VOGONS


First post, by Kahenraz

User metadata
Rank l33t
Rank
l33t

While examining my Roland Sound Canvas SC-55mkII while investigating a broken MIDI socket I spotted what appeared to be some minor corrosion around the DC filter capacitor. I might have ignored it completely had I not seen an episode of Adrian's Digital Basement where he discovered an adhesive that turned corrosive and destroyed part of a PCB he was repairing.

Unlike other adhesives I've seen used to secure large capacitors which remain somewhat pliable, this one had hardened and begun to crack. I don't know if this is a result of the chemical composition or its exposure to high temperatures near the VRM circuitry but it can be identified as an opaque dark brown that is consistent in color throughout.

After removing the capacitor and scraping away the adhesive, I found that the area underneath had become discolored with bits of corrosion just barely visible. While the board still looked serviceable on its face, whatever reaction had taken place had penetrated the lamented surface and begun to dissolve the copper ground plane underneath as well as some nearby traces.

IMG_20220117_190512_resize_8.jpg
Filename
IMG_20220117_190512_resize_8.jpg
File size
264.05 KiB
Views
3868 views
File license
CC-BY-4.0
IMG_20220117_195206_resize_67.jpg
Filename
IMG_20220117_195206_resize_67.jpg
File size
118.5 KiB
Views
3868 views
File license
CC-BY-4.0
IMG_20220117_195226_resize_63.jpg
Filename
IMG_20220117_195226_resize_63.jpg
File size
326.67 KiB
Views
3868 views
File license
CC-BY-4.0
IMG_20220117_195354_resize_56.jpg
Filename
IMG_20220117_195354_resize_56.jpg
File size
480.68 KiB
Views
3868 views
File license
CC-BY-4.0
Last edited by Kahenraz on 2022-01-19, 05:39. Edited 1 time in total.

Reply 1 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

The surface had corroded to such an extent that applying only a minor pressure with the dull tip of a probe would cause it to crumble. To prevent any further decomposition, I scraped away as much of the surface as I could, revealing large gaps where copper had simply disappeared.

While cleaning the area, I could feel my probe fall through the surface where a nearby trace should have been and glide along its path like an empty trench.

IMG_20220117_200908_resize_84.jpg
Filename
IMG_20220117_200908_resize_84.jpg
File size
229.34 KiB
Views
3847 views
File license
CC-BY-4.0
IMG_20220117_205302_resize_68.jpg
Filename
IMG_20220117_205302_resize_68.jpg
File size
300.83 KiB
Views
3847 views
File license
CC-BY-4.0
Last edited by Kahenraz on 2022-01-19, 04:48. Edited 5 times in total.

Reply 2 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I bridged the broken trace using a resistor lead and applied a bodge wire to the trace next to it whose via had been destroyed. I also applied a bit of solder to the exposed copper to provide some minor defense against oxidation before covering the whole surface with a protective layer of Kapton.

I didn't have an exact match for this capacitor and it tested fine in my tester so I cleaned it off and soldered it back into place.

IMG_20220118_141810_resize_10.jpg
Filename
IMG_20220118_141810_resize_10.jpg
File size
297.67 KiB
Views
3838 views
File license
CC-BY-4.0
IMG_20220118_141624_resize_98.jpg
Filename
IMG_20220118_141624_resize_98.jpg
File size
475.36 KiB
Views
3838 views
File license
CC-BY-4.0
IMG_20220118_142854_resize_90.jpg
Filename
IMG_20220118_142854_resize_90.jpg
File size
417.27 KiB
Views
3838 views
File license
CC-BY-4.0
IMG_20220118_145651_resize_69.jpg
Filename
IMG_20220118_145651_resize_69.jpg
File size
228.29 KiB
Views
3838 views
File license
CC-BY-4.0
Last edited by Kahenraz on 2022-01-19, 05:49. Edited 2 times in total.

Reply 3 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

I noticed that there was a similar brown adhesive used to secure the bodge wires found on my mainboard. I decided to scrape all of these away on the off change that it was the same material but I think it was actually something different. Only one place appeared to have some very slight corrosion around a via and scraping away the mask beneath the entire surface of contact did not reveal any corrosion.

I suspect that when the PCB was passed on for rework it was performed at a station with a different bottle of adhesive that was not as corrosive or was something else entirely. The only similarity is that it was a shade of brown, however it was clearly translucent and not opaque like the kind used to secure the capacitor.

IMG_20220117_202452_resize_98.jpg
Filename
IMG_20220117_202452_resize_98.jpg
File size
644.21 KiB
Views
3836 views
File license
CC-BY-4.0
IMG_20220118_141346_resize_70.jpg
Filename
IMG_20220118_141346_resize_70.jpg
File size
454.55 KiB
Views
3836 views
File license
CC-BY-4.0
IMG_20220118_141228_resize_11.jpg
Filename
IMG_20220118_141228_resize_11.jpg
File size
453.93 KiB
Views
3836 views
File license
CC-BY-4.0
IMG_20220118_141329_resize_43.jpg
Filename
IMG_20220118_141329_resize_43.jpg
File size
476.44 KiB
Views
3836 views
File license
CC-BY-4.0
IMG_20220118_141310_resize_88.jpg
Filename
IMG_20220118_141310_resize_88.jpg
File size
448.27 KiB
Views
3836 views
File license
CC-BY-4.0

Reply 4 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

As an addendum to my previous observation, the bodge wire with adhesive which did had some corrosion was in fact opaque. The amount of corrosion was very minor so it's possible that heat plays some role in the volatility of its decomposition.

Notice how there is only as very minor corrosion at the via and the copper beneath the solder mask is intact.

IMG_20220119_000628.jpg
Filename
IMG_20220119_000628.jpg
File size
42.99 KiB
Views
3826 views
File license
CC-BY-4.0
IMG_20220118_141437_resize_72.jpg
Filename
IMG_20220118_141437_resize_72.jpg
File size
524.96 KiB
Views
3826 views
File license
CC-BY-4.0
Last edited by Kahenraz on 2022-01-19, 05:10. Edited 1 time in total.

Reply 5 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Despite the damage caused by the brown plague I found in my device, it was luckily isolated to a very small part of the board that was mostly ground plane with only two small traces being damaged. If this had been a multi-layer board then the damage might have otherwise been severe but this repair was thankfully very straightforward.

My Roland Sound Canvas SC-55mkII has been rescued just in time but I would encourage anyone with any Roland Sound Canvas to inspect their synthesizers to see if they have signs of the same caustic adhesive.

IMG_20220118_200307_resize_17.jpg
Filename
IMG_20220118_200307_resize_17.jpg
File size
1.07 MiB
Views
3821 views
File license
CC-BY-4.0

Reply 6 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

One further addendum. I inspected all of my synthesizers which include the following:

Roland MT-32 (old)
Roland CM-32L
Roland SC-55 (V1.21)
Roland SC-55mkII (V1.00)
Roland SC-88VL (V1.04)
Roland SC-7
Roland SC-55ST
Roland SC-88ST Pro
Yamaha MU-50 (V1.06)

I found an adhesive used to secure large capacitors in the CM-32L, SC-55, SC-55mkII, and SC-88VL, however only my SC-55mkII of which I have only a single specimen exhibited this corrosive behavior. Even the SC-88VL which had a brown adhesive in a similar location with similar thermal characteristics was still pliable and was not corroding the PCB.

Unlike the adhesive in my SC-55mkII which was hard and brittle, all of the other adhesives were still somewhat pliable indicating a different material or composition.

IMG_20220118_211724_resize_77.jpg
Filename
IMG_20220118_211724_resize_77.jpg
File size
414.65 KiB
Views
3813 views
File license
CC-BY-4.0
IMG_20220118_211711_resize_69.jpg
Filename
IMG_20220118_211711_resize_69.jpg
File size
413.27 KiB
Views
3813 views
File license
CC-BY-4.0
IMG_20220119_001545.jpg
Filename
IMG_20220119_001545.jpg
File size
138.06 KiB
Views
3813 views
File license
CC-BY-4.0
IMG_20220118_214649_resize_0.jpg
Filename
IMG_20220118_214649_resize_0.jpg
File size
283.31 KiB
Views
3813 views
File license
CC-BY-4.0
Last edited by Kahenraz on 2022-01-19, 21:53. Edited 1 time in total.

Reply 7 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Here is a link to the Adrian's Digital Basement episode where he found a similar adhesive that had corroded parts of a circuit board he was working on.

https://www.youtube.com/watch?v=mLVbmnViz7I&t=1089s

The adhesive in his video appears to be the same color and consistency as what I saw. The major difference between it and the other adhesives I found elsewhere is that it was hard and dry with an opaque brown color consistent throughout.

Reply 9 of 34, by darry

User metadata
Rank l33t++
Rank
l33t++

I wonder if the capacitor that was glued in the SC-55 might have leaked some electrolyte onto the adhesive and contributed to or even triggered/caused the issue .

Thank you for looking at all those synths .

Reply 10 of 34, by the3dfxdude

User metadata
Rank Oldbie
Rank
Oldbie

Good work.

I also saw Adrian's video yesterday before your post. I got really concerned while watching it since I have an IBM monitor from the same time that I want to open up now. It still works ok but I haven't checked for this problem, and given the timeframe, now is the time to check it.

Also, this is not the first video I've seen of this. There was an EEVBlog video from a couple years ago, that showed brown goop issues. But that didn't clue me as how much it could be a problem with aged devices with IBM especially.

Reply 11 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t
darry wrote on 2022-01-19, 21:26:

I wonder if the capacitor that was glued in the SC-55 might have leaked some electrolyte onto the adhesive and contributed to or even triggered/caused the issue .

The capacitor had not leaked and was in perfect condition. You can confirm from Adrian's video that the adhesive corroded in places where no capacitors were present.

Reply 12 of 34, by Shreddoc

User metadata
Rank Oldbie
Rank
Oldbie

In my SC-55MkII, that adhesive on that capacitor had clearly burned up. Could see the evidence of combustion when I cleaned it up, the carbon residue of scorching. The leftover adhesive was a harder, darker, very brittle remains which chipped off easily.

I agree that the range of special problems exhibited by that particular capacitor('s adhesive) in that particular device is likely heat-related.

Note it's close proximity to a heatsink.

Supporter of PicoGUS, PicoMEM, mt32-pi, WavetablePi, Throttle Blaster, Voltage Blaster, GBS-Control, GP2040-CE, RetroNAS.

Reply 13 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Even if it chips off cleanly, it may have attacked the copper beneath the solder mask. Did the area look corroded at all?

I'm not certain but I would guess that the reason I didn't experience any issues with my unit despite the broken traces is that the damaged area has something to do with the mini-DIN connector which I've never used.

Reply 15 of 34, by Tiido

User metadata
Rank l33t
Rank
l33t

This sort of corrosive glue is found all over in a lot of Japanese made hardware. I see a lot of it in amplifiers, tuners and other stuff that I regularly fix and it has always made something rusty whenever I see it.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 16 of 34, by keropi

User metadata
Rank l33t++
Rank
l33t++

I just checked my MK1 and MK2 SC-55s , glue was there ofcourse but it did not cause any corrosion. I did remove it for good measure - IPA made short work of it 😀

MK1:
NvvCJI9.png

MK2:
c0BJp5D.png

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 17 of 34, by Kahenraz

User metadata
Rank l33t
Rank
l33t
keropi wrote on 2022-01-20, 15:17:

I just checked my MK1 and MK2 SC-55s , glue was there ofcourse but it did not cause any corrosion.

Do you have any before photos? The stuff that Adrian and I encountered was a dark brown with a consistent opaque color. It was also dry and hard.

If what you found was translucent then it was a different adhesive. I found this kind in my SC-88VL as well and it was fine.

What firmware versions are your SC-55 mkIIs?

Reply 18 of 34, by keropi

User metadata
Rank l33t++
Rank
l33t++

I only have one pic, both looked like this - indeed both had the translucent variant

F7tbyPD.png

my MK2 is 1.01

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 19 of 34, by Bondi

User metadata
Rank Oldbie
Rank
Oldbie

Opened my SC-55K - no sign of that adhesive at all. According the dates on the chips it was made in mid 98, much later than OP's device.

PCMCIA Sound Cards chart
archive.org: PCMCIA software, manuals, drivers