What's new
By:
Pinball info

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

SAM IO board keeps blowing the same FET

Fubar

Fubar

Site Supporter
5Years
Joined
Feb 5, 2020
Messages
1,958
Location
Cambridge, UK
I have a SM which suddenly stopped working when I shot Doc Ock: the ball stayed there and after a moment the DMD read: "Operator alert: Doc Ock malfunction".

It turns out F8 was blown. This is the fuse for "50VDC B" which powers everything on J10-P9/10. After poking my DMM probes around the place, I've noticed 2 issues:
  1. Unplugging J10 from the PCB, measuring resistance between the cabinet wires for P9/10 and GND: P9 shows infinite resistance whilst P10 shows only a few ohms, suggesting something is shorting that line.
    • P9 serves #7 #8 #11 and #12
    • P10 serves #1 #2 #4 #9 and #10 -- notably the Doc Ock VUK, which is the coil supposed to fire when that fuse blew -- coincidence?! (Also turns out that coil was only held by a single loose screw -- not sure if relevant.)
      1601416166816.png
  2. Measuring any 50VDC line out of BRDG1 reads 75V. Maybe the bridge rectifier needs replacing?
So my question is .... where do I go from here?

I suspect I might have a fried coil which is causing (1). I could unsolder a coil and see if the fuse blows again. Might be a bit costly in fuses even if I do binary search.

The 150% voltage (2) is more worrying - couldn't it cause all kinds of havoc?
 
Last edited:
If you have a fried coil you will find its taken the MOSFET driver out. Or vice versa.

Most flipper power is 70-75v dc after rectified and smoothing. That is not an issue.

I doubt a transformer would cause double the voltage coming out of one winding. It coule be two 24v windings that across it read 48v ? Not really looked.

Best thing to do is take out the driver PCB and check the mosfets with a meter on ohms. If any of the legs are shorted, change it. Plus change the coil as well that is linked to the mosfet.
 
Thanks a lot for your reply @Pick Holder I didn't realise the voltage would normally be so far off.

Anyway... The bad news is the root cause must be elsewhere... I've put in:
  • New FET (Q4)
  • New coil (23-800 ND)
  • New fuse (F8, 3A)
The machine plays fine for a couple minutes, then the FET short-circuits and the fuse blows. Replacing the fuse, as soon as the machine is switched on the coil fires and stays engaged (logical since the FET is short-circuited).

I think the fault occurs as soon as that coil fires for the first time in play. However I did give it a go in the coil test beforehand (no ball on it though) and that seemed fine. Very weird.

I'm not too sure where to go from here - the PCB looks fairly clean and I don't have another one to swap in for a test. Any ideas?
 
Last edited:
Look at all the coils there will be evidence of charging on the coil wrap and observe when playing what blows the fuse when triggered you could also test each coil individually with your multimeter on ohms test
 
CHRIS B PINBALLS said:
Look at all the coils there will be evidence of charging on the coil wrap and observe when playing what blows the fuse when triggered you could also test each coil individually with your multimeter on ohms test
Click to expand...
It's always the same FET that blows (Q4), and I've replaced the coil linked to that one. (The previous coil read correct ohmage but replaced it anyway.)

Could one of the other coils be responsible for blowing Q4 up?
 
CHRIS B PINBALLS said:
I assume you have replaced the transistor too
Click to expand...
Yes I've replaced both the transistor and the coil. The transistor keeps frying though and I've replaced it 3x now... I'll check the other parts in that power circuit (D4, R41) in case one of those is blown.

CHRIS B PINBALLS said:
Did the coil have a diode on ?
Click to expand...
It's explicitly listed as "No Diode" in the manual. And I replaced it with that exact type (didn't ship with a diode on it).

1601828171147.png
 
no problem this should cure it if not maybe a problem further up the line which would be a lot more difficult to fix but I am confident the diode and fet will fix it
 
Yes if I'm lucky I can source a 1n4004 from work tomorrow, otherwise I've already got some on order from pinparts!
 
Pick Holder said:
I put in 1n4007 on everything.
Do not bother with 1n4004 or 1n4001s.....
Click to expand...
That is good advice. The higher rating of the 1N4007 means it would maybe not have failed and I'd have a working pinball machine!! *withdrawal symptoms intensify*

1601888398505.png
 
Replaced the diode (D4) and the mosfet (Q4). Still no good! :mad:

This time I paid special attention to how the failure occurs though,

Game plays fine so long as I don't put the ball in that specific VUK. As soon as the game kicks it out, the coil stays engaged (i.e. the plunger doesn't fall back down) and after a second or 2 the fuse blows. The mosfet's also dead (for the 4th time... I'm going to lift a trace soon).

I've measured the wire resistance from the IO board to the coil's terminals, they both read about 4 Ohms.

I guess the next suspect up the line has to be U3 (74HCT273)... Maybe one of its outputs stays stuck high?

1601930642277.png
 
Thinking about this a bit more last night... I wonder if the FETs aren't a bit underrated...

The 24NF10 has a Vds of 100V. My "50VDC" is actually more like 75V. I know there is a clamping diode, but that's still only 33% margin.

Anyway I have a 74HCT273 on order, will try replacing that chip next and report.
 
Check that there’s no chance that the coil solder tabs can contact the mounting bracket. If they can, then there’s a chance of applying a short across the transistor when energised.
 
with the mosfet out can you test the relevant output of the 74htc283? a logic probe is useful. but even a led would do the test


Sent from my iPhone using Tapatalk
 
If your using a 24NF10 then that is your problem. Been there, done that. For some reason this doesn't pulse but stays energised a lot longer. You need the equivalent to the 20N10L as listed on the Pinball Heaven shop.
 
pinballmania said:
Check that there’s no chance that the coil solder tabs can contact the mounting bracket. If they can, then there’s a chance of applying a short across the transistor when energised.
Click to expand...
Checked the mech, all looks good to me. Able to move it by hand with no resistance and nothing is coming in contact with the coil solder tabs.

AlanJ said:
with the mosfet out can you test the relevant output of the 74htc283? a logic probe is useful. but even a led would do the test
Click to expand...
I did take a look with my DMM on voltage mode, pins 18/19 (D7/Q7) both read 0V even after actuating the corresponding VUK switch. I didn't replace (and blow) a transistor just for this test though...

Nevin said:
If your using a 24NF10 then that is your problem. Been there, done that. For some reason this doesn't pulse but stays energised a lot longer. You need the equivalent to the 20N10L as listed on the Pinball Heaven shop.
Click to expand...
Thanks @Nevin that may well be my problem... I'm using 24NF10 based on the pinparts.co.uk shop listing it as a replacement, guess it's not. I also see people recommending the IRL540 - going to compare the datasheets of all 3 to try and understand exactly what's happening!
 
Right, so whilst waiting for the supplies from PH, I found a IPP65R074C6 lying around. It's rated for a whooping 650V - hah!

Using that makes my problem go away and my game is finally playable again!! :clap:

I'll still put in an appropriate replacement when I get them (future buyers may not like the above hamfistedness).

@pinballmania I suggest you alter the product page with at least a warning this it isn't a suitable replacement in all cases. https://pinparts.co.uk/products/24nf10-100v-26a-n-channel-mosfet

Any electronics nerds out there who are able to understand this issue btw? I'm really curious but my hobbyist knowledge is insufficient...
 
You must log in or register to reply here.
Back
Top Bottom