10/100 Ethernet Hub Dead

[Solder Boy]

I thought I would come here and pick your collective brains on this power supply from a big internet hub. Main issue is that it is dead won't turn on.

Here are my initial findings:

Main Cap= 168 VDC
Bridge Rectifier 123.5 VAC

Toshiba K2611 DC-DC Converter 168.3 VDC

Transformer - 169.7 VDC

The two Schottkey Diodes behind the transformer no voltage

Q7 No Voltage
Q8 No Voltage
Q2 00.2 VDC

I checked for voltages at the ATX plug found none.

Given that there essentially is no voltage on the output side.. would my gut instinct be correct that the Toshiba DC-DC converter has gone south? It has VDC coming into it but I don't think it's outputting anything.

I've got two of these to repair.

Thanks in advance!

[Krankshaft]

Re: 10/100 Ethernet Hub Dead

When my hub power supply went the PWM IC died replacing it fixed it.

Check the startup resistors on the PWM for an open. If they are okay check for a drive waveform to the base of the switching transistor.

If there is no drive and external components around the PWM look okay including the feedback circuits. Replace the PWM IC.

This is where having an oscilloscope is the difference between gathering the evidence and replacing a part and hoping .

[Solder Boy]

Re: 10/100 Ethernet Hub Dead

When my hub power supply went the PWM IC died replacing it fixed it.

This is where having an oscilloscope is the difference between gathering the evidence and replacing a part and hoping .

I hear ya about the "cilly" scope but it's not something I can afford at the moment. Last I looked even used ones were excessively expensive.. don't do enough repairs to justify the cost.

My next guess would have been the IC as you mentioned.

[momaka]

Re: 10/100 Ethernet Hub Dead

There's an 8-pin IC on the primary side. What part number is written on it?
It's possible that the stat-up cap (likely the one close to the MOSFET on the primary side) could be bad. Or it could be that the auxiliary winding rectifying diode is shorted.

[Agent24]

Re: 10/100 Ethernet Hub Dead

Transformer - 169.7 VDC

Shouldn't that be AC? Transformers don't work on DC...

If you are measuring DC (I assume on the primary side) wouldn't that mean the switcher isn't switching?

[rogfanther]

Re: 10/100 Ethernet Hub Dead

I suppose he measured that in the transformer after the toshiba ic. As it is after the bridge rectifier, there it would appear as DC pulses,

[Solder Boy]

Re: 10/100 Ethernet Hub Dead

There's an 8-pin IC on the primary side. What part number is written on it?
It's possible that the stat-up cap (likely the one close to the MOSFET on the primary side) could be bad. Or it could be that the auxiliary winding rectifying diode is shorted.

I'll get back to you later about the IC part number.. all the caps are good.. checked them all with my ESR and Capacitance meter.

@Agent yes it was DC I measured not AC.. will check for AC and report back.

[momaka]

Re: 10/100 Ethernet Hub Dead

Measuring AC at the transformer primary is pointless. Why? - because the signal is high-frequency square wave and a regular multimeter will not measure it properly. If you have a cheap multimeter, this may even destroy it.

[Agent24]

Re: 10/100 Ethernet Hub Dead

I guess you would need an oscilloscope there instead then?

[momaka]

Re: 10/100 Ethernet Hub Dead

I guess you would need an oscilloscope there instead then?

You could do that, but then you will most definitely need an isolation transformer. Everything on the primary side has a hot ground, so if you touch anything on there with a grounded object, you'll probably get nice fireworks .

Rather than checking AC at the primary, it's usually easier to just measure the voltage going to the PWM IC, then verify from its data sheet whether the voltage you measured meets the IC's threshold.