Soft start for debugging switch-mode power supplies

[torquil]

Hi!

With traditional linear powers supplies, an incadescent light bulb works nicely in series with the mains in order to debug PSU issues without blowing a lot of fuses/components.

But how do I do something similar with switch-mode power supplies? I've read several comments other places stating that this is not safe in the SMPS case. Is this true? And if so, what can I do to achieve a "soft start" for my SMPS so that I can detect problems without blowing fuses and/or components, other than disconnect it from the circuit that it is powering?

Specifically, I'm trying to fix the SMPS in a Tektronix 2213A. I have another thread on that, but I wanted to ask in general about SMPS (if such a generalization is possible, of course).

Best regards,
Torquil Sørensen, Norway

[fourtytwo42]

The potential problem arises from the large amount of energy stored in the rectifier or other front end reservoirs combined with an undervoltage lockout meaning typically the smps will not attempt to start until destructive energy is already stored. These same components assure the lethality of smps and no investigation should be countenanced by the inexperienced.

[minsik]

In general  SMPS  require an understanding, care and more care.
after 50 years servicing consumer products and high power broadcast transmitters I can say there is experience behind  these personal comments.

When you have experience on particular products, yes, the use of a light bulb can be a very useful service tool. I used a 60W 240V in  mains series to limit inrush current and show any specific over current problems.

Otherwise a variac (and isolation transformer combination - also goes without saying) is a really good service tool to use here. Sometimes a low AC like 40-80V AC will enable control ccts to operate and a very light load to generate output at a less explosive switch on. Unfortunately some designs dont lend themselves to those fault finding technique!

Again experience is the master and  theory a very good 2nd in command. You need both.

Good  luck

Minsik

[CaptDon]

The problem with trying to soft start a SMPS when creeping up from a lower input voltage is that once the undervoltage lockout releases the supply will use a very long duty cycle to make the output voltage correct with a low input voltage and that may stress parts beyond the normal design limit. Also, as mentioned in a previous post the input energy storage bank made up of low E.S.R. capacitors may have enough stored energy to blow things up if the circuit isn't working correctly. Sadly with most SMPS it's Go or Blow, not much middle ground. You need to do a lot of tests with components removed to be individually tested. Most importantly things like zener diodes, rectifier diodes, main switching FET and snubber circuit and the opto-isolator used in the regulation control circuit. It is very difficult to diagnose the transformer. I sometimes make an attempt to remove the main FET and drive the transformer with a sinewave signal of perhaps 50KHz starting at 0 volts of drive and creeping up to maybe 50vpp of drive to see how the secondaries respond or to see it the transformer is drawing seemingly excess current due to an internal short. This is a very specialized technique and is never done with mains voltage applied!!! Also, be aware that if the SMPS has a good fuse and failed to start there is a good chance the energy storage caps are fully charged possibly to around 300VDC and is waiting to give you a nasty shock or blow up your meter / scope!!