Keithley 2450 SMU, Power supply spark at Turn-ON (fixed)

[HighVoltage]

My Keithley 2450 SMU scared me today, when I turned it on.
I saw a flash and heard a spark at power turn-on, really loud and really bad.

So, I took the instrument apart to look for the problem and it seems like the problem was found right away at the power supply.
Right at the 230V input protection there is a part RT1 and it looks like a varistor / VDR but I am not sure.
This part definitely is broken and needs to be replaced.
It also seems all writing on the part is burned away.

Anyone here having a better idea of what that blown component is and why it would fail in a relatively new instrument (2 years old)
Thanks for any help.

[HighVoltage]

More detailed picture

[HighVoltage]

OK, PSU removed from the instrument.
Similar to Keysight, it seems Keithley is not making their own PSU.

[HighVoltage]

How interesting, this PSU is available from Farnell:

http://de.farnell.com/xp-power/ecs100us12/psu-offen-100w-12v/dp/1738297?mckv=sOUpHKACu_dc|pcrid|88251738663|kword|ecs100us12|match|p|plid|&CMP=KNC-GDE-GEN-SKU-MDC-XP_Power&DM_PersistentCookieCreated=true&gclid=CPa8pPiY99MCFYaT7QoduF8E6Q

[HighVoltage]

Taking a screen shot from the PSU datasheet, the part is market with an NTC 10

[HighVoltage]

Is it possible that this broken part is a EPCOS 10 Ohm NTC Thermistor?

[dr.diesel]

10 Ohm NTC Thermistor?

Yup, I'd say so!

If you end up replacing it I'd suggest testing/loading the PS outside of the 2450, just in case there are other issues with the PS.

[pmcouto]

Yes, it's possible.
10R NTC makes perfect sense in that PSU.

[pmcouto]

NTC usually explodes when there’s a short circuit on the primary side of the SMPS.
I would suggest to check it carefully, starting with the bridge rectifier and switching transistor.
Also, check the fuses – One of them is probably open.

[HighVoltage]

OK, thanks for the confirmation.

I ordered some 10 Ohm NTC and they should arrive in a couple of days.

Interestingly, although there was this explodes NTC, the 2450 still turned on.
But I will inspect the PSU carefully and take some measurements.

[HighVoltage]

The new NTC arrived today.
The old one had about 10 mm diamter
The new one has about 14 mm diameter NTC 10D 13

Although the new one is much larger, it fits well on the PCB.

The rest of the power supply showed no problems.
I could even load it to 8A with my DC load and everything seems perfect.

[HighVoltage]

Now I measured the inrush current on the 230V input side to the PSU

Inrush Current, MAX = 49.6 A
Idle current of the instrument = 718 mA
Current at a 1A DC output of the instrument: 1.3A

I am surprised by the high inrush current, but may be that is normal?

[HighVoltage]

Well, I thought to replace the NTC in my Keithley 2460 SMU but that one looks very differently inside.
Actually it looks like a completely different generation of SMU.

Here are some pictures.

[Gyro]

If you power cycle the unit too quickly the NTC doesn't have time to cool to ambient, so increases the inrush current significantly. I'm not sure if this was the problem with the original NTC but I've had it before when I've cycled SMPSs too often during testing. Do it too many times and I think it becomes a fatigue failure mode.