MXR M175 Digital Delay nonworking

[GibsonGM]

Hi Guys,

A friend gave me an old early 80s MXR "Digital Time Delay".  Think it's the M-175 (label worn off). It passes dry audio, but any setting of the wet just gives oscillations from police siren to motorboating (so hey, the speed range pushbutton controls work!  Speed pot seems to also).  I've checked out the PS, seems to be stable tho caps of unknown condition. 

And ideas where to start on this one? Easter-egging isn't doing much - diodes are ok, etc.   I'm wondering if it would be worth it to go thru and recap the whole damn thing, it's all I can think of that would be 'gone' as a matter of course.  Friend said it just started doing this one day, nothing happened to it.

Any ideas?  Can't find a schematic anywhere.  Thanks!

[GibsonGM]

Aaah, no dice.  Repurpose the rack and PS, then?

[11-90-an]

Probably start with swapping out either IC's (if socketed) and elctro caps

[Rob Strand]

Any ideas?  Can't find a schematic anywhere.  Thanks!

Maybe the feedback/regen  is passing full feedback even when the pot is on minimum?
You should be able to see that with an oscilloscope.

[11-90-an]

Wait... no schem? Trace time!

I'm guessing they're using those digital chips that are inaccessible to us, right?

No tracing probably gonna happen...

[Rob Strand]

Wait... no schem? Trace time!

I thought I actually had a schematic for one of those big MXR Delays.

Is it this one?

http://ampage.org/hammer/files/Mxr_digital_delay.pdf

Hmmm, maybe not, that's looking like the M-113.

[ElectricDruid]

It might be possible to debug it without tracing the whole thing. I mean, we know pretty much what's inside it by the era and the features, right? Ok, there are question marks on some bits, but I'd start with:

Input buffer
ADC (what type? this is a question mark)
Memory, probably a whole batch of DIPs like you used to have in your C64.
DAC (another question mark)
Output mixer

Meanwhile, the memory address is driven by a big counter, fed from a variable clock. The clock circuit is a high frequency VCO with an LFO attached to provide delay time modulation. There'll be some substantial amount of glue logic to deal with Read/Write cycles and chip select on the memory and latching data into and out of the ADC/DAC.

Finding the counter outputs or the address bus would let you check the address. Finding the Input buffer output, or ADC input would be good, as would finding the DAC output.

[GibsonGM]

Wait... no schem? Trace time!

I thought I actually had a schematic for one of those big MXR Delays.

Is it this one?

http://ampage.org/hammer/files/Mxr_digital_delay.pdf

Hmmm, maybe not, that's looking like the M-113.

No, it's sort of "1 generation newer", has the 3 speed select buttons.   It seems similar in design; that schematic is the closest I could come, too.

Shown here: https://www.worthpoint.com/worthopedia/vintage-mxr-digital-time-delay-model-153442920

Tom - Oww, LOL.  It's pretty complex.  I've traced the audio path as far as I can...things like the TL072s, 4558s appear to be ok, have good voltages.
It's that bank of what you call "C64 DIPS" that is a little intimidating (hey, it WAS a good computer for the time, ha ha!).

I will post a pic for grins when I have a moment.   Thanks for replying.   I've never monkeyed around inside of anything digital.   SOMETHING is happening, which leads me to think something is out of spec - dead/leaky cap, something like that.    I'd hate to toss it or tear the guts out if it can be fixed, it is an antique!

[ElectricDruid]

I'd love to see a picture of the insides, even if just for curiosity. Stuff from that era is amazing.

[GibsonGM]

Sorry for the delay, busy time of year here! 

Not great pics, but this is what I've got. Banks of what I assume are a 'bucket brigade'.  Something is working in there, but extremely 'out of sync'.  No idea where to look to get a handle on it.   I'd love to get this going, as you said Tom, amazing era!  I'm thinking '79 to '81 maybe; lots of electros/tantalums in there, I wonder if they could be involved.   

[Govmnt_Lacky]

Mike,

Can you post up some of the IC part numbers? I cannot see the pics right now. Wondering if they are the same chips that I have endless drawers full of here at my work.

[GibsonGM]

I think I found most of them, ha ha.  Given the sheer amount of them, and not knowing what function they play, I don't know how to troubleshoot them beyond basic voltages!

In the big bank upper left near filter caps we have:

SN74LS164N
DM74LS157N
AM2502PC
SN74LS32N
74LS393PC
SN74LS08N
SN74LS04N
DM74LS157N
TMS4164-20NL
DM74LS109AN
74LS283PC
74LS393N
74LS265P (? the 5 is a bit worn)
74LS109AN


Elsewhere,

CD4016BE
4558's
NE571N
LF356N
LM311
SN74LS624N

Not sure what use the numbers are, tho now I'll probably pull data sheets and go back thru and make sure the supply voltages are really correct and all that.

If I had a schema, I could probably find somewhere that mixes wet/dry and track backwards...oh well. 

[Govmnt_Lacky]

I think I found most of them, ha ha.  Given the sheer amount of them, and not knowing what function they play, I don't know how to troubleshoot them beyond basic voltages!

In the big bank upper left near filter caps we have:

SN74LS164N
DM74LS157N
AM2502PC
SN74LS32N
74LS393PC
SN74LS08N
SN74LS04N
DM74LS157N
TMS4164-20NL
DM74LS109AN
74LS283PC
74LS393N
74LS265P (? the 5 is a bit worn)
74LS109AN


Elsewhere,

CD4016BE
4558's
NE571N
LF356N
LM311
SN74LS624N

Not sure what use the numbers are, tho now I'll probably pull data sheets and go back thru and make sure the supply voltages are really correct and all that.

If I had a schema, I could probably find somewhere that mixes wet/dry and track backwards...oh well.

Other than the obvious, I can tell you that the TMS4164-20NL are the DRAM chips. Same storage chips used in the Digitech XP series pedals although, yours are the 200nS version.

[GibsonGM]

I'd like to locate the clock and scope it.  With the garble the thing does on input, it wouldn't surprise me if it's way out or something.  I have a 60MHz CRO that maybe I could do something with.  Never looked inside a delay before!   

EDIT:
I see that SN74LS624N is a VCO - maybe a good place to start?

[PRR]

> Banks of what I assume are a 'bucket brigade'.

No. BBD are garbage today; back then they were worse garbage. This is (by context and contents) onna dem fancy RAM based delays.

AM2502PC  - 12-bit 8 bit(!) successive approximation register (the gizmo for an ADC)
https://www.digchip.com/datasheets/download_datasheet.php?id=1502945&part-number=AM2502PC

TMS4164-20NL - 64k bit DRAM memory
http://pdf-html.ic37.com/pdf_file_B/20200531/pdf_pdf/PdfOld/icpdf_datasheet_5/TMS41_datasheet_51610/843849/TMS41_datasheet.pdf
Just ONE of those? Jez that's ancient.

All the 74** are small or medium glue logic. We see a couple opamps, a fast comparator (clock?), a switch, the compander.

[Scruffie]

AM2502PC  - 12-bit(!) successive approximation register (the gizmo for an ADC)

8-bit 

I'd love to see a picture of the insides, even if just for curiosity. Stuff from that era is amazing.

I can get you photos of an EHX 2 second digital delay, A2504 so that is 12-bit.

12 x 16k RAM 

[ElectricDruid]

N74LS164N
DM74LS157N
AM2502PC
SN74LS32N
74LS393PC
SN74LS08N
SN74LS04N
DM74LS157N
TMS4164-20NL
DM74LS109AN
74LS283PC
74LS393N
74LS265P (? the 5 is a bit worn)
74LS109AN


Elsewhere,

CD4016BE
4558's
NE571N
LF356N
LM311
SN74LS624N

I think this gets you quite a bit further. I agree with Gibson that the 74LS624 is probably the clock, since that IC is a VCO. The NE571 is the obvious compander to deal with the 12-bit depth and try and wring a bit more S/N ratio out of this thing. CD4016 switches are probably in/out effect switching, so no big deal there either. The '4164 is the RAM, so how many of those there are determines the maximum delay time. Since we can find a pin-out for that chip, we can check the address lines to see if the memory is being addressed correctly (I'd expect to see a simple incrementing count, but other options are possible).

Tom

Tom

[willienillie]

Tom

Tom

[ElectricDruid]

Tom

Tom

Is there an echo in here?

here?

here?

here?

[GibsonGM]

The '4164 is the RAM, so how many of those there are determines the maximum delay time. Since we can find a pin-out for that chip, we can check the address lines to see if the memory is being addressed correctly (I'd expect to see a simple incrementing count, but other options are possible).

Ok, going back thru the thing with an audio probe...there's definitely an 'audio area'...there's gotta be some sort of stuff in there that tells those fancy register chips what to do.   I can find clean audio, and also clock sounds.  From distorted thumping to high pitched oscillation depending on knob settings. It might in fact be the '624...gently messing with the trims by it increases/decreases the clock pulses (I did put them back where I found them ha ha). 

To check the '4164, do you think an LED w/resistor to ground would work? Not sure how my CRO would work in this application.

I think if I can find where a delayed signal comes back from the computer stuff and gets mixed, I'll be onto something.  Or will know the digital crap is dead.  Enough for one evening of hair pulling tho!

I did find a TL072 in the area I think digital and audio meet, seemed to be running warm to me.

1: -13.4
2: -13.4
3:  0
4: -14.8
5: -3.9
6 -3.9
7: -3.9
8: 15

Don't know what to make of that, +/- 15v supply.  Not a lot going on on pin 3...

  Thanks for the insights!