Plexi makeover Debugging, I Need You !!

[warioblast]

Here's the plan.

The based amp is a 50W JMP; 1987 circuit; plexi that I made from a DSL 50 head. The final amp will have a 'Shiva' clean channel a dirty channel based on the XTC blue; with a toggle switch to go from JMP to JCM. Channels will be footswitchable via relay. I also intend to elevate the heaters. I think that's about it.

This will be my 3rd build, I hope I didn't get too cocky.

[warioblast]

I cut the bias part of my old board and turned the rest into strips.

[purpletele]

Alright!

Looking pretty serious Costanza!!

[warioblast]

Heater is done.   

[warioblast]

I relocated the heater wires to their correct taps.

The buss wire for ground is actually 3-pieces, in order to facilitate potential mods.

I Did a bit of lacing. :)   

[warioblast]

Power supply is done: bias, elevated heaters & relay.

I have wasted a bunch of cable ties so I'm going to wait a little to clean wires up.   

[pompeiisneaks]

Starting to really come together!  Looking good, keep the pictures coming :)

~Phil

[purpletele]

Interesting layout!  I didn't expect to see the tubes pre amp and power tubes on opposite sides.

Looking good.

[warioblast]

The layout is loosely based on the SLO. Transformers on both sides, preamp tubes on the front and on power tubes on the back.
I spend some time figuring how to make it right with the parts I had and my limited, but growing, knowledge  . There are surely better ways to do it.

Today wasn't as productive as I had hoped.

[warioblast]

Episode 6

[purpletele]

You are making some good progress!

[warioblast]

I didn't get much done today.

As I was working on the input jack I realized I put the sockets of the preamp tubes backwards. So I had to rewire the preamp part of the heater.

Before that I tested the power supply for the relay. I used this circuit provided by PRR http://el34world.com/Forum/index.php?topic=22124.msg235648#msg235648
I measured 29.4V so I guess I should be fine. The datasheets pretty much looks like chinese to me but I read max voltage 180%. It's a 24V Omron GV5-2-H1.
The power switch wasn't lit, so I looked at pictures I had taken before putting the amp to pieces. I had mixed up stand-by and power switches wires.
 
While juggling with the chassis and the sockets I realized the choke being very close to the PI tube. .. How come I didn't notice it before ? I gotta admit too, I didn't put too much thought on the placemcent of the choke...

Back to extensive google searches. I can't seem to find a definitive answer. Should I move the choke while I still can do it without too much hassles ?

Dayum I had to retype this message too. I can't wait for this day to be over :)

[bnwitt]

I didn't get much done today.

As I was working on the input jack I realized I put the sockets of the preamp tubes backwards. So I had to rewire the preamp part of the heater.

Before that I tested the power supply for the relay. I used this circuit
I measured 29.4V so I guess I should be fine. The datasheets pretty much looks like chinese to me but I read max voltage 180%. It's a 24V Omron GV5-2-H1.
The power switch wasn't lit, so I looked at pictures I had taken before putting the amp to pieces. I had mixed up stand-by and power switches wires.
 
While juggling with the chassis and the sockets I realized the choke being very close to the PI tube. .. How come I didn't notice it before ? I gotta admit too, I didn't put too much thought on the placemcent of the choke...

Back to extensive google searches. I can't seem to find a definitive answer. Should I move the choke while I still can do it without too much hassles ?

Dayum I had to retype this message too. I can't wait for this day to be over :)

Don't let it bother you.  I've got an old saying..."If you're not making mistakes, you're not doing anything."  This is one of the reasons I do my own layouts down to the nth degree.  Not every tube socket is aligned the same way (mounting holes to pin numbers) so sockets might not be in the same orientation with a particular layout anyway so as long as how they are oriented doesn't cause a wiring mess then it really doesn't matter.  I always do a chassis layout showing transformer positions, a board layout with all component values and a complete wiring layout showing ever connection before I build an amp.  I find that drawing the amp completely reduces mistakes.  I started this draw first build second way back in 1995 when my wife had me moving furniture around seeing how things would fit...hmm, try it this way, no, move the couch over there, no........  Then I went to Visio and drew the layout several ways, let her pick and did one move.  My back loved me.   It's what I call the 5 Ps  Prior planning prevents poor performance.  I'm just no good at winging it I guess.

[warioblast]

I decided on moving the choke a bit.

I'm short of a few turrets so I cut a piece of strip board for the relay.

Put the pots and switches in. It's a bit crowded around V2.
  A thought crossed my mind; I could build the amp without wiring the toggles. But it might be tricky to add them later. I'll sleep on it.

[warioblast]

I added the ground buss, prepared the toggle switches and wired V4.

[warioblast]

Another slow day. Wired V1.
I dusted off the breadboard and tried to figure how to wire the footswitch.

[warioblast]

1 tube and 3 switches to go. 

[warioblast]

 

Soldering is done. I will be firing up the amp tomorrow.
I will make a faceplate once I will have decided if I need all those switches. 

[warioblast]

I found a little moment to fire it up. Before that I checked all the connexions and found out I had the presence control wrongly wired.

Well, despite my last examinations, the amp isn't working right.

I'll start with the good points:
All the toggles switches work. All the controls on the dirty channel do what they are supposed to do.
The channel switcher, with the relay, works.

Now the bad parts:
The clean channel only gives a loud hum.
The dirty channel has a background noise, like a washing sound, not as loud as the clean channel. It doesn't have all the gain that it should.

The voltages looked in the ballpark. I didn't have the time to start the debugging process. If you guys have ideas, I'm all ears.

[pompeiisneaks]

Bummer, if you've got a layout and schematic, that would help a ton.  Seeing what's expected and comparing to the photos may help someone see the problem.

~Phil

[warioblast]

I don't have a scope. I checked for continuity on the wires and the connexions. Everything looked fine.
I poked around with a sharpie.
I swapped tubes. I took them out one by one. Hum went away when I took V4, the phase inverter tube.

I poked around again and I found a bad connexion at V6 ground that needs to be fix.

Now I have signal on the clean channel. The loud hum is still here but I do hear a clean sound that reacts to the volume pot, bass & treble pots works too. The volume pot has no effect on the hum.

On the dirty channel, it can go from loud hum, "washing" sound & cracklings and all of a sudden I'm left with just a hum not far away from being normal.

The JCM 800 side is notably gainier than the JMP side which is fairly clean.


Here are my DC voltages:
After rectifier 487V

Bias -45V

Elevated Heaters 42V (5.9V AC between the filaments)

V1
pin 1: 112V
pin 2: signal
pin 3: 1.13V
pin 4:
pin 5:
pin 6: 176V
pin 7: signal
pin 8: 1.40V
pin 9:

V2
pin 1: 133V
pin 2: signal
pin 3: 0.85V
pin 4:
pin 5:
pin 6: 189V
pin 7: signal
pin 8: 2.08V
pin 9:

V3
pin 1: 160V
pin 2: signal
pin 3: 1.1V
pin 4:
pin 5:
pin 6: 278V
pin 7: 160V
pin 8: 168V
pin 9:

V4
pin 1: 218V
pin 2: signal
pin 3: 49V
pin 4:
pin 5:
pin 6: 206V
pin 7: signal
pin 8: 48V
pin 9:

V5
pin 1: ground
pin 2:
pin 3: 491V
pin 4: 490V
pin 5: -37V
pin 6: 491V
pin 7:
pin 8: ground

V6
pin 1: ground
pin 2:
pin 3: 491V
pin 4: 486V
pin 5: -37V
pin 6: 492V
pin 7:
pin 8: ground

They look fine to me.

I took the grounding scheme from this guy http://el34world.com/Forum/index.php?topic=9400.msg85360#msg85360I have 3 points of connexion to the chassis.
1 = input, tone controls, all other pre-PI grounds
2 = all the rest; PI, power tubes, power caps, OT common, rectifier, presence
3 = earth from power jack.

Tomorrow I'm going to reflow every ground points. I think there is a big chance this will fix the hum on the dirty channel. As for the clean channel I have no idea.

[warioblast]

My layout.

[warioblast]

I fixed up my bad connexion at V6. Reflowed the entire ground connexions. Reflowed the clean channel.
I bypassed the relay by wiring the clean preamp directly to the PI. I also wired the clean channel on a separate input jack.

Nothing changed. The clean channel gives a loud hum. And the dirty channel vary from almost normal to bad craklings. At times, the entire amp becomes microphonic.
When I turn the amp off, you can hear the hum slowly fading.

[warioblast]

Those schematics might be clearer.

[pompeiisneaks]

Try chopsticking around, especially the OT wires, and such, I had a recent hum in my AC30/4 that was just my ot wires being too close to the tone stack wiring, a slight shift and it was gone. 

I don't have time to peek yet at the rest, hopefully someone else will, if not I'll try to look back later myself.  That's a good idea on the reflowing grounds too, but chopsticking the ground wires/points will make pops/hum increase/decrease which could help ensure that's it.

~Phil

[warioblast]

I swapped all the tubes one more time and the dirty channel is acting normally. Fingers crossed.

On the clean channel, I tried to stack a bigger resistor and a bigger cap on the cathodes with no results.
There are several different schematics of the Shiva floating around. Some of them have a 470K resistor between C17 (as per the last shiva schematic I posted) and the relay. I bypassed that resistor and the hum cut down by 50%. Now the hum is as loud as the signal 

[warioblast]

Dirty channel is up and running. The washing sound came from the presence cap.

I modded the clean channel to be like the Landry LS100, basically I moved the tonestack after the 2nd half of the tube. The hum is less predominent but still there. I'm wondering if my layout might not be the problem. V1 used to be closer to the OT but in a different angle.

[pompeiisneaks]

Hum could very much be lead dress.  I'd suggest looking at the heaters near the input tube, or any strong signal wires to/from the OT near any part of that channel.  Do you have a scope?  If so you could try and trace out where the noise enters the circuit.  The voltages seem pretty decent to me, as far as I can tell, I don't know the circuit super well, V3 looks like a DC Coupled cathode follower with that high voltage on the grid right?

~Phil

[warioblast]

V3 is indeed the CF. It's a tweaked plexi.

I don't have a scope. 

I don't really have an OT related wire close to the clean preamp part. But V1 plate resistor is the closest thing to the OT. I was thinking about moving it an inch farther. I will also try to unbolt the OT and took it outside of the chassis.

[pompeiisneaks]

If you just try and move that wire with a chopstick does it impact the hum?  It should if you move it closer or a bit farther away.  If not, it may not be that issue. 

~Phil

[DummyLoad]

you have silver mica signal caps laying on top of filament string, you have filament string passing directly under signal switches, under pots, input jacks, etc..

IMO, respectfully, it's a poor layout with respect to the front panel and filament string routing...

move flament string away from switches, caps, pots etc.

as a first, try pulling the two blue mini toggles out of the front panel and tape them off away from filament string & move that silver mica cap that's on top of the filament wires (in front of the barcode) or just clip it out the circuit temporarily if it's a bright cap.

--pete

[warioblast]

The 3 close mini switches are only active in the dirty channel.

Like I said the layout is loosely based on the SLO. Only his have the input on the side away from the filaments, no mini switches and filaments using buss wire.

The whole layout was a headache. You have to put this there, but you can't put that close, but some do it and get away with it... I think you what I mean.

I hesitated for the filaments; use the buss wire, put them on top like Fender, route them along the chassis or not. I could move them under the turrets strips. 

Ultimately I think I'm going to get rid of the mini switches, build a stock fine working plexi and go from there. In regards to the dirty channel.

As for the clean channel, I didn't pay attention to it, but with the Landry clean, the hum reacts to the volume knob. And with the volume back off, I only hear your little normal background hum. That means the massive hum appears at the 2nd half of the tube, right ?

[pompeiisneaks]

I'm definitely at a loss as well, usually it ends up being positional related.  If you can move the OT a bit, even rotate it, and see if the sound diminishes that may help too.  It's never fun chasing hum ghosts.

~Phil

[warioblast]

I didn't have much time today. But moving the OT is next for sure. I could try losing the elevated heaters.
If it doesn't change a thing I'll consider moving the filaments, and maybe swap places between the preamp tubes and the turrets strips. How close can the preamp tubes be to the power ones ?

[warioblast]

So I gave the  OT a 90° turn = 0 effect
Took the OT out of the chassis = 0 effect
Got rid of the elevated heaters = 0 effect

Before messing with the filaments wiring, the preamp tubes and the turrets I decided to modify the grounding scheme. I went back to the Larry grounding.

  Most of the hum is gone. 

I got rid of the buss wire. Right now I have the following ground nodes:
1: input, V1 & V2, heater CT
2: gain from dirty channel, V3, preamp filter
3: V4, OT common, dirty mid & volume, clean volume, bias
4: V5, V6, main filter, screen filter, rectifier
5: power cord

I still have the Landry clean , I'm going to resolder the Shiva back in.

The amount of hum is low, I'm wondering if it is worth the hassle to move stuff around. 

[pompeiisneaks]

Oh yes a ground loop can definitely cause ugly hum, I've had to beat that demon on a few builds myself.  Glad you found it!

~Phil

[warioblast]

Thx for sticking with me. 

I rewired the clean to Shiva specs, the hum is back. I'll try a few mods tomorrow. At least I know that I'm good with the clean Landry.

[pompeiisneaks]

Yeah and that implies something with the shiva mod is not happy.  I hope you figure it out, I worked on a bogner shiva maybe 6 mos ago and they sound really great.

~Phil

[warioblast]

Did you have to work with / on a schematic ? If so, do you see something wrong with the schematic I posted ?

[pompeiisneaks]

I didn't have a schematic, and the problem was, as I recall, just that the caps were getting old, and there was a bit of a buzz.  It disappeared when I did a recap of the electrolytics.

~Phil

[warioblast]

I failed to make the Shiva clean hum free.

I tried the clean channel of the Orange Rocketverb and it was still humming.

I put the Landry clean back in but mainly with alligator clips, and it resulted in hum too. I soldered it properly and the hum was gone, but the problem is I don't want to add a gain pot and I couldn't get a pristine clean sound.

So I tried:
_ Shiva;            1/2 tube -> tonestack -> volume -> 1/2 tube              = hum
_ Friedman BE clean;  same topology                                                  = hum
_ Landry;           1/2 tube -> gain -> 1/2 tube -> tonestack -> volume = no hum
_ Orange RV;      1/2 tube -> volume -> 1/2 tube -> tonestack            = hum

I decided to try  1/2 tube -> tonestack -> 1/2 tube -> volume. I got no hum  and a nice clean sound 
The tonestack became less effective and needs to be modify though. What a journey !

[Ed_Chambley]

If it were me I would do as Dummyload suggested.

[warioblast]

Would you just rewire the heater or invert the tubes and turrets places, kinda make it like the Hiwatt layout ?  While I'm at it, do you guys see other improvements on the layout ?

[DummyLoad]

the original plexi layout is really hard to beat. i'd humbly suggest that you start there. try not to stray away from too much and you don't need to keep it as compact as the original - it'll most likely be a bit more stretched out due to your turret spacing and you have wider chassis. build the plexi plan with SSR and use the unused capacitor - you'll like the tone better with a bit more filter. also, try 1uF in place of the .68uF.


http://site.triodestore.com/JTM45LAYOUT.pdf


respectfully,


--pete

[Ed_Chambley]

I have done the same thing years back.  You cannot know what you do not know.  But I wanted my amps sound and quiet, so I would take them apart.  I would attempt to use chassis to small.  Put filter caps close to pots and had humming madness.  I still took them apart.  I did not think it possible amps would do this.


I kept on seeing the knowledgable folks write about using proven layouts.  Now almost every amp I build follows a proven layout if it has more than 1 preamp tube, but I do like to build simple point builds.  What you are building has a lot going on.  Follow the lead of the guys the know and you will receive all the help you need and most importantly you will have a great sounding amp that is stable and will last.


It will make you grin!

[warioblast]

So I changed my layout into something HIWATTish. From the front side to the back I have: pots - turrets - preamp tubes - power tubes.
I think I have a bit more hum.   

The cracklings and parasiting noises came from a faulty presence pot. I took it apart to have a look and I don't see why it doesn't behave normally. I ended up changing it.

I still have the same problem; massive hum; with the clean channel.
When I first fired up the amp; and it will turn out I had forgotten to wire the presence pot; the clean channel worked fine despite a low volume.
Wired the presence pot and the hum came back. I moved a bunch of grounded wires but with no luck. The mid pot from the dirty channel actually interacted with the volume of the clean channel. 
Right now I have my clean like this : 1/2 tube -> tonestack -> 1/2 tube -> volume. The hum is slightly higher than on the dirty channel. I will try to put a resistor between the tonestack and the 2nd half of the tube.

The dirty channel is fine. I got rid of the 'bright' switch. I still have the boost switch but it gets really noisy when active so I will take it off.
I have 2 toggles switches wired to pin8 of V2:
_ n°1_ I have a 4k7 resistor that's parallel to the 4k7 resistor on the turrets, so I can have 4k7 and 2k35
_ n°2 _ Also in parallel I can add a 2.2uf or a 0.68uf cap
It's interesting to realise what they do... and how much noise they can bring in. I'm going try 10k like the JCM800. 

About the layout, I could go back to the classic Marshall type but having this laydown output transformer doesn't facilitate things.
The pic shows my initial build. I had the heater wires running along the OT's. What would be a better option:
_ moving  the heater on the other side of the tubes, between the tubes and the board.
_ using a buss wire
_ standing-up heater like Fenders

[warioblast]

Update:

Last week I decided to gave up and bring the amp to my local tech. I reached out to him, and he promptly answered me he wasn't doing repairs anymore:  He has concentrated on selling his own amps; guitars and hifi.

So this is where I'm at.
A couple of weeks ago, I changed the heater wiring and put buss wire. I focused on building a straight-up JCM800 2204. I ended up with more hum than I wanted. I once again changed the heater, went back to twisted wires. I tried to put the wires away from the front pannel = same hum. Put the wires back in the corner of the chassis =  same hum.
I tried 3 different grounding schemes; one of them having only 1 connection to the chassis, I didn't hear any differences.
I tried a set of new tubes too. I reflowed all the soldering points that were accessible.

Could this hum be generated by an overheated component  ?

I'm considering putting the amp back to a Marshally layout. I have a few concerns though.
If I save a spot on the turret strips for a clean channel, I will have a bunch of medium long wires (+/- 30 cm -- 10-inch) going from the CF & PI tubes to the turrets. That worries me a bit.
If I get rid of the clean channel parts, I could move my separated bias circuit board back on the turrets though.

[pompeiisneaks]

If you're going to readjust and have long signal leads, use shielded wire, connecting the shield to ground near the start of the run not where it heads off to, so the noise picked up returns to the previous stage and can't couple into the next one.

~Phil

[Auke Jolman]

There's an article at the http://www.geofex.com/ site. Maybe it helps you identifying which kind of hum you're dealing with.