note: pictures of the unit attached.
Someone sent me an early EHX Memory Man hoping I can repair it. I'm a bit apprehensive to start working on it, so I hope someone with experience with these units can help me out a bit. It looks like there have been some modifications over the years, the transformer is dead and there's a loose wire coming from the boost switch. No clue if the delay chips are still good. Still looking for a verified schematic for this particular unit.
On the EHX forum I found that a replacement transformer for the DMM is the F-115X which Mouser carries. I'm going to compare some specs, but I expect it will work with the MM as well. I have a EHX 24V center positive wall wart, was wondering if I can use this to test the unit before buying the transformer.
Any other suggestions for testing the unit are very welcome.
You'll be wanting moi.
I can't see any mods in the photos? Although some of that wiring could be new, looks stranded instead of solid core.
Yes you can use a 24V supply, the minimum transformer EHX used was 12-0-12 for 24 volts (ish, AC 'n' diode drops 'n' all) it's regulated through a zener/transistor capacitance multiplier set up I think on this unit but some did use a 78L15.
Schematic attached, it's not 100% accurate, been a while since I worked on this particular version but I recall that being the case, but close enough. I've replaced one SAD1024 in the past year, on EHX stuff, check the Op Amps First.
Thanks for chiming in Scruffie :)
The reason I mentioned mods is that I saw some components that aren't on the circuit board but rather fly over it. I just assumed they where added later.
Which terminals should I attach the 24V supply to? Referring to picture 03 in my original post, I would say that the top wire is for the ground terminal (outer sleeve of adapter)? Center tap is for the LED, so -24V on the bottom connection in the picture? Power is then very crudely rectified by those 2 diodes, right?
The ones on the track side? Very common on old EHX :)
The top wire is ground/the centre tap as you say and then the other two are both AC which yeah, get half wave rectified by the diodes. Just stick positive to either of those two bottom wires and you get a polarity protection diode just in case.
Only had time for a quick power test. When powering the effect you clearly hear a high pitched whine that starts up almost immediately and persists. This is usually followed a few seconds later with static noise (from a delay line?). No audio probing yet, that's going to be the weekend.
[one thing I noticed is that approaching the -24V probe to the center tap produces a little spark between circuit board and probe. The outer tap does not]
Does the whine change with delay time? Sounds like the clock, the filtering on these ain't great so it bleeds out quite a bit.
Getting +15V from the regulator? Voltages for the two 4558 and the 4047 would tell a lot.
Static noise... could be a lot of things depending on the exact sound.
Whine changes with delay pot, goes supersonic when turning delay time down.
4558 closest to the jacks
Pin 1: 5.5 V
Pin 2: 5.5 V
Pin 3: 5.5 V
Pin 4: 0V
Pin 5: 0 V
Pin 6: 10.6 V
Pin 7: 11 V
Pin 8: 11.6 V
4558 middle of the board
Pin 1: 4.2 V
Pin 2: 4.2 V
Pin 3: 4.2 V
Pin 4: 0V
Pin 5: 5.8 V
Pin 6: 4.8 V
Pin 7: 11 V
Pin 8: 11.6 V
So that looks like both the b-sides of the chips are not doing what they should be doing.
Could very well both be dead.
But first things first, your supply voltage, assuming you have the transistor regulator, what voltage do you get on its 3 pins.
If you haven't tried already, give the entire unit the "chopstick test". Just get a chopstick and start poking around at various components and wires while the unit is on and making the whine. I've always been amazed how you can find dry solder joints or loose wires with this method, or how one component being poked will change the sound you're hearing (that gives you a pretty good idea what to focus on). It saves a lot troubleshooting time.
So without a load the EHX wall wart sits at 34V. Connecting it to the MM it drops to about 14V. The input of the regulator is at 12.6V, middle pin is 12V and the output is 11.5V. I'm guessing those numbers are off ;)
Just a tad :D
So something is pulling the voltage down, kind of tricky as those old EHX boards don't stand up to a lot of re-soldering so pulling things just to test is not an easy task but it's probably a safe bet to replace both those 4558.
But before, just to be sure, test the 4047 pins 10 & 11, they should have half the supply voltage, it's already whining with the delay time so it's probably okay.
pin 10 is at half supply (5.8V), pin 11 is at full supply (11.6V)
Oh, in that case, it's worth pulling it out as it's in a socket and seeing what happens to the supply voltage.
Thanks for all the help so far and apologies for taking so long to get back. I really only have a few minutes a day for this kind of thing at the moment...sigh.
Anyway, rechecked: I made a counting mistake. Pin 10 and pin 11 both at half supply after all!
Next baby step?
No need to apologise, we all go through busy patches.
I'm looking over your voltages again and Pin 5 of the 4558 closest to the jacks should be your input and have a bias from the 470k/390k resistors next to it so why is it measuring 0V? That Op Amp determines the voltages of the following one so check for internal shorts and make sure those two resistors are connected. But I suspect that one is dead and should be replaced, the one in the middle of the board might be okay.
In the mean time while you're doing that I would replace any power filtering caps, even if they still work they'll be past their best now and it's not going to affect the sound apart from removing noise in its various forms. On the schematic that would be the 100uF on the BC239 collector, the 220uF on its base and then the 100uF on the +15V line (coming off the emitter) and the 10uF connected to the 1N4001 diode.
I attached the wire from the boost switch to the board, replaced the top 4558 IC and the larger electrolytics. Sure did something. I can now blend between input signal and 'whine' signal and the boost also works. Here are some fresh measurements:
Voltage at board input: 11.6 V (as before), but supply voltages on the ICs have come down to about 9.5V.
4558 closest to the jacks
Pin 1: 4.9 V
Pin 2: 4.1 V
Pin 3: 4.1 V
Pin 4: 0V
Pin 5: 4.3 V
Pin 6: 4.4 V
Pin 7: 4.4 V
Pin 8: 9.7 V (no longer equal to supply?)
4558 middle of the board
Pin 1: 3.5 V
Pin 2: 3.5 V
Pin 3: 3.5 V
Pin 4: 0V
Pin 5: 2.4 V
Pin 6: 2.4 V
Pin 7: 5.5 V
Pin 8: 9.8 V
How odd... but those voltages look a lot healthier apart from the supply of course.
You say you get 11.6 at the board input, do you mean where the supply connects because before you were measuring 12.6V there.
It's time to take readings of the SAD1024 but I'm wondering if the power supply itself is possibly to blame for the low supply, it's a high sag unit and perhaps the circuit is just putting too much load on it as that regulator style isn't that efficient, a good 18V unit might be better. For a quick test you could try two 9V batteries, or you could just replace the zener regulator set up with a 78L15.
I'd like to keep the original regulator setup. I measure the voltage at the point where I clip it to the board. The old transformer is still attached. I was thinking that maybe, if the primary is broken but the secondary is still ok, that could be pulling down the supply because there's a lot of current flowing through the coil? I haven't noticed it heating up though.
I can connect it to a Strymon Ojai or PP2+ for 18V. Just to be double safe: I can use a 'regular' 18V supply, ground to ground and +18V to the power input?
Well if it's broken... cut the wires and test your theory ;)
Yeah a regular 18V supply is fine.
Quote from: Scruffie on February 25, 2019, 05:36:33 PM
Well if it's broken... cut the wires and test your theory ;)
Yeah it was just dawning on me while I was typing the post. Will test this soon, it makes sense.
One step forward, one step back...I should've paid better attention, but when replacing the power filtering electrolytic cap I installed it with reversed polarity. On the original cap, the plus-side was marked instead of the standard minus side these days. Could've spotted that, but I blew the cap. Replaced it and the wire from the board to the switch had become detached. Fixed that. Desoldered the transformer, replaced the 4558 with a socket and installed a fresh chip.
Good news: power is now at 28V from the EHX supply to the board, 1st opamp sits at 6.7V bias, 14.6V input. So that seems fine. The clean signal is now very very faint however and most likely only induced in the wires, not really going through the switch. Probably another silly mistake somewhere.
Whoops! Oh well, you've got a good supply voltage now and that's the main thing. Why did you replace the 4558 again though?
So I'm assuming that while your clean signal is low you're not yet hearing any delay on the wet side, so the voltages for the SAD1024 and sounds like it's audio probe time to see why you're clean signal is so low and to see if signal is reaching the BBD's and passing through any of them. Pins 15 & 2 are the inputs and pins 6 & 12 are the outputs apart from the last chip where you'll want to measure the middle pin of the 1k trimmer. Looks to me the one closest to the 4047 on pin 15 is where the signal first hits them.
The 4558 was looking good on the DMM but with no sound coming through I thought that maybe something else shorted out as that cap was losing its mind. So I decided to replace it once more with a socket (should've done that in the first place). That way, future repairs on the IC won't damage the board.
It's 1:00 AM right now. I'll continue tomorrow (vacation time, so let's hope I fix it this week).
Necro bump, but still. I had put this unit on the shelf for a bit and haven't looked at it in a while. Took it out today and miraculously it's working. Guess my last iteration of repairs did the job but I hadn't noticed it. Doh. There's now a bypass signal and a delay signal, so the SAD chips are ok. The only thing is there's still a low volume high pitched whine riding on the audio. I'm thinking that might be due to the walwart so I'm going to look for a replacement transformer and also replace all remaining electrolytics. Any other suggestions to combat the whine?
Set delay to max, unplug the input and adjust the 1k 'balance' trim while audio probing its middle pin.
Quote from: Scruffie on March 28, 2020, 07:30:18 PM
Set delay to max, unplug the input and adjust the 1k 'balance' trim while audio probing its middle pin.
Remind me, how much am I paying you again? A 1 degree turn of that trimmer and the whine is gone.
Now to find a suitable transformer. The F-115X is not the right one since being in Europe we use 230V, not 110V. Any suggestions for a suitable replacement with the same form factor?
Quote from: Rockhorst on March 28, 2020, 11:31:21 PM
Quote from: Scruffie on March 28, 2020, 07:30:18 PM
Set delay to max, unplug the input and adjust the 1k 'balance' trim while audio probing its middle pin.
Remind me, how much am I paying you again? A 1 degree turn of that trimmer and the whine is gone.
Didn't you get the notice ??? £100ph ;)
QuoteNow to find a suitable transformer. The F-115X is not the right one since being in Europe we use 230V, not 110V. Any suggestions for a suitable replacement with the same form factor?
That I can't help with, Maplin used to sell a match but then they went bankrupt and I'm down to my last two.
Here's the specs if it helps, let me know if you find a supplier!
12-0-12V
100mA output
Overall size: 37 x 31 x 30mm
Fixing centres: 45mm
Considering choosing the easy way out and just installing a modern pedal jack for a wall wart.
Also have little experience reading tranformer specs. I found this one (https://leeselectronic.com/en/product/102906.html), but can't find the full specs. It seems this is 25V input, so not suitable. Right?
If I go with the wall wart option should I (re)wire anything to the center tap on the circuit board?
https://www.ebay.co.uk/itm/EI28-1W-Power-Transformer-220V-TO-6V-9V-12V-15V-18V-24V-Output-AC-Single-Dual/173371131503?hash=item285db9866f:m:mn8Bzp74VrvTThqRSuinWtw&var=472080328400
This might fit, either the dual 12 or dual 15 could be used, you'll have to measure the mounting holes etc.
Nice find. Screw holes should be 45mm apart, sadly the ones in your link are quoted to be 42 apart. I can probably find a work around for that though. This 2W 12/15/18V one (https://www.ebay.co.uk/itm/EI35-2W-Power-Transformer-Input-AC-220V-50Hz-Output-AC-Single-Dual-6V-24V/173083937369?hash=item284c9b4a59:m:mQAak9Ia-1JO2-U2rVJ-ypA) from the same seller might be a better option?
Same group of units, different listing heading.
Look at the 3rd image, it has the datasheet for the various options.
The Maplin ones were 45mm and they didn't perfectly line up so measure the mounting holes on your units, the difference could be a bit smaller or larger.
Ordered, I'll be back here in about three weeks I suppose ;)
Thanks for all the help Scruffie!
I got the 15V and 18V transformers in the mail past wednesday. The eyelets on both are much wider than the datasheet had me suspect, so both would've been a good fit. I installed the 18V transformer and the Memory Man is working fine. Thanks to Scruffie! Only thing left to do is get two nuts and bolts to mount the transformer and it's good to go.
;D Glad I could bring some EHX back to proper standing.
But I will say one thing, 18V? Are your power filtering caps rated for that? (With both sides in use that'll be something like ~+30V in practice) I thought you were going to get either the 12 or 15V.
Glad to find another supplier of suitable transformers and thanks for being a guinea pig ;) I'll have to grab some spares.
One cap is rated at 50V and measures 26V across the terminals, the other one is rated at 25V and measures 15V across the terminals. So that should be all within safe limits although there's probably some unnecessary heat dissipation. Would you advise to swap it for the 15V?
Oh that's fine, I expected higher.