JTM45 Input stage problem

[JB]

Hi,

I built a JTM 45 a while ago.  Had some issues with it and now coming back to debug.

It's basically like this:

http://el34world.com/charts/Schematics/files/marshall/Marshall_jtm45-readable.pdf

But I've installed a power scaling board for the output stage (KT66's) and hence a post PI drive control.

The problem is in the front end.  It's noisy at times and I noticed the guitar volume pots being scratchy - yep, DC on the input jacks.  I did have a shared/split cathode switch which was intermittent.  I've removed that and linked the cathodes - still the same.

I've been all over it, re-soldered, checked earth paths but can't find it.  I've checked the input jack wiring, all looks good.  From the tip of a guitar lead I measure 34K in series to grid and 1M to ground on the high sensitvity inputs and 68K series to grid/134K to ground in the low sensitivity inputs.

I was measuring a little DC (low tens of mV) on the volume pots, so for now I've unsoldered the pot ends of the coupling caps and have V2 removed.

Voltages are all a little high at the moment as I've put in an old pair of 6L6 for testing purposes and they're running a bt cold.  I'm measuring:

V1 B+ node = 389V

Normal channel triode

Va = 254V
Vk = 2.1V
Vg = 7mV


Bright input triode

Va = 260V
Vk = 2.1V
Vg = 15mV


If I plug a guitar lead (no guitar connected) into the lo input of the normal channel then Va/Vk stay the same.  Vg goes upto about 20mV and I'm measuring the same on the tip of the guitar lead.  Doing the same on the Hi input of the normal channel gives me Va= 226V/Vk = 2.2, Vg = 0.6V, guitar lead tip = 0.6V

For the Bright channel lo input with a lead plugged in I get similar results, Va/Vk stay the same, Vg goes upto 38mV.  For the Bright channel Hi input, Va drops to about 220V, Vk goes upto 2.3V, Vg measures 0.8V and also on the tip of the lead.

With V1 removed I'm still measuring similar volts on the guitar lead tip!

Feels like a grounding issue somewhere but I can't see it.  Jack's and cathode resistor/cap are grounded together just under the jacks.  The PSU decoupling cap for that stage is down to chassis about 2" away, along with the pots.  I'm not using a pot buss ground.  If I jumper a lead from the ground end of V1 cathode resistor/cap to the decoupling cap ground lead it makes no difference.

Any thoughts on what to try next?

[shooter]

It's noisy at times and I noticed the guitar volume pots being scratchy

does it do the same thing with nothing plugged in?

[HotBluePlates]

Feels like a grounding issue somewhere but I can't see it.  Jack's and cathode resistor/cap are grounded together just under the jacks.

Plug a guitar cable in. Measure continuity from the cord's ground all the way back to your 1st filter cap ground (the one right after the rectifier).

In continuity mode of most meters, you should get a beep. Once you do, switch over to measure Ohms. Should be less than 1Ω (many meters will register 0.2-0.6Ω).

[JB]

I can hold the sleeve of the guitar plug against the chassis and still measure dc on the tip.

A croc clip lead from v1 cathode resistor to the decoupling cap for that stage doesn't change the dc reading.

Weird.

[VMS]

Sometimes the board gets conductive, try if lifting your input resistors from the board helps.

[HotBluePlates]

I can hold the sleeve of the guitar plug against the chassis and still measure dc on the tip.

Understood.

I'm asking you to verify ground integrity from the cable ground all the way back through the amp, to determine if grounding is your issue.

[JB]

From the sleeve of the plug, right through he cord to the first filter ground point is 0.4 ohm. About 0.3 of that is from the cord itself.

I know some of the old Fender boards can get conductive. This is a brand new board so hopefully ok!

I think I may have to undo the pots and jacks and lift the board and look for anything wrong underneath.  Maybe also measure for dc on the input again with it lifted.

[JB]

Ok, properly weird now.

I lifted the jack ends of the bright channel 68K resistors.  Valve removed, power up, measure the grid, approx 12V and gradually decaying through the DMM.

Pull the corresponding grid wire off the valve base in case there is some tracking there. I now measure 0V on the grid pin and 12V or so on the wire I've just lifted!  So 3" piece of wire connected to the two 68K resistors, the other ends of which are floating, has lots of DC on it.

Measuring continuity from the floating grid wire to every turret and source of volts in the amp - nothing, all open.

I've not lifted the normal channel 68K's, but the initial symptoms are the same for that channel.  So the same weird issue twice??

It's a pain in the rear to lift the board, and nothing I'm measuring says I need to, but I can't think what else to try.

Anyone got any suggestions before I buckle up and start dismantling?

[sluckey]

Measuring continuity from the floating grid wire to every turret and source of volts in the amp - nothing, all open.

Don't just check continuity. Check actual resistance. Show us a hi rez pic.

[JB]

Hi,

Sorry, should have said resistance, which is what I measured.  All off the scale of my DMM - many meg ohm.

I've made a bit more progress since then.  I undid the mounting bolts for the board and lifted it an inch or so.  I removed the bright channel grid wire completely and re-ran it a different route - soldered it topside, dropped it through a hole behind the turret and then down under the board and back up in a U shape, so it's definitely flat against the chassis.  Cured the problem!

Here's a photo link:

https://www.dropbox.com/s/hegy9vc9wl9en5f/2015-08-31%2014.15.22.jpg?dl=0

The bright channel 68K's are the right hand pair.  Immediately next to them are the anode resistors, so maybe it was picking it up from there somehow?  I'm doing the same for the normal channel which has the same issue (but no immediately adjacent HT).  However with that grid wire now on the turret and re-run under the board I'm still measuring DC there.  I'm going to lift the jack end of those 68K's next.

I'm getting some DC on the volume pots too.  Don't know if that's a similar issue from the grid of the next stage through the mix resistors.

On the rear panel is the now bypassed split/shared cathode switch, hence the temorarily joined yellow cathode wires with the black sleeve.  It did all like nice and tidy - will need some straightening up when I'm done!

[sluckey]

Nice pic. I think your board material may be conductive. Try this... Unsolder the ends of the two 68Ks and green wire from that common terminal. Lift the ends up slightly and solder the two ends to the green wire so they are just floating in air and not touching the terminal. Do this for both pairs of 68Ks. Does the mystery voltage disappear?

[JB]

I've just re-routed the normal channel grid wire and I'm down to around 100mV of DC now.

If I take the +ve meter probe and push it into a turret hole, so it's touching the inside material rather than the shiny top surface, I can measure from a few tens of mV upto around 200mV in places.  Near those 68K/grid junctions it's around 150 to 200mV.  That's not ideal but enough to worry about?  Re-making the board at this stage is not top of my list of things I want to do!

I get a little bit of DC noise as I turn up both volume pots.  I'm measuring around 100mV on each when turned full up.  I may try re-routing the wire from the mix resistors to the next grid, same method as I've done on the first two grids.

Un-related question - that silver mica treble bypass cap on the second mix resistor is fairly microphonic.  Not sure I have another of the right value.  Could that be a noise source and hence worth replacing?



UPDATE - re-routed the second stage grid wire. The amp is a little quieter now.  Still have 100mV or so on the high sensitvity inputs of both channels, and similar on the two volume pots.  So a bit of pot noise from them and the guitar volume but way improved from where it was.  That bit of conductivity through the board is going to bug me now!  Anyone with a JTM 45 or 5F6A care to measure what if any DC you get on the tip of a guitar lead?