SUN MGA 1200 EXHAUST GAS ANALYSER NEWBIE DUMPSTER CHALLENGE

[obd.tech]

Hi Folks 
In the true spirit of the great Dave's Dumpster Youtube videos, which brought me to this great forum 
I found a great bit of old 1991 pro kit (regarded by some as probably the best 4 gas analyser made) that was in a skip for the great crusher  and i had to save it.
Being a gentleman  , I gave the owner a beer token and now it's my challenge to get it working again.
He told me 'it booted up last week to try, but when i turned it on yesterday....it went "woof". 
After a 'lets look in side' and blow the dust out session.....nothing nasty visible.....i plugged her in....BANG! FLASH!!    
5amp plug fuse blown....and my reactions tested 
It has a toroidal transformer with a power supply pc board, containing 5 reasonable size electrolytic caps....and this was where the flash came from. As a relative newbie (but keen to learn!) i'd like some advice on whats the best way to test/repair and wake up these Caps(and the ones on the infrared gas bench), that probably have been unused for about 10 to 15 years?
Caps image attached(how do i insert it here, or is it best to attach?)
will post more pics when i get the pcb out and my phone charged.
I have a china ESR meter, solder station etc. i did read that its possible/best to apply the rated cap voltage to old caps via a resistor to wake them from there dormant state, do you do this first before testing them? I dont have a 35v dc source but could link some 12v car batteries if this would work?
Any help/advice welcome 

[German_EE]

I know nothing about this unit but if something is flashing over and going bang then you should be seeing physical damage, where is this damage located? Normally when electrolytic capacitors go bad the cases bulge on the top and/or the bottom and it's easy to spot but from your picture they look OK.

Just take care around the high voltages please.

[obd.tech]

Yes i was expecting visible damage too, maybe 1 cap looks at a raised angle from the others?
I'll wear my wellie boots 
Maybe the 5amp fuse saved some damage?, not sure what the mains plug rating should be, info for this gas analyser was pre internet days and hard to find, tho i did speak with a guy today that was taught at college in Kings Lynn. U.K. by the guy that was the chief engineer on this Tool (sadly, he too is suffering old age  )
I look forward to more replies 

[obd.tech]

Picture of the Gas analyser for people to see attached, and also the other Caps, visible on the side of the heated NDIR gas bench that detects the gas makeup

[poot36]

I would start by replacing the fuse and connecting the unit to a dim bulb tester (a light bulb in series with the plug lead) using a 40W bulb as a start.  If there is a major short in the primary section of the transformer the bulb will glow brightly but the fuse will not blow.  If the unit draws more then 80 or so watts use a light bulb that is roughly half of the units wattage for the test.  If the bulb glows medium or dim then the primary side of the transformer is most likely fine.  Use a multimeter on volts to see if there is any voltage on the various capacitors in the power supply when it is unpluged and if there is not set the meter to ohms mode and check the capacitors for shorts.  If there is voltage across the capacitors then there is probably no short but they could have become leaky at higher voltages (hard to test unless you have a high voltage capable capacitor tester).  I do not think that you will have to worry about refreshing the capacitors due to the unit been from 1991 or so.  I have used a dim bulb tester to power up tube radios from the 40s that the capacitors still worked fine (no reforming necessary)!

[SeanB]

I would probably look for the shorted bridge rectifiers that are now there, if one diode went short or leaky over time it now well have a big hole blown in it.

[obd.tech]

Thanks for the replies guy's.....please remember i'm a newbie at this electronics stuff.....my knowledge and learning time has been spent on all things automotive......i believe that most things can be fixed if you have 4 ingredients. Knowledge, tools, interest and time......thats why i'm here  to borrow some of that from you guys  .......and return some if you have auto questions.

SeanB..I've heard of of bridge rectifier....whats it do/look like?......what happens when it goes wrong?
poot36....not sure what you meant about powering it up? what parts connected? it's got a separate toroidal transformer with wires going out to different boards.....and to the cap pcb one attached.

Any tips on what/how to test the components on this board? i know to discharge the caps first. i have a china esr meter and also the $20 transistor gadget that is followed on this great forum 
What are the 4 gizmos on the heatsink? has a +~- on them. and the 3 legged transistor things on the back of the pcb which are insulated from the case ground (heatsink)with the film and compound.
No visible damage??
Do i test the caps in circuit or off? I also rescued a later version tester which has a similar pcb with the 5 caps on to compare readings/rob parts from?
All help welcome 

[PA0PBZ]

SeanB..I've heard of of bridge rectifier....whats it do/look like?......what happens when it goes wrong?
...
What are the 4 gizmos on the heatsink? has a +~- on them.

You found them  There is BR2, BR4 written on the pcb, that is a giveaway. If you disconnected the transformer use an ohm meter between the 4 pins and see if there is a dead short between any of the 4 pins for each BR, if so you found the problem.

[obd.tech]

BR of course    what an idiot! thanks for that 
no shorts between the 4 pins on all of them, but 2 have readings from one half to the other of say 150kohm and 300kohm depending on which pins you check....but the other 2 'BR's (i'm learning 
have nothing on them.....is that because they are in circuit and affected by it and the caps etc?

[PA0PBZ]

...is that because they are in circuit and affected by it and the caps etc?

Yes, and it also depends on what meter you use. They contain diodes which will start conducting at 0.6Volts.

If the primary fuse blows, is there maybe an AC filter between the power entry and the transformer? Either a boxed one or one built with descrete caps and coils?
The caps in these filters tend to die and can give you the fireworks...

[Rerouter]

From what i can see in this circuit, check the diodes, and those 3 transistors on the back of the board, the transistors on the back check with an ohm meter across all 3 pins, if none are shorted, use the diode check, and from the base you should measure ~0.7V to the other 2 pins (this should only be valid in one polarity, e.g. only 2 pins should conduct like this to any other 2 pins),

Next up for the diodes, use the continuity check across them (In both directions, leaky zeners are best found this way), then a diode check across them, only 1 direction should conduct, just be aware if there is anything else in parrellel to them in circuit you may need to lift 1 leg of the diode to properly measure it,

[obd.tech]

Yes PAOPBZ  m8 there is one of the ali boxed ones, i nearly took a pic of it...because it has a sort of 'bulge' in the top of it. I did read the printed symbols on the box and saw it had some uF cap rating on it, it got me wondering? It also houses a built in fuse, but the mains plug fuse blew.
Does this mains filter box need a load/ how do i test it please?  Can i bypass it?
I don't know, but i'm guessing the less amount of current surges/test runs i give this & the electronics in the analyser the better?

Right i checked the caps on the psu pcb and the readings are here
cap- rating - volts -  ESR2501A         - reading (mastech meter)
C1   10K         25     0.08ohm              8.1mF   (this is 8.1k?? )
C5   10K         25     0.08ohm              8.2mF
C11  10K        25     0.06ohm              8.5mF
C13  3.3K       35     0.13ohm              2.5mF
C17  3.3K       35     0.10ohm              2.9mF

please excuse the cheap test gear.....i spent thousands on auto tools 
Thanks Guy's

[PA0PBZ]

Yes, just bypass the box and see what happens. Could be an easy repair after all 
The caps readings look ok.

[Rerouter]

The readings are likely lower as your tester would not be testing at the frequency the rating is for, even still that small of a drop should not cause explosive bangs,

Also i seriously recommend to you a $20 transistor tester off of ebay, (the one that does diodes caps etc), it has helped me out on so many unique failure modes it isnt funny, and for $20 posted plus a 9V battery it will be as handy as a multimeter (It can measure 0.01 of an ohm if you do the 3 lead calibration (connect all three leads together and press test)

[obd.tech]

Yes Rerouter,
i have one of the $20 transistor testers (posh one in the plastic box, darn the expense) i just don't know how to use it....yet
i'm trying to understand your description on the testing transistors.....base is common with the middle leg. what do i test from there?
gas analyser is at the auto workshop, but i hope to try later without the filter box connected?
can it do any harm.....how can i test the filter box?
Cheers guys

[obd.tech]

Rerouter, i dug out the $20 'super gadget' ...that i had to buy (big boy's toys, you'll understand  )
see picture attached.....am i doing right? if so i'll check them all.
There are 5 transistors i guess to work with the caps.....what do they do?
If you could please simply explain what it's telling 'newbie idiot' me
Thanks Guys

[PA0PBZ]

can it do any harm.....how can i test the filter box?

Just take the filter box out, if it then works it's clear that the box has died.
The base of a transistor is not always the middle leg, and you probably can't test the transistors without taking them out.

[obd.tech]

There's no quick test i can do on the box first? its a bit of a fiddle to get the pcb back in and be nice to see evidence/test failed....rather than 'parts darts' that most garages use as a form of 'diagnostics' 

[PA0PBZ]

Well, you could just connect the box to the mains and leave the output disconnected. Fireworks -> bad box 

[obd.tech]

Good plan 
Thats my logic too.
Looking at the pic again......the 'Box' is above the pcb .......maybe the flash did come from there?
I'm gonna have to drive 30k now 'to have a play' 

p.s. is there any real test that can be done on these filter boxes?

[obd.tech]

  Just tested the other transistor 3 legged things.
first 2 the same, then different readings, then faulty one? then ok one?
please enlighten me oh knowledgable ones .......i have no idea......yet 

Off to the workshop now to ........in Dave's youtube words....... "blow some Shit up!!!" 

[obd.tech]

mmm........test time....removed all the wiring from the filter box, new 5A fuse in mains plug --> On
Nothing 
tested voltage at the filter outputs P-N =230V     N-earth =0v     P-earth =230v
took box out and small burn hole lower side inline with P terminal in photo attached.

Now the question is??......what caused this?
Do i chance by passing the filter box? or is there a fault elsewhere.
Could the suspect transistor be a result/problem or cause?
Ideas please. Cheers guys 

[PA0PBZ]

https://www.eevblog.com/forum/repair/hp3458a-schaffner-mains-line-filter-failure/
https://www.eevblog.com/forum/testgear/just-got-3457a-and-then-boom/

And lots of other examples, so I'm quite sure it's gone. Test the analyser without the filter.

About the transistors: as I said you can't really test them in-circuit.

[SeanB]

Well, the filter had a fault and blew it open. The thing is those are not transistors on the board, but voltage regulators. Note the Ux designation, which says integrated circuit. You cannot really test them with the simple tester, but must note what device it is ( there will be 1 of 4 types in each position) and look up the device to test the input and output with a multimeter with the power applied but with no output connected to the board. You likely have a 5V regulator, a 12V regulator and a +15V and -15V regulator at a minimum there on the board. You will have to undo the screws and carefully look to see what each one is, and write it down so you can see what each one does, then look up the datasheet and then get the info and then use a voltmeter to check it is working correctly.

[PA0PBZ]

The thing is those are not transistors on the board, but voltage regulators. Note the Ux designation

Ah, I missed that!  But there is a good chance that the only fault is the filter, I'd certainly try that before going any further.