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

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
...

PS: check out Don Solaris' epic rebuild on youtube ().

I just watched the video. Impressive how he managed to do all this work with his soldering iron. Iβ€˜m glad I have this tweezers kind of iron where you can solder/desolder both sides of the SMD parts at once. That makes the work so much easier and faster.
I was sceptical about renewing all of the diodes. But the way he took them off was already destructive. Same with the ICs..
I really liked the latest OS Eprom he burned πŸ‘ŒπŸ» .. and what blew me away was the new display πŸ‘ŒπŸ» . (Does someone have a link? πŸ™‚) Someone must have exchanged the original green/yellow display thru a blue one before. That looked off (beside the dim light) Don replaced it for a new yellow one. πŸ‘ŒπŸ»
He replaced all Elcos with thru hole ones. But Iβ€˜m skeptical about the 3.3 uf between the regulators cause of the heat the regulators produce.. 😬
Originally I wanted to do the same (thru hole), but in the end I went for the SMD ones.(Except the big ones that you can also install thru hole)
The SMDs are easier to solder on and I doubt they are as crappy as they have been in the 90ties..
The unit looked awesome in the end! πŸ‘ŒπŸ»πŸ‘πŸ»πŸ‘πŸ»πŸ‘πŸ»
 
Last edited:
But Iβ€˜m skeptical about the 3.3 uf between the regulators cause of the heat the regulators produce..
+1
Skeptical about that too. Those caps should never have been placed there in the design stage, yet SMDs are lower in profile and as heat tends to go up .. I like the solution on the PSU board.

I too had thought about using regular elcos as in the video for my Unit A, but some of the pads just wouldn't have held them. Also traces/vias in this unit are generally so thin, they offer very little stability.

Any trace/via and board thru-holes turned black (in most cases there's still copper underneath) should be thoroughly cleaned and redone, I think. Question of time.

For wicking, I now use (Chinese) Amtech solder flux/paste (and not sparingly so), which has the added benefit of very strong 'cleaning' properties. That flux alone lifts up almost any kind of dirt and corrosion. And together with wick it's just splendid. -- (But I should get a glass fibre pen too nonetheless.)

Also I use this flux/paste for placing new components. Together with 'no-clean' SMD solder, the PCB is easy to clean afterwards.

That new display (Chinese, I guess) in the video after the recap is breathtaking. Almost too good for a 1994/95 unit 😊
 
Just something that had crossed my mind the very first time I looked at the 22uf cap platoon in the main board input sections.

Probably, to keep production cheaper (that is, more affordable for more people back then) they used same-value caps throughout the board. Makes sense.

But aren't two 22uf caps there each sit back to back , mimicking one 22uf bipolar cap? Or am I missing something here ?
 
+1
Skeptical about that too. Those caps should never have been placed there in the design stage, yet SMDs are lower in profile and as heat tends to go up .. I like the solution on the PSU board.
Thats right.. strange design. And yes Donβ€˜s solution on the PSU board actually is great.
On my last unit the big caps near the small 3.3 caps were lifted a bit (not directly attached to the pcb) That gives the thru hole caps even more space to be placed without touching the big caps.
If I would install these 3.3 thru hole caps on the analogue board between the regulators, I would try to leave the legs a bit longer and would try to angle them to get the caps out of the heat zone.
 
Just something that had crossed my mind the very first time I looked at the 22uf cap platoon in the main board input sections.

Probably, to keep production cheaper (that is, more affordable for more people back then) they used same-value caps throughout the board. Makes sense.

But aren't two 22uf caps there each sit back to back , mimicking one 22uf bipolar cap? Or am I missing something here ?
Good question... Is the value then still the same if you would replace it through a bipolar?
 
No, not the same. My bad. Was thinking HPF.

Placing two caps (of same uf and V) in series, even if back-to-back, halves the capacitance. Voltage rating stays the same.

(A BP of that value might change the HPF cutoff frequency. Not sure the ADC would like it.)
 
No, not the same. My bad. Was thinking HPF.

Placing two caps (of same uf and V) in series, even if back-to-back, halves the capacitance. Voltage rating stays the same.

(A BP of that value might change the HPF cutoff frequency. Not sure the ADC would like it.)
So that ment a typical 10uF/25V (BP (the original are 16V, I prefer to use a higher voltage mostly) would work instead of the two 22/16-back-to-back? The 10/25 or 10/16 are used so often :D
 
Yesterday I was at the new built studio of a friend of mine as he asked me to help him with the cableing.. He has a lot of nice vintage gear and I also saw his DP4+ ... I told him about the problem of this model and he told me, itβ€˜s in a great working condition.. I told him this unit has to be recapped no matter what. We took a look inside and it looked good indeed (without glasses) I showed him some black spots on the traces and told him more. When I left, he gave me the unit for a recap and later on my bench.. guess what..
View attachment IMG_8817.MOV

I sent him the video and he told me, he’s wondering as he hasnβ€˜t used the unit for years.. I told him β€žThe silent dead of this unit..β€œ
So I wonder if this unit will be another challenge or now a routine?
I showed him the difference how it should look like (outputsection without elcos) :
AEDD3C00-9C18-4D79-93FE-6296656362A0.jpeg
And how it should not (inputsection with leaked caps) 😬:
9E793ED3-EACD-42AB-8985-B2B63FCD0694.jpeg

I will post about the progress. Nothing gets through none of the channels ..
I start to clean and decontaminate the inputsection as even measurements are a waste of time here πŸ˜…
 
Well done. PCB does not look healthy. Very clear signs of leakage. Please report back.

Think about getting the Amtech flux from AliX. It makes process faster.

Can't wait to get back to my units myself but 'out and about' for couple of weeks..Maybe even longer...
 
Well done. PCB does not look healthy. Very clear signs of leakage. Please report back.

Think about getting the Amtech flux from AliX. It makes process faster.

Can't wait to get back to my units myself but 'out and about' for couple of weeks..Maybe even longer...
Yes, they look bad.. the typical spots/rails ...
Already took off all the parts in the inputsection before the opamps except some resistors I will go for in a minute..
And I wanted to write anyways.. Thanks for the tip about the cleaning effect of the flux.. got some from RS and it works pretty nice on the contaminated pads πŸ‘ŒπŸ»
FF0715A1-0D11-4D10-B280-BA175A731CC8.jpeg
The diodes looked bad in this unit underneath.. they were sticky πŸ˜…
This is a pic before the one above..

75B14CC1-083B-4CD4-88F3-601A261B9E53.jpeg
Also the connection dots to the underside of the pcb that look dull can be cleaned with the flux very nicely.. still not easy, but it works.. need to complete this with the rest of the dull dots..
I will post the proceding when I have cleaned and put the first batch of parts back on the board. ✌🏻😎 (I have to wait a few days for the Elcos, so be patient πŸ˜‰)
 
After taking off all parts in the inputsection and nice cleaning, now everything is back on the board except the elcos that Iβ€˜m waiting for to be delivered.
1D8C3A33-D4A9-47DB-8B6E-E57F75FF3F8C.jpeg
I also tried this this time... I installed normal elcos at the regulators to get them away from heat..

7FF97D00-8D71-4A30-9516-64627E279F41.jpeg
 
--- (Have an unexpected free day off at home. Time to get all caps off of Unit B and assess damage underneath.) ---

Audio board of Unit B looked overall pretty good. Some slightly corroded PCB side-to-side 'eye' (no longer shiny). But NO apparent cap leakage and, more importantly, NO corroded vias. Yet, leakage underneath caps nonetheless....

I removed all caps by grabbing them with pliers and gently turning them till the leads snap off. This is by far the fastest method, but it is not recommended when traces/vias are visibly compromised (i.e. already turned black), as this might lead to ripped off pads and vias !! --- On the other hand, if a pad comes off this way, it was sure compromised already and so needed attention anyway. Wouldn't want too many pads ripped off though.


RESULT
- The 100uf caps on the audio board show little, if any, signs of leakage. However, one 100uf cap sure had leaked and made adjacent solder joints look dull/corroded.
- The 22uf caps (all in audio path, I think) are mostly clean (still). Maybe there are first signs of leakage or it's flux residue from factory assembly (though I wouldn't bet on that).
- BUT ALL 3.3uf caps have partially leaked without exception (heat area or not). Those are the worst on my Unit B. Same on digital board (wet leakage) and same on front assembly board (wet drops).


WORKING THEORY
The 3.3uf caps are the first ones to leak (at least on this Unit B). The 3.3uf caps sit on the power rails.
The 100uf are the next to leak. The 22uf ones (all in the audio path, I think) might be the last to go.


AND SOME PICTURES

3.3uf caps leaked.jpg
Leakage underneath 3.3uf


100uf cap leaked__eyes compromised.jpg
Compromised 'eyes' -- despite no obvious leakage. Components near dull-looking solder joints ('eyes') better be retouched and all SMD components better be removed to clean underneath.


Cap near power transistor__pad ripped off.jpg
3.3uf cap near power transistor heat area clearly leaked. Pad came off in removal (twisting with pliers). Similar to a compromised pad of one 3.3uf on the front assembly board.


Cap near power transistors.jpg
Another 3.3uf in power transistor heat area


Leakage_1.jpg
Another leakage. area. See crusty line at bottom left. Adjacent components better be removed to clean underneath. Hot air station makes this a fast process.
 
Last edited:
!! NOTE TO ALL DP4+ OWNERS !!

Your unit might be still fully functional. But cap leakage is almost guaranteed, even if not visible. Better get down to fixing it now. The sooner the better, cos it will be faster -- once leakage and damage advance, it can turn into a lot of work, or even mean unit exitus.
 
Last edited:
Iβ€˜m back at the unit of a friends studio. After some measurements I again ended in the area where these polystyrene caps sit... I think two of them are faulty..(ch 1+2) From there the Sony IC gets a kind of rectangle signal .. normaly it looks nicely on the scope.. If it looks kind of gibberish.. something is wrong there and you get the DC offset error diplayed.
On the last unit I had two broken lines at these PS caps and the Error displayed.
This time no broken trace..
I already ordered new caps.
As these polystyrene caps are not easy to get anymore...
Can I just replace them with Wima caps? It needs 1500 pF min50V ..

The other thing I wanted to say again... It makes no sense to just recap the unit if you can see some dull solderjoints.. these areas have to be cleaned..
The area in the middle is like hell as you got these pots sitting there..
To get them out is a play of patience πŸ₯΅ Not to rip solderjoints is a challenge... This time I got the first four out without damage 😯.. Now Iβ€˜m going to clean up that section. If you look closely you can see a black spot on the trace underneath Pot 1.. screw you C608 for spilling 🀬 πŸ˜…
CABB11A5-12DA-4730-AE10-71B8169649B1.jpeg
 
I've just looked at my DP4+ that I have in storage, I have wanted to open it up and check the PCB since I saw this thread.

I looked at it and almost opened it up, then I was afraid of what I might see inside, so I just ran away

:ROFLMAO:
:ROFLMAO:
 
Last edited:
+1
Spillage areas better be cleaned thoroughly. Black traces better be scraped clean and restored with layer of solder. Solder in fully dull and corroded "eyes" better be replaced.

Those trim pots have bent leads for clipping in. If taken out once, check thoroughly ( might even consider replacing).

I am keeping track of the time I am spending on unit B over here. It's Herculean for sure. Once there is cap spillage, fixing these units is not a business model...
 
I've just looked at my DP4+ that I have in storage, I have wanted to open it up and check the PCB since I saw this thread.

I looked at it and almost opened it up, then I was afraid of what I might see inside, so I just ran away

:ROFLMAO:
:ROFLMAO:
πŸ˜‚πŸ˜‚πŸ˜‚ 😬 πŸ˜‚πŸ˜‚πŸ˜‚
If it works when you switch it on, you might have an easy job, just replacing the caps.. If it shows DC offset in the startup and some of the meters explode to max, I wish you a lot of fun restoring it πŸ˜¬πŸ˜…πŸ˜¬
I wish for you it just runs without issues 🀞🏻
 
Itβ€˜s been a while since I reported from the last unit I had on the table.. It finally workes. πŸ₯³πŸŽŠ After I found out the digital board works fine, I tried to encircle the error on the analogue board. I had no outputsignal from any output. And something was wrong in the area where the polystyrene caps are. In the end it was the U15 TL072 and U16 4565 Opamp...
E4F692AF-E8F0-4FE6-9144-4EC0CB94B57C.jpeg
0D13C7A0-B92C-4C6F-8769-657A0A2E35A8.jpeg
Then I found two other TL 072 Opamps defect sitting after the big U2 TDA1541 ... U6 and U7 sitting directly at the four elcos C45/46 C49/50 that might have destroyed the Opamps with electrolyte ..
5E49850A-B172-466A-BE02-7FD5EF7423E5.jpeg
4FE15903-59C9-48A7-A06F-DF0A087FCAEF.jpeg
I cleaned up all those sections before I found out they were defect.
Also cleaned up the area where the AD DA ics sit. There is a elco C238 as well. If this area is contaminated from C238 it might influence the digital signals that go to and come from the digital board over J9.
ED17C1D2-6CCC-4D8C-8E48-806324AC51E9.jpeg
Just adjusted the digital distortion and now I will run the described selftest.
This unit was not easy. It canβ€˜t be repeated often enough.. Exchange the elcos before they leak.. easy job.. after that.. horror show...
✌🏻😎
 
Last edited:
Back
Top