Ensoniq DP/4+ REPAIR (cap leakage, corrosion)

[Script]

Unit B -- recapped, working.

Literally, 90 percent of components came off the board with a hot air gun simply for cleaning and looking under ICs. Probably this was not necessary, but for me these units are like "carrots" to hone my still rather rudimentary SMD skills.

There were two green LEDs lit up on power on, but turning the DC offset trimmers turned them off. Now needs measuring of signal path to confirm / find cause, as well as for fine calibration.

Final step is to put UV-curing solder mask ink on some traces / vias.

 

[Script]

BOARD REVISIONS

The two units I have seem to be of different age:
- Unit A's digital board is Rev A from 1994.
- Unit B's digital board is Rev B from 1995. Same PCB green as audio board

The obvious difference is R200,:which is on- board SMD in Rev B.

Unit A, the older one, was a catastrophe right from the start. While Unit B had just started to leak.
____
Once Unit B is fully calibrated, it can serve to troubleshoot Unit A and/or swap boards for confirming.

 

[58Glammer]

BOARD REVISIONS

The two units I have seem to be of different age:
- Unit A's digital board is Rev A from 1994.
- Unit B's digital board is Rev B from 1995. Same PCB green as audio board

The obvious difference is R200,:which is on- board SMD in Rev B.

Unit A, the older one, was a catastrophe right from the start. While Unit B had just started to leak.
____
Once Unit B is fully calibrated, it can serve to troubleshoot Unit A and/or swap boards for confirming.

Congratulation! Now you have fixed unit B, I bet you will get unit A fixed as well!

 

[Script]

Another difference by which to determine age is the OS, unless unit has been upgraded.

Unit A has 2.00A
Unit B has 2.02
Latest OS available is 2.05.

 

[Script]

Swapped OS chips and boards between units. Some new symptoms on wrecked Unit A

(1) OS chips are OK. Good news

(2)i Digital board is faulty. Error message reads: "Unexpected event error type 45" (= ESP sync error). Not sure what that means. Need to study or someone to explain please. I am not a computer guy.

(3) Audio board channel 2 works. Channel 4 basically works but static/electric crackling noise on output (changes with volume knob) Channels 1 and 3 don't work at all. No input it seems.

That's all huge progress compared to what the wreck looked like on day one...

What I need now is time, which I have hardly had as of late. Always out and about.

 

[58Glammer]

Swapped OS chips and boards between units. Some new symptoms on wrecked Unit A

(1) OS chips are OK. Good news

(2)i Digital board is faulty. Error message reads: "Unexpected event error type 45" (= ESP sync error). Not sure what that means. Need to study or someone to explain please. I am not a computer guy.

(3) Audio board channel 2 works. Channel 4 basically works but static/electric crackling noise on output (changes with volume knob) Channels 1 and 3 don't work at all. No input it seems.

That's all huge progress compared to what the wreck looked like on day one...

What I need now is time, which I have hardly had as of late. Always out and about.

Wow, that’s good news! Ch 2 working! So you could compare certain parts of the channels and the signal path.. with Ch2

Silly question: Do you use a scope? At a certain point you have to.

Check the area with the AD DA converters as well.

Check my post, where I found faulty op amps ..

With a scope check the polystyrene caps… there is a kind of pulse, you should see on working Ch 2 (the second one from the left)
If you find the same signals/pulses on each of these caps, this area might be ok. (no audiosignal is needed) . As long these pulses are not there, you won‘t have a chance to get the audio signal through..

Also check all clocks at those 2 sony ICs and the AD DA converters..

The other tricky area is on the right side right beside the power regulators.. check the op amps there as well.

I took off the DA AD converters as well to clean underneath. But I guess, with your hot air unit you cleaned that area as well, right?

If one channel works, you‘re close..

The PDF is a great help. I could encircle the problems by marking spots with different colors.

And if you don‘t have a scope, I would suggest, to just blindly order the op amps (I think it’s two different types) used on the unit and exchange just all of them on the analog board used in the area where elcos sit.
Start with the ones in the input section, then the ones in the region the polystyrene caps sit and then the ones in the area right beside the powerregulators.
They are pretty cheap, just look for the right size (smd type) .

 

[Script]

Yes, that's the plan and steps. On the audio board of Unit A, I had replaced quite a number of op amps already -- not all though. Yes, they are cheap.

Having a working audio board from Unit B will make checking for connectivity easier.

That said, in my case, I'd rather love to get the digital board running first. Two units booting correctly would make for even fast(er) comparison.
---
On the road again from tomorrow on...

 

[Script]

Another weird symptom:
Today, faulty digi board into proven working audio board made CH 3 LEDs all light up . Audio board was and still is fine!
Might be random, faulty boot symptom ...

 

[Script]

(1) OS chips are OK. Good news

And forgot to mention, so are the four ESPR ICs (U1, 7, 13, 19) themselves, as well as the 63B03 (U25) that sits next to the battery. These are all the socketed ICs.

 

[Script]

Also, ESP can mean 'Ensoniq Sound Processor', a proprietary DSP. It's even on Wikipedia:

https://en.wikipedia.org/wiki/Ensoniq_Signal_Processor

 

[Script]

And another difference between Digi boards from 1994 and 1995 (Rev B).

Extra components are next to, but NOT part of, the footswitch circuitry. They connect to the 40-pin header leading to the front panel assembly.

 

[Script]

Finally some time for the units.
____
Unit B is fully functional (so that maintenance fix was successful by just swapping out all elcos and cleaning the boards thoroughly).
_____
Unit A, however, doesn't boot the digital board correctly. Spend an hour measuring around rather randomly and comparing with functional board of Unit B Observations:

(1) Some random connections (when powered off) read 10M but only 7M on the non-functional board. Need to pin down the area.

(2) When powered on (but OS and ESPR chips pulled), some connections on the non-functional board give a DC reading against ground. It repeatedly goes slowly up to 390mDC/500mDC or so and then starts at 0 again. The functional board doesn't do that at all. Interesting !

 

[Script]

Can a damaged SMD voltage regulator cause that ?

Or maybe it's CMOS chip U36 acting up. There was massive leakage there.

 

[58Glammer]

Finally some time for the units.
____
Unit B is fully functional (so that maintenance fix was successful by just swapping out all elcos and cleaning the boards thoroughly).
_____
Unit A, however, doesn't boot the digital board correctly. Spend an hour measuring around rather randomly and comparing with functional board of Unit B Observations:

(1) Some random connections (when powered off) read 10M but only 7M on the non-functional board. Need to pin down the area.

(2) When powered on (but OS and ESPR chips pulled), some connections on the non-functional board give a DC reading against ground. It repeatedly goes slowly up to 390mDC/500mDC or so and then starts at 0 again. The functional board doesn't do that at all. Interesting !

Congrats for Unit B
With the DC on unit A is strange. I can‘t give a specific tip for it.
What I had, was a little piece solder falling under the crystal, causing bad noise wherever I measured. Luckily I found it relative fast. It happened when I tried the analog board on another‘s digital board and I thought for a moment it was caused by it. But it was the solderpiece under the crystal.
Good luck with Unit A! If I have any further idea, I‘ll come back.

 

[Script]

Took out U36 and periphery for cleaning & re-soldering. Nope, that wasn't it. Unless U36 is fritz'ed.

Had found what looked like tiny solder debris between leads of Ensoniq chip. Nope.

 

[58Glammer]

Took out U36 and periphery for cleaning & re-soldering. Nope, that wasn't it. Unless U36 is fritz'ed.

Had found what looked like tiny solder debris between leads of Ensoniq chip. Nope.

I had U16 fritz‘ed (where does that term come from? ) that is also a 4565 .. I think it was the last ic I exchanged, before my unit worked again. Give it a shot . Hopefully you still have some left . Do you have the clocksignals at U17 and U55? (pin 7,10)
Do you have the same signals on the four polystyrene caps? (needle on one side and kind of scan signal on the other side.) They must all be the same on all four caps (without feeding audio)
Also check all clocks and signals that go to the digi board at J9 .

 

[Script]

Ordered 4565, the other day as I had run out. Haven't really looked at the audio board yet as it was severely damaged. Even stole a 4565 from it for another repair, oups.
___
'fritz', don't know, shredded potato? No, that would be frites (fried).
Fritz = Fredrich (WW I), huh ?
on the fritz (Am. colloq.)= kaputt

 

[58Glammer]

Here is another one …(actually my 4th one ) This time I got it from Ebay for a decent price for myself or maybe to sell later. Description was difficult, so it was a bit of a risk to buy it. „Technically not tested. Unit switches on and all buttons work“
Got the unit and .. surprise.. DC offset error on Ch 2… But NO audiosignal passing through at all Look at the pics and guess (But the seller was pretty cool and gave me a refund. So the price in the end was supernice, if you don‘t have to pay a technician to repair it )
Input section with the typical traces black… Some diodes really look bad.. I will definitely test them when I take them off. Not sure, if they still work..
Look at the traces near powerplug J10
Look at the cap beside C608 and the diode CR75
I also took pics this time of the signals you have on the styrene caps on the scope. Three of them were ok Only Ch 2 is missing the peaks.
Digital board looks pretty good. Just the El caps themselfes and a little bit of the surrounding parts .. but no black traces.

 

[58Glammer]

The inputsection (part 1) was pretty nasty. Krusty krust on some of the diodes… Some of them were stuck on the pcb hard to get soldered off.. One diode‘s leg kept on the pcb ..
I think I will replace all diodes of the inputsection as they look pretty wrecked .. as I have to order them anyways as one broke.
The last pic is the actual status… three pads came of .. one relevant and two irrelevant..
That‘s ok in the end according to how they looked before..
I will try to use only through hole caps this time.
Also thinking of replacing the tantals with through hole caps as well.

Btw: I also reflowed the contact holes from one side of the pcb to the other and it was a pita.. The electrolyte is like a heatshield on the solderjoints…

Hopefully the other sections are less filthy than this one..

 

[Script]

That's by far your worst-looking one.

Contact holes can be "cleaned" by getting the solder completely out first with flux, iron and wick. Refilling then should be a breeze -- if not easy to refill, the contact hole is compromised and needs to be wire-filled.