I have a Marshall JCM 900 100W tube amp head from 1990 that likely has an issue in its Reverb Drive solid state circuit.
JCM 900 Preamp Schematic from 1990
/\ The schematic of the JCM 900 preamp especially the "reverb drive" section referred to here.
Troubleshooting Progress
While the amp is on and feeding it a 1khz sine wave, an oscope successfully sees the sine wave up to and exiting the V2a preamp tube and can also be seen after the reverb section of the circuit following the dry path.
However, the sine signal is lost below the R23 resistor at the start of the reverb circuit (totally fine above it). Above it is a 400mv peak/peak sine wave. I would expect to see at least a 200mv sine wave below R23.
- While the amp is off/caps discharged, in circuit continuity and ohm checks using a DMM on every component in the reverb circuit pass/expected (cannot test the caps but none beep with continuity which is assumed good).
- While the amp is on, measuring DC below R23, the DMM is registering about -300mv.
- While the amp is on, measuring DC right of R28, left of R26, above C17, left of R27, right of R26 all measure roughly -14.25V.
This leads me to assume that "maybe" either R23 or C16 have issues and should be replaced, but since I am not a guitar amp tech guru, before I start desoldering components and adding even more variables/fail points to this issue, I am seeking some understanding and suggestions.
Questions
❔ Should I only desolder components, test them out of circuit and then get them back in as the only option to pursue?
❔ Can/should I see if R23 is the issue by temporarily jumpering another resistor of same/lower value on top of it in parallel?
❔ Can/should I see if C16 is the issue by temporarily jumpering another capacitor of same/lower value on top of it in parallel?
❔ Can/should I do either of the above by snipping one or both ends of the resistor or capacitor and then jumpering or "top soldering" the left over leads with new components and avoid temporarily having to disassemble the preamp board to under solder them?
❔ Is there a way to replicate this issue with a simulation? See my attempt at an LTSpice rendition of this part of the circuit below.
❔ Is it correct that if the reverb drive circuit was working I should be able to see a > 50mv sine wave through its components on a scope but doing so with DC coupling on the scope since the op amps have 12-15VDC on all of those lines?
❔ What affect does having the reverb tank being connected or not play into any values I would read?
❔ Would acquiring a newer/digital scope provide such a difference that I should seek that first?
Additional Info
Above is an overview of the LTSpice I attempted to create to provide a simulation of this circuit.
The reverb tank is not an exact replica as it has a much higher input impedance, the reverb pot is fixed R, and I took liberty excluding the channel switching IC7 M5201 and replacing that with a single TL701 (see the actual real link schematic above).
With this simulation, I achieve the expected results below and above R23 and above R20 in the dry area.
While the simulation does see -12V in the op amp areas of the drive circuit, it produces a very small and very distorted ac result which leads me to assume the simulation isn't great.
Thanks to Adrian Immler's post "spice reverb simulation?? This is now possible!" for providing a LTspice of a similar reverb tank.
For extensive additional background leading up to this point, please see this post:
Marshall JCM 900 Reverb Issues
I am aware that it is not ideal and borderline/crosses forum rules about not posting two threads about the same topic; however, when I posted to the valve/tube area I did so since a similar reverb thread was in the solid state area and I have a tube amp and other research implied that one of the more common faults with reverb were related to pre amp tubes, so made the assumption that group was appropriate and now see that forum is more specific to hi fi tube amps.
I believe I found the issue to the reverb drive circuit (see that reply in the next reply thread).
Since the replies by others here have been minimal on this topic, I am going to continue to lean into the "Y" of DIY and reply to my own questions to assist with some poor sob like myself months/years from now having similar issues and troubleshooting steps.
Please see this thread above and Marshall JCM 900 Reverb Issues that started this in Tubes / Valves for more context to the original questions answered to the best of my current insight below.
-------------------------------------------------
❔ I should be able to put the scope between R20 and R22 and see an ac only (since DC decoupled) on the scope and should see a clean sine wave right?
Yes
❔ I should also be able to see a clean sine wave on any of these points too right?
For the most part, "yes" you should see a sine wave on all of those connection points.
❔ Does the reverb tank have to be connected to test the "Reverb Drive" portion of the circuit?
No. I found Forgotten Gear Restorations - Marshall JCM 900 (4100) || Testing Reverb Without Sound which showed success not having the tank connected and after I replaced the TL071 I was able to do the same test.
❔ Please confirm that all three op amps are preparing current and that the two MC1458 aren't setup to do a push pull or inversion thing where one is handling one side of the polarity. I am basically just asking for ways to prove what works and doesn't and not necessarily a reason why this circuit is a bad design as outlined in other posts here on diyAudio.
I don't have a definitive answer, but I am going with the TL071 is a non-inverting buffer with a gain of 20+ and the two MC1458 are voltage followers and are there to set the impedance and higher current needed for the reverb tank.
3) I have read through many pages of the sticky on safety here and elsewhere. However... <snip>
You have to drain all caps. Just because you drain one doesn't mean the others later won't bleed into them.
❔ Should I only desolder components, test them out of circuit and then get them back in as the only option to pursue?
Short answer to this is yes, especially for a through hole PCB that is not multilayer. It actually was not as difficult/scary to pull things off the board as I first thought. Be safe and be brave.
While one could acquire an ESR tester, that has a series of pros and cons. I am unaware of any way to test an IC in circuit.
For guitar amplifiers, in many cases one can skip components after the problem areas and inject an audio waveform later in the circuit to assist in narrowing things down. For example, Stuart ukguitarampguy - How to Fix Marshall Valvestate No Sound Problem does this.
I felt I could not attempt this because of how the reverb drive circuit was designed a lot of things were all interconnected.
❔ Can/should I see if R23 is the issue by temporarily jumpering another resistor of same/lower value on top of it in parallel?
Because the issue ended up being the TL071 opamp, this would not have worked, but attaching a couple other resistors in series and then to ground on paper would have allowed a scope to see the signal prior to that opamp.
Typically you want to replace components and not create a whole new design especially if you are not "back of hand" knowledge of what the circuit is doing (be safe and don't burn up an amp).
❔ Can/should I see if C16 is the issue by temporarily jumpering another capacitor of same/lower value on top of it in parallel?
See above but replace resistor with cap for the most part.
Typically you want to replace components and not create a whole new design.
❔ Can/should I do either of the above by snipping one or both ends of the resistor or capacitor and then jumpering or "top soldering" the left over leads with new components and avoid temporarily having to disassemble the preamp board to under solder them?
Although this has some places that this could work, this question was a clear indication of my concern and hesitancy of touching the PCB from the under side. Even though I discharged the 2 filter one one large cap on the preamp PCB any time I got near it, "you never know".
Moreover, moving a 30+ year old PCB could add stress to components and solder joints and I was concerned I would introduce even more problems. I used care and was easily able to pull, then test, then replace the components back on the PCB.
❔ Is there a way to replicate this issue with a simulation? See my attempt at an LTSpice rendition of this part of the circuit below.
I am aware that in the image of the LTSpice I mocked up the R20 resistor is not going where it should, but I wasn't able to get a LTSpice model to work as good as I would like. There was some hope with it, but too many gotchas. And since I couldn't get it to work perfect any attempt to sabotage a component to mimic what I was seeing in the actual amp went nowhere. I still have a desire to LTSpice this entire amp to assist in my understanding of how preamp and power amps work.
❔ Is it correct that if the reverb drive circuit was working I should be able to see a > 50mv sine wave through its components on a scope but doing so with DC coupling on the scope since the op amps have 12-15VDC on all of those lines?
My best answer to this is that if a reading point had DC on it, then yes you can set the scope to filter out the DC, but on this reverb drive circuit the output should only have the A/C and if it has DC like it did when there was an issue with TL071 then maybe there is an issue with the circuit.
❔ What affect does having the reverb tank being connected or not play into any values I would read?
Because I was able to do readings after I replaced the TL071 I don't need this answer, but I did confirm the tank doesn't need to be connected to do the readings (Forgotten Gear Restorations - Marshall JCM 900 (4100) || Testing Reverb Without Sound )
❔ Would acquiring a newer/digital scope provide such a difference that I should seek that first?
My scope hasn't been the cause of the issues, but it needs serviced as the triggering mechanism basically doesn't work and I end up readings like I am looking at "The Matrix" sideways. I have acquired a FNIRSI 2C53P handheld model to replace the analog one I have been using.
Since the replies by others here have been minimal on this topic, I am going to continue to lean into the "Y" of DIY and reply to my own questions to assist with some poor sob like myself months/years from now having similar issues and troubleshooting steps.
Please see this thread above and Marshall JCM 900 Reverb Issues that started this in Tubes / Valves for more context to the original questions answered to the best of my current insight below.
-------------------------------------------------
❔ I should be able to put the scope between R20 and R22 and see an ac only (since DC decoupled) on the scope and should see a clean sine wave right?
Yes
❔ I should also be able to see a clean sine wave on any of these points too right?
For the most part, "yes" you should see a sine wave on all of those connection points.
❔ Does the reverb tank have to be connected to test the "Reverb Drive" portion of the circuit?
No. I found Forgotten Gear Restorations - Marshall JCM 900 (4100) || Testing Reverb Without Sound which showed success not having the tank connected and after I replaced the TL071 I was able to do the same test.
❔ Please confirm that all three op amps are preparing current and that the two MC1458 aren't setup to do a push pull or inversion thing where one is handling one side of the polarity. I am basically just asking for ways to prove what works and doesn't and not necessarily a reason why this circuit is a bad design as outlined in other posts here on diyAudio.
I don't have a definitive answer, but I am going with the TL071 is a non-inverting buffer with a gain of 20+ and the two MC1458 are voltage followers and are there to set the impedance and higher current needed for the reverb tank.
3) I have read through many pages of the sticky on safety here and elsewhere. However... <snip>
You have to drain all caps. Just because you drain one doesn't mean the others later won't bleed into them.
❔ Should I only desolder components, test them out of circuit and then get them back in as the only option to pursue?
Short answer to this is yes, especially for a through hole PCB that is not multilayer. It actually was not as difficult/scary to pull things off the board as I first thought. Be safe and be brave.
While one could acquire an ESR tester, that has a series of pros and cons. I am unaware of any way to test an IC in circuit.
For guitar amplifiers, in many cases one can skip components after the problem areas and inject an audio waveform later in the circuit to assist in narrowing things down. For example, Stuart ukguitarampguy - How to Fix Marshall Valvestate No Sound Problem does this.
I felt I could not attempt this because of how the reverb drive circuit was designed a lot of things were all interconnected.
❔ Can/should I see if R23 is the issue by temporarily jumpering another resistor of same/lower value on top of it in parallel?
Because the issue ended up being the TL071 opamp, this would not have worked, but attaching a couple other resistors in series and then to ground on paper would have allowed a scope to see the signal prior to that opamp.
Typically you want to replace components and not create a whole new design especially if you are not "back of hand" knowledge of what the circuit is doing (be safe and don't burn up an amp).
❔ Can/should I see if C16 is the issue by temporarily jumpering another capacitor of same/lower value on top of it in parallel?
See above but replace resistor with cap for the most part.
Typically you want to replace components and not create a whole new design.
❔ Can/should I do either of the above by snipping one or both ends of the resistor or capacitor and then jumpering or "top soldering" the left over leads with new components and avoid temporarily having to disassemble the preamp board to under solder them?
Although this has some places that this could work, this question was a clear indication of my concern and hesitancy of touching the PCB from the under side. Even though I discharged the 2 filter one one large cap on the preamp PCB any time I got near it, "you never know".
Moreover, moving a 30+ year old PCB could add stress to components and solder joints and I was concerned I would introduce even more problems. I used care and was easily able to pull, then test, then replace the components back on the PCB.
❔ Is there a way to replicate this issue with a simulation? See my attempt at an LTSpice rendition of this part of the circuit below.
I am aware that in the image of the LTSpice I mocked up the R20 resistor is not going where it should, but I wasn't able to get a LTSpice model to work as good as I would like. There was some hope with it, but too many gotchas. And since I couldn't get it to work perfect any attempt to sabotage a component to mimic what I was seeing in the actual amp went nowhere. I still have a desire to LTSpice this entire amp to assist in my understanding of how preamp and power amps work.
❔ Is it correct that if the reverb drive circuit was working I should be able to see a > 50mv sine wave through its components on a scope but doing so with DC coupling on the scope since the op amps have 12-15VDC on all of those lines?
My best answer to this is that if a reading point had DC on it, then yes you can set the scope to filter out the DC, but on this reverb drive circuit the output should only have the A/C and if it has DC like it did when there was an issue with TL071 then maybe there is an issue with the circuit.
❔ What affect does having the reverb tank being connected or not play into any values I would read?
Because I was able to do readings after I replaced the TL071 I don't need this answer, but I did confirm the tank doesn't need to be connected to do the readings (Forgotten Gear Restorations - Marshall JCM 900 (4100) || Testing Reverb Without Sound )
❔ Would acquiring a newer/digital scope provide such a difference that I should seek that first?
My scope hasn't been the cause of the issues, but it needs serviced as the triggering mechanism basically doesn't work and I end up readings like I am looking at "The Matrix" sideways. I have acquired a FNIRSI 2C53P handheld model to replace the analog one I have been using.
I am pretty sure I found a solution to the reverb drive issue on my JCM 900 4100.
Short Story
- The TL071 in the reverb drive circuit was malfunctioning and after replacing it I have full reverb flying through the JCM 900.
But as this cases points out, it can be other things too.
Long Story
- I acquired a new TL071 and made a couple simple voltage follower and non-inverting opamp circuits with it and a couple 9V batteries on a IC testing breadboard.
- I pulled the TL071 from the PCB and stuck it on the same breadboard and it did not work. It did not respond to the gain resistors and although it did show a waveform it was 1/10 the input.
- I pulled the three caps surrounding the reverb drive circuit and all of them when tested out of circuit within range (the electrolytic at the very bare minimum range as expected for a 30+ year old cap).
- I replaced the TL071 and the MC1458 even though the original one was working with new chips I snagged from digikey.After resoldering everything into place, I turned on the amp. Nothing smoked. I then took a sine wave reading off the black RCA which feeds the reverb drive and got a large clean sine wave.
I then hooked things back up and played a real guitar through the amp and woot woot gots some reverb back.
The test was under 5 mins so I am going to keep burning things in and see how things go, but sounds like I am good to go now.
Agreed dubadub and have some at the ready for that being my next project. When I made the digikey order for what turned out to be the replacement opamp, I threw in a bunch of caps from other areas of the PCB that need replace due to their 30+ year old age. But in my rush to get the order out quick so it would ship to me before last weekend, I didn't measure some and the one in question above I had to replace but I ordered one too big (rookie mistake). I also couldn't find the 50+50 big filter caps off the power transformer so I will likely get those F&T brand from amprepairparts.com
Caps get smaller as the tech progresses; selecting the right caps for an old amp is tricky. Don't hesitate to ask the folks here or over at Parts for advice.
Beware the Mouser Markup. Compare those prices against Antique Electronic Supply before you pull the trigger next time. TubeDepot has good prices on caps too. F&T are top of the line, but Mouser and Digi don't carry them so you'll get hit for both shipping charges 🫤
Beware the Mouser Markup. Compare those prices against Antique Electronic Supply before you pull the trigger next time. TubeDepot has good prices on caps too. F&T are top of the line, but Mouser and Digi don't carry them so you'll get hit for both shipping charges 🫤