MXR Analog Delay Noise

[Blues Lyne]

I know this isn't a build question, but I know some guys on here have experience with these pedals. 

I pulled my old big green MXR Analog Delay out and I'm getting a lot of noise.  The delay is strong.  With the delay time knob all the way down the noise goes away.  As the delay time is increased, the noise gets worse, and changes.  At the longest delay time there is a high pitched whistle like sound in with the hiss.  Also, with the mix knob all the way to the right, you get just delay, as you should, but this it all the way to the left, the delay is at least as loud as the undelayed signal, maybe louder. 

I know how to solder, and have modded and built some pedals, but I don't have scopes and don't really know how to trace a problem in the circuit.  I thought I'd see if this was a common problem in these pedals and maybe someone could point me towards some likely culprits.  I'm assuming the R5101 is OK, since I'm getting delay, but what do I know

Thanks

[Fret Wire]

Two things come to mind, the power supply electrolytic capacitors are failing. You haven't fooled around with the trimpots, have you. If it's been sitting unused for awhile, and now its noisy, it's probably bad caps.

[Blues Lyne]

I haven't touched the trimpots.  I know enough to know that I don't know enough to mess with the trimpots in a delay.

It should be easy enough to replace the electrolytics and see if that helps.  I've had the pedal since '79, so I know they haven't been replaced. 

Thanks

[Blues Lyne]

I take that back, I took a look inside and remembered that I replaced the one electrolytic about a year and a half ago when I replaced the transformer.

[MartyMart]

It's most likely the trim pot for "delay time"
I've done this to a couple of Boss DM2's and an Ibanez AD9
There's a point at which maximum "quiet" delay time is exceeded and you get
kind of "aliasing" from the delay chip.
back it off just a few "hairs" 'til the "squealing" stops.
It's how i tuned the above delay's for max time/least noise.
I adjusted the "regen" pots too on these, to get that feedback "dub" delay.

To be "safe" just mark the positions with a small marker pen before "tweaking" !

Marty.

[Mark Hammer]

The MXR Analog Delay's most distinctive feature is that it used a tracking filter to optimize filtering for the specific delay time.  Where other units of that era would pick a fixed filter rolloff that either gave bright short delays and some audible whine at long ones, or nice whine-free long delays but something a little too dull at short ones,  MXR's was bright when short and duller when long.

It did this by taking the HF clock driving the BBD chips and dividing it down.  That separate clock pulse would switch a series of CMOS gates (sections of a CD4016 quad switch chip) on and off to mimic a variable resistor.  At faster HF clock speeds, the gate would be on more often, mimicking a smaller-value resistance and raising the rolloff of the filter.

It is quite possible that smething is amiss with that particular chain of events and the filter is "stuck" somehow, higher up than where it normally is for that delay time.  I posted some schems of several issues of the pedal a while back.  Here are links to the relevant files.

http://img.photobucket.com/albums/v474/mhammer/MXRAnalogDelayschem1.png
http://img.photobucket.com/albums/v474/mhammer/MXRAnalogDelayschem2.png
http://img.photobucket.com/albums/v474/mhammer/MXRAnalogDelaylayout.png
http://img.photobucket.com/albums/v474/mhammer/mxr_analogdelay.png

[Blues Lyne]

Marty, thanks for the reply.  Would I be right in assuming that if the trim pot was off there would only be noise when the delay time was turned up?  This is noisy when there is any delay at all, it just gets louder and changes frequency when the delay time is increased.

Mark, thanks for the schematics.  While I understand what you are saying about the filtering, I'd have no idea how to check that.  I have a friend who used to design components for, and install radio studios.  He's very busy, but it looks like it may be time to take the schematics and pedal over and bug him.

Another odd thing is that there is a volume boost when the delay is turned on and at the minimum setting on the mix control, the delay time is at least as loud as the straight signal, maybe a bit louder.

One more question, I don't remember how the two outputs are supposed to work.  Right now with both plugged in, one a mix of the delay and the straight signal and one is just delay.  Is this correct?  This seems like an odd way to set it up.  Seems like one would be delay only and one straight signal with both plugged in.

Thanks.

[Mark Hammer]

Buy yourself a couple of CD4013 flip-flops, and the same again of CD4016 quad switches.  They're cheap and shouldn't cost you more than $5 for the whole lot. 

You can see here - http://img.photobucket.com/albums/v474/mhammer/mxr_analogdelay.png - that the unit only uses two 4013's to divide down the clock, and another pair to do the filter switching (everywhere you see a little box marked TG is 1/4 of a 4016, with 8 in total).  Pop the existing ones out, and the new ones in and see if makes a difference.  I'm hoping for your sake that they are socketed.

Somebody else here had an MXR unit that was giving them grief.  hey did a simple chip swap and that cured it.  I'm not saying this is necessarily your problem and cure, but it just might be, given the whining, and happily the chips are cheap enough that repair can sidestep the detective work and just involve a parts swap.  You will note that these chips are static sensitive, and don't like supply voltages above 15v.  In contrast, many of the other devices on the same board will happily coexist with +18v or other voltages arising from power line fluctuations.  That's the long way of saying that CMOS chips *do* fry often enough to be near the top of the list for things to check.

[Blues Lyne]

Thanks Mark,

I'll order some.  Unfortunately they are not socketed, but I'll put some in so I don't fry the new ones.

Mine has 2 CD4016's but only one MC14013B, which looks to be the same as a CD4013.  It also has a CD4001, should that be changed also? 

Thanks

[Mark Hammer]

The CD4001 is also CMOS so frying is within the realm of possibilities.  The one you are repairing is this one: http://img.photobucket.com/albums/v474/mhammer/MXRAnalogDelayschem1.png

From the looks of it, while the tracking filter is still used, the R5101 had its own on-board flip-flop section to produce complementary clock pulses, so the clock division was done a little differently.  In fact, in this instance, the 4001 becomes particularly critical in clocking the 4016s, and the lone 4013 appears to be exclusive to clocking the BBD.  Indeed, if you can achieve different delay times, then the takehome message is that your 4013 is alive and well. Which shifts our glance to the 4001 and 4016s.  So, a complete overhaul involves a single 4001 and two 4016s, and their respective sockets.

HOWEVER, before you do ANY of that, take note that there is a balance trimpot on the output of the 5101.  That trimpot is for adjusting the relative levels of the two complementary BBD signals so that they....wait for it....minimize clock whine.

Given the age of the pedal and the oft-mentioned fact that they live their lives getting kicked around, it IS within the realm of possibility that this trimpot needs some adjustment.  Misadjusting it won't hurt anything except for your ears.  You can easily mark off where it was set before you started futzing around.

The other thing I just realized is your wet-dry balance issue.  The schematic shows it is a bleed-to-ground type of balance control, similar to what's found on the MXR Blue Box and millions of home stereos.  Were it a dual-ganged type, I'd suggest something amiss with the wet pot half, but that isn't the case here....unless that side of the pot is simply dirty.  If you haven't cleaned it out, you may want to do that and see if it makes a difference.

[Blues Lyne]

Mark,

I really appreciate all of your help.  I tried to adjust that trimpot (R60 on the layout you linked to).  By just jiggling it a lot of the hiss went away and the mix knob started working again.  I don't see any broken solder joints or anything, but all of the trimpots are pretty gunked up with the decaying foam MXR stuck inside their boxes.  It seems to have melted into a sticky goo.

Now there is just a slight bit of hiss at low delay times.  Probably acceptable for a vintage pedal, but it gets worse the higher you set the time.  It's mostly hiss, and it gets progressively worse as you turn the time up, not just at the max setting.  At the max setting there is a bit of a whine or high pitched sound.  Sounds like this could be fixed with the adjustment you and Marty mentioned.

As for the trimpots, they don't seem to turn.  There is a dab of solder on each one.  Would they have soldered them into position to keep them from moving?  I tried forcing one a bit, but only the  screw turned, not the part that moves with it.  They haven't been moved in close to 30 years, so I suppose they could just be stuck.

Also, I'm still getting a big jump in the volume of the dry signal when turning the pedal on.  Is this normal?

I also remembered about turning the mix to all dry when using both outputs.    Oh well, we all have to embarrass ourselves sometimes.

Thanks

[Blues Lyne]

I got a chance to play a bit this weekend, so I hooked the MXR analog delay up.  The noise isn't too bad now until the delay time knob reaches about 3 o'clock.  However there is still a big boost in volume when it is turned on.  Is this normal?  I don't really remember it doing this before.

I looked at the trim pots again and it definitely looks like they are soldered so that they won't turn.  I'll have to get my soldering iron out and see if I can adjust the noise out of the upper delay times.  I'm wondering if the noise is being exaggerated by the volume boost.

[cjlectronics]

I've repaired many of these and based on your symptoms I would suggest cleaning the Mix pot.  If you feel comfortable doing this, dismantle the pot by removing the back shell  and cleaning the element with a little pot cleaner and soft cloth. Also, while you are in there clean'em all like this.

You can desolder the shaft of the trimpots by using solder wick to remove the solder that secures the trimpot's shaft to the post.  If there is foam all over the components, get an old toothbrush and brush the board clean.  That foam is nasty when it deteriorates. I would discourage the use of a spray solvent because it will only make things worse plus you'll remove the color coding stripes from the resistors and/or capacitors. In a worse case scenario, I have seen R5101's cause a noisy output but this was a very rare case.  90% of all of my repairs on vintage equipment is related to mechanical failures. I even clean the pins on the jack that connect the pots to the main PC board using a fine wire brush.

CJLectronics

[Blues Lyne]

Thanks,

I think the problem with the mix pot is worked out.  It seems to be functioning properly now.  The volume boost in on the whole signal.  If I turn the mix knob so there is no delay, I can use the pedal as a clean boost. 

I'll clean it up and adjust the trimpot and see what happens.