midiverb II fettling

Alesis Midiverb II front panel

I bought a Midiverb II a while back, and when it arrived, it was dead.

It was advertised as working so I should’ve sent it back, but it was thirty quid and I thought I’m a super–DIY-electronics-pillock who totally knows what he’s doing – I’ll have this fixed before bedtime.

Four years later, and the Midiverb had fallen into a tiny gap between the guitar amp I got when I was fifteen and an Ikea shelving unit, still waiting for me to resurrect it.

Here it is in its dead state:

Midiverb dead state - all LEDs on, nothing on the numeric LED display

The dust! You’d think I’d have given it a bit of a wipe before taking a photo.

And inside – the board takes up about three-quarters of the rack.

Overview of the Midiverb II circuit board

The big square chip with the sticker on it is the magic Alesis reverb chip, and the chip to the left is the DAC.

The ICs above are for memory, and on the left-hand side there are some old-fashioned silver-topped TL084 quad op-amps, some of which might be used for output.

I’m guessing because there’s no schematic available, which makes it more exciting/fruitless, although the Midiverb III service manual might be similar in some ways. Anyway, for the Midiverb II this page on wolzow.com gives a far better impression of the insides than I ever could.

Here’s a random picture of the power regulation bit – the Alesis external PSU is a 9V AC adaptor, which gets rectified into +12V, -12V and +5V.

Closeup of Midiverb II power supply

I optimistically replaced the capacitors in this section a while back, plugged it in, and when it still didn’t work I thought, oh well, that’s me fresh out of ideas.

Looking at the date codes on the ICs this seems like it was probably a good thing to do anyway.

I rechecked the +5V power supply after reading this post on anlage-e – seems like some of the decoupling capacitors go bad after a while – no shorts, seemed fine.

Recently I plugged it in again and boringly got the same result, all LEDs on, no LED display. Maybe something to do with the microprocessor? I wondered if the reset line on the 80C31 wasn’t going high – measured it on the oscilloscope, it seemed to be blipping high when plugging the MVII in.

Alright, next thing – clock on the 80C31, it should be ticking along at 12MHz, what have we got?

FLAT LINE. (I should’ve taken a photo, although it wouldn’t have been very interesting). At this point I dragged the probe slightly, LEDs flickered, and the front panel 7-segment display burst into life.

Tears of joy.

I worked out that if I flicked the resonator next to the 80C31 hard enough with my finger it would jump-start the clock – restarting the Midiverb would produce the reset signal and get it going.

This wasn’t much of a fix, so I replaced the resonator with a brand new 12MHz crystal.

And it’s been working happily ever since. Here’s a picture of it under my Quadraverb, on the infinitely large patch 29.

Here’s an edit of the second thing I did with it, which might be an actual track eventually. Soz about the quality, it was onto cassette tape…

The main high-ish sound is the System 100 102 expander through patch 29 on the Midiverb, the bonky bass is the 100M clone, drums are a 606, and the pad-ish synth at the end is the Poly-61M. All running off the MC-4.

Anyway, what have I learned from this? The main thing is that digital things aren’t totally unfixable. Also, the lack of a service manual isn’t the end of the world.

Tagged under , , ,

31 comments

  1. 30th June 2020Peter Dodge says:

    I have the same problem many I’ll try what you did

  2. 1st July 2020ua726 says:

    I’ve not seen this mentioned anywhere else, so I don’t know how prevalent it is. Worth checking the power supply as well.

    FWIW mine is still working happily six months later.

  3. 16th November 2020geeky says:

    I had the same problem a few months ago with a unit bought on ebay.
    The power supply was working fine, found a few shorted caps (and then replaced nearly all of them to be on the safe side).
    Noticed the leds seemed to be flickering and then checked the clock and found the resonator not working properly

  4. 17th November 2020ua726 says:

    Nice, glad you got it sorted out.

  5. 23rd February 2021Chris says:

    Hi,
    Any specific specs you have to look for when you replace that resonator, apart from frequency?

  6. 24th February 2021ua726 says:

    Honestly don’t know – I think I used this one, which in case that link goes dead in the future is a 12MHz crystal with a load capacitance of 30pF, product code CX016.

  7. 18th August 2021Jaap says:

    My Midiverb 2 shows the same symptoms (random leds) but it gets a clock from the resonator.
    The clock signal oscillates between 1.5V and 3V.

    If I read the SC80c31 datasheet correctly, the signal should be between 0.85V (0.2VCC-0.15) 3.6V (0.7VCC+0.1);

    Do you remember what the clock signal voltages were when you put in the crystal?

  8. 19th August 2021ua726 says:

    Just checked through my notes and my photos, and I can’t find anything. Bit surprising because I usually include gratuitous oscilloscope shots with blogposts whenever I can.

    I’ve had a look at some datasheets for the 80c31, and they all quote similar figures.

    Like you suggest, it sounds as if the resonator no longer has enough oomph to clock the processor. Given that the cost and effort involved is minimal, I’d be tempted to replace the resonator with a new crystal and then see where that gets you.

  9. 19th August 2021Jaap says:

    Thanks for checking, I really appreciate that!

    I replaced the resonator with a crystal, hoping that would help, but the scope readings were the same. Then I replaced U26 (74HC573 latch) and that did the trick: the MV is up and running now!

    I think I’ll replace the caps now to be on the safe side.

  10. 19th August 2021ua726 says:

    Nice! Glad it’s working now. One to bear in mind. Interested to know how you tracked that one down as well, were you seeing stuck inputs/outputs on the latch?

  11. 20th August 2021Jaap says:

    The outputs were stuck in a “random” state (different after each reboot). The inputs behaved like what I was expecting (changing state). I had a replacement 573 and thought putting that in wouldn’t hurt : )

  12. 20th August 2021ua726 says:

    Excellent, thanks for that Jaap.

  13. 31st October 2021Golgoth says:

    Hi there,

    Great article!

    I just picked up a Midiverb II unit and it is unfortunately faulty, very noisy audio without any bass. Patches are switching correctly but are not affecting the audio too much. The input pot seems to be extremely noisy compared to the other two pots…

    Screen and leds are working as intented tho so I hope that this is a power issue..

    I don’t have another power supply to try unfortunately, I bought the unit with it’s origin PSU.

    I have a DMM but not no schematics, could you point out where to take the measurements on the pictures you’ve taken?

    That would be of great help!

    All the best,
    g

  14. 31st October 2021ua726 says:

    Looking at the picture of the circuit board from the front of the unit (as above), the left-most pin on each of the large transistors on the bottom right (LM340T5, LM7912CT, and LM340T12) should show +5V, -12V, and +12V.

    Now I look at it again, there’s a troubleshooting guide in the Midiverb III service manual which might help a bit, although the designators for the chips are different. It mentions that distorted audio could be down to dodgy power rails, bad opamps, or a bad analogue switch. It mentions a faulty ASIC being a potential problem, but hopefully it’s not that because it sounds like it’d be time to chuck it in the bin if that’s the case.

  15. 31st October 2021Golgoth says:

    Thank you very much for the quick answer!

    I am getting all the correct values on the transistors, which leads me closer to digital issues 🙁

    I have hope that the potentiometers could be at fault here tho, they are very very scratchy. In some positions I get white noise and in some other it sounds like a 1/4″ jack is half plugged. The gangs look pretty dirty and corroded, I will buy some contact cleaner and try to clean them as best as I can and report back!

  16. 1st November 2021Golgoth says:

    Contact cleaner half fixed the issue, no more noisy audio nor contact issue when turning knobs. Audio is clean!

    Except the dry signal still has no bass 🙁 Wet signal with mix at 100% sounds exactly like the input source tho so we can isolate the issue to the input audio path or the input potentiometer itself.

    When I power off/on again sometimes the bass is back on the dry signal and everything works as it should.. But it’s gone 80% of the time if I turn the unit off and on again..

    Every other functionnality is working as expected, patches, leds, screen etc..

    Any idea about where to look next? Are the U1,U2,U3 and U4 opamps dedicated to audio inputs/outputs?

    Thanks again,
    G

  17. 1st November 2021Golgoth says:

    w̶e̶ ̶c̶a̶n̶ ̶i̶s̶o̶l̶a̶t̶e̶ ̶t̶h̶e̶ ̶i̶s̶s̶u̶e̶ ̶t̶o̶ ̶t̶h̶e̶ ̶i̶n̶p̶u̶t̶ ̶a̶u̶d̶i̶o̶ ̶p̶a̶t̶h̶ ̶o̶r̶ ̶t̶h̶e̶ ̶i̶n̶p̶u̶t̶ ̶p̶o̶t̶e̶n̶t̶i̶o̶m̶e̶t̶e̶r̶ ̶i̶t̶s̶e̶l̶f̶ obviously went ahead of myself there and meant that we can isolate the issue to the dry audio path, need to figure where the input signal is cloned and what happens to the dry signal until the mix section!

  18. 1st November 2021Golgoth says:

    Contact cleaner half fixed the issue, but the dry signal still has no bass at all. 100% wet signal sounds exactly like the input tho… I believe we can isolate the issue to the dry audio path, need to figure where the input signal is cloned and what happens to the dry signal until the mix section!

  19. 1st November 2021ua726 says:

    No bass? weird.

    Like I say, maybe have a look at the Midiverb III service manual, I suspect the analogue bit will be fairly similar. You might have to use the continuity setting on your multimeter to check if it’s the same as the M3. Then, see if you can follow the audio through from the wiper of the input pot through the opamps.

    I tend to use an oscilloscope but I’ve also carefully probed the audio amplified through an old mixer and some headphones in the past, I’ve also seen other people recommending just trying carefully with a pair of old headphones (with the sleeve of the jack clipped to ground, and perhaps the tip as the probe, connected through an electrolytic capacitor to remove any DC voltage.

  20. 2nd November 2021Golgoth says:

    Hi,

    Sorry for the multiple messages sent yesterday, they disappeared after posting so I posted again, before they were up again!

    Fortunately the bass issue was just me misusing a new piece of gear haha! I was sending a mono signal to the Left input, which resulted in a phase problem. The Right channel seems to be meant for mono sources!

    Thanks again for your help,
    G

  21. 5th December 2021Peter says:

    I picked up a MIDIVerb II from Reverb a couple weeks ago shipped without a 9VAC PS and found a nice 9VAC PS on Amazon that shipped from within the USA.

    Upon ignition, the OVLD LED is illuminated constantly. Ran signal through it and very over-driven and distorted dry or wet, but the effects do work. I checked the 5, 12 and -12VDC voltages and all within around 200mVDC. Checked ripple on the output of each regulator with the scope and I’m seeing maybe 10-15 mVAC, probably more noise then anything even with bandwidth limit on.

    It appears a schematic/service manual does not exist online since it must contain some secret launch codes or the whereabouts of Jimmy Hoffa, so I’m going to start the shotgun approach and start replacing parts in the audio chain. I ordered all the electrolytic caps and a bunch of TL074 op amps from Mouser. From the datasheet for the original National LF347N op amp the TI TL074 will work and is a better op-amp IMHO.

    I’ll further reverse engineer the audio section once I get all the caps replaced and report back. Parts will be here early next week, ya gotta whole lotta love Mouser UPS 2nd day air shipping for $7.99 as default.

    Over and out.

    P

  22. 19th December 2021Robert Staurovsky says:

    Hello, I have a Midiverb II, the overload light stays on in certain program modes but not all. Any idea would help, ALL Caps have been replaced, U1, U2, U3 and U4 replaced and U17, U18, U19 are replaced. Originally the Overload Light stayed on all the time, any ideas ??? Thanks, Bob

  23. 19th December 2021ua726 says:

    Hmm. Does it work (pass dry audio + correct effected sound) for those programs that don’t have the overload light on? Which programs show the overload light? Is it always the same programs? Or is it just one type of effect that doesn’t work? Is there any pattern in the numbering of the broken patterns?

    The MVIII schematic might provide some hints, although like I say, can’t guarantee it’ll be the same. Wonder if it could be a logic chip that has a dead input or output on it.

  24. 26th November 2023Thorsten Rösch says:

    Hi everyone,

    I have a Midiverb II which I managed to get back to running by replacing a shortened capacitor. Now it starts fairly normal, but after approximately 30 seconds it shows „1“ on the display and freezes up. I replaced the mentioned 12MHz oscillator to no effect. Any ideas as where to look next?

    Appreciate your help!

    Thorsten

  25. 27th November 2023ua726 says:

    Hello Thorsten – haven’t seen this error (…in my sample of one), this one sounds a lot trickier to diagnose. I’d suggest confirming general troubleshooting: check the power supplies, both for stability and for the correct voltages. Do any chips get hot, apart from the regulators, perhaps?

    Can you switch patches, does the error still occur for a different patch? Does it pass audio and the correct effect (as far as you can tell) up to the crash?

    I’m presuming there’s nothing plugged into the MIDI input?

    30 seconds sounds like a long time for it to be a memory error; can’t help but thinking it would have already cycled over that a few times by that point.

    Does the reset pin on the 8031 stay high? Does the 8031 stop sending signals after the the crash?

    It sounds like fun anyway, good luck

  26. 28th November 2023Thorsten says:

    Thanks for your reply!

    I will start with further investigations next week.
    Could you kindly let me know where the reset circuit is located.
    I only have a MV 3 service manual which has a modified circuit.

    Thanks,

    Thorsten

  27. 29th November 2023ua726 says:

    Both the MV2 and 3 use a 80C31, so the reset pin (active low) would be pin 9.

  28. 13th July 2024Matrix12x says:

    My Midiverb 2 shows the similar symptoms (all 4 leds) but with some life on the display after the reset pulse happens (after I fixed the reset ckt). The display shows (dimly) a “11”” and if I hit any keys the display changes to “__” and if I hit 1 it changes to”11″ again. If I remove the eprom or the 8031 I get no display at all and no reaction from the display when I hit keys. So I’m assuming the 8031 and the eprom are ok.

    The 4.7uf cap for the reset was bad, I replaced that. Now when I power on the Reset pin on the 8031 starts high and goes low. (seems correct)
    I replaced all of the PSU caps and regulators and the voltages are correct.

    I replaced the resonator with a crystal (with caps on both legs like in the crystal datasheet). The 12MHz is there and looks good. The 8031 seems to be running. When I check the ALE pin on the 8031 it had a narrow pulse at about 2MHz, which seems right.
    I replaced U26 (74HC573 latch) and no differences.

    I have not tried putting audio thru the MV2.

    Any ides of where to check next?
    I wonder if U24 and U25 (74HCT574) are bad. U24 seems to control the display U25 controls the 4 LEDs.

  29. 15th July 2024ua726 says:

    Yeah, tricky – me fixing mine was more through luck than judgement, I’m no expert at fixing these particularly.

    Presuming it’s similar to the Midiverb 3 I’d be looking at the D0-D7 pins on the 8031 to see if anything is half-way in-between 0 and 5V or stuck. I imagine there’s a chance that one of the latches has developed a duff input and could be holding the line. Hard to tell which one it would be though. Also worth checking if the latches are receiving a clock.

    Personally I’d be tempted to put some audio through it anyway, just to see what happens, see if the patch changes audibly if you type in another number.

    Hope you fix it, good luck

  30. 19th July 2024Matrix12x says:

    Just wanted to provide an update, I replaced U24 and U25 (74HCT574) and everything works now.

  31. 21st July 2024ua726 says:

    hoorah, good news!

Write a comment: