METRIX OSCILLOSCOPE DEAD

[tautech]

Slightly off topic, be aware SMPS has been around for a good while and many of the last CRO's used it exclusively. A combination of linear for LV and switching for CRT HV has been used in CRO's since the 70's, maybe even earlier. Those that derive HV from a secondary winding on the mains transformer are very few, only some Hameg's come to mind.
The required CRT HV has always been problematic, only very good design and quality componentry ensures reliable service.
Even though one might think DSO's don't have these HV problems, be aware LCD backlight circuits commonly require 1 kV+ and this is always supplied with switching circuitry, albeit more modern and these days very very reliable.

[daverave56]

i connected up a direct DC 18v supply to the UC 3842 chip and these are the results.

Pin 1 - 0.09v
Pin 2 - 0.06v
Pin 3 - 0v
Pin 4 - 1.91v
Pin 5 - ground
Pin 6 - no osc
Pin 7 - 18v
Pin 8 - 5v

[daverave56]

been looking at a schematic of the UC 3842 chip and it shows a resistor of 56k 1W from the 320v supply to pin 7 which i guess is the startup resistor but i cannot find any 1W resistor around the UC 3842 chip,all resistors are small surface mounted types   

[SeanB]

been looking at a schematic of the UC 3842 chip and it shows a resistor of 56k 1W from the 320v supply to pin 7 which i guess is the startup resistor but i cannot find any 1W resistor around the UC 3842 chip,all resistors are small surface mounted types   

It will be made up from a series connected string of smaller lower value resistors. Trace from pin 7 and you will find the set of typically 4 resistors, and one or more is either open circuit or high in value. Replace all 4 with new ones if that is so. You could also have a broken trace.

[tautech]

i connected up a direct DC 18v supply to the UC 3842 chip and these are the results.

Pin 1 - 0.09v
Pin 2 - 0.06v
Pin 3 - 0v
Pin 4 - 1.91v
Pin 5 - ground
Pin 6 - no osc
Pin 7 - 18v
Pin 8 - 5v

In my Onsemi datasheet, the operating current is listed as less than 20 mA.
Substantially more might will indicate a faulty IC.
But when powered on with an external PSU, also check for an output waveform for confirmation of operation.

[Ketturi]

Was there ever any conclusions about fault? I have the same oscilloscope, with very same fault. It won't get on every time, and I have to pump power switch back and fourth until it starts (and even then sometimes CPU board locks up and won't work properly).
I have changed C112 and C138 to corresponding 50V electrolytics, but that did not help anyway. If it goes to fault mode, US3842 vcc is only 10v and it takes only 1mA power, no output or any life in circuit. When it properly starts up, VCC is 16V with 15mA of current and everything seems to work as meant. Also, if running properly, if I touch error amplifier pin with multimeter PSU stops working and goes back to fault mode. Also, in fault mode, large 400V cap wont discharge, but in normal operation it's voltage drops quickly to around 60V after cutting power.
I have checked every component in primary side, and all resistors and diodes seems to be just fine, no open or short circuits. I'm pretty bummed about this, just can't figure what the problem is.
How can one power supply just be so frustrating  . Should I just replace whole primary side with another switching mode controller...

[tautech]

Was there ever any conclusions about fault? I have the same oscilloscope, with very same fault. It won't get on every time, and I have to pump power switch back and fourth until it starts (and even then sometimes CPU board locks up and won't work properly).
I have changed C112 and C138 to corresponding 50V electrolytics, but that did not help anyway. If it goes to fault mode, US3842 vcc is only 10v and it takes only 1mA power, no output or any life in circuit. When it properly starts up, VCC is 16V with 15mA of current and everything seems to work as meant. Also, if running properly, if I touch error amplifier pin with multimeter PSU stops working and goes back to fault mode. Also, in fault mode, large 400V cap wont discharge, but in normal operation it's voltage drops quickly to around 60V after cutting power.
I have checked every component in primary side, and all resistors and diodes seems to be just fine, no open or short circuits. I'm pretty bummed about this, just can't figure what the problem is.
How can one power supply just be so frustrating  . Should I just replace whole primary side with another switching mode controller...

SMPS can do that to you. LOL

Some can continue to operate with all the primary side perfect and very bad smoothing caps on the secondary, others will give incorrect FB and produce faults like yours.

In a case such as yours I would check all secondary caps are close to spec, that is within 5% of labeled values and better than 100 milli Ohms ESR if you can test ESR.

If you find low capacitance values, replace them all.

[Ketturi]

SMPS can do that to you. LOL

Some can continue to operate with all the primary side perfect and very bad smoothing caps on the secondary, others will give incorrect FB and produce faults like yours.

In a case such as yours I would check all secondary caps are close to spec, that is within 5% of labeled values and better than 100 milli Ohms ESR if you can test ESR.

If you find low capacitance values, replace them all.

Yup, I would do that, but my meter only goes to 22µF and without working oscilloscope I can't measure their ESR. So options are a) replace all capacitors at once b) get capacitance and ESR meters 

[daverave56]

Hi everyone....my scope is still not working...i just decided to put it away in a box for another time...ive just decided to have another go at repairing it and noted many of your past advice posts...that guy with the same metrix scope as me....did you find the problem when i first bought the scope i also had  to keep pumping the power button until the unit fired up but then one day nothing...100% dead

[vk6zgo]

Hi everyone....my scope is still not working...i just decided to put it away in a box for another time...ive just decided to have another go at repairing it and noted many of your past advice posts...that guy with the same metrix scope as me....did you find the problem when i first bought the scope i also had  to keep pumping the power button until the unit fired up but then one day nothing...100% dead

This was a very common problem with domestic TV sets.

A popular circuit used a 220k (or thereabouts) resistor from the incoming Mains to kick the  SMPS into operation.
The resistor would go high,& the TV would refuse to start.

People found out that by "pumping" the switch,they could get it to go,so they would do this until they killed the switch,then would send the set in labelled "faulty switch".

The "noobs" would replace the switch & be astounded when the TV set still "refused duty".

[daverave56]

i will check resistor again...i really want this time to get this beast of a scope working !!! i need a challenge !!! in my photo of the pcb it shows two tant capacitors and i forgot to solder them back in circuit before storing the scope...so not sure of the values or voltage of the caps    any idea anyone 

[vk6zgo]

Googling brought up this:
https://www.google.com.au/?gws_rd=ssl#q=uc3842+power+supply+schematic

Click on "more images" & you will be overwhelmed with 3842 circuits.

You will note that the schematics are almost identical,apart for the few being used for applications other than
AC to DC SMPS.
If you trace through your supply using these schematics,you should find one that is most like it.
You can probably cheat with a few capacitor values.
If you can't remember what they were,use one from these circuits.

I also found:
http://noel.feld.cvut.cz/hw/philips/acrobat/8014.pdf

Note: the 14 pin DIL changes the Vcc pin from 7 to 12.

P.S---I doubt if there would be an external "soft-start" circuit involved with pin 7,as the chip does that internally.

[SAUL BRITTO]

Test the optocoupler! If that marked as Z105 is an optocoupler it can be the culprit of your headpain!!! Replace it.

[oldway]

i connected up a direct DC 18v supply to the UC 3842 chip and these are the results.

Pin 1 - 0.09v
Pin 2 - 0.06v
Pin 3 - 0v
Pin 4 - 1.91v
Pin 5 - ground
Pin 6 - no osc
Pin 7 - 18v
Pin 8 - 5v

In this condition, UC3842 MUST oscillate.
Replace it.

[tautech]

i will check resistor again...i really want this time to get this beast of a scope working !!! i need a challenge !!! in my photo of the pcb it shows two tant capacitors and i forgot to solder them back in circuit before storing the scope...so not sure of the values or voltage of the caps    any idea anyone 

With study of the App Note vk6zgo linked and the Datasheet I linked in reply #16 AND your pics, the caps should be 100uF (C112) and 10uF (C118?) (but I'd use 22uF) and all at least 50V rated and only ELECTROLYTIC.
BTW, meter out both these caps footprints and check against the overlay which is the GND pin. (to be sure, to be sure  )

You've still got more to work though IMO, pics are barely enough detail, there's parts of the circuit we can't see, the switching Mosfet/BJT for one and trying to identify the flyback VDD diode.
Please confirm if the via next to pin 8 is GND. (3842)
Where does pin 5 go?
What component is the TO-92?
As SAUL BRITTO says replace the opto too.

Another pic from a bit further out and lit in the currently dark areas would be a great help.

Try as I might to find a "nice" circuit example similar to what we can see in the OP's images WITH opto feedback, a flyback PSU winding, multi PSU caps and multi outputs, this is the closest I could:

[Ketturi]

This scope is cursed, after using far too much time to repair it, the front panel board gave up, and now all buttons and leds are dead and do nothing, scope just shows channel 1 voltage and timing and trace slowly sweeps outside the image area creating large flare on screen. Probably the MCU in the front panel is shot, and it may be as well mask rom one, so fixing it could be nearly impossible. I gave up and retrieved my old analog kenwood cs-5140 from other side of country (400km and 5 hours at one direction just to get one working scope... ).

[SAUL BRITTO]

Stay calm! 

[SAUL BRITTO]

Ketturi>> Has this scope a 3V battery type coin? If yes, measure it.

[daverave56]

been doing some testing on the metrix scope tonight....tested capacitors again with esr meter but all seem ok...the mosfet is a buk 456....i did a resistance check between the positive of the 400v capacitor and VCC of the UC 3842 and i got a reading of about 600k....so i connected a 16v supply to the UC 3842 and the 5v reference was now present...and a saw tooth waveform on pin 4...but alas no oscillator output from pin 6..i did replace the UC 3842 some time ago when the scope first stopped working so it should not be the chip faulty but not sure now...i will post photos soon as requested.   

[SAUL BRITTO]

And the optocoupler?

[tautech]

been doing some testing on the metrix scope tonight....tested capacitors again with esr meter but all seem ok...the mosfet is a buk 456....i did a resistance check between the positive of the 400v capacitor and VCC of the UC 3842 and i got a reading of about 600k....so i connected a 16v supply to the UC 3842 and the 5v reference was now present...and a saw tooth waveform on pin 4...but alas no oscillator output from pin 6..i did replace the UC 3842 some time ago when the scope first stopped working so it should not be the chip faulty but not sure now...i will post photos soon as requested.   

Seems much too high, I'd expect 150K max. It seems in the pic there is a chain of same value resistors from HV DC+ to UC 3842 VDD (pin 7), have you checked they all are the same and none have gone high?

Good progress, keep it up. 

[daverave56]

still cannot find the fault...this power supply is complecated compared to other schematics ive seem using the same UC 3842 chip...i made a mistake today by not  putting the insulator back around mosfet and tab touched ground while i was working on it and we had some smoke which took out three low value resistors        i nearly thought about binning it but im still persisting although i will need now to get another mosfet and small resistors.

[tautech]

still cannot find the fault...this power supply is complecated compared to other schematics ive seem using the same UC 3842 chip...i made a mistake today by not  putting the insulator back around mosfet and tab touched ground while i was working on it and we had some smoke which took out three low value resistors        i nearly thought about binning it but im still persisting although i will need now to get another mosfet and small resistors.

From your pic (without not being able to meter it out), there'll be a dropper resistor chain from HV DC+ and  an electrolytic for this (IC startup supply), AND then a series diode from the transformer primary flyback winding with a further electrolytic (IC run supply) AND a ceramic SMD cap at pin 7 for local IC decoupling.
Hope that makes sense.

That's how I interpret your pic. 

Having 2 electrolytics for the IC VDD is a little confusing but I'd also like to know what the TO-92 is close to the opto-isolator, is it a BJT or a V-ref?

I'd suggest you make a dim bulb tester, it uses an incandescent bulb in series with the mains to limit the current available in case of  moments.
Google dim bulb tester.

[daverave56]

had a little break from trying to repair oscilloscope...been doing a blue tooth audio project for my metal detector....hopefully will start looking at scope tonight...i would really love to see this scope light up again.

[Agent24]

You can contact Metrix through http://www.chauvin-arnoux.com/en/contact-us/international and they may just help you out with a schematic.. (they did for me, anyway)