Valvestate 8200: effectsloop: dry signal comes through even when on full wet

[Van Esch]

Hello everyone,

The parallel loop on all of these valvestates Iīs as far as I know(and I have an 8080 as well) shouldnīt let any preloop signal pass when the knob is set at full wet. My 8200 does so and itīs starting to be problem for certain setups. The dry signal strength seems to be as strong on dry as on wet.

Can anyone confirm that this is not normal and might know whatīs wrong? So far Iīve checked and replaced the following components:

- replaced the effects dry/wet pot (as well as most others)
- replaced the bigger elcoīs on pre-amp pcb (still waitng for the 4700uFīs for the power-pcb to show up)
- checked soldering on all boards

Amp works fine except for this problem and since I use some stereo effects the loop is essential. Iīm inclined to think the problem may be IC2B on the reverb/loop-pcb, but I donīt know enough about ICīs to see if itīs faulty. How do measure it with a multimeter and what should I expect?

Schematics are of the preamp and the reverb/effectloop pcbīs


Thanks!

[J M Fahey]

Maybe I'm going blind but can't find ant dry/wet pot in any of the schematics posted.
Can you please point to it?

[Van Esch]

The pot is tandem called VR11A and VR11B. The schematics below have coloured lines as what to my limited understanding of electronics(studying a little each day now though!) is the signal path:

Starting in red, coming from clean/boost channel, to the reverb/loop pcb,to the series loop and in case of a jack in the series return jack through IC2B, which I think ought to separate the dry series loop from the parallel loop(?), elsewise the signal returns to preamp and changes to blue, goes to IC5A, is split into reverbsignal(purple) and a normal signal and goes to final amplification(not sure thatīs 100% true). Then through the parallel loop, coming back in yellow and green back to the preamp. The 2 parallel signals now go through their parts of IC5, the dry/wet pot, level switch(pedal/rack switch, which I suspected but have tested thoroughly and is not faulty) and then to final amplification.

[J M Fahey]

What a MESSY signal path.
The one on the preamp too.
No wonder why it does not work as intended.
Looking at it I feel the URGE to rip it off and start again with a clean sheet of paper.
What were they thinking?
Klunky.

[Van Esch]

That is IF my theories on the signal path are correct. Any ideas on where the problem could be?

[g1]

You really need to confirm that there is a problem. You can't compare it to the 8080 as the FX circuits are different. You must try another 8200 or talk to someone who owns one.
Some Marshall owners say their models of Marshall only get to maximum 50% wet. In those cases you would never get rid of the dry signal.
Just the name alone "parallel FX loop" implies to me that there will always be some dry signal present.

[52 Bill]

I hate these schematics. VR11 a&b only controls the levels of the right and left parallel return signals. The straight output from the preamps and series loop appear at the output of IC5a which is mixed with the outputs from VR11 through all of those 100K mixing resistors. So unlike the other versions, there always will be straight signal in the mix.

If you look at the owner's manual for these amps, there is a warning about turning off the complete signal when there is no processor in the loop on only some models.

[Van Esch]

The manual is pretty unclear about the 8200/8280 model: "Effects mix: for mixing the effects to dry signal." The completely wet situation like on an 8080 seems an obvious way to build it because there are many stereo effects in my effect unit that I use together with certain distortion effects from the same unit. None of which I usually want to mix with unfiltered clean signal.

Iīll try and confirm the workings with other owners and Marshall. Your insights in the schematics tells you that 100% wet is indeed impossible? If so: modding?!

[J M Fahey]

What about .... replacing?

[Van Esch]

? A valvestate I with a solid boost from letīs say uhhhhm, my Boss EQ-20? No sir. Let us stick to fact and see if anything is wrong. I know many would like to say why donīt you like this or that sound better, itīs an endless/useless discussion. Iīve send some messages to 8200/8280 owners and Marshall England to see whatīs what, and will post there answers asap.

[Enzo]

Um, I don't speak for Juan, but I do believe he meant replacing the FX loop circuit, not the entire amp.

The serial loop on this model does break the signal path, so no bleed past there. Looking at the circuit for the parallel loops, there is a continuous path through to the output to the power amp. I see no way for the dry signal to be turned off. The two returns mix into the left and right power amps, and that dual R11 adjusts the level of those returns. SO basically you can have the dry signal with as much wet signal mixed in as you like. No killing the dry.

Look at the schematic. The parallel send is after IC2b, the path does continue from there through IC5a which feeds the chorus circuit, the reverb driver, and the dry signal to the output. That dry feed goes through C44. Lift C44 out of circuit and you break the dry path. Now the only dry signal coming out of the master volume wil be whatever dry is included coming in the parallel returns. Simple enough. The downside is that there is no longer a dry connection to the output, so you will have to always have a path for your dry through the parallel loops. Either with a jumper cord, a switching circuit you invent, or some connected external gear that includes a dry path if desired.

[J M Fahey]

Yes, the idea is bypassing this mess by picking signal just before it gets into this uselessly complex loop , send it to basically 2 jacks and 1 pot and back to the next stage.
Will be faster and cleaner than tracing that criscrossing paths nightmare.
Passive will not be perfect, but easy and good enough.
If you want to, you can copy the old simple Valvestate active one (just 1 IC) on perfboard.
But with normal modern pedals you don't *really* need buffers, since most (all?) are high impedance in; low impedance out.
Didn't utter a single word about tone, nor the classic (in *other* Forums) "drop that SS junk and get a Toob amp"

[Van Esch]

Havenīt received any confirmation yet wether full wet is possible or not with the knob, but I think you are right. I never really considered this possibility because it renders the loop useless for some sounds. Looking at a bigger picture: maybe when these amps came out there werenīt many stereo multieffects yet? Plenty of stompboxes though..

I never use a dry signal, so Iīm thinking a simple audio switch on the back. Will this not cause any problems or unwanted side effects on the rest of the board? Iīve been studying the possibility of changing the pot to an 8080 type situation, but I suspect that will not be a case of mere signal routing...

Many thanks so far for all of your input!

Also a big sorry to J. M. Fahey for the misinterpretation! I should not be allowed to touch my computer when coming home drunk in the middle of the night..

[Van Esch]

Hereīs an idea that seems logical to me, but might well be stupidity: take out r72/r73 and r30/r31 and link them directly. Wouldnīt that bleed dry signal completely to ground when the pot is on full wet? Combined with a higher resistance pot?

Or: using the pot on the far side between r67 and r72 with the arrow/slide part linked to ground?

This one seems best though:

A link from before after C44 to the side of the pot that is now ground, ground to the slide/arrow, and after C20 a link to before r72/r73. Resistors around the pot like in the 8080. Likewise for the other one. That makes it more or less the same as the 8080?

The pic is the 8080 situation. The second pic is a how the last idea would look. It doesnīt have the resistors though.

[Enzo]

I highly recommend you try the simple lifting of C44 first. Doesn't it get you where you want to be?