Frankensteining a power supply and device together

[sweetbeats]

I have an opportunity to purchase a VGC TimeLine Micro Lynx synchronizer system. The seller is a very trustworthy source. Unbeknownst to either of us the PSU is hosed. The system was a backup system that he purchased a couple years ago and never powered it up.

The Micro Lynx feeds on a three rail PSU, +12, -12 and +5.

The interconnect is via 5-pin DIN plug (like on the old pre PS2 computer keyboards).

I have tested the two ground pins on the stock supply connector and they are common to each other so I know the stock PSU doesn't have isolated grounds for the +5 and the +/-12V rails which, in my mind, is good (more better simpler).

Here we come to the crux: I have sitting here in front of me a standard 250W ATX computer power supply that has almost no hours on it. It supplies +/-12, +/-5 and +3.3V rails. I am assuming it provides relatively clean power considering its intended application though I'm going to scope the rails tonight to have a look. I have tested it and all of its grounds are common as well. Good. My only concern is that the -12V rail is running at just over -11V. The +12V and +5V rails look excellent in terms of the voltage they are supplying. Do I need to be concerned about the -12V rail running low? I realize that question depends on many things about the Micro Lynx, but in a general sense does it sound like a bad idea?

Anything else that sounds like a bad idea to doing this?

I have an old keyboard from which to pirate the DIN plug and umbilical lead and barring any red flags anybody can provide, this might be a really nice economical and quick solution to the issue that comes in a neat fan-cooled steel housing not much bigger than the stock unit...

 

[sweetbeats]

Hm...and I suppose some fueses would be good.

The stock supply is rated at 3A, 1A and 3A for the +12V, -12V and +5V rails respectively.

The ATX supply I'm using is rated at 28A, 0.6A and 32A respectively for the same rails...

I could just use a couple 12V 3A automotive fuses for the 12V rails and maybe a 6V 1A fuse for the 5V supply?

 

[sweetbeats]

I tested two different supplies. One of them seemed really clean. I could see no discernable ripple in any of the three rails when set to DC, 0.2V/DIV, and the X10 switched in. Am I right that that would be 20mV/DIV with the X10 function engaged?

So the only pieces unresolved is how to fuse protect the DC rails to limit current draw over stock specs, and whether or not I'm unwise to utilize the -12V rail when its producing about -11V.

 

[sweetbeats]

Wellllll....

I tried it out. Seems to work.

Still would like input as to:

1. Fuse protection on the output side since the PSU is (for the most part) oversize for the device.
2. Is -11VDC an abnormal/unhealthy product for a -12VDC rail?
3. Am I understanding correctly that when the "X10" function is engaged on my Tek 455 scope, and the V/DIV is set at 0.2, that the vertical divisions represent 20mV?

 

[dgatwood]

Good. My only concern is that the -12V rail is running at just over -11V. The +12V and +5V rails look excellent in terms of the voltage they are supplying. Do I need to be concerned about the -12V rail running low? I realize that question depends on many things about the Micro Lynx, but in a general sense does it sound like a bad idea?

I'd be concerned, yes. That means that the device is drawing a metric crapload of power, and almost certainly way more than it should. Probably a short circuit on the board somewhere---unless you mean that the PSU provides 11V unloaded, in which case somebody probably set a trim pot in the supply incorrectly.

 

[sweetbeats]

Yes, -11V unloaded...I'll have to see if I can find a schematic for the PSU...funny thing is I tested two of them and they were BOTH running -11V on the -12V rail unloaded.

EDIT

I just looked and there are no trimmers in the PSU so...hm.

 

[dementedchord]

check that 12 with a simple resistive load... alotta switching supplies put out essentially nothing when not under load... so only way to be sure is test it again...

 

[sweetbeats]

You were right...

You were right, demented...in fact, even having the PSU connected to the DUT with the DUT still switched off pulled the -12V rail up to (er...down to) -11.5V, and then when I switched the DUT on it pulled down to -11.75V. So all is good, yes? [when I measured -11V before the PSU was disconnected from the DUT]

Also, the rail that had some AC ripple in it before stabilized when the DUT was switched on...in fact there is absolutely no discernable ripple.

The last remaining question is about protection...any suggestions on conventions?

 

[dementedchord]

where did the other 12v rail settle to???

 

[sweetbeats]

Oh, I think it was, like, 12.07V or something close to that...it was 12V dead on unloaded.

 

[sweetbeats]

Can anybody comment on my desire to craft some output protection on this rigged power supply?

Is it necessary?

If so, would it just be a matter of wiring some fuses in between the outputs of the PSU and DUT? (i.e. maybe 1A 12V fuses for the 12V rails and a 1A 6V fuse for the 5V rail)

I apologize if this is electronics 099 but this is new territory for me and I want to protect the DUT.

 

[Blue Jinn]

(OK, I'm a liberal arts guy, not an engineer, or Dammit Jim I'm (not a doctor) but you get the picture...)

Is the concern having the PSU fail and send too much voltage to the MicroLynx? If so, something like a zener diode might be more appropriate rather than a fuse.

http://www.kpsec.freeuk.com/components/diode.htm#zener

I would expect the PSU to be fuse protected on its own, and that a failure of the PSU regulation I don't see would necessarily result in an increased current demand from the MicroLynx that would blow a fuse, so a fuse wouldn't offer any protection to the MicroLynx. I would also assume that an overcurrent demand would cause a thermal shutdown in the PSU regulators. If the concern is that the PSU is rated for much higher current, so that it won't shutdown soon enough? Someone who know more about this though please chime in...

 

[sweetbeats]

If the concern is that the PSU is rated for much higher current, so that it won't shutdown soon enough?

Yeah...that's the concern. The stock PSU is rated at no more than 3A on any of the power rails (1A on the -12V) but the franken supply is rated at up to 32A , so I'm concerned that if there is a failure in the DUT the PSU is going to keep pumping up to a current that will take out a whole lot more of the DUT than would be the case when using the stock supply...the franken supply is indeed fused on the input side but I'm assuming that the fuse rating is matched for the max current ratings of the supply which, again, are much higher than the stock PSU.

This is all foggy new territory for me as well and if there really is no reason to need additional protection I'm good with that...just need assistance seeing how that works.

I do know that the short circuit protection works really good in the franken supply...found that out accidentally......accidentally shoved one of the temporary wires into the wrong socket and *pop* the PSU shut off instantaneously to no apparent harm of the DUT.

 

[evm1024]

A 32 amp supply will happily supply 30 amps all day. 30 x 12 is 360 watts. Something will fry. Over current protection is to protect the supply not the device.

If you want to protect the device you should put proper sized fuses in line with the the supply rails.

-Ethan

 

[sweetbeats]

Ohhhkay. That's what my gut said.

So was I on-track above when I said this?:

The stock supply is rated at 3A, 1A and 3A for the +12V, -12V and +5V rails respectively.

I could just use a couple 12V 3A automotive fuses for the 12V rails and maybe a 6V 1A fuse for the 5V supply?

 

[evm1024]

Ohhhkay. That's what my gut said.

So was I on-track above when I said this?:

Just use standard fuses (250 volt) at 3 amp for the +- 12 v and 1 amp for the 5 volt.

Fuse voltage specs are not too significant in the realm we are talking about so long as you don't exceed the voltage. (no 1000v through a 12v fuse. It will arc over and not be a fuse at all).

--Ethan

 

[sweetbeats]

Right-right-right!

Just went through this trying to figure out the voltage rating for the fuses on my MM-1000 supplies...wasn't thinking.

BUT...

Why 3A on the +/-12V and 1A on the 5V when the stock 3-rail supply is 3A for the +12V and 5V and 1A for the -12V??