Capturing Power supply turn on glitch.

[fqahmad66]

Hi,

I have a DS1102E and trying to capture turn on glitch (overshoot) of a home made power supply. What is the proper procedure and scope setting to do this?

Best Regards
F

[Psi]

Setup approx V/div and timebase settings and connect the scope probe to the powersupply.

Set the scope to single shot capture mode  (trigger menu / sweep --> single)

Adjust the trigger level to say 0.5V above ground so it triggers as soon as the voltage rises (the knob labeled level)

Check that the top right button is green indicating it's waiting for a capture event (If it's red press that button to reset for another trigger event)

Turn on the powersupply

The scope should single shot capture the waveform and then stop (red led)

To repeat, turn off the powersupply, press the red button (so it turns green) and turn on the powersupply again

[BravoV]

Once you already used to above methods from Psi with your scope and verified that your PSU is not acting up when turned on, once passed, then suggesting to go further to see if its good enough at transient response using this simple to build circuit -> Dynamic Electronic Load Project designed by Jay_Diddy_B. 

If you passed this 2nd test with dynamic load with a good score without excessive over & under-shoot, then you should be proud of your diy psu. 

[fqahmad66]

Thanks...I will post the results soon.

[neslekkim]

and a video here
have been recommended this, but not had time to watch it yet

[fqahmad66]

just did it...setting 5 volts......overshoot of 8Volts ...stabilize after 180usec. Should never switch on power cycle when connected to a circuit.

[rdl]

So, what causes a power supply to do this?

[Psi]

A good lab psu should have a low voltage on/off button (with it defaulting to off when mains is connected).

If you test a psu without this by turning the mains on you can get weird things happening because the voltage on the control IC starts well below its min volts where its operation is unstable.

If you have a low voltage on/off button you give it time to start up before connecting the output to the output terminals.

[tszaboo]

Yes, it looks like someone forgot to add under-voltage protection and turn-on delay (so called reset circuit) to this PSU. The problem with these home-breed stuff (in general), is that they sacrifice features for BOM cost. And I would like to have features like soft start and not overshoot

[BravoV]

just did it...setting 5 volts......overshoot of 8Volts ...stabilize after 180usec. Should never switch on when connected to circuit.

For sure this PS is not good for chip with "ABSOLUTE" max ratings at 5 Volt.

Not only at power on at the power supply, if it has very poor transient response, a power cycle "at the device" may fry them too.

Btw, don't mean to scare you, some poor design also can fry the circuit during the powering off moment. 

[fqahmad66]

It is a classic three opamp design with 2N3055 at output. I made it about 13 years ago..served me well. I will post pictures,  schematic and turnoff pulse tonight. BTW I am thinking about buying a Hantek PPS2116A Programmable DC Power Supply as a replacement as space and budget is limited.