Putting two power supplies in series for +/- rails...

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Ego Tripper

Well-known member
Joined
Jun 12, 2010
Messages
61
Location
Lombard, IL
So I'm trying to build a power supply for my Soundcraft Ghost by using two Power One HE24-7.2-A linear power supplies for the positive & negative rails. My first attempt at wiring them into the same enclosure ended in a bang (literally) when I forgot to tie their 0V to chassis ground. So I replaced the fried power transformer in the positive rail and ended up replacing the other PSU since a wire arced to one of the diode pairs while un-mounting it. I started mounting and wiring the new/fixed power supplies in the master chassis, this time with fuse holders installed. So I figured that I'd just double check that the (rearmost,) negative supply was working fine before I mounted and wired the positive supply. Unfortunately, the PSU won't "go negative" with it's positive output tied to chassis ground. The output hovered near ground and the unit produced a disconcerting hum. So I disconnected the positive-to-ground wire (the only output that was connected to anything (aside from my scope)), and ran the power supply as a regular single-ended positive supply & it worked fine, not even humming like earlier. I used a continuity tester to ensure that the power supplies' negative output is not grounded to their chassis, and they're not, so that rules out the most obvious possibility.

So my question then, is whether this is normal or not, i.e. do both power supplies need to be connected in series before the negative supply will "go negative"? Obviously, I could test this theory by simply mounting and wiring the second power supply into the master chassis as the positive rail, but I've already fried enough of these power supplies and I'd prefer not to ruin any more if that experiment were to fail. Or maybe it's something else entirely-- your input is appreciated.
 
I have tied 2 psu's before to get my +/- rails needed on a project.  I used to do that more often as my local electronics junk yard had unused lambda psu's in the voltages I need, brand new in boxes for like 10 bucks.


lets do basic's. On the input side, I will usually do parallel connections from the AC to each PSU, I have never had a problem doing it this wa. Just make sure your transformers are jumpered so that they are set for the correct input voltage.


the output of your power supply will have a V+and a V- per each individual supply. If you run a dmm  set to read DCacross  the V+ and V- pins with RED lead on V+ and black Lead on V- you will get a positive DC voltage reading, if you flip the the test leads you end up with a negative DC Voltage reading.  That is if the supplies are in working order. Before you start to ground one end to give a postive or negative swing, I would make sure they are working first.

Anyway  on one supply if you tie V- to ground on one supply and on the second supply tie V+ to the same common ground, you should have your bipolar voltages you need when referencing to ground.

If you had a negative voltage reading before mounting one of the psu's and then had no voltage, I would suspect that your mounting has caused the psu to either short out or that the output is being grounded to chassis. I would unmount and remove the psu and see if it's still working, if it is then double check how your mounting it making sure it's not grounding to chassis and also double check your wiring scheme.


oh but to answer your question yes you can do it that way. It should work unless there is a problem with one or both of the supplies.



Edit(more info for ya)
if your really concerned

Connect back-biased diodes across the power supply terminals as shown below.
Rate these diodes at the same output current as the power supplies.
In the event both power supplies do not turn on at the same time, or if the load becomes a short circuit, then the diodes will protect the power supplies from any applied reverse voltage.
Do not exceed the output to ground/chassis voltage rating. Inside most power supplies are noise filter capacitors connected from the output to ground. It is possible to exceed the operating voltage of those capacitors, particularly when configuring several units in series.
Avoid using “fold-back style” current limited power supplies as these may lock up the power supply during initial switch on.

Now I never bothered with the diodes but then again I am usually doing low amperage stuff.

 

Attachments

  • 4.jpg
    4.jpg
    8.6 KB
Thanks for the advice pucho. I'm pretty positive that it's not the wiring, 'cause I double checked the hell out of that. I'll take the power supply out of the chassis and test it with it's positive output grounded and ungrounded when I get to the studio tomorrow.

I'll definitely want to implement that diode scheme for my own peace of mind, so thanks for that. It looks similar to what's in the schematic for the Soundcraft PSU, but Soundcraft places their diodes in series with the + & - outputs. Your arrangement looks better as it would give any surges a path to ground, if I'm not mistaken. I have some 1N4007s waiting to go into a guitar amp build-- would that diode work? Looking at it's datasheet (http://www.fairchildsemi.com/ds/1N%2F1N4001.pdf), it looks like "Non-Repetitive Peak Forward Surge Current" is the relevant spec here, no? If it is the appropriate spec, then I'll be fine with it's 30 A rating.

OK, I'm off to drill some holes in an enclosure for a UTC-based DI!
 
While this is commonly done, you need to be certain that the supply outputs are floating. If the low side of both PS are grounded, connecting the high side of the - supply to ground will short circuit it's output.

JR
 
Thanks John, though that was the first thing I checked when these power supplies arrived for this project, and also the first thing I checked after experiencing my problem, it is a good thing to be sure of.
 
It turns out that the sub-chassis wasn't being grounded properly via the mounts alone (even though my continuity tester said it was...) and running a wire from the sub-chassis to the ground bolt's circuit on my terminal block did the trick!

Thanks again for the suggestions!
 
Back
Top