Help identifying Chernobyled components on CMM WiFi board

[Marc M.]

I walked into work to find this unidentified board sitting on my bench.  Turns out it came from our fabrication plant and is out of a portable CMM* (a Hexagon Metrology Roamer).  For some reason you can't have it clamped to a welding table while actually welding  .  It's going to cost the company close to $4000.00 USD to replace the board and right now we can't afford to buy hand soap or rags.  I recently saved them $150,000.00 repairing a balancer that all parts are obsolete and unavailable (if anybody needs a new keyboard for a Hofmann Polar Graphics HD balancer control I have 5 more I can build) so I've become the goto guy for big $$$ electronics repair.  Hexagon refused to offer any assistance or information regarding this unit claiming it's unrepairable and must be replaced  .

It appears the TVS (transient voltage suppressor diode) did its job correctly and failed short.  By doing so it converted the 2 inductors L6, L7 to carbon.  Apparently, the design engineer didn't feel installing a fuse would be prudent to this design  .  I have been unable to find anything about this TVS marked with 327.  I don't recognize the odd B looking logo so I can't even limit my search to a specific manufacturer.  Without knowing what the specs. are for this device I'm pretty much dead in the water.  Any help ID'ing this part would be greatly appreciated.  I also need help with an educated guess regarding the inductors.  I'm assuming they're there for noise suppression but again this is a bit out of my sphere of knowledge.  Any input regarding what would be a reasonable value would also be very welcomed.  I've attached pics. - a close-up of the power input section and a very close-up of the mystery VTS diode.  Didn't have a macro lens with me but they didn't turn out too bad.

*CMM: For those who aren't familiar is a Coordinate Measuring Machine used in Metrology labs to very precisely measure objects with some going to the sub-micron level.  They are extensively used in the metalworking industry in QC labs to verify that parts/molds/etc. are machined correctly, incoming QC to verify parts conform to specs., or to digitize incoming parts to determine wear, create 3D models to replicate parts, etc.  Normally they run on huge slabs of granite for dimensional stability but this Roamer is interesting in that it is portable. 

[android]

The B logo could be Bourns.
http://www.bourns.com/ProductLine.aspx?name=tvs_diodes
They have a TVS in an SMA package with a UM marking at least...don't know what the 327  means - probably the series number.

[slashguitar]

Hi Mark,

The B logo on the diodes comes from Bourns.( http://www.bourns.com/data/global/pdfs/SMAJ.pdf ).
You might be able to find the marking in the datasheets of the various series.
 I think L6, L7, C36 and C37 form a common mode choke so I think they can be replaced with any ferrite bead or short.

[Fraser]

You beat me to it.

+1 on Bourns 

If you cannot identify the exact TVS on the Bourns web site I would look at the likely damage voltage of the components being protected by the TVS. That is what sets it activation voltage maxima. If it is protecting a chip input, find the specs for that input from the datasheet.

Inductors are often used for noise filtering of supply rail RF isolation. No harm in experimenting with some values and monitoring effect with an oscilloscope.

Aurora

[dmills]

I am thinking that the transorb is a Bourns part from the logo.
Can you measure the voltage across those power pins? It will make a reasonable guess on the transorb easier.

Regards, Dan.

[mauroh]

Out of curiosity have you tried to remove the TVS and replace the 2 ferrite beads with shorts?

If not, the proper identification of this components will be marginal

Mauro

[Fraser]

Is 'L8' (the square white component) in the failure path ? If so, it could be damaged/degraded by the overload. Worth checking it out thoroughly.

Can you provide the details written on the IC adjacent to 'L8' titled 'M5'. MOSFET switch ? This needs to be checked for transient damage as well. A TVS does not guarantee protection of sensitive components so some investigation of the supply path is a good idea.

I agree that bypassing the two blown inductors and removal of the TVS is a first step to seeing whether the unit will fire up on a current limited power supply.

Aurora

[dmills]

Got it.
The transorb is a  probably a SMAJ15CA (Part labeling is UM).
I would fit SMT fuses or low value fusible resistors in place of those inductors, but that is just me.

The question is what else did the welding current take with it?

Regards, Dan.

[daqq]

You'll be pretty lucky if this is all that croaked - those chokes (L6, L7) can be replaced by most any other similar choke assuming a similar current rating - most of the time they are there just for EMC filtering.
Inspect the board very thoroughly before powering it up though - in the case of a catastrophic failure like this there's a good chance that the nasty traveled all the way through the power supply before vaporizing the inductor.

Also check whatever equipment was powering this unit - the place where the nasty came from.

[Marc M.]

Man, you guys rock!!  Thank you all for your input, it's very much appreciated .

I never would have guessed Bourns since the logo looks like a combination of a capital B and P.  The only Bourns logo I'm familiar with is the one they use on their pots.  That gives me a good starting point.

I found the power requirements are 15vdc @ 3 amps.  On the back side of the board are some terminals for quick disconnects.  Presumably some of the MOSFET's on this board are powering some higher current stuff.  I'm doing this all by remote control since the CMM is at another plant so I'm unable to see what else may connect to this board. 

I know less than nothing about spec'ing TVS diodes.  I suspect there's probably a lot more to it than simply I/R ratings, likely in the time domain, but I'm not sure.  I'll sniff around Bourns and see if I can narrow things down.

Regarding the inductors, I understand that they're there for noise suppression, just not sure of what value to stick in.  I know I can simply bridge them but considering the welding environment this unit is operating in (I'm guessing huge amounts of broad spectrum garbage - aka spark gap transmitters) I'm thinking it would be prudent to replace them with something.  I'm just don't know what value of something would be appropriate.

Is 'L8' (the square white component) in the failure path ? If so, it could be damaged/degraded by the overload....

L8 is indeed in the path, just before the TVS.  My guess was that it's a common mode choke so I simply checked it for continuity thru each leg and that was OK.  I've already removed the TVS to confirm that it was the cause of the short and it was.  I now measure about 100k across the power rail initially, but that climbs up as the caps charge.

...Can you provide the details written on the IC adjacent to 'L8' titled 'M5'. MOSFET switch ?

There are a bunch of them on the board.  I've attached a close-up of the best pic I have of it.  The top line is 4825D, the second an S with a diagonal slash thru it, AA, then a Delta (open Triangle), bottom line is W05C.  I didn't find anything online about this puppy either.   I don't know of any easy way to check them so I simply checked the resistance across the source and drains and didn't find any that appeared to be shorted.

Got it.
The transorb is a  probably a SMAJ15CA (Part labeling is UM).
I would fit SMT fuses or low value fusible resistors in place of those inductors, but that is just me.

I think you are right Dan, I'll take a look at it but considering it's a 15v supply that would sound correct .  Regarding fusing, I have every intention on doing just that.  It really pisses me off when manufacturers intentionally design products to fail catastrophically so they can sell you replacements at big $$$$  .  Absolutely no excuse for not incorporating some sort of fuseable element - oh wait, silly me, they designed in inductors  .

I've already pulled the TVS and soldered a couple of pieces of fusable link wire (that was a lot of fun) in place of the 2 inductors.   I keep a Kikusui PAL 35-10 here at work.  It's not the best for low current work but good enough for a quick test.  I powered up the board and it's only drawing 3mA @ 15v.  There is a spot on the board marked 3.3v and I had nothing there but I suspect that likely comes from the main board via the large board to board connector.  I've attached a picture of the entire board (I didn't have space on the 1st post nor the time to write a second).  Since the board has passed the power-up test I'll send it back to the other plant and have them reinstall it for a quick test of basic functionality.

It's taken me close to 10 hours to get this written due to continuous maintenance calls tonight so I apologize if it's a bit scattered

[lm3baker]

Man, you guys rock!!  Thank you all for your input, it's very much appreciated .


There are a bunch of them on the board.  I've attached a close-up of the best pic I have of it.  The top line is 4825D, the second an S with a diagonal slash thru it, AA, then a Delta (open Triangle), bottom line is W05C.  I didn't find anything online about this puppy either.   I don't know of any easy way to check them so I simply checked the resistance across the source and drains and didn't find any that appeared to be shorted.

I found the power requirements are 15vdc @ 3 amps.  On the back side of the board are some terminals for quick disconnects.  Presumably some of the MOSFET's on this board are powering some higher current stuff.  I'm doing this all by remote control since the CMM is at another plant so I'm unable to see what else may connect to this board. 

This may be the device you are looking at http://www.vishay.com/docs/71291/si4825dy.pdf
However I cannot be certain as the datasheet provides no product marking guide.

It's a 30V, 8A P-Channel MOSFET in an SO8 Package.

Edit: found the part marking guide - looks like I am on the money as the marking on the bottom left looks like the siliconX logo.
http://www.vishay.com/docs/72762/72762.pdf
Bad news: this is an obsolete part.

[Marc M.]

Well folks, they reinstalled the board and it still doesn't work  .  As you suspected, the TVS wasn't able to fully protect everything downstream.  Further troubleshooting will be very difficult since much of this board interacts with the main unit thru the b2b connector.  I'm sure it cost more than my car and one slip of the probe and it's  .

Once again, a very big thank you to all who contributed to this thread. 

Marc -