2
\$\begingroup\$

I am trying to upcycle an old bit of tech I had lying about. The main part I am looking to recycle is the power supply however I do not have specs of the power supply itself. I have measured an output of about 5.5V with a multimeter but what more can be done to calculate/determine, the maximum load of this component.

Below you will see a photograph of the said PSU, the device actually has dual PSU (presumably for redundancy as the two were wired in parallel).

Photograph of PSU which I am trying to find out more about

I have googled any distinguishing marks on the PCB with no avail: this was probably a proprietary board built for this machine.

With regards to components there are your standard capacitors, transformers, fuses and such as well as two components which I have been successful in finding datasheets for: PC123 and 30GWJ2CZ.

I am hoping someone can suggest which components I should look at to determine what the maximum load is or a practical method to test.

In terms of what I would like to achieve I would like to use this to power two raspberry Pis that would fit nicely in the chassis of this device. I am fairly certain there is enough power but I would like to determine more about this PSU before hooking it up and possibly look into ways of protecting the Pis from over voltage/current.

\$\endgroup\$
  • 1
    \$\begingroup\$ You are referred to an SMPS because of the photos. With a look at output capacitor and coil value, we can approximate the output current, if chopping frequency and output ripple is known. \$\endgroup\$ – GR Tech Nov 26 '14 at 13:46
4
\$\begingroup\$

Most modern power supplies are not damaged by loading them more than intended. They will usually go into current limiting, shut down for a while, or at worst, blow a fuse. Some high volume OEM supplies designed for a specific use might blow up, but encountering one of them is unlikely. If you do, oh well, you're not worse off than before since you never actually had a power supply you could reliably use anyway.

Therefore, start at a relatively low load and keep increasing it (meaning drawing higher current or connecting lower resistance). While you do this watch the output on a scope. At some point it will start to sag or the supply will otherwise obviously object, like shutting down for a short while. Since the supply will be designed to the worst case, you will see this limiting behavior at a bit higher load that whatever the rating really is. I'd probably back off 20-25% or so and consider that the maximum rating. For example, if things start acting flaky at 10 A, then consider 7.5 or 8 A the maximum for your purposes.

If the supply has multiple output voltages then it gets more tricky. However, from the looks of it, your supplies have a single output. I'm guessing that each pair of wires (red and black) at the output are actually tied together.

From the general look and size of these supplies and the size of the output wires, I'm guessing around 100 W each.

\$\endgroup\$
  • 1
    \$\begingroup\$ Okay, I will give that a go. The supply does just have the one output and the red and black of the outputs are joined together, again I feel this is for redundancy more than due to the load that would be drawn. Or maybe I am underestimating how much power the original device would have used. If it is around 100W then that would be more than sufficient for my needs. Is there a device that can be connected to safely vary the load without shorting? \$\endgroup\$ – flungo Nov 25 '14 at 21:34
  • 1
    \$\begingroup\$ @flungo: The easiest way is to probably to get a bunch of 3-5 Ohm 10 Watt power resistors and keep connecting more in parallel until the supply objects. Do this on only one of the supplies at a time. In the unlikely case the supply is damaged, you still have the other one with some idea of what load current is safe. I expect the dual supplies are for the extra current capability, not redundancy. They are clearly switchers, which can be designed to load share without too much additional burden. \$\endgroup\$ – Olin Lathrop Nov 25 '14 at 21:48
  • \$\begingroup\$ Just to check, are these the kind of load resistors you would recommend? ebay.co.uk/itm/191197440813 Based on the calculations I have done, with 5 I should be able to get 0.6 Ohm in parrallel which at 6V is effectivly 60W (more than enough for my application). \$\endgroup\$ – flungo Nov 26 '14 at 8:52
  • \$\begingroup\$ @flungo: Those appear to be 3 Ohm 10 W, which is OK for the 5 V I thought your supply put out. They are not OK for 6 V, since each resistor would dissipate 12 W with 6 V applied. If you have 6 V, use 5 Ohm resistors. They would dissipate 7.2 W at 6 V, so you'd need 9 of them to draw over 60 W. \$\endgroup\$ – Olin Lathrop Nov 26 '14 at 13:38
  • \$\begingroup\$ I measured more accurately today and its 5.2V, I did do the similar calculations and noticed that same thing. I will see if I can find 5 Ohm to be safe or a higher wattage rating. However, they are the correct devices that you talk about? \$\endgroup\$ – flungo Nov 26 '14 at 17:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.