Epson EB-U42 projector / Caps with 0.7 ohms ?!

[narf]

hi all,

first of all: i'm a (motivated) newbie, so please forgive me any stupid question i may or may not ask

i'm currently repairing a projector: Epson EB-U42 that doesn't start (no lights, nothing).

what i've found so far:
- the PSU does its job and supplies 12V to the mainboard via the connector.
- shortly thereafter, there is a mosfet (TS420-600B) where i can test 12V on the one side (in the data sheet this side is called Anode), Gate 0V, Kathode shorted to ground
- not far away, there's the first cap with a resistance of 0.7 ohms. i've seen many around the processors with 0.7 - 1 ohms.

(i'm uploading a picture where i enter all the findings)

questions:
- are 0.7 ohms normal around the processor? i've heard in some video, that modern graphics cards have a resistance < 1 ohm, processors a few ohms... but are 0.7 ohms normal in a projector?! or does that mean that probably one of the big chips is shorted?
- do you know a magic place where i can find the schematics to this thing? google didn't show me any good results.
- is there even a possibility that the Kathode on the mosfet shorted to ground is "normal"?!
- do you see where the Gate of the mosfet should get its power from? (the pin marked with 0V is the Gate)

thanks for reading so far.

cheers,
narf

[CapLeaker]

Usually this symptom of yours is a dead shorted MLCC capacitor on the main board backside near the connector and it kills one of the two power rails. Find the ground pin on the connector, then unplug the connector from the main board, then check all the pins and compare connected to the main board vs not connected.
As for shorted caps showing resistance, I found them not always being a dead short. A thermal camera or 99% alcohol plus voltage injection will show a shorted cap fast.

[narf]

thanks for the reply, CapLeaker.

i found 2 pins that are connected to ground (pins 5+8 from the left). but i wasn't able to measure the voltage on the connector while it is disconnected. when it's connected, pins 6+7 show 12V. when it's disconnected, i get irratic readings between -1V and +1V on all pins. that's probably some standby mode of the PSU (the internet said). I'm not sure how i can verify if one of the ground pins is actually no ground when i can't measure the connector unplugged.
btw: unplugged, there's no connection between ground and the potential ground pins (or any others) on the connector.

but i will try the alcohol way that's a good info that shorted caps don't need to be 100% shorted.

[CapLeaker]

There has to be a ground pin on that connector. Plug the plug into the main board, but do not power the unit. Then find a ground on the main board and look for continuity on one of the pins.

[narf]

i did that part and found that pins 5+8 are leading to ground. but i can't check if the connector would show 12V on one of those pins if it's not connected. that would be the confirmation that something got shorted.

[CapLeaker]

One of the other pins got to have some voltage on it when plugged into the main board. Other than that, this doesn't make any sense that there is nothing on any pins if it's unplugged from the main board. There has to be some sort of standby voltage somewhere.

[narf]

when the connector is unplugged, i can only measure some very irratic voltage that's jumping in between -1V and +1V. maybe the ground that i'm using for measurement doesn't work as long as the board is not plugged in?! (i tried different screws, metal from the housing and the actual ground wire that later goes to the mainboard.)

ok, well... i've injected some voltage into the board (1,5V max, limited to 2A) and the only thing that got hot was one buck converter IC (Texas Instruments TPS54335 - https://www.ti.com/lit/ds/slusck3/slusck3.pdf). i saw that pin 3 (source of the internal mosfet) had 1 ohms to ground like some of the suspicious caps. so, i'm not sure if this is a symptom or the reason.

i ordered some thermo-active foil that shows areas that get warmer. let's see if that can help a bit as soon as it arrives.

[Diah]

Man, epson will never troubleshoot in this way... sorry... there are pules Freq Mhz need to send to power to send back low voltage then LED indicator will turn on.. so here need Pulse analyzer
you have 12V on the socket when all together connected... this is fine... but if there short as you suspected at caps or any other parts this 12V will dropping down before parts will toast warm.

could you show the power boards where the socket to tell howmany Opto C. installed 3 or 4and the label of them

[narf]

thanks for the reply, diah. as i said: i'm a newbie, but i'm eager to learn

i'm uploading pics of the PSU + a pic of the 5 opto couplers with labels "731 P785F 4GR"

i also saw that a label on the power connector saying 17V (not 12V that i'm getting). i saw this in one of the videos but i think, the guy just changed the PSU altogether.

[Diah]

be careful... just the PSU and Lamp driver boards together with AC wall... check if you have on the main big caps >+350V. and on the Lamp driver boards 5V..

you can see this PSU working with pules signal via RX & TX. , with out analyzer very hard to tell... but let just try..
show the label where written 17V

also check F401 close to the socket.. you need to check all SMT R which are low value if they went high or open

[Diah]

follow this video to know if there caps on MB shorted C3016

[narf]

Thank you for your reply.

I've seen this video already (and many more ).. interestingly enough, this guy only measures 12V at the PSU, too, even though he reads 17V written on the board.

I must have destroyed something during voltage injection or the existing short got worse because the caps i've identified before are now down to 0,3 ohms shorted to ground but from the looks of it, i couldn't find any cap that was exploded or damaged. i desoldered one that looked suspicious but it was ok.
and during that voltage injection, the only thing that gets hot is that tiny IC mentioned earlier (in post #7). pins 1-3 are ground. if i look at the schematic, pin 1 is VIN. so i guess, the chip is only a victim, not the problem. after a while, the big chip got "hot", too. i could see it in the thermo foil (maybe 30°C after a few minutes of injecting 1,5V and 2A max on some coil)

what i've also found that when i just slightly press on the board in some areas, the injected power dropped to zero. is the board mechanically broken and creates a barrier (open line) when it moves apart? or is this a "normal" phenomenon and just my fantasy is limited at the moment?

[Diah]

pushing the surface and yours measurement went out... could be broken at boards or yours DM probe lost contact

i just referred to YT link to see the caps on the downside to have them in check... i do believe over 75% YT videos fake and had cutting.at time

could you tell the history of this projector?

[CapLeaker]

0.3 ohms should draw quite a bit of current. Sometimes the voltage injection method doesn't work well if the defective part has a good “heatsink”. Pushing on the board and there is no current draw suggests a cold solder joint, broken trace / PCB.

[narf]

thanks again for your replies...

ok, i was stupid. the current drop to zero when i pressed the board was because my ground connection during injection is not sitting too well on the board.

unfortunately, i don't know anything about the projector's history. i just bought it to repair

and i checked once more to see if something heats up. i may have found 2 caps where one might get warmer than the other. i'm gonna take this one out the next time i'm at home. but from the looks of it, all caps look ok. there are no explosion marks or cracks on any cap anywhere on the board.

[Diah]

MLCC Caps, get shorted / leaked with out damage of the surface too, so don't give attention to the shape

[CapLeaker]

maybe start depopulating suspects...

[narf]

here's a quick update because i haven't written in a while...

i'm uploading pics (front, back) with all the caps with <= 1 ohms. some i desoldered and re-soldered because they were not faulty and the short was readable on the bare board. and yes, they are looking ugly i might have to check some more videos on hot air solder guns or buy better flux - i dunno.
marked in red are all caps that i've identified. a green marker on the side means, i have desoldered and checked them.

they are basically all around the big chip. do you have any better ideas for me than to test all of those caps seperately? because sooner or later, my soldering skills might destroy something is there any other component i might have to check?

[narf]

update:
i checked all of the marked caps without success. and from all the coils i see, just the one below the big chip marked as L3000 is shorted to ground. so this must be the power rail that's affected. what else can i check before i have to accept that the short is coming from the big chip?

[narf]

ok, seems like i can't edit my posts - so post #3 in a row

i fiiiinally managed to get the buck converter out. really, i have to work on my solder skills. in videos, those kind of chips are out in 5 seconds and i held my gun on this 3x3mm chip for like 10 minutes on 420°C (tried 380 first), swimming in flux, until it went out. board might be dead now, i don't know

buuuut i've got no more short on the caps... so i should've trusted the thermal foil (post #7)

--> Texas Instruments TPS54335 - https://www.ti.com/lit/ds/slusck3/slusck3.pdf

i'm trying to get my hands on one of them and resolder it.