Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Power supply

Status
Not open for further replies.

Flyback

Well-Known Member
Hi,

We are working on a 50W power supply.
We need it to be isolated to 5000V... do you know how?
 
Last edited:
the whole eight of them would share with no other circuitry required…and their feedback loops would not fight each other....what do you say?
There is no mention of that level of paralleling on the DCM datasheet; it mentions paralleling 4 but shows only three. This brings up the First Rule Of Vicor - If it isn't on the datasheet explicitly, it is not supported.

All Vicor generations have problems in large parallel systems. My guess is that at a minimum Vicor will want a diode between each group of DCM's and the common load point.

The last Vicor-based project I did was maybe 10 years ago, but Vicor used to have excellent tech support. For something like this, I wouldn't think of ordering a part until I had sign-off from them on the full system schematic.

There is absolutely nothing "normal" about high-density DC/DC converters, but, even in that universe, Vicor's are weird. My company switched to SynQor for the last multi-brick project I did. They are far more normal than Vicor, but still have ... personalities. But some of them state a max paralleling limit on the datasheet, something I wish all DC/DC manufacturers did.

ak
 
Last edited:
….in this way we would get “droop” sharing,
That's not what droop sharing is. In droop sharing there is no feedback or control from the common sharing point (where all of the individual power source build-out (or droop) resistors sum into the load. It is very much like the paralleled output power transistors in an audio power amplifier, where there is a ballast resistor in series with each emitter. The benefit is that there are no control loop conflicts, but the bad news is that the output sags (droops) with variations in load currenr.

Vicor used to have an app note for this using the 1st gen and 2nd gen bricks, but it did not come up in a search. However, a Vicor app note mentions that the DCM has a "unique built-in droop sharing scheme" without giving any details. Hmmm ...

ak
 
I was working from incorrect memory. I re-read the datasheet and edited my initial response.

ak
 
Since you do not have extra converters in the design for redundancy in case one fails, the multi-converter approach has a lower MTTF than one with fewer switching stages. Note that your Vicor schematic has two isolation stages in series. You'd be better off growing an all-discrete non-isolated buck converter to bring the 800 V down to a regulated-but-non-isolated 48 V, and feed that to the DCMs.

OR - keep the BCM to get certified isolation, and replace the DCMs with a multi-phase buck regulator. IIRC, National and/or Linear Tech have app notes for 100 A designs.

Are there severe size and/or weight restrictions?

ak
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top