT primaries definitely wired up correctly...

[Structo]

The feedback tames the beast a bit.
Feedback will increase headroom to an extent and is a neat way to add or take away some sparkle (presence).

On my amp, I used a .1uf electrolytic.
I may have to try a bigger cap and see what that does for me.

My D'Lite has a 1uf.

I also put a DPDT switch on the cathode bypass caps. That really makes a difference.
It adds a lot of gain to have the caps in and cleaner without them.
This is a mish-mash that started out as a Deluxe hybrid but has slowly turned more into a plexi type amp.

What is weird about your amp is that it shows up after a couple minutes which would lead me to believe it may be related to heat.
Possibly a bad solder joint or ground.

[wjdunham]

That's another possibility, although I rebuilt it once thinking I had some leads too long that was causing the problem, with the same results.

On another note, I changed all the cathode caps to 2.5uF tantalum electrolytics, I think the first thing I'm going to do once I have everything else done is increase one or both of the clean channel caps, it's definitely a little thin sounding. Then I was going to try some higher plate/cathode resistors on the OD channel, but this is my main gigging amp and from what I've read here it sounds like 100K is the best option for live situations in a crowded mix.

One other question - I have a .01uF OD coupling cap, 220K 100K trimmer, 68k grid input resistor and the OD channel seems to suffer from a very flabby/loose bottom end. I then added the .05uF/10M OD entrance filter and it seemed to have no effect. By my calculations, .05u/10M has a 0.6Hz cutoff ..., am I not interpreting how this is supposed to work correctly?
Bill

[RB]

I just looked at a schematic for the prosonic and sure enough - no NFB. Saying that it is wired correctly from pictures and schematics is a moot point in this case as you have no point of reference of a circuit working with NFB. It could just as likely that the OT is wired so that you NFB is in phase and not out of phase. The first thing every body tells you is switch the OT primaries. This sometimes can be problematic because one wire is too short to move to the other tube socket. You can accomplish the same thing much easier and quicker by switching the wires that feed the power tube grids at the point of connection at the phase inverter output. This is where I go when I need to switch phase to check the NFB phase. I have worked on amps that were wired in phase that do not go over the top with squealing but act like you are describing. I would suggest switching the phase and making sure you have it correct or you could chase this problem for a very long time.

Regards Randy

[Structo]

Yeah, good info, I did it exactly as you describe on my amp.
But as soon as it I turned it on, it howled like a banshee, so I knew the other way was correct.

[wjdunham]

Thanks Randy. It's easy enough to swap the OT primaries and see what happens. I'm a little unclear at to your first point though. I checked to make sure the actual was wired like the schematic (ie, that there wasn't a wiring problem when fender built the amp), and that I connected my FB input to the proper half of the inverter. I can't find a datasheet for that particular OT - how would I tell if it's wired up with the proper phase. If you're not using NFB does it matter?

[Structo]

I haven't tried tantulums for bypass caps so I don't know how those sound but I don't recall ever seeing them used there.

I have all 5uf (4.7uf) bypass on my D amp.

I have a 1uf tant cap on my D'Lite presence.

Are your phase inverter values stock at this point? 100K and 91K?

I found with my amp that the PI plate resistors can really influence the tone especially on the OD channel.
I went to 110K and 120K but I'm not saying you should.

[heisthl]

Thanks Randy. It's easy enough to swap the OT primaries and see what happens. I'm a little unclear at to your first point though. I checked to make sure the actual was wired like the schematic (ie, that there wasn't a wiring problem when fender built the amp), and that I connected my FB input to the proper half of the inverter. I can't find a datasheet for that particular OT - how would I tell if it's wired up with the proper phase. If you're not using NFB does it matter?

nope - you only care if you're using a presence circuit(GNFB).

[RB]

Wjdunham wrote:

Thanks Randy. It's easy enough to swap the OT primaries and see what happens. I'm a little unclear at to your first point though. I checked to make sure the actual was wired like the schematic (ie, that there wasn't a wiring problem when fender built the amp), and that I connected my FB input to the proper half of the inverter. I can't find a datasheet for that particular OT - how would I tell if it's wired up with the proper phase. If you're not using NFB does it matter?

In answer to your last question: In practical terms It doesn’t matter how you wire it if there is no NFB (in phase or out). Now having said that I am sure there are some OCD types out there that would argue the point. But for this trouble shooting example that’s the position I take.

As to my first point: Doing the research and verification that you did was a good thing. Its just in this amp there is no way of knowing the phase based on it working from the factory. If it had NFB in the original design the phase would have been sorted out in QC before being shipped. But it didn’t have NFB so no squeals in QC wired one way or the other. You have to verify it yourself by hooking up the NFB both ways and blowing out your ear drums and speakers.

Parasitic oscillations are the hardest things to trouble shoot. The NFB circuit is the easiest of the oscillation bugs to test and fix. Make sure its wired correctly then you can trouble shoot knowing for sure its not the problem and have it send you off on a wild goose chase.

Regards Randy

[wjdunham]

And the answer is... Transformer primaries hooked up in the wrong phase for NFB application! Thanks so very much to Randy and everyone else who chimed in here. I never would have checked that, I did not think it though enough to realize that the OT primary phase only really matters if you're using feedback, in which case it matters a lot. I was lucky enough that the primary wires were long enough to just swap them at the power tubes, left it on for 10 minutes with no squealing.

I'm sure I will have a lot of questions once I really start tweaking the tone, but I still have to reconfigure the tone stack. There's a lot going on in this thread, but to answer Tom's question, the Prosonic PI plates are 100k/91k. I'm still a bit away from playing with things like that, but thanks for the info.

Thanks again you guys, I greatly appreciate all of your help. I'll post some followups as I get farther along.

Bill