I'm just gonna go ahead and share my work. My first big mod project. Fender PA 100. Be gentle.

[2L man]

I think you are ready to use this loadline calculator Believe me it is great when you see how median tube loadline look using anode and stage supply voltages and anode resistances you measure from your amps.

Because tubes are different and they change when they are used hundreds and often thousands hours there come differences to grid (bias) voltages. If difference is high to ULLC loadline you should test another tube.

Compare Cold Clipper stage and input stage and you see their difference and that cold clipper loadline is steep but bias point is very low close the zero current and thats why it clip other half of the signal.

Point cursor to grid lines which cross the loadline and you can calculate what comes anode voltage sweep when drid (drive) voltage change for example 2V. I use 2V sweep for pre amp tubes. When you divide anode sweep using grid sweep the result is that stage gain in numbers.

Then if you want to tune the stage you can test new anode resistor in ULLC and set anode current and then use grid voltage to calculate cathode resistor. Then you know that your actual rebias go to the direction you want it to go before you change the components. When double triode use common cathode resistor just divide resistance using 2. This work accurately when both triode halfs have same anode resistors and stage feed voltage. When anode resistor values are very different often triodes have own cathode resistors too.

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

 

[Whatizitman]

This depends on your ground scheme. If all of the other grounds are willy nilly attached to the chassis, this *middle* can be done so as well. Back of pot is usually connected to the chassis when the pot is screwed to the chassis.
If you have a preamp ground scheme, it should be attached with these grounds.

Have you changed the old filter caps? They could have something to do with motor boating.

I did change the caps before modding. Motorboating went away once I got my input jacks (kinda) worked out. Now it sounds good at low preamp crunch settings. Just lots of noise and oscillation that sounds like a bad tremolo.

I am still working out the overall scheme. Lots for me to learn there since I'm not building from scratch. I have to figure how to integrate well with what's already there and working well.

I'm hoping once I get V1 signal levels and input grounding nailed it will improve. Then it's just the rest of my spaghetti work and cold solders to clean up. That's the current plan, anyway

 

[Whatizitman]

Questions about voltage dividers and attenuation.

I want to decrease some signal level by increasing attenuator resistor values. Per Marshall 2203 schem, the preamp pot before 2nd triode has 470k in series for the attenuator, as well as the 3rd triode grid leak. Does changing the attenuator (top) resistors in the dividers change the grid leak potential? Good? Bad? Any thoughts would be helpful.

EDIT: ok, I'm starting to figure it out https://robrobinette.com/Tube_Guitar_Amp_Overdrive.htm#Controlling_Grid_Current

 

[Whatizitman]

Still confused. According to @robrob decreasing the 2nd triode attenuator will dampen signal. Not the other way around. Is this only due to dc cap value? I currently have .0022.

I guess I could experiment with adding a grid stopper to V1B, after the divider. Not sure that will make much difference, though. I think the signal is just too hot. I'm not necessarily hearing blocking distortion. OTOH, it distorts so much even at low level, it kinda feels like I'm gonna blow a tube. On that note, V1 does have some microphonics, even when swapping tubes around. Hmmmm. What am I missing?

 

[Lowerleftcoast]

ok, I'm starting to figure it out

They are all voltage dividers. There are voltage divider calculators available online. If you plug in *10* as the voltage *in* it is easy to determine the percentage of signal reduction. Of course if R1 is 470k and R2 is 470k the signal reduction is 50%.

Bypassing R1 with a cap will allow more highs to make it to the next gain stage. Marshall tended to use 470pF for those bypass caps and imo that value is a sweet spot for Marshall voiced amps. It seems to work best.

Since you have too much gain, my starting point would be to use 1M pot as a grid leak/voltage divider in each of the later stages to get the signal voltage to a manageable level. These can be replaced with resistors when you get the pots tweaked to the preferred levels.

You will have to determine what kind of distortion you want on tap. You may or may not want Preamp distortion, Phase Inverter distortion, Power Tube distortion. You are calling the shots with the amount of gain and where.

The reduction in gain may solve the oscillation and other troubles. If not, Fender and Marshall had answers.

Let me work up a schematic to show you what they did.

 

[Lowerleftcoast]

To reduce some ultrasonic highs... to keep oscillation from occurring, Fender and Marshall used the caps in green (0.01uF in parallel with the plate resistor) and blue (47pF) on some of their amps. (I always include the blue cap (100pF) on any build with a LTP Phase Inverter.)

Marshall would alternately use the red cap (.01uF ?) on some of their builds rather than the green cap. Rob mentions this on one of his sites. Essentially it serves the same function as the green cap, but you wouldn't use both. Sometimes you will see this on more than just the first gain stage, and sometimes it is used to reduce highs in the audible range by using a larger cap value.

Marshall will use larger grid stoppers on the power tubes (yellow) which also help to keep oscillations from happening.

 

[Whatizitman]

@Lowerleftcoast this is really helpful. Thanks! I'm gonna take some time to absorb the info. I don't have any more 1M pots to try. But I do have 1M resistors that could at least gimme an idea of the difference.

EDIT: oh, and as far as distortion, I'm just looking for some preamp distortion like a typical 2203. I won't be cranking the amp too much, so output tube and PI distortion are not priorties. Right now the distortion I get with the preamp pot at ~1 is almost too much, but it's definitely what I like with the asymmetry, etc. Just a hair more and it sounds like a DSL red channel, though LOL.

EDIT: And this is with an MIM strat with stock pups!

 

[Whatizitman]

Well... after some poking and prodding it's looking more like the oscillation is due to bad dc and/or tone caps, or at least something in the TMB wiring. Thanks again @Lowerleftcoast for the tips. I have a mix of old/reused and new dc and tone caps, various voltage ratings. Most likely a few are voltage underrated. I'm gonna change them all to new proper rated caps. Maybe orange drops, or whatever. I'm also reusing some stock 250K pots for the TMB because they have terminal strips mounted on the back, with one terminal grounded to the pot back. I'm using the terminals for other duties. But there could be ground loops or whatever as a result. I'll figure that all out eventually.

Now the input jacks I'm embarrassed to say I just don't get. I never was good at logic games LOL. They have 4 terminals, with switching, and I believe isolation. The fourth terminal and isolation is throwing me for a loop. I have verified the switching works and that they are able to ground to chassis despite isolation(?). Here's a bad pic of one. On the original circuit, the ones like this that had the ground tab connected were the lower input jacks on the channels. The grid leak (47k on original schem) connects to this jack. Most of the schem copies online cutoff the far left where it has lower input as B, and upper as T. I couldn't hear any difference on either jack in stock form, even when jumped with other channels, BTW. I do know that though they are switching jacks, they do not switch each other. I think that's where I'm getting crossed up.

How do I use these types of jacks for a standard 2203 hi and lo setup? Is that even possible?

I have hi working now, but lo gives no guitar signal. The signal from V1A is getting to the lo jack as it goes from there to the preamp volume. So something is connected right LOL. I would like to have lo working, at least so I can hear some possible gain difference.

Can someone kindly hold my hand with the input jack wiring? My brain just can't handle more than one level of booleans and the like.

 

[Lowerleftcoast]

The jack is known as a Switchcraft 13E.

To make it act like the 12A the *lifter terminal* would connect with the *tip terminal*.

Like the 12A, the 13E is not isolated.

 

[Whatizitman]

Yes! Yes! Thanks @Lowerleftcoast !!

EDIT: ok lessee if I got this straight. On the 13E schematic symbol, the top terminal is the lifter, followed down by the switch, tip, and sleeve in that order.

Does the lifter work like a switch with another jack? To make it work like a 12A, does the lifter connect to the other jack, or to the tip on the same jack? On the PA100 circuit the lifters on the T jacks connected to the V5A mixer resistors and treble pots. I really don’t understand the lifter purpose, to be honest.

EDIT: ok I may have it. V1A plate connected to the tip terminal, jumper from tip to lifter. Then shunt to preamp volume pot. I’ll try that next.

 

[Lowerleftcoast]

Yes the *lifter* and the *switch* make contact when there is no plug inserted and break contact *open* when a plug is inserted. Since there is no electrical connection to the tip and/or sleeve, this switch can be used to operate devices other than this 13E jack.

 

[Whatizitman]

Well, I’m just trying to get them to work like a typical 2203 input hi and lo. I rewired the lo as per last post, but lo still doesn’t work. Hi channel does, and clearly signal is getting to Lo pin terminal from V1A plate. Really no change from before.

EDIT: I’ll figure it all out eventually. I did realize that since I left the hi Jack shunt and ground jumpered per the PA100 circuit, it may not be shunting the 1M grid leak when lo input is plugged in. Probably one of several issues. But in the right direction?

 

[Lowerleftcoast]

From what I see on the 2203 schematic the High jack has 2 switches. I am not sure of all of the available switching jacks but a Switchcraft 14B should work.

doh! Marshall used Cliff jacks and many of those have the *double closed* switching.

 

[Whatizitman]

From what I see on the 2203 schematic the High jack has 2 switches. I am not sure of all of the available switching jacks but a Switchcraft 14B should work.

doh! Marshall used Cliff jacks and many of those have the *double closed* switching.

Right. Which just adds to my confusion. @robrob 6V6 800 layout has standard jacks, though. If I can wire my jacks in this manner I’m golden. Otherwise I may just buy a couple 12A jacks and be done with it. They’re cheap LOL.

 

[Lowerleftcoast]

You can wire them like Rob's version.

 

[Whatizitman]

You can wire them like Rob's version.

Yup! Finally figured it out! I never until now really understood Jack wiring, so that was part of the issue. With the addition of the isolating tip and lifter I was totally lost.

It works now! The Lo sounds great. Much less gain and distortion, and more like a real 2203 hi channel. I suspect the extra diode from the original PA100 circuit (V5) is the culprit. BUT, this means I get two Marshally hi channels for the price of one! One hi (lo) and one super duper hi (hi).

An all turret vintage DSL, with spring reverb!

Thanks so much @Lowerleftcoast and @2L man for your help and patience. And as always thanks @robrob for making so much info free for all to use! His page was is a major resource for every level of this project.

Now on to the oscillation issues. I did move some plate and screen wires a little which seemed to help a bit. But I also suspect crappy tone and dc caps. I will update when I tackle that. For now I’m gonna enjoy noodling a bit before I get around to ordering more parts.

 

[Whatizitman]

OK, I fixed the inputs again. They were working, but not as the marshall gawds intended. It took me forever for it all to finally click. But basically the switch is only tied to the lifter, so for cascaded lo and hi I needed hi to ground itself when plugged into lo. That meant making the switch act like an actual switch LOL. It didn't kill the tremelo type oscillating at all. But it did lower the idle noise floor at the hi input quite a lot.

Back to oscillation. This is going to take a long time, and I'm in the for the ride. To clarify, by oscillation, I mean audible. All within audible frequencies, so it can sound like a tremolo. If I hit a chord or note at sufficient volume it will sound good. But let it sustain for a moment, and it will eventually start "tremolo-ing". Cutting lower frequencies does help. A little.

Trying to read up on oscillation as well as I can from a non-engineer's POV I at least can understand that signal is getting mixed back in somewhere in a feedback loop. I know my wiring is highly suspect on that. Lots of long signal wires, crossing and parallel randomly with original turret board components and such. I also have back of pots used as grounding points for some of the preamp circuitry, but not for the pots themselves. All of that on top of the old Fender style brass plate that is factory soldered to the chassis at several points. Etc... Etc....

But I just came across something very relevant to my "design" that despite all the above might warrant a complete rethinking of the my preamp "layout", such as it is. On the Fender PA 100, all of the channels and the 2nd stage (after channel mix) are tied to the same last two filter caps. The caps are connected via 10k, but both make up the last filter stage after the PI node. Well, plexi and bassman this is the case. But not on a 2203. It's there as plain as day on the schematic - an extra filter stage feeding only the V1A and V1B. Doh!

Remember that my amp has extra extra high gain, and not by my intent. I think it's because it has an extra triode after the tone stack. Soooo.... Of course my lead dress needs to be tended to. But do I need to do anything about the filtering? It has two caps in parallel. Should I tap the 10k for V2? That triode now has the cathode follower, BTW. Should I go ahead and tackle that? Or should I clean things up first?

Thoughts? Any insight or recommendations would be helpful. Thanks!

EDIT: I guess I could change it back to plate driven tone stack, and just not use V2. Or at least not use V2A. I'd rather not. But it would make things a lot easier at this point. Leaving out V2 would be V1A and V1B (hi cascaded) --> tone stack --> V5 stock fender values. Certainly that would lower the distortion. And hopefully a bit of the oscillation.

Hmmm. The more I think about it, this is a good way to go since I'd rather not mess with anything else stock on the reverb side of the preamp circuitry. I'll first try by taking out V2A, leaving the cathode follower. If that doesn't help enough, back to plate driven stack.

Live and learn, amirite. It's all been for the learning experience first and foremost.

 

[mountainhick]

OK, I fixed the inputs again. They were working, but not as the marshall gawds intended. It took me forever for it all to finally click. But basically the switch is only tied to the lifter, so for cascaded lo and hi I needed hi to ground itself when plugged into lo. That meant making the switch act like an actual switch LOL. It didn't kill the tremelo type oscillating at all. But it did lower the idle noise floor at the hi input quite a lot.

Back to oscillation. This is going to take a long time, and I'm in the for the ride. To clarify, by oscillation, I mean audible. All within audible frequencies, so it can sound like a tremolo. If I hit a chord or note at sufficient volume it will sound good. But let it sustain for a moment, and it will eventually start "tremolo-ing". Cutting lower frequencies does help. A little.

Trying to read up on oscillation as well as I can from a non-engineer's POV I at least can understand that signal is getting mixed back in somewhere in a feedback loop. I know my wiring is highly suspect on that. Lots of long signal wires, crossing and parallel randomly with original turret board components and such. I also have back of pots used as grounding points for some of the preamp circuitry, but not for the pots themselves. All of that on top of the old Fender style brass plate that is factory soldered to the chassis at several points. Etc... Etc....

But I just came across something very relevant to my "design" that despite all the above might warrant a complete rethinking of the my preamp "layout", such as it is. On the Fender PA 100, all of the channels and the 2nd stage (after channel mix) are tied to the same last two filter caps. The caps are connected via 10k, but both make up the last filter stage after the PI node. Well, plexi and bassman this is the case. But not on a 2203. It's there as plain as day on the schematic - an extra filter stage feeding only the V1A and V1B. Doh!

Remember that my amp has extra extra high gain, and not by my intent. I think it's because it has an extra triode after the tone stack. Soooo.... Of course my lead dress needs to be tended to. But do I need to do anything about the filtering? It has two caps in parallel. Should I tap the 10k for V2? That triode now has the cathode follower, BTW. Should I go ahead and tackle that? Or should I clean things up first?

Thoughts? Any insight or recommendations would be helpful. Thanks!

EDIT: I guess I could change it back to plate driven tone stack, and just not use V2. Or at least not use V2A. I'd rather not. But it would make things a lot easier at this point. Leaving out V2 would be V1A and V1B (hi cascaded) --> tone stack --> V5 stock fender values. Certainly that would lower the distortion. And hopefully a bit of the oscillation.

Hmmm. The more I think about it, this is a good way to go since I'd rather not mess with anything else stock on the reverb side of the preamp circuitry.

Live and learn, amirite. It's all been for the learning experience first and foremost.

It would be really good for you to draw out your circuits out in schematic fashion and share that here. I can not follow your verbal descriptions.

 

[Whatizitman]

It would be really good for you to draw out your circuits out in schematic fashion and share that here. I can not follow your verbal descriptions.

I don't have my mods on paper. I'm following the typical 2203 pre amp design schems, and the PA 100 schems for reference. Probably about time I attempt to document what I have. I can't really draw schems well, though. I'll give it a try.

 

[2L man]

Sometimes high gain amps have negative feedback loop between pre amp stages. If there is and you have cancelled stages perhaps it has turned to positive feedback which can cause oscillation?