NTC help needed for Fluke PM3082

[RNA]

I was desperate for a 4 channel scope but I could only afford a broken one off eBay.  I ended up with a PM3082 which blows its T1.6A fuse.  I have found a shorted transistor and a couple of short diodes in the initial input section.   Every thing after this appears to be ok.   The NTC15 inrush resistor looks like it might have a couple of pin holes on one side.  The service manual says it is rated 2A.  Will I be able to use this one I have lying around? http://docs-europe.electrocomponents.com/webdocs/13c0/0900766b813c0d16.pdf

[RNA]

In a turn for the worse I have just spotted a blown BRY26.  This is a silicone switching package but I can not find one anywhere except'ebay', its really expensive and listed as a thyristor.
Anyone know of an equivilent?

[RNA]

Turns out it is a triggering thyristor.  I bought the one on ebay even though its looking like a bit of a dud.

[pigrew]

In a turn for the worse I have just spotted a blown BRY26.  This is a silicone switching package but I can not find one anywhere except'ebay', its really expensive and listed as a thyristor.
Anyone know of an equivilent?

Such a weird device. I'm not old enough to know the history of them. (By the way, it's BRY62, not BRY26, according to a schematic I found on another website.

I only found one other device that is of the same type, the Bourns TISP83121D, though I'm not sure if its characteristics match well enough. Perhaps with some resistor values tweaked....  It's probably easier to try for old stock of the BRY62.

[SeanB]

You can substitute it with a small NPN and PNP transistor though, wired like a SCR, which is what that is, you can trigger with either gate as desired. Here it is being used as a small signal SCR, so there are plenty of options if you want to substitute it, but just the 2 transistor method will drop in and work.

https://www.digchip.com/datasheets/parts/datasheet/364/BRY62-pdf.php

If that blew then pretty sure the BUW12 is toasted as well, along with most of the other transistors there, along with R1023 and R1024.

[RNA]

Thanks for that, R1023 and R1024 still measure 1R but it took out v1023.  Here is a picture indicating faulty parts so far.It comes with a trigger warning and is not for the faint at heart.  I do not have an isolation transformer so I can only work on this powered down and hope I have caught everything so pointing out any other things I should check is much appreciated. 
Also R1019 which is a 1.96? 100ppm resistor is proving impossible to find.  Do you think a pair parallel 3.9? will suffice?
I have all replaced/ordered all the faulty parts found so far.  Today I will double check the last remaining smd stuff.
Now would be a good time to place bets lol

[RNA]

On the plus side the other smd silicon BAS16 and bzx84-c15 are easily available so I might as well refresh them.

[RNA]

And for those searching for this fault in the future 2 of the 3 BAS16's were open circuit so it is def worth pulling them out.
All the SMD caps measured fine for farads.The only SMD resistor which measured incorrectly was R1017, a 100ohm which had a big hole in it.
The BZX84 managed to ping off into eternity so I did not get to measure it.  Luckily its easy to find new ones.

EDIT: also be aware that the smd parts appear to be glued down for the wave soldering process.  Under V1014 is a trace which could easily be pulled of the board.  Mine had just started to come of the board.

[RNA]

Oh well, after replacing everything it still blows the fuse.  This now has to become a back burner.
(hits fail button lol)
I have multiples of most of the parts so its quite easy to redo when I get back to it.

[Toasty]

See attached picture.

Use a incandescent 120V, 60W lamp in series with the mains input as a current limiter.  If the lamp glows brightly when you apply power, your short remains.
 If it pulses bright and goes dim, then you (likely) have solved the short.

Attached image shows a 100W lamp (0.83A @ 120V).  Better to start with the 60W if you can (0.5A @ 120V).  Use the correct voltage lamp for your mains.

My suspect in this would be the blue 2n2, 2kV cap and that it is shorted. (0.0022µF)

T

[RNA]

The 2n2 2kv still measures fine.  But tonight I will try that set up and see what happens.

[Toasty]

I've seen those caps ohm out okay, but under power (300V) they have a high leakage current.

Are all the components you have replaced still good?

Since I don't see the fuse you have blown on the schematic, I'm guessing that it is prior to this board...?

If you can, lift the output leg of the L1001, the one that connects to C1009, and apply power with the series lamp installed.  You want to make sure the short isn't in the DC supply section.

Can you post clear top bottom views?  Just would like to see the overall layout.

T

[RNA]

I made myself a test jig, however it will have to wait as a couple of parts still blew.
A lot of it survived including (hopefully) the BRY26 in which the diode junction seems ok.
I have to order another BUW12a so I might as well well order all of the high voltage caps in this section.
I measured 3 ohms across the base pad and the emitter pad after I removed the BUW12a.
I also took some pics of the layout but with the high voltage caps removed.

[RNA]

the pics

[Toasty]

I made myself a test jig, however it will have to wait as a couple of parts still blew.

 

Using the tester the parts blew?  Or they were blown from before?

Did the fuse blow also?

The light bulb you're using looks to be a quartz-halogen.  On startup, they pass too much current for this purpose.  You need a "standard" old-style incandescent.

I can't tell from your setup if that is wired correctly.  With the black cord plugged in and the red disconnect closed, does the lamp illuminate when you throw the switch on the white box with nothing plugged into the outlet?  If so, it's not correct.

T

[RNA]

I have not used the tester yet.  It works as you described.  If I plug a table lamp into the socket ,and only then, both lamps light up.  It is very hard to find any proper light bulbs in the uk.  I shall have to dig deeper before I use it.
The fuse blew on the previous attempt to power up.

[RNA]

That wasn't too bad I just ordered a pack of 5 60w incandescents for £1 each of eBay.

[Toasty]

the pics

Thank you for the pics.

The 3 caps around the mains connector are not an issue.  They aren't the problem. They are just X & Y caps for filtering noise & interference and to provide the line triggering.  Had they failed, you'd have a hole in one of them.

T

[RNA]

The 3 ohm between base and emitter pads of the BUW12a seems correct. 

[RNA]

I can see how the 2n2 going leaky would screw things up.  Although measures open circuit and read 2.2nf on my cap meter I am hoping you are correct about the high voltage use of this cap. It would make sense why just this later section gave up.

[Toasty]

While you're waiting for parts, go over the output side semiconductors with the DMM using diode check mode.  Let's be sure there isn't a short on the output that could be causing this.

Also, are you testing this out of the scope, on the bench?

T

[oldway]

See attached picture.

Use a incandescent 120V, 60W lamp in series with the mains input as a current limiter.  If the lamp glows brightly when you apply power, your short remains.
 If it pulses bright and goes dim, then you (likely) have solved the short.

Attached image shows a 100W lamp (0.83A @ 120V).  Better to start with the 60W if you can (0.5A @ 120V).  Use the correct voltage lamp for your mains.

My suspect in this would be the blue 2n2, 2kV cap and that it is shorted. (0.0022µF)

T

This kind of power supply don't work with such a current limiter.
To oscillate, voltage upon R1023 and R1024 must be higher than 1.2V, that's mean that the current must be higher than 2.4A to trigger V1014 (BRY62).

With a current limiter, this value of current will not be reached and the power supply will never oscillate.
The lamp will stay fully on even with a good working power supply.

The lamp current limiter is useless with this kind of power supply.

If you want to test this power supply safely using a 100W 240V current limiting lamp, You should remove R1023 and R1024 and replace them by one 5.6R resistor.
You will be able to check as the power supply is oscilating or not.

NB: with this modification, the power supply will oscilate at more or less 200Khz, that's far too high frequency.

To lower this frequency, change value of C1011 from 47nF to 470nF.
Power delivery with these modifications is very low and it will work as an heavy overloaded flyback converter.

Output voltages will be very low, that's normal.

[Toasty]

Is there any progress on the repair?

T