Mitsubishi CNC switching power supply board dead / relay board bad diode failure

[sam_sam_sam]

I was working on this CNC machine today found no keyboard functioning no control relay powering on
The screen powered on and was giving an operation error but the manual was not very clear about what the error exactly was but with a little bit of troubleshooting and finding out that the control relays not powering on and no keyboard functions we narrowed down to this one switching power supply which of course does not have any indicator LED light they are on the keyboard interface/relay controller board

Found shorted diodes on main controller relays there are 3 of them that are shorted
One note the 3 shorted diodes one for each relay one of the relay that was suppose to power on had the shorted diode the other two are powered from another power supply from somewhere else

Diode test on meter showed 0.010 when checked where as a good one would show 0.500 to 0.556

So my guess is that the shorted diode what caused the switching power supply to fail he was telling that it has been giving trouble powering on a few times before but he would check a few things and all of sudden it would power on like nothing was wrong with it

Luckily he had a spare switching power supply and a relay controller board from another machine that he had an “E” Stop failure of some type even the service technician from Mitsubishi could not find the failure and recommended parting it out it was from the 1970s ( I had also troubleshooted this machine and I could not figure out exactly what the issue was either)

I do have the machine up and running but now I have to fixes the relay control board and figure out what failure is on the switching power supply board the main fuse is good but that is far as I gotten today I look at it tomorrow morning and see what can find out

I will post some pictures of the boards sometime tomorrow

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Usually these diodes are just forming an AND/OR gate. With multiple (logic) diodes shorted, I think there would be a short at one of the switches or wiring, causing an overcurrent through the diodes and also overloading the PSU. I wonder why no fuse is there or popped.

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

I should have been more descriptive about the diodes they are parallel across the relay coil

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

The switching power supply board has quite a few issues
One being that at one capacitor was leaking from the bottom of it possibly the other one right next to it and damage the circuit board traces found one resistor burn up found one SCR shorted it has two voltage regulators one positive and one negative unsure if they got damaged or not

Also found a diode that appears to be shorted markings are MBR1045 meter reading is 0.154 this seems to be a little bit low

All diodes checked in diode mode on meter

MBR1045 - Motorola Schottky Diode B1045 (TO-220) This description was one of EBay seller in Croydon, VIC, Australia

Here are some pictures of the switching power supply board

The UC3842AN is a 30 volt 1 amp current mode PWM controller

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Here is a picture of the relay board and where the three diodes are shorted using the diode mode on the meter

These are the capacitors used and one of them you can clearly see from the bottom that it has been leaking

The person that I repairing this for is going to see if he has a another switching power supply for this machine and if so I going to recap the switching power supply board that we put in yesterday and if he finds another one that one will be recap as well

[petehall347]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

you cant test the flyback diodes in circuit as they will look shorted .

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

you cant test the flyback diodes in circuit as they will look shorted .

The diode was tested after removing it from the board

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Oh, the diodes are across relay coils? Strange failure, they don't get much stress.
On the PSU the Rifa 4700pF have to also be replaced, they crack and short causing a big mess.
Considering the Marcon electrolytics are from 1990, they did OK.

[stj]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Also found a diode that appears to be shorted markings are MBR1045 meter reading is 0.154 this seems to be a little bit low

schottky's will be 150-190 on average.
some meters cant even get a stable reading on them.

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

I was going to say, check for a bad relay as well. Where the coil has shorted to the contacts, insulation failed. That could give high voltage and pop those diodes. It's kind of grasping at straws but there's no reason yet for the three diodes to short.

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

schottky's will be 150-190 on average.
some meters cant even get a stable reading on them.

Okay thanks for the information

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

I was going to say, check for a bad relay as well. Where the coil has shorted to the contacts, insulation failed. That could give high voltage and pop those diodes. It's kind of grasping at straws but there's no reason yet for the three diodes to short.

I checked them out on a tool battery that was 20 volts and seem work correctly

Yes I found it strange that all 3 diodes where shorted but here in Jacksonville Florida power supply company has issues with 3 phase power always running high on voltage for example 480 volts might be as high as 510 volts yea it is not supposed to be this high but sometimes it is and sometimes momentary brown outs as well very short duration which play hell on electronic equipment

In fact this afternoon it happen where the brown out was so short that it caused a CD player to power on and start playing a CD

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Well I just had to recap the spare switching power supply for this machine yesterday because when he went to turn it back on it took several try’s before it came on

So I went to recap 1 of 475uf@25 volt was so bad that it did not even have an ESR resistance value ( <— bloated top vent ) at all another one had a value of 13.0 the last one was 0.05 replacement are 470uf@25 volt

The 470uf@35 volt ones ( 2 of them ) was 0.07 which is a little bit high than a new one replacement 470uf@50 volt

The 47uf@63 volt with a ESR value of 22.0 resistance this is not a misprint replacement 47uf@63 volt

Unfortunately this board also had signs of capacitors leaking which I had to clean up some of the pads and touch of some other corrosion on some other soldering joints on the secondary side of the switching power transformer pins

Install the recap board and now it starts up each time power is applied to it

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Don't forget to replace the Rifa's, they burn up badly.

I think this CNC site has problems with ground and neutral from the distribution transformer sometimes kicking up during a surge and damaging equipment. Those 3 diodes blowing is not normal.

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

One strange thing about this switching power supply is that it pulse ate depending on which combination of terminals are used also on the keyboard/control relay interface board has 3 LEDs flashing when switching power supply is activated

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Well if looks like this was a successful recapping because he power up the machine this morning and it power up again with no issues even though it was about 40*F this morning before I recapped this board is when it had been giving trouble powering on

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

The diodes on the relay turned to not be bad because you have to remove all relays because they are interconnected with other relays I found this out later when I removed all the relays all the diodes testing good

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

That makes more sense. I think the Marcon electrolytic caps are end of life, and the Rifa's you will see their epoxy cracking before they short and burn up.
It's so surprising to see $20 worth of parts give a many $10,000's machine more life.

[sam_sam_sam]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

Marcon electrolyte capacitors are they good brand

Where are these located Rifa's I am not sure exactly what they are

[redwire]

Re: Mitsubishi CNC switching power supply board dead / relay board bad diode failure

The Rifa caps have clear epoxy that cracks and they bulge too. Even new old stock sitting in a parts bin, I had some that got plump and cracked.
That looks like C4, C5 4,700pF Y-caps. I replace them on sight - many incidents where they short and burn because they are on mains. But you have a fuse so any damage would be limited.
They are metallized paper dielectric. You can still buy them as PME271Y series like P271HL472M250 LS=10mm. I'm using Kemet or Vishay instead.

Big X-cap C3 is a different make but these go low value with age, so I would test as well. The board has no MOV's and I think these caps help a little with mains spikes. Their EMI filtering is kind of ignored.