TM36 FX Loop Issue

Hello all from the not-so-sunny Isle of Man!

I come seeking advice please...

I recently picked up a used TM36 and overall, I’m really impressed with it. My playing style varies wildly, from detuned black metal to country and I find this amp is versatile enough to cover most of my needs.

The only issue I have is that the FX loop isn’t functioning correctly. The FX on/off switch doesn’t seem to have any effect at all; whether it’s engaged or disengaged, the signal passes right through. Furthermore, when effects are placed in the loop, they don’t affect the signal at all; the signal is 100% dry.

The amp still has it’s stock Chinese tubes which seem to be working fine. The TSC indicates that everything is functioning normally at least. Could faulty tubes affect the function of the Fx loop?

I’ve included additional notes below which may help to narrow the issue down:

> Factory reset yielded no change (using updated manual method - hard copy manual had a typo)
> No signal when plugging directly into Return jack (whether FX engaged or not)
> Sending preamp signal to another amp via TM36 Send works
> Reverb switch also doesn’t work. Reverb is permanently engaged (but the Reverb mix knob still functions okay)

Being based on a small Island has its perks, but access to decent amp techs isn’t one of them, I’m afraid! I appreciate these types of issues are hard to diagnose remotely, but any advice / help would be massively appreciated

Thanks in advance,

Chris

Hi CBS.  This is very odd as the Fx Send/Return system on your TM should be very straightforward.  I have access to data on the GM36 and the TM18 but not the TM36 I am afraid.  While they differ in their loop setup, there are usually definite family resemblances in the way these amps work so I believe your TM36 will follow the TM18 more than the GM36 and that is pretty much bog standard.

In the TM18 the preamp signal is fed to the terminal of the Send socket and also to the switched wafer of the Return socket.  The Return socket live terminal feeds the signal on to the power amp section.  If nothing is plugged into the Return socket its switch remains connected so the normal preamp signal is fed through it.  When a plug is inserted it breaks the connection and the signal on the plug takes over and becomes the power amp signal.  In normal use this signal will have been taken out of the Send socket and fed through various pedals before being fed back in.  However your TM36 has the push switch to engage the loop.  It really sounds as though you have a problem in that area and the "No signal when plugging directly into Return jack" point you added also is in keeping with that.

I would guess your best bet would be to open the amp from the underneath panel, there is no warranty to worry about now I assume, and simply look to see how easy it is to access that switch for replacement.  If you can get a picture of that switch I can certainly suggest a replacement model, these types of switch are not too expensive.  If you can find a normally decent tech he should be able to drop that in for you at not too much cost.

Thanks Bordenbert; I really appreciate you taking the time to come back to me! I’ll get a pic of the switch uploaded shortly.

I spent a little more time with it after my initial post, and I think I’ve stumbled across some more penitent info. I fed another amp into the TM36 return and cracked the gain, and low and behold, I could hear it through the TM36 power section and cab. Intrigued, I then put an obnoxious delay/reverb pedal in the loop with a boost in front and I can just about make out the repeats.

That led to be plugging a guitar into the boost pedal straight into the return and I can now hear signal, albeit extremely fuzzy, intermittent and crackled. Could it be that the fx loop buffer has gone? Or perhaps a dodgy power tube?

Thanks again

Aha!  That's useful.  Your new info shows a few things.  Do bear in mind that I am having to guess as to the correct configuration of this part of the circuit as I do not have schematics for the TM36.  It is a significant step on from the TM18 but not necessarily as sophisticated as the GM36.  "Somewhere in between" is the best I can guess at.

It won't be a power valve, the amp wouldn't work as it should in some modes as it does from your info and not others.  The power amp is in circuit without any changes to its setup in all cases.

You could be right though, it does sound as though it could possibly involve circuitry after the loop such as a a loop buffer.  The TM18 has a loop driver and a dedicated two transistor amp after the Return which lifts the signal level to what is needed to drive the valve power amp.  The GM36 has the same dedicated loop driver but on the return the processed signal is simply put back in the chain at the same point the Send is taken from as there is the whole Gain/Volume/Eq section to pass through.  As you say the amp works correctly when the Send is fed into another amp we can assume any loop driver before the Send which may be in there is working as it should.  The fault could be in a loop Return buffer if the TM36 has one, (which I reckon it will have), AS LONG AS this is dedicated to the loop signal and it does not pass the normal preamp signal with the loop switched out of circuit.  A setup as I described in the first post, where the preamp signal is routed across the loop sockets by their own switching, would not work like this however yours has a loop engaging switch to perform that task so it could well be different.

I know this is a bit technical but it may give some help to anyone you manage to get to look at the amp.  In the meantime I have asked H&K tech support if that switch is available as a spare part or at least if they can tell us who is the manufacturer of the one they use.  I don't have the actual part number as I don't have the schematic but the other switches used in the GM36 and the Switchblade are in the same series so it will go something like this:  "CIC WPML-2YL-NL" or "CIC WPML-4YL-SL".  That seems to mean it is a CIC WPML series, 2/4 pole, Red/Blue/Yellow (R/BL/YL) coloured, and the SL/NL is a mystery!  I can't find any reference to those numbers online so it may be that H&K have altered them to suit their own numbering system.  Switchcraft have a similar looking switch which may be the original and should show you what you are looking for:  Mouser version of Well Buying Switch It is an industry standard mounting footprint so there may well be other manufacturers out there supplying similar switches which would fit perfectly.

That led to me plugging a guitar into the boost pedal straight into the return and I can now hear signal, albeit extremely fuzzy, intermittent and crackled.

Now, if that is the case how about when you play like that and push the switch on and off fairly sharply and repeatedly? Does the switch alter the sound in any way?

Just for the record, I have found a picture of the TM36 insides.  Here is the switch we are talking about ringed in red.  You can now see that it is a DPDT version, (double pole rather than the bigger 4PDT 4 pole versions), and it is the industry standard footprint.  Being just inside the base of the amp it is easy for either yourself if you are capable, or for a tech if you are not, to remove and replace.  If it is the switch which is faulty it should be easily available and a straightforward repair.

If you can get the bottom off yours and take a few clear pictures of that area we can maybe see if there is some other obvious issue though I don't think there will be too much that would be visible.

Hi CBS!

I have the same problem.
Did you find what was broken?

Hi Vlad.  Have you tried the tests recommended in the earlier posts?  If not then try putting a signal direct into the Fx Return socket.  This is best from the Fx Send/Preamp out of a different amp or an output from a boost/overdrive pedal with the guitar plugged directly into that.  It really needs to be a largish signal, bigger than what a guitar usually puts out though you should hear something fairly low level with just a guitar.  If that is fed directly into the Fx Return then it is cutting out any effect of the preamp circuitry right up to the Send socket, including any loop switching and buffer.

With that setup then try popping the switch in and out fairly firmly.  Do you hear anything of the guitar signal with that setup, and does the switching action produce any clicking or crackling noise at all?



(EDIT: Just checked up and _CBS joined 26th Dec '20, posted his problem, and last visited on 29th Dec '20. I don't think it's likely that he will be back to respond with how he fixed his own problem Vlad.)

Hi bordonbert,

I took the test and it's 100% the same as CBS

> No signal when plugging directly into Return jack (whether FX engaged or not)
> Sending preamp signal to another amp via TM36 Send works
> Reverb switch also doesn’t work. Reverb is permanently engaged (but the Reverb mix knob still functions okay)
>Switching produces a click and when Fx is active you hear crackling noises but no guitar signal

This has been a difficult problem to diagnose both times it has occurred Vlad. I don't have a schematic for the TM36 so I can't see what is really happening after that Fx Return socket. I do have the TM18 to refer to but that has significant differences which are right in this area. For one the Reverb is permanently engaged with no switching involved. It is connected to the Fx Return line direct and permanently. In your TM36 that is not the case. The best I can do is to hope that the TM36 follows the GM36 which I know well. So I'm afraid all the advice I can give is based pretty much on guesswork.

The TM36 has some MIDI capabilities. If you take a look at the user manual page 16 Section 3.3 you will see what the default Factory Settings for the MIDI are. In the bigger more MIDI functional amps the MIDI side can get its head in a whirl and become corrupted. Even though the TM36 is much simpler I wonder if that might be what is going on here. As the Reverb is one function controlled by MIDI I would suggest you try performing a factory reset. If you have nothing stored which is crucial to your setup then try this: "To reset the amp to the factory default settings, press and hold the REVERB ON and FX LOOP ON while powering up simultaneously until the MIDI LEARN button lights up. After you release the ON buttons, MIDI LEARN will flash once to confirm that the reset has been executed properly." In Section 3.3 there is a list of the default settings for the MIDI which will show you how it is now programmed.

I think that any work from this point on would require opening the amp and testing stuff inside. That honestly does require specialist experience and test equipment. The voltages in there are at best destructive to the amp and at worst genuinely lethal to your health. People keep saying that but less experienced people do tend to think they will "be fine" as long as they take care. Unfortunately, sometimes even the simplest tasks can lead to disaster. This work is not just the same as changing out a set of valves, it needs much more knowledge and skill than that. So, all that said, is taking it to a tech to have the fault investigated an option?

Another idea and it's simple. The Fx Loop can be activated by the rear switch, by MIDI or by the external footswitch socket. We've considered the idea of a faulty rear switch, we've considered the MIDI side and suggested a factory reset, but the footswitch socket could just be involved. Try putting in a TRS stereo type jack plug or cable, (that's the 3 terminal jack plug). Then if you short out the terminals for the Tip to the Sleeve you should activate the Reverb. Shorting out the Ring to the Sleeve should activate the Fx Loop. There is no danger of any damage doing this, it simply mimics what the stomp switch in the footswitch is doing. Can you try that and see if they act as they should that way?

Hi

I found the solution on YouTube, it's exactly the same

(Tubemeister 36 reverb and fx loop problem (fixed))

This Hugues and Kettner Tubemeister 36 amp head had a non-working effect loop, and the reverb was always on.
After diagnostic, I suspected the analog multiplexer (CD4053), so changed it, but it was not enough. The solution is on the powering of this multiplexer, which is a 15VDC stepped down to 7.5V through a zener diode.
You have to replace the Zener Diode and the capacitor parallel to it, and there it goes.

U4 chip (CD4053) and the zener diode D15 (UDZS7.5B)

That's wonderful! Well done for getting to that. I wonder if the guy got his hands on the schematics. It'e either that or he just probed around with a meter to find an oddity in the power lines. Well anyway, that seems to be the way to go. I would try just changing the zener alone first off. It's very possible that he didn't really need to swap out the chip. That would depend on whether the zener went short and pulled the line down or went open and the voltage went up which would fry the chip.

Anyway, that sounds like the solution and we have a new fix in the bank! Thanks for getting back to us with the info. Are you going to put this in to a tech to have it fixed?

I'm an electronics engineer myself, I'll first replace the diode myself, if not then the chip will help.
I'll report back if I managed to do it.

Here are a couple of things which might help here.

DO NOT remove the HCF4053 chip, it is probably not damaged.  It is not a case of the 4053 blowing and taking the zener with it.  I have looked at the circuitry of the GM36 and the chip seems to be part of the selection mechanism just as you say.  There are 3 chips of this type in the GM36 as it has much more switching capability. They have +15V on Vdd and 0V on Vee just as the datasheet shows.  It also shows that the absolute maximum for (Vdd - Vss) is +22V and we are running significantly below that.  The zener only supplies a half-rail voltage, 7.5V, as a reference to the chips pin 7 Vss.  In the datasheet this is shown with voltages on it as low as 0V, (the same voltage as is on Vee), and it describes having input voltages between Vdd and Vss with protection diodes in place.  I don't believe it is likely that this chip would be damaged by the zener supplying the +7.5V going either short or open.

You should first look at the voltage on the zener//cap to see if it is damaged.  If the zener has blown open you should find approximately +15V on it and you will have your culprit.  It may possibly be at 0V if it has shorted out which is even better.  The cap is a 2.2uF/25V electrolytic so would survive +15V and feel nothing with 0V.  The chances are good that the zener is the only damaged part though it would not hurt to replace the cap too.  The extra faffing around removing the chip and replacing may be stressful to the board and actually be totally unnecessary.  In the GM36 these are through hole devices so will need cutting out then removal of each pin individually with a general PCB clean up afterwards.

Good luck with the job, it shouldn't be too difficult to find the zener and check it out though it would mean running the amp up while it is open and with boards loose.  Not for the faint hearted or under-experienced.  Keep us up to date with your progress and even give us a few pics of the innards when you have it open, there is very little info about the TM36 available unlike its littler and bigger brothers.

Keep an eye on your private messages Vlad. You have one with info in it.

Everything went well, amplifier works properly again,
first unsoldered Zener D15 and tested it was ok, then I replaced D15, U4 and capacitor anyway, had all the parts there, didn't have a CD4053 and installed 74HC4053D.
It's really not easy, D15 is really small and very little space, it probably won't work at home with a soldering iron.

Many thanks for that info Vlad, it should be a big help to others.  I'll try to find a way to make it more prominent for internet searching.  If this is a known issue, (though a very rare one apparently with only 2 people coming here in years), having the solution out in the wild is important.

That's really difficult work too, I'm genuinely impressed.  I know the difficulties of working with surface mount SMD devices using only hand held "home soldering" gear myself.  I've done it in the past and that convinced me to get a hot air setup too.  It isn't ever an easy task, especially if the components are tucked away in spaces amongst much bigger ones.  That is the price we pay for more and more miniature boutique gear with greater and greater functionality I am afraid.  Keep an eye on those socket switch wafers, they look as though they could be a little corroded though it might only be reflections in the pic.  But knowing that you can take on that level of PCB work, I'm sure you've got that under scrutiny.

As to whys and wherefores, I do wonder why such a simple area of design has given up the ghost like this a couple of times.  I wouldn't expect a zener producing a reference voltage on what is I think a high impedance line, supplying no significant current to any other device, to be under any thermal stress at all.  Did you trace back and check out the value of its feeding current limit resistor?  I wonder if there is a case for that lowering in value as it ages increasing the standing current?  In the GM36 the resistor is 470R setting the zener current to a max of 16mA which would give a maximum of 120mW dissipation, which would be less if any current is taken by the devices on the 7.5V line, Vss + Inhibit.  The ZMM7.5 zener is a 500mW device, plenty of slack there, but in the datasheet its breakdown voltage is defined at only 5mA so why choose a 16mA standing current?  The breakdown curve shows there is not much difference in voltage between 5mA and 16mA and the voltage is only making sure that there is significant difference between Vss and Vdd for switching purposes.  It seems odd, granted without full understanding of the 4053 chip.

Anyway, I'm rambling on again.  Great fix!  Time to get back to using that amp.  Make some noise!!!  



EDIT: Just occurred to me. I wanted to confirm, (though you can really see in the pictures), this fix is on the Rearboard which is the lowest level outer one and easiest to remove with the bottom off the amp?

I hope it will help other people having the same problem.