TMB + Negative Feedback Loop = Soiled Underwear.

[Budso]

[quote="zaphod_phil . . . if you use a 25uF cap to bypass the 100 ohm resistor, that will ensure that only the very lowest frequencies get fed back and all the rest of the lows, mids and highs will go through the amp just like before. But maybe that's what you want - tightened up low end and everything else going through full throttle....[/quote]

My first thought as well. But I have to hand it to JA. He's not afraid to try wild values to get great results.

----------------------------


New wiring twist. It seems apparent that the TMB channel benefits greatly from the NFB, while the modestly gained normal channel suffers. Here's a fix for that.

Wire the thing up as per the latest drawing which on Phil's recommendation does away with the 220K summing resistors. Wire the switch in as well. Now the NFB is switchable either "on" for both channels or "off" for both channels.

Now here's the twist. Simply run a grounding wire from the tail resistor to the grounding prong on the TMB input jack. Now when you're plugged into the normal channel, the tail is grounded as stock (no NFB). However, when you plug into the TMB channel, the male cable jack on the guitar cord lifts the ground and puts the full NFB circuit into play. And it's still switchable on or off. Just a thought.

Bud

[Plexi]

""""Wire the thing up as per the latest drawing which on Phil's recommendation does away with the 220K summing resistors. Wire the switch in as well. Now the NFB is switchable either "on" for both channels or "off" for both channels. """"


If you look back you'll see li posted a switch so you could use it on either channel.

Richie

[Budso]

""""Wire the thing up as per the latest drawing which on Phil's recommendation does away with the 220K summing resistors. Wire the switch in as well. Now the NFB is switchable either "on" for both channels or "off" for both channels. """"


If you look back you'll see li posted a switch so you could use it on either channel.

Richie

Yeah, I just thought since the NFB wasn't really useful on the normal channel, the switch should be wired to only affect the TMB channel, switching NFB either in or out. The ground wire to the TMB input jack then just keeps the NFB permanently disabled on the normal channel side.

Different strokes for different folks.

[zaphod_phil]

OK Budso, are you ready to build yourself an 18W/TMB/NFB now...?

[Budso]

OK Budso, are you ready to build yourself an 18W/TMB/NFB now...?

OK Phil, now my cover is blown! A couple of days ago I was in the middle of building up my tag board for a TMB when I went back to this board to confirm a couple of component values. Imagine my surprise when I landed squarely in the middle of this switchable NFB, I want a presence control, hook me up to the TMB gain stage discussion.

So I had to throttle back and rethink what I wanted to put on my tag board, maybe another hole for a switchable NFB, etc. I feel very fortunate that I tuned in just in time for all of this creative experimentation going on around here.

For me, I'm going with the switchable NFB with a fixed presence cap. Probably try Richie's trick and hook up the normal channel to the second gain stage of the TMB.

Jersey's idea of a large (25u) cap in the presence circuit has got me thinking that there is some very real active tone control potential in a properly thought out NFB. Alright, where's my cold coffee . . .

Regards,

Budso

[Wayne_Alexander]

I'm confused now about how to wire this. In version 1 of the "Dookie Mod," the .01 was connected to the end of the tail resistor, and the input from the channel 1 volume was put through a 220K resistor into the .022. Now in Version 2 of the Dookie Mod, the input of the channel 1 volume is left as it was, going straight into the .01 capacitor, which is not connected to the tail resistor. So far I understand that. However, the drawing of the switch was done based on Version 1 of the Dookie Mod, so I'm confused about where to wire the leads for Dookie Mods version 2.0. I see that the 820 feedback resistor is wired from one side to the other of the DPDT switch, with one end connected to the impedance selector and the other end connected to the wire that connects to the end of the tail resistor, but where does the wire to the center lug on the DPDT switch go now? Before it was wired to the end of the .01 which was connected to the end of the tail resistor, now I don't know which one of them to connect it to. Could one of you clarify that, and maybe even draw the switch into the Dookie Mod version 2.0 drawing to help those of us builders who are weak in electronics theory?

[Plexi]

Does this method work? Or has it been tried. If you attach the feedback to the center or tail without the cap..then there is nothing there to block the DC feedback voltage? Thats why the feedback is only attached to one side,The cap keeps the DC from entering the tube.Or maybe i'm completely off here..Just wondered if it worked.

Richie

[Budso]

Richie, once that OT does its inducting thing and induces a current into the secondary, those 'trons come out with high current, low voltage AC to the speakers. So in my mind, there is no DC to block. The negative feed back is pure AC signal voltage. I've hooked up NFB right to the bottom of the cathode resistor on a preamp gain stage. No ill effects.

Of course, I've been wrong before . . .

Bud

[zaphod_phil]

... the drawing of the switch was done based on Version 1 of the Dookie Mod, so I'm confused about where to wire the leads for Dookie Mods version 2.0.

Just put a simple on/off (SPST) switch in series between the 820 ohm resistor and the 100 ohm resistor. That will do it.

[Budso]

If I may add to what Phil said above, the switch drawing may have confused some since it effectively uses a couple of the switch lugs as a "terminal strip". Very handy if you're outta space and outta turrets, eyelets, etc. But it does the same thing as Phil laid out with a simple single pole switch.

[Plexi]

"""Richie, once that OT does its inducting thing and induces a current into the secondary, those 'trons come out with high current, low voltage AC to the speakers. So in my mind, there is no DC to block. The negative feed back is pure AC signal voltage. I've hooked up NFB right to the bottom of the cathode resistor on a preamp gain stage. No ill effects. """

Hmm.. so no dc voltage is on a presence pot? Which is in the feedback circuit. the phase has alot to do with how the feedback works,and where its injected or used. I'm no EE so i'll just keep out of the way..

Richie

[BrianH]

So if one was to try the negative feedback mod without the added switch on their TMB, Jersey Aaron's Dookie Mod 2.0 drawing is the one to go with? I'm guessing some might suggest first trying a .1uf instead of the 25uf, though? Being of mediocre electronics skill, I have a bit of trouble following the discussion at times.

[mcv]

New wiring twist. It seems apparent that the TMB channel benefits greatly from the NFB, while the modestly gained normal channel suffers. Here's a fix for that.

Wire the thing up as per the latest drawing which on Phil's recommendation does away with the 220K summing resistors. Wire the switch in as well. Now the NFB is switchable either "on" for both channels or "off" for both channels.

Now here's the twist. Simply run a grounding wire from the tail resistor to the grounding prong on the TMB input jack. Now when you're plugged into the normal channel, the tail is grounded as stock (no NFB). However, when you plug into the TMB channel, the male cable jack on the guitar cord lifts the ground and puts the full NFB circuit into play. And it's still switchable on or off. Just a thought.

can you please specify where exactly does the wire from ground goes at the input jack? (newbee here!)
will i be able to jumper both channels?

[zaphod_phil]

So if one was to try the negative feedback mod without the added switch on their TMB, Jersey Aaron's Dookie Mod 2.0 drawing is the one to go with? I'm guessing some might suggest first trying a .1uf instead of the 25uf, though?

Sure. Try it and see how you like it....

[wornadaptor]

I don't have a TMB, but in looking at the Dookie Mod 2 it seems to me that the biggest problem lies in the fact that the signal is nearly exhausted by the time it gets to the ground end of the 47K tail resistor. That is why the resistor values have to be so low and the capacitance has to be so high. This leaves no room for a presence control.

My thought is to make the presence control part of the 47K path by replacing it with a 42K resistor and a 5K pot. The resistance on the NFB resistor will now be a larger value (to be determined), and the capacitance can probably be smaller. The presence control would be in effect all the time. If you turn the presence to zero it will be the standard 47K tail resistance without any presence effect.

I apologize for the hasty and crude sketch, hope you can see what I mean.

[zaphod_phil]

... it seems to me that the biggest problem lies in the fact that the signal is nearly exhausted by the time it gets to the ground end of the 47K tail resistor.

I'm not sure that explanation is actually valid in EE terms! However, you may be right in terms of the ratio of 100 ohms to 47k being very small, so that only a very small amount of NFB takes place. And with a 25uF cap across it, the NFB will only happen at low frequencies, but don't forget this combination of values has been carefully tuned by JA. So it will sound good. Also the effect of the NFB resistor being 820 ohms, means that a higher proportion of the output signal is being applied to compensate for the other two. However, what you're suggesting is closer to Richie's version of the NFB mod....

[Budso]

... it seems to me that the biggest problem lies in the fact that the signal is nearly exhausted by the time it gets to the ground end of the 47K tail resistor.

I'm not sure that explanation is actually valid in EE terms! However, you may be right in terms of the ratio of 100 ohms to 47k being very small, so that only a very small amount of NFB takes place. And with a 25uF cap across it, the NFB will only happen at low frequencies, but don't forget this combination of values has been carefully tuned by JA. So it will sound good. Also the effect of the NFB resistor being 820 ohms, means that a higher proportion of the output signal is being applied to compensate for the other two. However, what you're suggesting is closer to Richie's version of the NFB mod....

Another thing to consider is that this 820/100 ohm ratio was typically used by Fender feeding off the 2ohm OT tap. JA is using the same ratio from the 16ohm OT tap. That's a huge (8X?) jump in NFB signal voltage. So I think that tends to equal things out a bit. Normally, the plexi, with its use of a 10K tail resistor, would probably requite a different setup.

[Budso]

New wiring twist. It seems apparent that the TMB channel benefits greatly from the NFB, while the modestly gained normal channel suffers. Here's a fix for that.

Wire the thing up as per the latest drawing which on Phil's recommendation does away with the 220K summing resistors. Wire the switch in as well. Now the NFB is switchable either "on" for both channels or "off" for both channels.

Now here's the twist. Simply run a grounding wire from the tail resistor to the grounding prong on the TMB input jack. Now when you're plugged into the normal channel, the tail is grounded as stock (no NFB). However, when you plug into the TMB channel, the male cable jack on the guitar cord lifts the ground and puts the full NFB circuit into play. And it's still switchable on or off. Just a thought.

can you please specify where exactly does the wire from ground goes at the input jack? (newbee here!)
will i be able to jumper both channels?

It might make more sense to look at wornadaptor's drawing. My idea then is to run a wire from between the tail resistor and smaller feedback resistor over to the grounding lug on the input jack. I guess it depends on the type of input jack you're using. I had in mind the three prong swithcraft jack. It includes a center switchable grounding lug that you would attach the wire to. On the Marshall type jacks, if you were facing the input jack from the front, you would attach the wire to the left rear lug. Then ensure that a wire runs from this lug to ground.

Also, I don't see why you still couldn't jumper the channels with this setup. Again, as you plug in the male jack, it pushes the spring steel up, disconnecting the grounding lug. Hope this makes sense. Without a draw, it's awfully tough.

[Plexi]

"""Another thing to consider is that this 820/100 ohm ratio was typically used by Fender feeding off the 2ohm OT tap"""

In what fender with a 2 ohm tap,if its the 5F6A bassman its a different feedback circuit.most with the 820/100 or 820/47ohm are all 4 ohm trannys.
Alot of this is probably confusing more people,there are so many ways of doing the feedback, and if your trying to do a preset type without presence,or a curcuit that uses a presence control.
Look at many different amps..the Atomic by soldano which has a presence control. Look at the pro jr,thats about what i am useing in values..but without the bypass cap.. The list is long.. but to jump and say this one works and this one doesn't is a matter of opinion,or maybe you just don't have it wired up correct. They work fine in many many other amps.
Richie

[Foster]

I've got a TMB kit soon to be on the way, and I've been doing my best to follow this thread. Even so, once y'all work through this, would anybody out there be willing to write up a "negative feedback for dummies" for us noobs?

-Tim