Epilog Power supply

[andrew zen]

Was Q9 the failing component?

Q9 looks like a power transistor connected to the heat sink. This things can be expensive. Actually almost everything except for the caps, resistors, and diodes are expensive in this supplies. Can you take a picture of it by itself so we can see how you identified it as the cause?

I think most of us are worried about power supply issues (or should be).

Did the power supply fail during use or did it just not turn on or did it turn on and then you heard a pop and smelled electrical parts burning?

[Graham Symns]

I have included a picture of the two failed devices.Photography is not my strong suit but I think that you can see that the failure mode is quite plain.
On the the left is Q8(the signal transistor) and on the right is Q9(the mosfet device).The part numbers are 2MPS2222 and IRF820 respectively.
Not shown is the resistor R62 ( 5.1kohms 0.6W watt).
As it happens these were available locally at several dollars each..not expensive at all, the mosfets that is.
A few seconds after switch-on the fans in the power supply came on followed by the control panel display lighting up and at this point there was a loud crack and everything died so to speak.
I found that the garage supply circuit breaker had popped.(20Amp.)
At this point I decided that the immediately obvious source of the noise was the 48 volt power supply.
So I started stripping down the supplies ,both of them, and found the above devices failed in both the supplies.
I have yet to test the second unit .The first unit has been installed ,is producing 48Volts (no bangs!) and I am aligning the laser beam optics at present.
Hope that this is of interest.

STA40328.jpg

Regards Graham

[Dan Hintz]

The 2222 is a standard signal switching transistor, probably one of the most common over the last 20-30 years. I use them extensively in my own products, and versions are manufactured by quite a few companies.

The 820 is also manufactured by quite a few companies, though I don't use those nearly as much. Nothing fancy schmancy about it, just a N-channel FET good for a few amps.

Both could be replaced by a multitude of similar devices, though I doubt you'd have any trouble finding these units again.

[James Stokes]

On mine, I had the power supplies rebuilt, installed them and the mother board was blown. I was told that was what took out the power supplies.

[Graham Symns]

Hi James
I have to say that I think that in retrospect I was lucky the the fault seemed to be confined to the power supplies.
As I said before I had never encountered an image fault in twinned equipment.
Sorry about your experience though.
Regards

[Graham Symns]

Andrew,
I re-read your posts particularly the reference to ESR measurement.
I have done a bit more research and have found the Blue ESR meter which had its origins in a Silicon Chip magazine of 2004.
In the text for the Dick Smith kitset there is reference to the fact that that the ESR is at its worst when the capacitors are cold.
My second supply has not been re-tried for the reason that I felt that some capacitors had slighlty domed tops leaving me apprehesive about powering this supply.
I am in the process of acquiring the Blue ESR meter from Australia so as to test them.
I'll keep you posted as to the test results.
I figure it is worth it to me to be able to fix my own gear when I can.
Regards

[Dan Hintz]

Graham,

As with almost all resistances (ignoring NTC items), ESR increases with temperature, so it's at its worst at high temperatures, not low. I'm not familiar with this Dick Smith you mentioned.

If the caps have domed tops, their electrolyte has already boiled at some point... most likely over a period of time, which means those caps are probably pretty dry at this point. Time to replace them. Without the smoothing capabilities of the caps, components further down the line will suffer greatly and eventually pop themselves.

[AL Ursich]

Bad Caps.... http://www.badcaps.net/forum/index.php?

AL

[Graham Symns]

Hi Dan,
Nice to have your input.
The Dick Smith I referred is an Australian icon in the consumer electronics field who marketed the ESR kitset designed by Bob Parker (also an Australian).The Blue ESR meter marketing was subsequently undertaken by Anatek,a Portugese outfit.
It was in the Assembly Manual for the ESR Meter MK2 for the Dick Smith kit Cat No.K7214 that the remark about the "ESR increasing rapidly as the temperature drops" was taken.
I have done some more research regarding this characteristic and the only source I found regarding ESR rising with temperature was from "Radio Electronics.com" which I would be think to be quite authoritative.
To be honest I am now left uncertain as to which is the absolute truth in this matter.
I rather favoured the increase with cooling as this might explain why my supplies, which were in a quite cold(our winter) workshop, failed at switch-on so catastrophically.As I have noted the second supply had only slightly domed caps wth no signs whatsoever of electrolyte leakage.
What I do not know is what happens to the smoothness of the DC supply voltage when the ESR rises.I have never had an oscilloscope conveniently attached to the supply output at the time of failure.(wry smile)
I hope you don't think that I am nit-picking but I really would like to be able to have a complete answer to this whole issue now out of sheer b...y mindedness.

Al,
I have explored the site you referred to but that is not to say that I recall it all.
I'll go back and have another look.

Regards and thanks to you both.

[AL Ursich]

Graham,

I made a ton of overtime money replacing bad caps when working at a Sony Service Center about 9 years ago....

So I always look at the Caps first.

Good Luck,

AL

[Dan Hintz]

I rather favoured the increase with cooling as this might explain why my supplies, which were in a quite cold(our winter) workshop, failed at switch-on so catastrophically.

At switch-on, the caps see a surge in energy... they are empty and look like near-shorts due to their low ESR. You will see a higher failure rate at turn-on, regardless of temperature, similar to the way an incandescent bulb will blow 95% of the time when you first turn it on... the filament is cold, offers a much lower resistance to the incoming current surge, and thermal shock weakens it.

[Graham Symns]

Update on power supply repair.
It turns out finally for me that the ESR situatin was a non starter as I tested the filter capacitors with the ESR meter and it revealed that there was no problem in this regard as the source of the problem.
I finally got round to replacing the signal transistor,the FET and the resistor in the 2nd supply as replaced in the first supply.
While stripping the supply I noticed a melted portion of the insulating layer between the underside of the AC-DC convertor circuit board and the aluminium case.
Tracing the source of heat revealed that a diode had overheated for some reason and reached a temperature sufficient to melt a soldered joint and this allowed the device to protrude into contact with the insulating material.The offending device was hidden from view
The attached pic shows that to get at the device I had to remove a high value capacitor.
This revealed that the failed device was a rectifying diode IN5402.
It had ruptured along its length.
I replaced this with the nearest ,better, rated device IN5406 (1000v 3A)
After reassembling the power supply and installing it in the chassis and the application of power I was rewarded with 2 green LEDs and 48V on the meter.
I hope that this provides some help for anyone else who has a similar problem with a power supply and who needs a bit of encouragement will take something from this exercise.
I am having a bit of bother attaching the pix so I will close for the moment
and send another post with them
Regards

[Graham Symns]

Further to my previous post here is the pix.
There is a diode to the right side of the pwb and its mate is to the left in the area of the exposed copper where the solder mask has been removed to allow for soldering of the replacement device.
The diode to the right was also replaced as it could have been compromised by the failure of the left hand unit.
To the right of the circuit board can just be seen the failed device.

STA40338.jpg

Regards

[andrew zen]

Yowsa!

You have balls of steel for investigating this and doing the work yourself. Of course I am referring to your IKO bearings and their patented "C-lube" rolling guide series using Japanese steel for the bearings.

Please keep us updated as you turn this power supplies on. Maybe wear rubber shoes.

[Graham Symns]

Hi Andrew,
I sure do not have balls of steel anywhere in a personal sense but I sure did have my fingers crossed at switch on,I can tell you!
Having recently retired I am time rich but still have a few things on the go.
A non engraver exercise is building a mobile unit to carry my De Walt thicknesser/planer and for it to double as a worktable/tablesaw extension.
Anyway thanks for your comments..much appreciated.

Regards