Lots of Memory Man info I compiled; credit goes to Aaron, Analog Man and Howard Davis for the posts below.
Analog Mike:
The Japanese Panasonic MN30XX chips are the best ever made, a little better than the more recent (but also discontinued) Japanese MN32XX chips in the DM3, AD9 etc. and much better than the BL32XX Chinese chips. These MN chips are all discontinued, only the BL Chinese chips are still made. Don't know how much longer EH will be able to make the memory man with the good old Japanese chips
3X05 = about 4000 delay stages, 3X08 = 2000 stages, 3X07 = 1000 stages.
In about early 2006 the DMM changed to relay true bypass switching, and four MN3008 chips instead of two MN3005 chips. The new DMM has (4 * 2000 = 8000 stages), The older one had (2 * 4000 = same 8000 stages) for the same delay time.
4000 stages is good for about 300mS at reasonable quality (AD9 etc) or 200mS at high quality (DMM).
1000 stages is more than enough for a chorus (we use new old stock MN3007 chips in our chorus). There were also 512 stage chips used fo chorus and flangers (best for a flanger).
To check if it the new version with the relay, turn it on and play, so you are hearing the delays (effect on). Then unplug the power from the back of the DMM. If it turns off and you hear your normal guitar sound, then it has a relay. If you hear nothing then you don't have the relay. You can also tell as the stomp switch only has two wires on it if you have the relay.
I guess they ran out of the MN3005 chips. Theoretically, the less delay chips you have, each with higher number of memory positions to add up to the same amount of memory, the better the sound. But in reality there is little sound difference between two 3005 chips and four 3008 chips. But there are twice as many calibration points to dial in if you have 4 chips. Some delays have EIGHT CHIPS (ad999?) that must be a nightmare to calibrate...
We can install four MN3005 chips on the newer model Deluxe Memory Man to double the delay time to about 800mS. But it's a VERY expensive mod as the chips are about $100 and about the same for our labor in calibration. And after this mod the pedal can no longer be used as just a chorus as the short delay times are now too long for a normal sounding chorus pedal.
Memory Man op-amp IC chip upgrades
There are a few web sites and posts about improving the sound of an EH deluxe memory man, by replacing the normal 4558 op amps with some high tech chips. The noise and tone in a memory man comes from the BBD chips, not the op amps. The BBD chips have about 10,000 times more noise, and very low fidelity. That is normal in an analog delay and nothing can be done about that, it's why people love their sounds. So replacing the op amps will have little benefit, or so I thought.
After writing the above I got some new DMM pedals, they are now using ST electronics MC4558CN dual op amp chips. I replaced all five of these chips while I was in there (those knobs are a pain to remove) with new Texas Instruments TL072 chips.
Dry sound when playing clean - it is improved. Put the mix to pure DRY (no echo) and the tone when playing clean is a little purer. I was surprised.
ECHO SOUND: also clearer when you listen to 100% delayed sound, much clearer and closer to a digital delay sort of tone. Seems to be more difference at lower delay times.
When I turn up the MIX for about 50% delay I tested for noise by playing a low string gently. The modified pedal is a bit noisier, may be a unit to unit discrepancy or could be the TL072 chips allow more high end through which contains the noise.
OVERDRIVEN sound : running a good OD pedal into the DMMs, with the gain knob set for unity gain (LED just glowing) the sound of the stock one seems a little warmer to me, both dry and echo sounds. But the modified one is clearer again.
DISTORTION : Running a good distortion pedal into the units, the modified one is clearer, the repeats are almost too clear. The stock one has repeats which are darker and more in the background. At this point in the test I brought out a '59 reissue Les Paul with burstbuckers 1 and 2 and played through the Maxon SD9/808/silver, into the DMM into a '66 blackface Deluxe Reverb plugged into a '73 Marshall 4x12 cabinet and just played for 45 minutes as it was a MAGIC combination! Gilmour, Hackett, Page sounds were flowing... best tones I ever got for leads.
Anyway, the difference with changing the op amps is more than I thought it would be. But not sure whether I like it enough to go through the hassle of pulling the board out... depends on what you want, clarity or a warm background echo. Most people will prefer the upgraded op amps I think. Also the new 2006 pedals do not use sockets so the chips need to be desoldered, a lot more work. These are the EC2002_REV_E boards which use four MN3008 chips and a relay to switch the pedal on and off. You can test for this one by turning it on, playing, then unplug the power jack. If it turns off (you get normal OFF sound) then it has a relay and is the new version.
Aaron:
Simple Deluxe Memory Man Reissue Mods
WARNING! This probably voids your warranty! I am not responsible for any problems you may have doing these mods. Do these mods at your own risk.
________________________________________
Apparently, one of our friends has tried this first mod on his reissue Deluxe Electric Mistress (change the first op amp to TL072) and found that the difference was substantial as well.
"The whole unit is quieter, the phasing effect is stronger without a lot of buzz, and the flange effect is deeper"
________________________________________
Smoother, better clean signal
I recently bought a new Deluxe Memory Man (DMM) reissue for my delay needs. I remembered that nice warm tone that the DMM gave me. It was pricey but well worth it right?
Anyway, once I got it I noticed that it changed the tone of my Shaka Braddah 3 when the DMM was placed at the end of my effects chain (which is where you usually put the delay pedal). Here's what the chain looked like:
Guitar->Shaka Braddah 3->Deluxe Memory Man->Bassman
I noticed that the Shaka Braddah 3 took on a brittle edge and that it just sounded "edgy" and not warm like it should.
After a myriad of potential mods went through my mind, I noticed that the schematic showed that the DMM used a JRC4558D for the input op amp as well as for other parts of the signal chain. The clean signal clearly goes through the JRC4558D and into the blend pot. I remembered that the JRC4558D didn't sound as good as a TL072 for clean signals in my Shaka Braddah 3 so I figured, why not change it?
The JRC4558D is mounted in an IC socket so changing it was easy. The result? FIXED!!! The DMM sounds smoother and warmer like it should be.
To mod yours:
1: UNPLUG the unit first. Take off the black screws holding the back panel onto the DMM.
2: Take the knobs off and the washer and nut holding the pots to the chassis.
3: Locate U1 on the PCB. It will be located next to the 1Meg input pot. Note the orientation of the JRC4558D, you want your TL072 to be plugged in with the same orientation. Simply replace the JRC4558D located there with a TL072. While you are there, I suppose you could replace every RC4558D with a TL072, but I haven't tried it yet.
4: Put everything back and enjoy your new warmer, smoother DMM!!!
________________________________________
DPDT Conversion
The Deluxe Memory Man (DMM) comes with a Carling DPDT switch installed but the DMM is not true bypass. This relatively simple mod can be done if you simply have to have true bypass on your DMM. The DMM is really a free buffer/booster when the effect is off and not a bad one at that with the TL072 mod mentioned above. Notice that the input level still works when the effect is bypassed? The straight signal is still buffered through the input op amp.
For reference, here is the original wiring:
DPDT Switch
[] [] ------- T5 - Brown wire
[] [] ------- T7 Red wire
[] [] ------- T4 Grey wire
T8 to input jack
T6 to output jack
Brown is the buffered direct signal
Red goes to the PCB which goes to the echo out jack.
Grey is the blended signal direct from the blend pot.
To convert your DMM to true bypass:
Follow steps 1 and 2 from the previous mod.
1: Disconnect the brown wire (T5) from the switch and tape off the end (you will not use this wire).
2: Disconnect the grey and red wire from the switch and solder them together. Insulate the connection.
3: Disconnect the white wire (T8) from the input jack and connect it to the DPDT (Illustration 1)
4: Discnnect the T6 from the output jack and connect it to the DPDT (Illustration 1)
5: Connect the middle lug (left) of the switch to the input jack.
6: Connect the middle lug (right) of the switch to the output jack.
7: Connect the bottom two lugs together on the DPDT switch.
Illustration 1:
New DPDT Switch wiring
T8 ---- white from input of PCB ----[] []---- white from output of PCB T6
from input jack ------------------------[] []---- to the output jack
[]_[] <--- connect the bottom two lugs together on the DPDT jack.
Look at the output jack, there might be a cap across it; if so, consider removing this cap (it seems to get rid of highs and makes the bypassed sound muddy) - you can always put it back if you need to.
Close up your DMM and enjoy your DPDT DMM.
Since i have read many posts about squealing EH MemoryMan's and Electric Mistress', i decicded to give some info (even for beginners) on how the trimpots have to be set up in BBD-based circuits. I know, not all circuits offer the possibilities to adjust EVERY aspect of the BBD, but this should only be a guide to get the best out of them and what to look for
To tell you the things to be done, check out the Deluxe ElectricMistress schematic at
http://smoke.csoft.net/~stompbox/pedals ... deluxe.jpg. Luckily, this circuit has 5 trimpots for all adjustments as seen on the Memory Man as well.
Be sure you have a scope and a sine-wave low-frequency generator since this will give the best results...
Here we go:
Step 1: BBD balance
settings: Feedback, Rate, Range & Levels to CCW
This is a scope job if you want best results...
Connect the scope tip to the taper of the 1K 'BAL. ADJ.' pot that is connected between pin 6 & pin 12 (outputs) of the SAD1024 and adjust for minimum bleed-through of the clock signal. You can't adjust this to 'ideal zero', so the best is a setting which shows both clock-wave components at equal minimum level. The ideal balance setting varies with delay time, so maybe re-adjust for best results over the entire delay time range.
Step 2: BBD offset/bias
settings: Feedback, Rate, Range & Levels to CCW
This is a scope job if you want best results, but could be done by ear, too. This is an important setting since it determines the maximum undistorted output level of the BBD.
Connect scope tip to taper of the 1K 'BAL ADJ.' pot. Feed 400Hz sinewave into input. Increase output level of generator until the signal shows clipping. Adjust the 10K 'BIAS ADJ.' pot for symetrical clipping / proper sine wave. IMPORTANT:
Most BBD- based circuits have a pre-emphasis network in front and a de-emphasis network behind the BBD. This means, treble is boosted in front of the BBD and cut down the same amound behind the BBD to keep BBD-noise level low. This maybe affect the ideal offset setting. For best offset setting, check the adjustment of the 10K 'BIAS ADJ.' pot with different input frequencies! The ideal offset setting varies with delay time, so maybe re-adjust for best results over the entire delay time range. BBD offset and BBD balance interact a little with each other, so check back!
Step 3: BBD gain
setting: Feedback, Rates, Range & CCW, Delay Level to MAX
This could be done by scope or by ear. This setting sets the level of the delayed signal in relation to the dry signal. This could be adjusted to taste, but the most common setting is to have equal volumes on delayed and dry signal. On short delay times (as on flangers) there's a scope trick to achieve equal volumes: Connect scope tip to pin 7 of the 4558 (right of the SAD1024). (REMEMBER: Don't connect to the taper of the 1K 'BAL. ADJ.' pot this time, because we need to cancel out the effects of the pre/de-emphasis network this time!) Feed 400Hz sinewave into input. If you now change the delay time knob, you can see the sinewave going up an down in volume as long as you turn the knob. Because you vary the delay time, there are boosts and cuts at certain frequencies due to phase cancelations. Adjust the delay time knob to a position where a maximum cut/cancelation occurs. Now adjust the 10K 'GAIN ADJ.' pot for zero output. Now, delayed and dry signal are in equal volume, because the sine wave is fully canceled. This adjustment interacts with the 'BBD balance' setting, so maybe you have to re-adjust it after setting BBD gain. Due to the tolerances of the resistors and caps in the pre/de-emphasis networks, the ideal gain setting should be re-adjusted with differnt input frequencies. Delay time setting affects this adjustment too, so check back with different settings or just adjust to taste
Step 4: BBD clock
setting: Feedback, Rates, Range & CCW
This is mostly an uncritical setting, since it just determines the minimum/maximum delay time covered by the delay time knob. This pot affects the clock circuit that drives the BBD. On the schematic, it's the unnamed 100K trimpot connected to Ub and the 3k9 resistor left of the CD4013. Only on long delays, this setting might be critical. If the clock frequency is adjusted too low (high clock frequency = short delay time, low clock frequency=long delay time), the clock signal might come down to audio range and will be heard as 'faint' signal. So adjust this trimpot until no clock is heard or adjust to taste and get some weird sounds
Step 5: FEEDBACK
setting: Feedback to MAX, Rate to CCW, Range to MAX
This is no direct BBD-related adjustment, but is found on most delays, flangers and phasers.
This trimpot (the 100K 'F.B. TRIM') limits the range of the Feedback knob. While feedback knob in MAX position, adjust for maximum possible feedback with no self-oscillation. Re-adjust with different delay times, some circuits intend to get into self-oscillation at shorter delay times....or adjust to taste for driving the circuit into self-oscillation for wild feedback orgy
A post to alt.guitar.effects dated 3/7 re. the Electro-Harmonix Deluxe
Memory Man reissue mentioned a drop in signal volume when the effect is
engaged.
To date I have only observed this problem in units equipped for true bypass.
Some of these units use a 3PDT footswitch with an LED by the power-on switch
that actually functions as an EFFECT ON indicator, not a power indicator.
By bringing up the LEVEL control, the effect level can be adjusted to match
the bypass level, but doing this can drive the unit into undesired
distortion.
This alteration of the design to include the true bypass switch was done
without consulting me as to the possible consequences.
I HAVE DESIGNED A MOD THAT FIXES THIS PROBLEM
If interested, please e-mail me at
howard.davis2@worldnet.att.net
Howard Davis, EE
Designer of the Memory Man series.
howard.davis2@worldnet.att.net