Noise Gate problem

Hi, i'm new to this forum and new to the GM40. I recently bought a secondhand model. The amp sounds great and pleased with the purchase. However, the Noise Gate isn't working. I was hoping to progress a home repair if possible.

When not engaged the amplifier plays normally. When the NG is engaged the LED lights and the output volume drops to almost zero (it sounds like the NG is permanently fully on) - and adjusting the NG control has no affect. I've tried adjusting other controls.... power/gain level and power soak etc., to try and tease out where the problem might emanate. But nothing changes, so i must assume their is a fault around the noise gate circuitry.

I have the GM36 schematic and i'm guessing that the NG circuit will be very similar to the GM40 circuit. I was wondering whether anyone has experienced similar problems, or can describe the NG circuit operation? there are a few test points mentioned on the circuit, and it would be helpful to know what might be expected on them. A long shot but hopefully someone might have just a tiny bit of information that might help out.
Thank you for any feedback, Andy

Hi Andy.  Do you have experience of working inside modern valve amps like your GM40D?  That includes working with surface mount components.  If not then I would strongly advise you to take a step back before you go in to do anything hands on.  As you say you already have the GM36 schematics this makes thing a little easier.  However, even though they are out there in the wild please don't post too much here in open forum about it or where to get it, we do recognise  H&K's intellectual property and we value their helpfulness with us here so any direct info has to be removed.  (People can find this info freely available on the internet anyway so I'm sure it isn't a problem for anyone, it's just a matter of respect for H&K's business.)

You are right in that the basic blocks of the GM36 will most likely be shared with the GM40D and the Noise Gate is very likely one of them.  Electronics engineers tend not to reinvent the wheel if something is working well.  The circuitry may well be the same but the PCB layout might be subtly different and I would guess that the component numbering will have changed.  A change in anything circuit related adding or removing parts can domino up through the other components if they update the numbering to match.

I think a good starting point is to remove other easily checked possible factors before you start thinking of chasing circuit faults on the boards themselves.  Have you tried a factory reset of the amp?  If not then start there.  The Noise Gate, like most other functions, is tied in with the microcontroller system.  A glitch in that and you may get problems like the ones you are experiencing.  TAKE CARE, a reset will delete any presets you have created so be sure to back these up in some way before you delete them.  The switch LED voltage which also controls the gate itself, and the switch selection signal, (it's just +5V or 0V DC level) are on separate lines and handled by different processors so the fact that the LED is working does not mean that the Gate itself is being controlled correctly although the switch action is most likely to not be a problem.

Hi bordonbert , thank you for taking the time to respond. I understand what you say in respect of H&K intellectual property but yes the information is freely available, and my only way to try and address the problem. My preferred option in respect of a repair would be to diagnose the likely problem-area before doing anything, by understanding circuit operation in that area. I was hoping someone might have a shred of information that might help. I can read some of the circuit operation but really need a bit more of the ethos.

Yes, the logic switching side of the circuit seems to function ok, so I can only think that one of the discreet surface mount components has failed. This might be a show stopper in itself. At least if it had been one of the high voltage components it might have visibly failed and easier to diagnose/replace. I don’t have access to a scope only a multimeter currently.

Having said this I did disassemble some of the amp chassis to take an initial peek at what the circuit looked like. I got as far as removing the rear pcb containing the NG adjustment (to check the pot itself). So that’s where I stand at the moment. I keep going back to the schematic to build on what i think is happening.

I still would advise you to try a reset first.  It costs nothing and takes only a few seconds.  It will rule out the software as a source of the problem.  It doesn't just clear out the stored presets it cleans up the actual control software too.  Glitches in the software are a problem with smaller digital systems like our amps and reinstalling everything that can be restored is a good thing to do.  Some really fussy professional guys would do this as a regular maintenance task.  It is impossible to be absolutely sure that the control side is working as it should under all circumstances without doing it.

Ok, let's look at the circuit.

The switch is basically a paralleled pair of switch wafers, (dual redundancy), which either allows a +5V High signal from the power line through a resistor to the master processor on the N-Gate-SW line, or they ground the resistor end and hence the pin of the processor.  This simply tells the processor whether you want the Noise Gate on or off.  The processor then outputs its control signal on the LED-N-Gate line.  This line also controls the Noise Gate circuitry.

The Noise Gate block on the uC and Digi-Pot board is basically a pair of JFETs which turn on when the audio signal is to be quenched and short the Send-EXT-FX line to ground killing it.  The Input-Signal line feeds into the preamp section but also acts as an input to the Noise Gate control block on the Rearboard which sets its output high or low according to the signal level on the NF-to-Gate line.  That block returns a control signal on the Gate-Signal line in response to that level.   A Low level on either the LED-N-Gate line, (noise gate off), or the Noise-Gate line, (signal high enough to be playing), will turn off the JFETs allowing the signal to play.  A High level on both of these will allow the JFETs to turn on on killing the sound.  With the Noise Gate switch Off the noise gate is prevented from working no matter what the Rearboard block is telling it.  With the noise gate On the gate is controlled by the signal from the Rearboard control block in response to the signal level, which is noise gate action.

You should be looking for the signal levels on TP204.  I can't find that in the PCB layout pics.  There are a lot of components which don't show up in that so it looks to me as though we are only seeing one side of the boards with many components on the far side not showing.  A High level on TP204 will make the noise gate kick in muting your sound and a Low level will turn it off allowing you to play.  The NG Switch takes precedence over the signal level of course, and the Mute control overrides them both but isn't in our thinking yet.  You can see that the only situation which allows the Noise Gate to silence the amp is when the NG is switched On and the signal level is low. The level on TP204 should be as follows:

NG Switch	Audio	Level TP204
Off	Silent	Low
Off	Playing	Low
On	Silent	High
On	Playing	Low

.
If the internal mute is activated it will pull TP204 up no matter what the other two control lines are doing.  I hope I have all of that correct, it's late here and this is all pretty quick and off the top of my head.  Have a think about this and don't be too ambitious too soon!!!  Remember - THE HIGH VOLTAGES INSIDE THESE AMPS KILL.  THAT IS GENUINELY - "KILL"




(Incidentally, that's my favourite track from Layla.  Great choice!)

Thank you, that’s a good and detailed description, I can only imagine you have been involved in the design and, or repair of the H&K amps. I have read and re-read your description there’s a lot to digest, and gives a fighting chance to hone in on a likely problem area.

I did try re-setting the amp when I first got it home, but i will try this again before anything else. Yes, I’d like at least a few days to think things through on the diagnosis front, before deep diving. But TP204 seems key to working out whether it’s the control bit of the NG not working, or the actual clamping action of Q9, Q10. If I’ve understood correctly. I’m guessing the Mute signal and associated Q2 are ok, else the amp wouldn’t work - with or without the NG activated.

Again, thanks that’s very helpful. I’ll come back later in the week when i’ve had a chance to think things through and test out the above.

Yes, lovely song Bellbottom Blues, inspiration provided by Patti Boyd must have been overwhelming for this, Layla and maybe other of his songs of 70s, 80s. What a time to have all that music coming out fresh!

Ok, good to hear you have reset the amp yourself when you first bought it.  I would guess that should have taken care of that side of things but it wouldn't hurt to try it again if there is no loss of presets involved.

My background is actually in circuit design originally for the communications industry and later on for the seismic surveying industry which has surprising parallels with this one.  What is generally not believed by guitarists is that the design of our precious valve amplifiers is very basic and simple and undemanding.  It's a done deal nowadays and was back in the 50s when the engineers were genuine valve experts!  They attribute magical properties to designs laid down by people who were not engineers, (Jim Marshall was a drummer when he started and Leo Fender was trained in accountancy and was self taught in electronics), which do not stand up in the real world of "proper" design.  Unfortunately this leads to some very ropey circuits, (particularly pedals which are in general dire in design), sold for ridiculous amounts of money which only boosts their reputations.  "Like, it cost a fortune so it must be radically good dood" seems to apply.  And the attitudes are dictated by people who actually have little experience of design, particularly at the deeper levels.  Techs are wonderful guys doing a great job but they are not engineers trained to understand circuit and component action to a high level.  Some of the statements and claims made on YouTube and in forums are laughably incorrect.  Now in their praise, I have no idea what are the common problems of the range of say Fender amps.  A good tech with their experience can home in on what is wrong before he even opens the case and hence fix it in minutes.  I may have to probe around for a long time before I find the area.  Likewise, I can suggest where there is poor design in almost any of our precious amps, even the GM36/40D, in ways that they will just deny are real - because it works!  I know this attitude prevails because I've spent years arguing with people who clearly have no understanding of the designs they own or work on.  There is a good reason why electronics engineers don't take part in many discussions about circuitry with musicians and their techs.  "Never attempt to teach a pig to sing.  It will only waste your time and annoys the pig."  They plain won't listen to sound verifiable engineering fact if it contradicts what they want to be the truth!  Anyway, enough of my rambling.

Yes, Q9/10 do simply clamp the signal to ground ACwise.  Their source terminals (with the arrow) are connected to 7.5V DC via the resistor and to AC ground via the capacitor nearby.  Flipping their gate voltage significantly above or below 7.5V effectively turns them on and off as switches.  I made the same assumption as yourself, the Mute should be working as the amp works in normal mode.  That isn't strictly true but you have to start somewhere.  Once we know what is happening around TP204 we can look into each connected area more closely.  The mute control is the upper left section, the Gate On/Off action is the lower left section, the signal level sensing is the block on the Rearboard.  They are all each separate in action and can be chased through individually.  Hopefully the TP204 check will show which if any of these is going wrong.  My money is on the signal sensing section on the rearboard as the On/Off switching seems to work fine but, once that is switched on, the gate is on permanently which suggests that signal level sensing is going wrong.

Just a thought, which I maybe should have mentioned earlier.  You could check the connectors on the cable between the two PCBs.  A bad connection feeding the audio signal to the rearboard or the control signal back to the controller board could do this.  Just gently pull the plug out of the socket and refit it at both ends.  In particular you are looking at pins 1 and 3.  You could also look for the DC level on pin 3 at both ends with your meter.  You can usually touch the metal terminal pin inside the plug with it in place to get a reading but do that gently.  It should read the same at both ends with the Noise Gate switched On of course and it should then flip up and down as you play.  If your meter can do it, you could check out the AC levels at the top and wiper of the NG pot.  I'm sure you know how to check out the pot action this way.  I'd guess you will need a friend to do the guitar work while you probe.


EDIT: Incidentally, I think we can assume the actual gate circuitry around the 2 JFETs is working as the switch does turn the action on and off correctly. It seems to be only the control aspect once the gate is on which is faulty.

Thank you for the update. It sounds like you enjoy a varied career in valve technology, amongst other things, as you talk of the 50s era. Yes, I do remember some of the big names in amps, weren’t engineers but more enthusiasts, but also take a guess that engineers back then may not have taken guitar amplification as a serious career choice, as compared to today - in the early days of amplified music venues? Yes, I can imagine some of the early amp designs were trial and error endeavours. Before i bought the H&K, I looked at some of the more recent valve amps such as the British Victory range but the price would make a hard purchase with less functionality, but they did look like they are well designed (maybe borne out by few breakdowns). I had to laugh at the ‘teaching a pig to sing’ comment, i haven’t heard that before. Ramblings maybe, but good to hear what really went on in past times, so often the history is being written by people that were never there.

Cut-to-the-chase. I may have found the problem before even getting to TP204. I dismantled the amp again this morning with the intention of spending some time this week actually carrying out the testing. When I checked the NG potentiometer (thought I had previously). I found there was indeed 10K across the pins, but the wiper was open circuit. I tried an on-PCB repair, which didn’t work as not enough access without fear of damaging something. In the end I removed the faulty pot. I have ordered something that will fit in the space left behind and will hopefully ‘conveniently’ fasten in to the little tin shroud - possibly see this from the pictures. The replacement pot isn’t due till later this week, so I won’t know if this is a complete fix until then. It looks like this is a 10k log pot too. I think this also falls in line with your thought that the signal sensing section is likely the fault area.

Incidentally i did spend a few years working for what was-then (90s) STC in Greenwich, not long after I joined it was bought by the French Alcatel. No valves involved but wondered whether your communication role touched on the STC era, I think they were a big name in the business around the 70s and 80s. I don’t know whether the Greenwich site is still going, but it was all subsea fibre systems at the time.

I was looking through older posts on this forum and found one discussing getting a thicker low end, or a slightly less harsh sound from the Grandmeister. I believe this was something you had been investigating with various cab and speaker choices. Something I have been using on various amp set ups, including the GM40 in my short ownership so far is this: I put a TC Electronic IMPULSE IR Loader in the Fx loop. There are various different types out there. But it’s quite effective at modifying the dynamics of the audio output. I’m only mentioning this here as a side thing, i know it would be better placed as separate subject. My initial use of this pedal was to tame humbuckers that seemed to scream everything. Various IRs come loaded and also are freely available on the net and can be loaded to give a diverse use of the pedal. You tube suggestion… h-t-t-p-s://youtu.be/goFCMsMQxbg?si=brVwXhD_OtHde04h  . You may have already explored these.

Aha!!!  Sounds like you have got to the root of the problem, good catch there, well done.  The pictures are interesting, (and unusually focused and clear for a guitar forum site!   )  I'm not sure whether your dismantling of the pot itself will help as they are usually supplied as a solid complete structure.  The metal shroud mount is not a separate item that you can fit other pots into - as far as I am aware and I have been wrong before.   In any case you will certainly have to unsolder and remove the three pin base from the PCB.  The best way to do this is to carefully split it through the brown insulator board into three separate pins and remove them one at a time.  It is a pain to have to unsolder three spaced pins all at the same time to pull it out without eventually damaging the PCB pads.

A solder sucker to remove solder freeing the pin and to clean the hole up afterwards is a big help, or maybe just solder wick which is cheap.  I know these are not in everyone's kitchen draw so just do the best you can but be careful of overheating the PCB itself as the tracks and pads will lift off easily.  Also don't stress the PCB too much around the pot by twisting or bending it while you are working.  Despite what every guitarist "knows", surface mount fitting with lead free solder performed by machines is absolutely reliable, the statistical test data shows this clearly!  However, its one weakness is that physical force on a joint can crack it or even the component itself and this is almost impossible to see even with a magnifying glass.  Don't get worried about this, just take sensible care when you work.

If you really want to check this out while you wait for your replacement pot it is an easy thing to do assuming you have access to basic components.   You can simply temporarily replace the pot with either a 10k resistor across the outer pins with the wiper terminal shorted to the top end putting the pot on its full setting, or fit a pair of resistors, one from each end to the wiper centre, as if the pot was somewhere in its mid travel.  There is no need to be absolutely fussy about them adding up to 10k, anywhere in the vicinity will do for a test.  [EDIT: Thinking about this and checking out the schematic, don't even bother with the resistor for a quick test, you can just connect the pads for the top end and the wiper together temporarily. This will cause no stress to the circuit supplying the signal.] Anyway, I'm sure the pot will turn up fairly soon so you can get the amp up and running again.  The circuitry of the whole Noise Gate section has proven to be very reliable and the pot being damaged should not have kicked the damage to any other components.  This really sounds like the solution.

ha, yes the pictures were lovely and clear but i had to downscale them, as it wouldn't let me upload the images direct from the phone. The quality disappeared!

Unfortunately it doesn't look possible to unsolder the shroud due to a tiny pcb soldered on the underside, right in the 'wrong' place. This makes underside access virtually impossible, hence leaving the shroud in place. I have ordered a physically smaller pot that will mount within the current constraints. I don't want to risk damaging surrounding components unnecessarily. I hope my hi-fi picture will post this time ;-> (i tried a different tack)

Thanks for taking the time to work through the circuit. It's been an interesting experience, even if it turned out to be simpler than first thought.

Aaargh!  Yes, I wasn't aware of that pcb covering it and it is a pain.  It looks to me as though that is a major difference between the GM36 and the GM40D.  The GM36 info shows that pot is not lying under any other board and that that pcb is fitted to the microcontroller board unlike the GM40D, assuming it is the effects pcb.  That is useful information for the future.  These shrouds or "mounting brackets" are sold separately too and sized to fit a standard pot so removal might not be necessary.

However...  It may be possible to remove it from the upside if it were needed, (which I hope it isn't).  Even refitting only really requires that a rock solid joint is made on one or two of the legs, a tack on most of them as you would get when heating from above and pushing through would be enough to hold the shroud in place permanently.  If metal is tinned before heated insertion that would be enough to ensure the joint gets made with what is already there.  Notice that the bottom pads extend through and onto the top side where the pot is fitted.  With double sided PCBs and a grounded part like the pot bracket it makes no difference which side the joint is made.  If you have a reasonably good iron the heat would be enough to loosen it right through and you have no component to damage, only the PCB pads to be careful with.  Keep that approach in mind just as a backup to your original plan so if you can't fit a new pot into the existing bracket.

As long as the pot you have ordered will fit or can be made to fit in there that seems a great way to repair this to me and certainly the best choice.  I'm pretty confident that the bracket will be the right size but if it should be too small, if you are careful about filings and cover the job you could also file or grind out the hole to fit if necessary with a Dremel type mini-tool or jeweller's needle files.  I would use strips of paper and masking tape to cover the surrounding PCB, it will only be a mm or so.  Drilling it out may just be too stressful on the surrounding area but that depends on how much would need to be taken out.  Anyway, I'm pretty sure it will fit if it is a standard type.  As I said before, take care removing the old pot terminals.  Try to separate them into individual ones first.  Solder the new pot in from above, the terminals will be accessible.

Make sure to keep us up to date as to how you get on and if it fixes the issue.  Others may just benefit from your experience as they will already have done.



[EDIT: A tip here. If you need to clear a hole of solder from above because you can't access it from below, you can insert something like a needle of the correct size, as close to the hole size as is possible. A needle is stainless steel and will not solder. Heat the pad from above while you feed the needle in then wiggle it up and down quickly while the solder cools. It will be easy to remove the needle leaving the hole pretty much clear for the new component to be fitted. I've even managed to do this with a resistor leg if I keep it moving quickly in the hole as things cool. ]

Thanks, yes i will post again when I fit the replacement pot and had a chance to test the noise gate - and I'll include a hi-fi picture. I also had to repair the footswitch, which had been previously poorly repaired but that's another story. I hope you saw the bit about the IR loader a couple of posts back, as it may be helpful (when in the right forum subject of course), to modify output tones.

Hi Bardonbert & anyone else following
The new pot arrived today. This afternoon I managed to fit it in the very tight space available. Not something I'd like to do too often. As well as space being premium there's many components to avoid whilst implanting, such as surface mount bits fairly close. I have reassembled and now tested the amp and all sounds good! I was surprised how affective the NG is. Not having experienced it so far, i found it removed most unwanted noise. It seems to have made a good amp sound better.

You also asked me if I had worked on valve amps previously, well no, not really - only as a kid making radios and amplifiers . But I have done some design work using the Z80 and had too many years as a field engineer on all manner of equipment. But didn't expect I would be investing in a valve amp in the 2020s. There still hard to beat for a big sound (accepting venue amplification which is amazing these days).

I've said it several times, but I appreciate your time given in getting things sorted.
Regards, Andy

EDIT: I put the pot in on wires as the pcb space just didn't allow a direct connection, but it seemed to work out ok.

Hi Andy. Well done on a great fix there. I was genuinely worried about how you were going to get the original terminals out and the new pot in place. The idea of leaving them in place, fitting the pot sideways and just patching through to the terminals is genius. I take my hat off to your creativity.

The only suggestion I could offer is that, in cases like this, always cut back on the length of the wires as much as you can. One is generally a ground and won't matter but the other two may be susceptible to picking up or radiating noise signals. In this Noise Gate control we have line level signals into a simple AC/DC conversion circuit at about 10k impedance, so pickup shouldn't be a big issue but it shouldn't be ignored. With no low level circuitry nearby to pick up radiation (as far as we know), spraying out noise isn't too critical either. However, in some circuits it could lead to increased noise and even instability. You can see the high wattage resistors of the Power Soak nearby and they WILL have high current high voltage signals passing around them. That could lead to pick up in those wires. It won't come through into the guitar signal line of course and shouldn't make much difference if any in other terms but if you ever get back in there it may help. You could shorten them and, considering the method of mounting that pot back from the front face, maybe even take them tight over the front of the pot rather than round the back. As I said, it's not a huge deal but may just make sure of lower noise. (I used to be pedantic but I'm ok now. When I say "ok" I really mean... )

Great job and an easy fix to get to the bottom of eventually. As I said, that "good tech" with his experience of diagnostic and repair work would have gone for the pot straight away while we were aware of it but had fun tossing around the whole circuit action. It got repaired though. And I'm glad you like the action. The H&K Noise Gate has always been praised by myself. I have another couple of independent gate pedals knocking around but it easily betters them.

Back to playing then...