S+ Barcrest MPU4 PS missing negative 12 volts

[Harvs]

I have an MPU4 power supply that works as long as I am running LED's in the 'playfield'.. I tried to revert back to bulb and it couldn't handle the load.. Anywho, I simply swapped the MPU4 with a spare MPU5 to get the machine running with the OEM bulbs.. My questions are:

Does anyone have a schematic for the MPU4 Power Supply?

Or, do any of you Barcrest power supply fixer gurus have any advice on this missing the negative 12volt (-12v) line?


Bench testing, it has a faint 'click,click,click........' and my meter will react to that while testing that pin. The supply is clean and no signs of leaking caps or anything fried..

Hopefully someone has some tips or a schematic, these are hard to come by and would like to get it back to running the best machines on earth, the S+ Barcrests!

Thanks! 

[Trisail]

I've worked on quite a few power supplies. With the clicking, I would suggest replacing the electrolytic caps on the 2 small PCB's mounted vertically. These are probably the PWM IC PCB's. The cap in that circuit is critical.

Almost all power supplies can be fixed by replacing the caps. Even if they physically look OK.

If just the -12 V is missing, there may be a 7912 voltage regulator or such on the right hand side of the PCB that could be bad.


Just in case - caps can physically look OK and still be bad. An in circuit tester is the easiest way to test them. But you can replace a lot of caps for the price of the tester.


Be sure to measure the AC ripple voltage when you are checking the voltage outputs.
As a general rule - If the AC voltage is higher than about 20 millivolts (mV), the caps are in need of replacing - some power supplies will have ripple of 50 mV to 100 mV and still be OK. Just depends on the filtering circuits and Amp rating, higher the amp rating the higher the ripple. Most supplies I work on I tend to get them below 10 mV AC ripple. especially without a load. Without a load the AC ripple should be very low, if not replace the caps.

[Harvs]

I've worked on quite a few power supplies. With the clicking, I would suggest replacing the electrolytic caps on the 2 small PCB's mounted vertically. These are probably the PWM IC PCB's. The cap in that circuit is critical.

Almost all power supplies can be fixed by replacing the caps. Even if they physically look OK.

If just the -12 V is missing, there may be a 7912 voltage regulator or such on the right hand side of the PCB that could be bad.


Just in case - caps can physically look OK and still be bad. An in circuit tester is the easiest way to test them. But you can replace a lot of caps for the price of the tester.


Be sure to measure the AC ripple voltage when you are checking the voltage outputs.
As a general rule - If the AC voltage is higher than about 20 millivolts (mV), the caps are in need of replacing - some power supplies will have ripple of 50 mV to 100 mV and still be OK. Just depends on the filtering circuits and Amp rating, higher the amp rating the higher the ripple. Most supplies I work on I tend to get them below 10 mV AC ripple. especially without a load. Without a load the AC ripple should be very low, if not replace the caps.

Great info, I'll dig in shortly. Luckily I have a local place that stocks TONS of caps etc so I should be good there. (I ain'tint no pro though)

Thank you Trisail, will update. 

[Harvs]

Well, guess I didn't look good enough off the bat. Toasty!

[Trisail]

Couple things to look at and replace. Hopes it works.

[Harvs]

Well, after a brutal 2 week wait for some Caps I bought on feebay, the problem remains...

I replaced all caps and spent hours testing measurements of each component (in circuit) on this broken supply vs a working one. Still cant find the issue.

The power supply makes timed clicks coming from the 2 transformers, like its trying to power up but must be sensing a short or something and it shuts off, 'click', back on.. The measurements taken for each line show the voltages fluctuate wildly, the 36v and +12v lines are the most stable and only bounce around about 10%, while the -12v, 5v and 24v literally go from like 0 to a bit over their max V and everywhere in between.

I still have not found a schematic for this supply but did notice that there is an IC near that hot spot on the vert board (V regulator IC?) where the leg of the large resistor popped off the board. I ordered one of them up for sh!ts n giggles.. (Oh, and the working supply has the same hot spot there, but it works fine). I'm into this turd for about $70 in caps so there's no going back, I'll buy and replace every thingy on this if needed! Ha

Just an update. 

 *edit - I did watch some vids to try and narrow down common issues on a PSU and how to test things.. One was Mr Carlson's Lab. After about 20min into one of his vids, I found myself hunched over drooling all over.. - That guy is an Alien, or at the very least a Sorcerer.

[Trisail]

You are right, no schematics out there. There are a lot of different manufacturers that made power supplies like the MPU4. (A LOT) You said you got a replacement working. That's great. I know you would like to repair this one (especially with the money and time invested already). I will suggest letting it go. Usually when you have a supply working and are missing a voltage or have a lot of AC ripple on the outputs the caps usually fix it. The problem you had originally was the -12 V. Now it appears half (or so) of the voltages are 'out of control'. Just by looking at the pictures you supplied I will give you my best shot at what to look at:

What we should know :
1 # -12 VDC missing, but P/S worked with LED's, did not work with filament bulbs. LED's don't use -12 VDC, neither do incandescent bulbs.
2 #  Found one of the Pulse-Width-Modulator PCB's had bad components. The other one appears to be OK, at least no fried components.

What we may be able to ascertain:
From 1#  Because the LED's worked (less current demand) incandescent bulbs did not work (more current demand) the missing -12 VDC was the starting place to let you know the supply was bad.
From 1#   P/S cannot supply rated current to circuit demand as it should. It used to before it died.
From 1#  The voltage that supplies current to the incandescent bulbs is the most probable problem.

From 2#  The ‘good’ PWM PCB may be OK, or at the least better. It may be the one controlling the 36V and +12 V that are acting somewhat better. The voltage that supplies the incandescent bulbs is probably controlled by the PWM PCB that was fried. It may be the one controlling the 5V and the other two voltages 24 V and -12V. Probably not the -12 V as that should be generated by a negative voltage regulator that needs a negative voltage at least 2 volts more negative to operate. There would be a 3-pin device (transistor looking device i.e. 7912) that would supply the negative voltage. This P/S does not have any other negative supply, I believe the -12 V is generated by the +12 V and isolated in such a manner that it is hooked up ‘backwards’. Meaning any positive P/S voltage can become a negative voltage by referencing + 12 V to ground and the ‘ground’ (return reference) to a pin labeled -12 VDC. (research if confusing) This makes the most sense as I believe the current rating of the -12 V on this P/S is 0.15 Amps. I just could not get a clear picture of the ratings on any of the supplies I goggled. (A LOT)

You said you have tried to compare a known good P/S with this bad one and cannot find the obvious problem. That makes sense because the bad one has the fluctuating voltages. The good one has stable voltages.

From reading the below mentioned book, you will learn that all SMPS’s need a ‘start’ voltage applied to the power pin (VCC) of the PWM IC. This comes from the rectified DC from the main AC input rectifier. The large DC (120 VAC = ~ 150 to 160 VDC, 220 VAC = ~ 300 VDC), that voltage is divided down with resistors to the DC level that the PWM IC needs (12 – 36 VDC pin-16, see the datasheet for the Mitsubishi PWM IC M51995) to start. Once it is started the ‘run’ voltage is supplied by one of the secondaries of the Hi frequency transformer. In other words, the circuit must start and run for the output voltages to be present. (usually from a Mosfet transistor that the output of the PWM IC supplies pin-1)

The way the SMPS controls the voltage (current) is from the feedback of one of the primary voltages. Usually +5 VDC with high current rating, or the voltage with the highest current rating. I believe the feedback circuit may be working as you have some semi-stable voltage from one of the PWM PCB’s, but the other PWM PCB is probably ‘not running’ or the feedback circuit is not working.

1st thing to check – recheck the work you did on the bad PWM PCB. I’m sorry, but after the work on the bad one, the P/S is worse than before that work was done. Maybe the capacitor replacement work should be checked also. Look close at the solder connections and make sure nothing is shorted (bridged with solder), if any of the ‘pads’ were lifted (gone or damaged) you have to be sure the original connection from any of these is still intact (making the completed circuit trace). Add wires where needed to complete the original traces. If the trace was damaged where a component is sitting, meaning you cannot see it because the ‘cap’ is covering it, do your best to determine where it should go. You have 2 of the same PWM PCB’s if needed to compare. Trace/point to point. Hopefully you don’t need to do that.

I will say I think the start voltage is getting to both PWM PCB’s. The run voltage is getting to both PWM PCB’s. The feedback voltage is not getting to the bad PWM PCB. You can check this on the M51995 IC if you can get to the pins. The feedback voltage is coming back through the optocouplers on the main board – the 4 IC’s with 6-pins (TCDT1102). You can look up the datasheet and check those to be sure they are OK. (The feedback input comes in pin-1 the output is on pin-5). We can see there are 4 optocouplers all installed the same, as feedback devices, not output devices. Why 4 feedback devices with only 2 PWM IC’s? I will assume one each are for the feedback (pin-8) and one each for the OVP feedback (pin-6). Which I hope is the problem. If not, the device has a pin for ON/OFF (pin-5). That could be a problem.

Note: When checking resistors in circuit with an Ohm meter – they should read a stable Ohm reading, if not, and the reading is fluctuating - especially when you reverse the leads, you are probably reading through a capacitor. The resistor is most likely open. You will need to lift at least one lead from the circuit and check the resistance. I have found some open resistors and fixed bad supplies after doing the caps, to no avail.

I will let you soak all this in. I’m not sure if you wanted a class on SMPS troubleshooting. I am like you. I believe I can fix anything, giving enough time and resources.

You – “I'm into this t**d for about $70 in caps so there's no going back, I'll buy and replace every thingy on this if needed! Ha”

Up to you, the P/S’s are reasonable. I’d let it go, but then again, I can fix anything too. That's why I wrote all this.


Goggle this    Troubleshooting & Repairing Switch Mode Power Supplies   and download the PDF. I used this to help me understand the basics of SMPS. I found it very good, I was able to use the information to trace back from known components on P/S's without schematics. It is on standard computer P/S's, but all SMPS's are fundamentally the same.

[Harvs]

I think I am throwing in the towel on this. I can always have a Pinball collector buddy take a look at it, he's way better when it comes to troubleshooting this stuff.

I tried what I could to find a major difference between the working MPU4 to the nonworking MPU4, nothing major, just a couple readings were like a hair different.. (But even testing new components out of circuit had tiny differences)

Anywho, for now I will just keep using one of my spare MPU5's in the machine. I just hate to toss or forget about stuff that I know can be fixed.. Plus the MPU4's are tough to find. - The MPU5's are way more reliable and it's just a simple pin change to adapt them to the MPU4, I just like keeping stuff OEM if possible.

I really appreciate your help here! Thanks!! - If/when I find out what the issue is on this, I will update this and make fun of myself because I'm sure its like a $1 part causing the chaos.

Thanks again! 

[Trisail]

Cool. Good luck.