Electro-Harmonix Memory Man Deluxe not self-oscillating

[nullpainter]

Hi,

I own a rather beat-up second-hand Electro-Harmonix Memory Man Deluxe and all seems reasonably well, however the unit doesn't go into self-oscillation if the feedback is turned up sufficiently (the manual states that it's supposed to, and the previous owner has confirmed that it used to). This is my first experience with a Memory Man Deluxe so I don't know what to expect, but the delay/echo also appears to drop off quicker than I would have anticipated.

Could a capacitor discharging too quickly be to blame here? I'm no electronics guru so would like to tread as carefully as I could before attempting to replace all the caps.

Any suggestions or advice would be greatly appreciated.

M

[Matt T.]

If it's like my fairly recent Memory Man, there are 5 trimpots inside it and you can make an adjustment to one of them to get it to oscillate. Let me check my files when I get home, I might be able to help out more.

[Sock Puppet]

Hi,

Using the forum search for "Memory Man" should get the thread you're after.

S.

[nullpainter]

Hi,

Thanks for your comments. Now that I know it's to do with the trimpots, I've found the thread in question. Unfortunately I don't own an oscilloscope and electronics is still a bit of a black art for me, so I might have to refer the unit to somebody who knows what they're doing.

Is it possible to do irreparable damage to the unit by adjusting the trimpots?

M

[Matt T.]

nullpainter,
I don't know what I'm doing, let me just get that out there first. Secondly, there may be something wrong with your MM and that may be why it doesn't oscillate, so adjusting a trimpot might not work. I adjusted the trimpots without having any idea what they did but, I marked their original position and measured the voltage at each lug before I touched anything. I've gone so far as to identify each trimpot on the schematic but I don't understand most of what's going on in there anyway.

In any case, you can adjust the gain of one of the op-amps by adjusting a particular trimpot and it will make the 'Blend' knob more sensitive. If you unplug the unit and remove the bottom cover (careful, there is a ground wire connected to the bottom cover so lift the cover slowly) and orient the unit such that the jacks/power cord are facing away from you, the second 'hole' from the left (it will be near the center of the PCB) is the access hole to the trimpot I'm talking about (there are 5 total). Get a flashlight and note the original orientation if you decide to try adjusting it. Good luck.

[Mark Hammer]

There will be several trimpots inside, none of which will do any irreperable harm to your MM if you tweak them, but all of which may make the sound less than optimal if set wrong.

First, every BBD chip needs to see the AC audio signal riding on top of a steady DC bias voltage. If the bias is too low or too high, the delay chip will pass no signal (hence no audible delay). If it is a bit lower or a bit higher, there will be delay but very poor quality sound. So, while perfect adjustment likely comes with a scope, many people find they can nail the sweet spot in adjusting that trimpot by ear. There will usually be one of these trimpots per BBD chip, and no damage to the chip itself will occur via misadjustment.

Another comonly found trimpot concerns adjusting the balance between the two outputs of the BBD. The two outputs carry complementary signals and when they are in perfect balance, the sound quality is improved but more importantly, clock noise is cancelled or dramatically reduced. Again, misadjustment will bugger up the sound but not produce any risk of damage.

Some companies prefer to simply use fixed resistors to do this biasing and mixing, while others go for the precision and use adjustable trimpots to reallynail things. Maxon uses no less than 20 trimpots in their current issue analog delay pedal.

Another commonly occurring trimpot in delays is one which adjusts the clock rate to achieve the target delay range. The tiny value capacitors in the clock circuit can have substantial tolerances/variation, and we know that the resistors are generally 5% tolerance, so the trimpot allows the company to just shove in unselected 220pf (or whatever value) caps and nominal value resistors into the clock circuit part of the board, and then just tweak the trimpot to get a specific minimum/maximum delay time. Again, this trimpot does no damage but may bump you out of the delay range you want.

Finally, again because of component tolerances, turning up the regen/feedback control on a delay may result in irritating runaway feedback on some units ell before you hit max, while on others you can dime things and still not get any runaway screeching. So, many companies use a trimpot in the feedback loop to adjust. The regen control is maxed and the trim-pot is set to achieve howling feedback, then backed off gently until the howling disappears. This strategy is used not only on delays but on flangers and phasers.

The bias and balance trimpots will be found near the BBDs themselves. The clock trimpot will be found near any CMOS chips (start with "CD4xxx" usually), and the feedback adjust trimpot will be found somewhere else on the board. If you have a digital camera, take a picture of where things were set before you start monkeying around, then....start monkeying around. You should be able to tweak it to within inches of perfection completely by ear.

[nullpainter]

Thank you all so much for your advice. Mark; thank you particularly for your very useful breakdown of what each of the trimpots are responsible for.

The one thing to be aware of when taking photos of the trimpots before adjustment - and one thing that only dawned on me after the fact - is that an unmarked trimpot rotated 360 degrees looks exactly the same as one that isn't

Regardless, I adjusted all trimpots by ear with a relatively pure sine wave input signal from my synth and now all is happy. The chorus still sounds wrong (it currently sounds more like a sped-up vibrato), but I'm not too fussed about that.

Thanks again. I'm very happy

[Mark Hammer]

Excellent. Happy to help....particularly when it DOES help.

[lowell]

Hey guys I've got a mem man here that has a high screech underneath the signal. It's far less loud than the signal but still quite annoying. It changes when the delay knob is moved and also when the 3 wires connected to the chorus/vibrato switch are moved w/ my hand. This gets FAR worse when I put the back plate on...? I tried adjusting the trim pot for the 2 outputs of the BBD with no change.

[cjlectronics]

I've seen a few things cause this:
1. Noisy SAD1024 - replace it
2. Dirty/bad pots. - Clean or replace
3. Bad resistor - Rare but I've seen it happen. Most of the time its a hiss, though
4. Leaky capacitors
5. Noisy clock circuit - rare because typically the clock circuit just dies.

Hope this helps,

CJ

[lowell]

right on... which chips in the new memory man are the SAD1024? The MN3005?

[lowell]

right on... which chips in the new memory man are the SAD1024? The MN3005?

cjl you there? Is the MN3005 the new SAD1024? If so what's the new MN3005? Cannot find one.

[Mark Hammer]

There IS no "new MN3005", and there are probably more surviving borscht-belt comedians these days than there are SAD1024s that are not yet spoken for. However, Coolaudio makes the V3205SD that is functionally equivalent to the MN3005. :: Official website of Coolaudio Semiconductors (COOLAUDIO International Limited) - www.coolaudio.com

Unfortunately, you can't just pop one in. The MN30xx series of chips used a positive ground, while the 32xx series used/uses a negative ground. Some additional pin routing would be needed, even though it could be made to work.

[lowell]

Ok thanks Mark!

[lowell]

Mark or anyone have a couple of these? I don't want 10 of them, maybe 4. Can do paypal.