Modulus "Deepfried Marshbar" build debugging!

[Iain T]

Hi All,

New here, building my first amp using a kit form Modulus here in the UK, the Deepfried Marshbar.

It's a variant of the 18W with more tone stack options (T/M/B, Pres, Reso) and a 3-way toggle between 'normal', '800' and "#34" (whatever that is!) Yes, I should have built something simple first time but it's a good way to learn, right?

Been through the pre-flight checks, no shorts, so many times checking the right components are in the right places on the board, the flying leads go to the right places, etc.

Powered up through my bulb limiter with no valves and all looked good except no -ive voltage to pin 2 on the EL84s. The path for this is ground/0V through a 220k resistor and then an 8.2k resistor to pin 2 (see attached layout PDF), zoomed in, note the underboard dotted-line to ground:



Voltages on all the tubes look good apart from the bias, v4 and v5 have almost identical voltages:

2: 0v (bias/control grid)
3: 28.6v (k)
6: 369v (connect 470R to #9)
7: 375v (a)
9: 372v (screen)

I'm at a bit of a loss where to look for the issue as the bias circuit looks incredibly simple!

Apologies for the longish first post - not sure how much info is too much!

[JMPGuitars]

Why do you think there's an issue?

What voltage do you expect to see there, especially without a tube present?

Thanks,
Josh

[Iain T]

Hi Josh,

Sorry, in editing the post for clarity, I removed the key info - no output! Not even hiss at full volume. Testing voltage on V5 will cause a small "rustling" sound. Voltages above were with all valves in place and plugged into a 4 ohm speaker.

I spent some time last night scoping (1kHz @ 150mV) along the signal path but, as a rank amature, I lost sight of where I should be seeing the signal replicated to pin down the issue.

With power valves installed there is still no bias voltage - everything seems happy (no red-plating, no magic smoke, arcing or other obvious signs of distress.

Spent an hour today re-checking the wiring - continuity where expected, appropriate resistances where I'd expect it and can't see anything I've done wrong.

With this more complex tone-stack, I'm unsure where I should see the scope trace after V2.

[JMPGuitars]

Ah, your first post sounded like you had no tubes in, and were surprised that there was no voltage at the grids of the output tubes. Near 0V is expected at the power tube grids in a cathode biased amp, or at least in 18W variants anyway. Too much voltage at the power tube grids in these amps usually implies poor soldering somewhere between the grids and ground, but that doesn't appear to be an issue here. Measure the grid to ground voltage in the low mV range and see if you get anything there. I doubt it's 0V, but I wouldn't be surprised if it's somewhere between 10mV and 100mV.

When you say "spent some time last night scoping (1kHz @ 150mV) along the signal path" - do you mean you had a 1kHz sinewave @ 150mV at the input jack, and then probed from there forward? I assume so, but need to be clear as your statements thus far are a little vague.

I don't like when kits don't have a schematic. Schematics are 10000x better for troubleshooting than layouts. Look at some TMB schematics and voltage charts from our downloads section to get an idea of what you're looking at. Also try contacting the kit supplier to see if they have a schematic you can use. That will make it much easier for you to understand the signal path.

Another tip is to avoid any underboard connections. All wires being visible makes inspection much easier. Highlighter check the whole build, and hopefully you can do that with a schematic. If not, use the layout. Looking at the layout, it would be really easy to miss a connection with those dotted lines.

Thanks,
Josh

[Iain T]

Thanks Josh!

Scoping - yes, you interpreted me correctly - 1kHz sine wave @ 150mV on the input, traced with an Oscilloscope from input through output to gain control and on to V2.

I've actually created a schematic from the layout in KiCAD as I wanted to understand the circuit better. Obviously there is a risk of transcribing errors in doing that although I've been through it against the layout multiple times.

There was a small voltage on the grids for V4/5 to ground but it wasn't consistent so I assumed it's the meter being inaccurate!

I have traced all the underboard ground connections for continuity and don't appear to have missed any. I agree about the layout diagram and under-board connections. I'll contact Modulus and see if they have a schematic available.

Thanks again!

[Bieworm]

Hard troubleshooting without schematic.

[JMPGuitars]

Hard troubleshooting without schematic.
Stupid question: if it’s a fixed bias amp it is a very common error towire the electrolytic cap(s) in the bias supply backwards. Positive to ground is what it should be.

It appears to be an almost normal 18W style cathode bias. They put the 8.2k grid stoppers at the socket, and they used 220k grid leaks instead of 470k, but it's still the common arrangement otherwise with the 150Ω cathode bias resistor that should probably be 180Ω.

[JMPGuitars]

I've actually created a schematic from the layout in KiCAD as I wanted to understand the circuit better. Obviously there is a risk of transcribing errors in doing that although I've been through it against the layout multiple times.

Share your schematic if you can't get an official one from them.

Try probing at the PI grids and see if you hear anything from the speaker. Also adjust the volume knobs and the Rule #34 switch. Then probe the PI grids again. You can work from the input, or you can work backwards from the output section to see if you get any signs of life from the amp.

[Iain T]

It lives!

Josh - you were spot on about checking grounds and dislike for the layout diagram over a schematic... I'd missed ground to the 5W 150R resistor/cap from pins 3 on V4 and V5! This was despite going through with the multimeter and thinking I'd verified all grounds!

Immediate life after fixing.

I actually stumbled on this when verifying signal to pin 2 on V5 and accidentally touched the probe to pin 3 - I could faintly hear the sine wave and then rechecked the path from there! Checking back on here I see this is exactly what you suggested - thanks, it's appreciated.

The schematic I put together if anyone's interested, PDF attached... Spotted a "small" error - missing connection to ground after the res/cap on V2 Pin 3...

[Iain T]

... and so, on to the next problem appears to be DC on the Presence/Resonance pots - they emit a dull rumble when turned!

9.5V on the wiper of the Reso pot and pin 3 of the presence pot... on with the debugging!

[Crabman]

Interesting schematic. A lot of classic 800 mod ideas in there. This one moves the gain pot after the second stage when plugged into the high input. I'm in the process on moving on 3 stage + CF 800 type modded preamp with an 18watt PI and power amp. Similar schematic ... sourced a lot of ideas from Headfirst amps who have made it all available online. I'm curious to know how the Marshbar sounds? The concern I have is that most of the 800 mods are intended to drive EL34s and it may be a bit too much when pushing EL84s since a lot of 18W grind comes from the power amp. I guess I will find out. I know that some folks switch the EL84s out for 6V6s to tame a bit of that.

Note that your 3rd stage cathode isn't connected to ground in your schematic.

[JMPGuitars]

When you touched your probe to pin 3 (cathode) of the power tube, you created a path to ground, that's why you finally heard something. This is another reason we tell people to be exceptionally careful while probing an amp, as YOU don't want to be that path to ground.

When you create schematics with a program like Kicad, you need to be careful about the symbols you use. For example, both references to HT+ are input tags. There should be an output tag coming from the rectifier, or you can have them both rectangular (non-directional). This is important because running ERC (electrical rules checker) can also help you find issues with your schematic.

Thanks,
Josh

[Iain T]

Note that your 3rd stage cathode isn't connected to ground in your schematic.

Thanks, already spotted (added a note in my post) and have resolved in the schematic!

[Iain T]

When you touched your probe to pin 3 (cathode) of the power tube, you created a path to ground, that's why you finally heard something. This is another reason we tell people to be exceptionally careful while probing an amp, as YOU don't want to be that path to ground.

Yes, I was being careful but clumsy! Using one hand to scope, the other in my lap. I actually slipped up and touched pin 3 trying to get the probe off pin 2 - the grabby end on it was being a bit too grabby.

When you create schematics with a program like Kicad, you need to be careful about the symbols you use. For example, both references to HT+ are input tags. There should be an output tag coming from the rectifier, or you can have them both rectangular (non-directional). This is important because running ERC (electrical rules checker) can also help you find issues with your schematic.

I'm learning KiCad as I go, I'll update that tomorrow. I'll probably change the schematic to run the rectifier output to the +HT rail anyway.

Really enjoying learning how this works.

[Iain T]

... and so, on to the next problem appears to be DC on the Presence/Resonance pots - they emit a dull rumble when turned!

9.5V on the wiper of the Reso pot and pin 3 of the presence pot... on with the debugging!

A bit of digging and I think that this is normal/expected for this circuit. A couple of comments on ampgarage.com indicating that this is seen on Friedman amps and old style Fender/Marshall presence circuits.

[JMPGuitars]

... and so, on to the next problem appears to be DC on the Presence/Resonance pots - they emit a dull rumble when turned!

9.5V on the wiper of the Reso pot and pin 3 of the presence pot... on with the debugging!

A bit of digging and I think that this is normal/expected for this circuit. A couple of comments on ampgarage.com indicating that this is seen on Friedman amps and old style Fender/Marshall presence circuits.

It's not a part of the tone circuit. The better question would be, looking at the schematic you made, why wouldn't there be voltage there?

Follow the cathode voltage path from V3, and see what makes sense to you.

[Iain T]

It's not a part of the tone circuit. The better question would be, looking at the schematic you made, why wouldn't there be voltage there?

Follow the cathode voltage path from V3, and see what makes sense to you.

Yes, spot on.

From the schematic/layout it makes absolute sense - there's only a 470R and 22k resistor between the Cathode of V3b and the presence pot. No cap to block DC.

Thanks for your patience with my newbie questions and floundering.

I'm kinda glad I didn't pick a simple build for my first - way more learning going on this way!

[Iain T]

I'm curious to know how the Marshbar sounds?

Loud!

I'm actually running it into a 4 ohm 1 x 10 bass speaker and it's very loud. Going to drag it down to a local rehearsal space with a guitarist friend to crank it hard.

Initial thoughts:
1. Hard to get a clean tone out of.
2. Gain pot is very sensitive in the low portion.
3. The DF setting is going to work well for a thrash metal tone, #34 will do most heavy rock well.
4. Presence is very strong - a small back-off has a big change and all the way to 0 sounds like playing through wool.
5. Quite noisy on higher gain - going to have a look at improving the grounding scheme - splitting ground for power and pre stages might be an improvement.

[Bieworm]

Interesting schematic. A lot of classic 800 mod ideas in there. This one moves the gain pot after the second stage when plugged into the high input. I'm in the process on moving on 3 stage + CF 800 type modded preamp with an 18watt PI and power amp. Similar schematic ... sourced a lot of ideas from Headfirst amps who have made it all available online. I'm curious to know how the Marshbar sounds? The concern I have is that most of the 800 mods are intended to drive EL34s and it may be a bit too much when pushing EL84s since a lot of 18W grind comes from the power amp. I guess I will find out. I know that some folks switch the EL84s out for 6V6s to tame a bit of that.

Note that your 3rd stage cathode isn't connected to ground in your schematic.

In case you run into the overkill problem with EL84’s you can always try EL90 tubes. They’re 9 pin versions that have a lot in common with 6V6GT tubes. You’ll have to rewire the sockets and watch out for the B+ though. I did this in the past with great success. Bias resistor needs to be changed too ofcourse

[Iain T]

Been doing a bit more testing and I'm pretty sure I have a parasitic oscillation going on in the power amp section!

Testing methodology:

• 1kHz Sine wave @ 150mV into input of the low-gain channel.
• Oscilloscope attached to speaker out.
• Master Vol on full. Middle full. Treble and bass on zero.

With NO power tubes installed pre-amp output is looking great.

I'm able to get a clean sine wave in the low-gain channel up to about 8/10. Obviously, lower gain on the high-gain channel before the waveform starts to square off. Output is clean after the PI and verified on the inputs to the EL84s.

Adding the power tubes back in sees a lot more noise on the output signal and unpleasant squeal with he master at anything over 2 and gain still 8/10. The output on oscilloscope goes crazy!

The layout diagram suggests that: "The two transformer wires going to pin 7 on the EL84 power valves may have to be swapped over if the amp squeels like a banshee when you turn it on for the first time."

Does this sound like a likely culprit? It's a fairly quick fix so I'll probably try it anyway.