SOLVED: Outboard PSU for Alesis 3630 backfeeding -15VDC to AC line

Blue Jinn

Rider of the ARPocalypse
I did the gate disconnect and socketed VCA mod to an Alesis 3630. I decided to just power it from an external PSU and removed all the power supply components and wired +/- 15VDC to a 3 conductor plug to the +15v, -15v, 0v where the regulators would have been. For some reason I thought this would be easier than locating a beefier 9v wall wart and swapping out diodes & filter capacitors (a bit less soldering). The external PSU is just a prebuilt eBay bipolar supply installed in a repurposed PC PSU box. (It does look like the reverse diode over the regulators has been cut out of the circuit though.)

I had some problem and noticed a spark between the chassis and the rack. I measured -15VDC between the chassis of the 3630 (measured from one of the top screws on the 3630 the do tie to ground on the 3630) and the rack, as well as a random AC line measured from a different power strip. Putting the PSU and the 3630 on the bench, the same, I measured -15vdc at the chassis of the power supply and at the hot and neutral of the IEC plug.

I've never seen this before, and would like some insight as to what's causing this and the best remedy. It seemed to go away if the PSU board was left not grounded to the chassis box, but any info appreciated.
 
I did the gate disconnect and socketed VCA mod to an Alesis 3630. I decided to just power it from an external PSU and removed all the power supply components and wired +/- 15VDC to a 3 conductor plug to the +15v, -15v, 0v where the regulators would have been. For some reason I thought this would be easier than locating a beefier 9v wall wart and swapping out diodes & filter capacitors (a bit less soldering). The external PSU is just a prebuilt eBay bipolar supply installed in a repurposed PC PSU box. (It does look like the reverse diode over the regulators has been cut out of the circuit though.)

I had some problem and noticed a spark between the chassis and the rack. I measured -15VDC between the chassis of the 3630 (measured from one of the top screws on the 3630 the do tie to ground on the 3630) and the rack, as well as a random AC line measured from a different power strip. Putting the PSU and the 3630 on the bench, the same, I measured -15vdc at the chassis of the power supply and at the hot and neutral of the IEC plug.

I've never seen this before, and would like some insight as to what's causing this and the best remedy. It seemed to go away if the PSU board was left not grounded to the chassis box, but any info appreciated.
So what 3 conductor connector you are using?
Also, it sounds like you have either connected the -15 incorrectly connected or it is shorted to a ground.
Theoretically, it should not matter if the external power supply is grounded to 0V, however, in this case, it should be an isolated power supply and the only ground connection should be an internal connection of the 3630 and not its chassis or the power supply box. 0v should be connected only at an internal ground point of the Alesis.

The stock wall wart was a 9VAC double insulated transformer they voltage doubled and then derived +/- 15V from it and some of these types of power supplies float the 0V and ground the -15V . I wish I can find the right schematic to confirm this, instead of guessing the Alesis circuit. But a lot of this cheap stuff they made was designed to throw away instead of repairing it. That is why I try not to purchase things until I know I can get schematics for it. So I have very few things made by Alesis and Behringer because of this.

So whats the point of wasting money on an external DC supply?
 
Hi. Thanks for replying. It turned out to be a faulty connector on the power cable. Not sure why that was creating the -15v backfeed, but it works now.

As for the external supply, one of the mods recommends upgrading some of the PSU parts, and the heatsinks on the regulators are a bit dodgy. So it was just as easy to pull all the PSU parts and wire for an external supply as I have some other DIY stuff that has the same power requirements. Interestingly enough I went to post this on GroupDIY and there was another 3630 thread that had the schematics.
 

Attachments

That schematic is about as basic as it gets. For some extra protection I suggest a reverse biased diode across each regulator.
Also the pins of the 7915 are wrongly marked. The 78, left to right is in, grnd, out but the 7915 is GROUND then in then out. I fell foul of assuming they were the same many years ago!

Dave.
 
That schematic is about as basic as it gets. For some extra protection I suggest a reverse biased diode across each regulator.
Also the pins of the 7915 are wrongly marked. The 78, left to right is in, grnd, out but the 7915 is GROUND then in then out. I fell foul of assuming they were the same many years ago!

Dave.

I stripped the internal power supply in the 3630 in favor of outboard supply. It's a generic China board, and was pre-built with stout diodes, good heatsinks, and those diodes. The internal supply on the 3630 seemed dodgy
 
I had one of those. I got it because it had all the features I wanted, but I found that it had a feature I didn't want. There seemed to be no way to make it sound good. Will changing the power supply remedy that? I'm skeptical.
 
I've never seen this before, and would like some insight as to what's causing this and the best remedy. It seemed to go away if the PSU board was left not grounded to the chassis box, but any info appreciated.
what ever you’ve done - you have done it wrong - you are shorting something.
 
I had one of those. I got it because it had all the features I wanted, but I found that it had a feature I didn't want. There seemed to be no way to make it sound good. Will changing the power supply remedy that? I'm skeptical.

The recommended mods upgrade the VCA -although the stock pre-trimmed VCA specs, vs the stock VCA aren't that much better, but using a newer trimmable VCA requires replacing resisters and calibration of the trim. I think you need a distortion analyzer to do that. They also recommend upgrades to power supply components (diodes and capacitors) and a wall wart with a higher current rating. The PSU seemed a weak spot to me and instead of upgrading it on the board, I just decided to take it out of hte box altogether. They also recommend installing a switch for the gate function which I did. I might have a 2180C VCA in there for testing purposes. I ran a drum machine through it (Kawai XD-5) and it made a noticable improvement. It also has the sidechain insert which is a nice feature if I actually had to the time to setup something like that.

I only paid US$30 for it, then consider another $30 for the new VCA's and potentiometers with switches, so it wasn't much of an investment to mod it. I just got the schematic but it looks like a lot like dbx 266.

I have a bad habit of buying cheap gear to fool around with modifications.

I finally invested in a halfway decent de-soldering kit which makes monkeying about a lot easier.
 
Last edited:
Back
Top