+24Vdc from +/-15Vdc

i need +24Vdc from a api rack - how is it done properly?

-max

how much current does the rack provide? How much do you need? That will determine if you can do it or not.

i want to power a siemens module, similar to neve class a single ended, current drain ~80mA, and stay within api 500 specs ( +/-16v @ 130mA per rail)

Well, that's a tough one--you'll need to use both rails then, essentially a 30 V-to-24 V DC-DC with floating output.

Samuel

oh man, I'd say good luck!

[quote author="ioaudio"]i want to power a siemens module, similar to neve class a single ended, current drain ~80mA, and stay within api 500 specs ( +/-16v @ 130mA per rail)[/quote]

Is the module transformer-coupled on the input and output?

As long as the unit is floating with transformerbalanced input and output it will work to use the -16 as 0v and the as the B+ but many launchboxes with grappy switching PSU's can't handle the power-up current. There is a current limiting trimmer inside most of them.

thanks for your replies!
so floating is the way to go - keeping the chassis ground seperated - perfect.

hi there. my 2 cents.
surely you would really need to convert only the plus side of the lunchbox up to 24 volts ? somehow.
surely better to keep the ground in the zero in the middle earth?

if you try to make the negative bottom side into a fake ground
then the plus minus power transformer and regulators surely wont know which way the current is flowing.ha ha

if it does work then you gonna need a isolated enclosure. or maybe im crazy.

how did it go.

hi ho,

the chassis ground is seperated from psu ground, see connector
e.g.
http://www.twin-x.com/groupdiy/albums/userpics/meq500_layout.JPG

-max

that (awesome btw) meq500 pcb plugs straight into the plus/gnd/minus.

so i still cant see that your lunchbox will power a 24 volt single end supplied siemens module.

i reckon that chassis - gnd connection choice is only about being able to choose to connect the circuit gnd to the chassis/screen somewhere else ... on the device. like the place closest to the smallest expected signal level . or whatever.

i was really just thinking that the siemens
would go better with its own 24v positive/ground supply.

not to mention the other lunchbox toys who are about to suddenly find their rails being kind of shorted out by mr siemen sucking direct from the negative to the positive. smacks of angry buzzy supply noises to me.

i thort that 'floating' was good to relieve simple signal ground loops.
and that maybe its not so good for trying to correct a complex power supply miss match work around.

i did try a similar thing once and found that there became new sound fidelity issues with all the gears connected to the split-up power supply... the whole bloody room actually. either new ground buzzes. strange grumblings or rf sensitivities. the resulting problems just werent worth the price of a new 24v 200ma supply.
sorry. shoot me. im raving now.

anyone else any experiences with this?

That is exactly how we do it in the MP-500NV. I'll talk to Dan today at work and see if he doesn't mind posting a partial schematic and part number for the converter.

I found the CC6-2412DF-E from TDK/Lambda to do the job I wanted, in a smaller space and less noise than I was able to get rolling my own, so it's what I use in the MP-500NV.

It's available from DigiKey, P/N: 445-2492-ND

It'll take some filtering on the input side to not feed high frequency noise back onto the supply bus, but otherwise they're really quite quiet.

I also used a simple transistor turn-on circuit that keeps the converter off until the rack is near full supply, or it's initial surge will cause the regulators of the rack to current limit.

straight from the horse´s mouth, thanks wes & dan !
-max

So, looks like the TDK/Lambda CC6-2412DF-E DC/DC converter can be wired with floating input, and is fine with either side of the output grounded, to achieve SE PSU voltage, either plus or minus.  Looks useful, if I'm interpreting correctly.  Not an idea I've delved into before.  The addition of turn-on circuit looks essential.    I'm thinking, possibly foolishly, about some -30VDC transformer coupled stuff in a 500 series rack.  May play with it, and blow some stuff up.  Any DC/DC PSU wizards have anything to add? 

Data sheet:

http://us.tdk-lambda.com/lp/ftp/specs/CC6-xF-E_spc_en.pdf

I've used these a couple time.  They do indeed work pretty well.  Just make sure to check the outputs with a scope to make sure you filter any ripple good enough but these are pretty quiet anyway.  Add lots of bulk decoupling because these don't track transients that quickly.

I guessed you might be first responder; thanks.  I'll be sure to look at decoupling/ripple/input filtering.

Gents,
Is this the kind of thing to turn the old girl on at leisure?

peter purpose said:

Is this the kind of thing to turn the old girl on at leisure?

Click to expand...

With that cap to ground instead of to the emitter of Q1 it'll be more of a rude awakening (assuming the cap's empty at turn-on). Also if V1/V2 drop rapidly at turn-off Q1 BE might zener.

Otherwise... maybe. The TDK data sheet doesn't give enough information here (other than that the RC pin may source up to 1mA). The slew rate on the collector of Q1 looks awfully slow, and the CC6 may well not like that. Best case it might work, worst case it could go boom (literally).

A single OC-out comparator with some hysteresis is a safer (but more expensive) bet.

JDB.

Maybe look at a TL431.  It will turn on more sharply.  Not very expensive.