SMPS in parallel for 24V

[delicado]

Hi

I'm trying to power a 24V 10A motor with a 500W ATX PSU by boosting the 12V rail to 24V. Is there anything I should be aware of? I've heard that having two SMPS interacting with each other may not be a good idea because of potential oscillation events, but is there a way to mitigate that? I know I could use a Meanwell instead, but having exposed mains wires isn't a good idea in my current setup.

Thanks

[tooki]

Do you mean just using a DC-DC boost converter? That’s a totally normal thing to do.

[delicado]

Do you mean just using a DC-DC boost converter? That’s a totally normal thing to do.

Yes. AFAIK most DC-DC boost converters are switching regulators. If the PSU and the boost converter aren't synchronised, won't something nasty happen?

[PKTKS]

Hi

I'm trying to power a 24V 10A motor with a 500W ATX PSU by boosting the 12V rail to 24V. Is there anything I should be aware of? I've heard that having two SMPS interacting with each other may not be a good idea because of potential oscillation events, but is there a way to mitigate that? I know I could use a Meanwell instead, but having exposed mains wires isn't a good idea in my current setup.

Thanks

You  SHOULD NOT USE ATX for this.

I usually (daily and for a fairly long time) deal with ATX and
SMPS  PSUs.  Your problem is very typical.

The relevant points are:
- ATX  PSUs are **VERY DIFFERENT** from each other
- none of them are suited to drive MOTORS.
- Although you can use 2 or 3 ATX on "COMPUTERS"...
- THEY MUST BE SEPARATED - one for CPU - one for GPU - one for RAID
- They should also be ISOLATED from the mains using proper equipo

This will ensure safe proper operation

Driving MOTORs is a fairly different problem

SMPS  are sensitive to LOAD (type of load)

You just ca not drive a too inductive or too capacitive load
with some type of SMPS ( example: pure boost or pure buck)
they will probably go wacko oscillating or misbehaving

DO NOT USE A BOOST FOR THIS.

Try to buy a FLYBACK type (ISOLATED TOPOLOGY) with
decent  supervisor (vary among fabs.) capable of driving
inductive loads.

SIMPLE EXAMPLE
https://www.aliexpress.com/item/4000151682100.html

Get a proper one and hope for the better as  some
will not drive highly inductive loads that easy

Paul

[tooki]

Do you mean just using a DC-DC boost converter? That’s a totally normal thing to do.

Yes. AFAIK most DC-DC boost converters are switching regulators. If the PSU and the boost converter aren't synchronised, won't something nasty happen?

I don't think so. That topology is extremely common in modern electronics. In a modern computer power supply, nearly all the power is supplied as 12V, which then uses DC-DC converters right at the load ("local regulation") to drop it down to the various voltages needed. In fact, Intel is pushing a standard (ATX12VO) where the power supply is 12V only, meaning that essentially all components will be using a local DC-DC converter.

Similarly, in every single laptop, you've got an external switch-mode power supply (charger), and in the computer, DC-DC converters produce all the various voltages.

I'm sure there are some types of SMPSs and DC-DC converters that could induce oscillation between them, but I really don't think it's an issue with anything that's even halfway decently designed. And I'd guess that a few extra caps would probably solve the vast majority of them.

You  SHOULD NOT USE ATX for this.

I usually (daily and for a fairly long time) deal with ATX and
SMPS  PSUs.  Your problem is very typical.

The relevant points are:
- ATX  PSUs are **VERY DIFFERENT** from each other
- none of them are suited to drive MOTORS.
- Although you can use 2 or 3 ATX on "COMPUTERS"...
- THEY MUST BE SEPARATED - one for CPU - one for GPU - one for RAID
- They should also be ISOLATED from the mains using proper equipo
[snip]

But the question here is about using an ATX PSU to feed a DC-DC boost converter, in which case it's the output characteristics of the boost converter (and not of the ATX PSU) that need to be happy with the motor load, right??

The question is not about using multiple ATX PSUs in series to produce 24V. (The thread title is indeed misleading, since the question involves neither putting multiple PSUs in series nor parallel, but rather of daisy-chaining a DC-DC converter onto the output of the PSU.)

[PKTKS]

The question is not about using multiple ATX PSUs in series to produce 24V. (The thread title is indeed misleading, since the question involves neither putting multiple PSUs in series nor parallel, but rather of daisy-chaining a DC-DC converter onto the output of the PSU.)

yes I have got the point very well.

I had a problem with  several s-360 and s-540 and s-1000
trying to have some economy and smaller bench footprint
by just adding BOOSTs to them

Example I tried to put a 24/10A boost to a s-360-30A
which typically drives my BENCH GIZMOS for repair.

They do not do soft start when INDUCTIVE LOADS are
"overloading" the inserted boost.

It is safer and cheaper just to buy  a proper s-360-24V
and directly  try to soft start the motor

Paul