09/sep/2026 08/39 - Roland EDIROL V-8 - repair
< BACK
Roland EDIROL V-8
Power input:
9V DC, 2A, center-negative
Video inputs:
8 composite channels
channels 5 through 8 can be s-video
channel 8 can be vga
Video outputs:
2 composite outputs
1 s-video output
1 composite preview(osd) output
Roland EDIROL V-8 Owner's Manual
Roland EDIROL V-8 Service Notes
The issue:
I've been having intermittent problems with it for a long time,
the picture would distort on bright image,
the feedback would trip the TBC causing glitches, dropouts, and stuttering if it got too bright.
The root cause turned out to be that LM2662M ic(Jack1 board, -5volt supply) has failed,
i can assume that it was able to get a negative voltage out only when fully cold.
By the time i decided take the mixer apart completely, the ic has failed fully,
to the point of discoloration of its package, the board, and adjacent capacitors around it.
This ic might've also taken out the power supply due to added strain.
Other regulators upstream from LM2662M before the power supply seem unharmed.
08/sep/2026 22/04 - Roland PSB-1U
Update 22/sep/2026 23/12
After installing a replacement LM2662, it let out ⭑࿔⋆⭒˚⋆｡°✩˙࣪the magic smoke˙⋆✮⋆˚٠⭑ within 10 seconds,
so i decided to do what i should have the first time - check everything downstream on -5volt line.
After checking all 3 Jack boards, i zeroed in on these 2 MA147's on Jack 1,
these are double diodes with a center tap, probably for ESD protection.
Both of these have very low reverse resistance compared to others on Jack boards 2 and 3 - under 2000ohm,
while healthy ones read as basically open.
Update 26/sep/2026 22/32
Replacing both of the MA147's with a close equivalent BAV99 LT1G's resolved the issue.
My guess for why these specific diodes failed is:
- They are on the very top(Jack 1 board), where the heat rises(which there's a lot of).
- They're also always under load, since they are on the 2 main outputs.
Putting these two together the diodes experience a naturally shortened lifespan.
The same type of diodes didn't fail on the vga input, probably only because it wasn't used as much if at all
The BAV99 LT1G comes in a bigger package than MA147(SOT-23 vs S-mini),
but the pads are just close enough to be soldered without weird bodges.
Overall the repair was made difficult only with my lack of experience, not the circuit design or part sourcing.
Taking apart the EDIROL V-8:
Remove 6 screws from the sides and 6 screws from the bottom(ones in the bottom are different shape)
Then the bottom case can be lifted up
- Disconnect all the cables from the main board(there are lots of them)
- Disconnect the wide ribbon cable that goes from Scan Conv to Jack1,
if you can't get to it, be careful not to rip it when removing the Main Board bracket
(if you remove the bracket slowly, it will give you more space, or the ribbon will disconnect itself)
- Remove all the screws that hold in the Main Board/Scan Conv bracket:
-- 4 screws that connect it to the outer casing
-- 3 screws that hold it to the Jack3 board
Do not remove the last 4 screws from the Jack3 board
- Slide the bracket back and guide all the cables through the holes.
Don't try to remove or bend the plastic protectors.
- After the Main Board bracket is removed, undo all the screws from the back panel,
including the vga standoffs, excluding the grounding terminal
Take your time to not miss any, as it can damage some of the jacks.
- After you've removed all the screws, the Jack board assembly should be free,
slide it back guiding all the jacks through their holes
(be aware that bnc jacks have studs on their sides)