EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Atreidae on September 03, 2013, 11:45:11 am
-
Wow. Ive just had a browse around and Im so noob, that I make the newbies look experienced.
Anyway, Ill run down my electronics skills. I'm an avid fan of Dave's youtube channel and it has re-invigorated my love for electronics.
My skills are not much more than what you would expect of a high school student, I know what caps.diodes and most common semi conductors are and I can follow along with most of Daves content.
Ive got a SMPS from a printer that died 13 months into service (yeah, just out of warranty) and its decided to blow the internals of a dipped cap all over the main filter cap. My guess would be a tantulum, but they are blue (well.. this one is ka-boom coloured) and I thought they were orange. (a google suggests 2,200pf ceramics http://www.made-in-china.com/showroom/dgdihao/product-detailRepngYCDsqWm/China-Ceramic-Capacitor-222M-AC250V-.html (http://www.made-in-china.com/showroom/dgdihao/product-detailRepngYCDsqWm/China-Ceramic-Capacitor-222M-AC250V-.html) )
Anyway, I'm tempted to just replace the cap and see how it all goes.. but something must have caused the cap to blow.. its not a minor failure, yes there is a big power resistor nearby and you can see the heat damage to the board
The cap in question is between the source on a 3678 Power FET and a tap on the transformer, there are a bunch of small passives on the same trace
The FET doesnt appear to be shorted Gate to Drain or anything strikingly obvious..
Should I just replace the cap and see if anything else blows up? What else should I be looking at? Does anyone in the Melbourne area want to take a look for some cash / some beer? are there some good reading materials on the matter? Ive seen some youtube videos of guys diagnosing things (and just shorting things out for testing) and TBH I dont feel like that much of a cowboy. Been bitten by 240vac before.. didnt like the experience.
Even if i was just to replace the cap, can only guess the value is the same as other ceramics on the board to keep the BOM down..
Tools I have at my disposal are a DMM, Iron and a dual trace scope. I've gotta get this printer back together as the wife wants the space in the Study back... Help!
Also. Ive got a Melbourne made Scope from the 70's that I think would be great for Dave to do a teardown.. but he hasn't replied to my enquiry email.. I'm guessing because hes sick.
http://www.radiomuseum.org/r/bwd_oscilloscope_509b.html (http://www.radiomuseum.org/r/bwd_oscilloscope_509b.html)
-
The guys at badcaps.net forum (http://www.badcaps.net/forum/index.php) do a lot of this kind of thing. Their instructions apply here too: provide make and model numbers, and good quality photos of both sides of the board so that we can see what is up (we want to be able to read part numbers on chips, etc. but more resolution than that is not needed).
-
Ive done more digging on the board and found some small.. almost silicon looking packages that are fuses.. one of these along with the main fuse popped.. so whatever it was was a big short. Testing shows the cap failed shorted.
I should point out I have a *little* understanding on how these work (i think) basically as i understand it the voltage reg is in a feedback loop to drive the FET a PWM signal to the transformer so we can control the amount of current out the power supply taps and match the load. but hey, I could be totally wrong!
-
We need pictures, it's close to impossible to guess what's going on there.
-
Ive got a SMPS from a printer that died 13 months into service (yeah, just out of warranty) and its decided to blow the internals of a dipped cap all over the main filter cap. My guess would be a tantulum, but they are blue (well.. this one is ka-boom coloured) and I thought they were orange. (a google suggests 2,200pf ceramics http://www.made-in-china.com/showroom/dgdihao/product-detailRepngYCDsqWm/China-Ceramic-Capacitor-222M-AC250V-.html (http://www.made-in-china.com/showroom/dgdihao/product-detailRepngYCDsqWm/China-Ceramic-Capacitor-222M-AC250V-.html) )
Anyway, I'm tempted to just replace the cap and see how it all goes.. but something must have caused the cap to blow.. its not a minor failure, yes there is a big power resistor nearby and you can see the heat damage to the board
The cap in question is between the source on a 3678 Power FET and a tap on the transformer, there are a bunch of small passives on the same trace
If the cap is between primary and secondary or a primary side and ground then it's a Y-rated safety capacitor for EMI (the blue colour is common for this type).
Pictures of the board (both sides) are required for more assistance.
-
The device is a Kyocera FSC1020MFP and the PSU board seems to be sourced from Ricoh
More testing on my end here, Both fuses on board have blown. The socket type and the small Black through hole one of gone.
Not sure where I would get a replacement for the smaller fuse, but I'm sure if I look hard enough at farnell I can find it
I've attached pictures to the post, sorry about the split but my SLR is still at the office and this was the best focus I could get at the time.
Bits you cant see in the pictures are the Fets etc. ill list them here..
As I mentioned earlier, I've got some Ahsahi/some cash for anyone willing to look at it direct (Melbourne SE suburbs) as I can fully appreciate diagnosis via forum posts is quite hard!
Trying to stop the Wife going and buying some consumer rubbish and sending this to landfill
Left heat sink, top to bottom
3 pins, YG906C2 (a picture of 2 diodes) +1DP
5 pins 2 outers and middle bent out
2 pins, looks like a big arse diode, I9137 YG8 I2S04
5 Pins, bent as before, SHARP PQ1CG203 15H B8
2 Pins, looks like a diode again 19137 YG8 I2S04
Device on independent heatsink on secondary side
3 pins J494
Non heatsinked device, near right side bent electro's
I7LC4
K2808
Right Heatsink, top to bottom
Top device, 3 pins, center pin bent out as before. MB166(I) 19
Middle device (connected to cap in question)
3 pins, 0N074 01 K3678 (i should point out, the outer 2 pins have been bent towards the chokes instead of using the 3 holes next to the heatsink)
This is isnulated with a Silpad
Bottom Device, 4 pin device, D15XB 60 B2119
Thanks for all your help guys... SMPS's are a bit beyond me and as I said, I dont want to cowboy this as it could be dangerous
I've uploaded the images here as they are too big for the forum
www.chibiko.net/SMPS (http://www.chibiko.net/SMPS)
Also, Happy voting day!
-
what appears to have happened is the snubber capacitor failed. this is due to a number of issues from the beginning imo.
its probably 2nf? when it failed the switch failed, shorted, blowing the fuse.
-
Yeah other caps nearby at 2.2nf so guessing thats the way to go.
But my question here is obviously the cap failed for a reason, if it was an electro I'd replace it and be on my way but I was under the impression that ceramics were pretty indestructible unless you over volted them.
Do I just replace it, replace the fuses and see what happens or is there something obvious I should be checking first? (ie could the FET have blown it apart?)
Ugh, Jaycar dont carry them either.. But I'm LOVING the Farnel search tools!
-
Actually come to think of it... We did have a power outage at the time it failed... Maybe there was a huge transient at the same time?
-
that 5 pin part is a mystery because the way they have it hooked up, is might be one of those current sense mosfets with 5 leads, and they have two of them shorted out, but that doesn't work because its clear one of those extra pins is shorted to the gate. maybe its two mosfets in one package? common drain? that doesn't make sense either.
anyhow, that capacitor that blew up is directly conencted across the transformer and that is not a normal thing to do. when it failed it most surely blew up that 5 pin switch.
the rest of the circuitry is probably unharmed. you might be able to replace that 5 pin part with a generic 500 volt mosfet. i doubt its an npn transistor.
-
if its the part I think your referring to that's coupled straight into the cap in question, its a normal FET. its only using 3 of the 5 pins on the board.
I guess maybe they were using a higher current one that needed 2 pins for source? or maybe some sort of weird optocoupled gate? (who knows)
Anyway the device is only actually 3 pins, as I said in my original post
The cap in question is between the source on a 3678 Power FET and a tap on the transformer, there are a bunch of small passives on the same trace
Heres the datasheet http://www.alldatasheet.com/datasheet-pdf/pdf/241948/FUJI/2SK3678-01.html (http://www.alldatasheet.com/datasheet-pdf/pdf/241948/FUJI/2SK3678-01.html)
I've done the usual diode style testing on it and it appears to be fine, I could always pop it out and drive it to see what happens.
-
ah, well they just had extra holes to allow for a t0-220 instead of a 247 package.
anyhow, replace the cap, with a 2nf cap in series with a half watt 10 ohm resistor, replace the fuses and turn it on, since its a 900 volt fet it probably survived whatever caused the fuzes to blow.
-
I've already ordered the cap and fuses from element14 (except the glass one, that was a min order of 10@$20. Picked some up at Jaycar for $3.
What is the resistor adding that the circuit didn't have before? Is it just so if the cap fails shorted there is still 10ohms of resistance there potentially saving something else?
Obviously I'm clearly missing something here.
-
I realized this morning that my cover for the toner/drums wasn't exactly what you would call dark.. (Bed sheet) Here's hoping I haven't damaged them. I've put the printer back together enough that I can reinstall the drums to prevent any further damage.
If I have damaged them I'm sure it will make for some interesting effects
-
the capacitor is a snubber. they should have had a resistor in series with it.
-
Well I'd just like to say to everyone thanks for help.. Finally got the Cap from Element 14 in the UK.. as the aussie warehouse didnt have stock..
Powered it all wait waiting for the magic smoke (I had smoke catching apparatus on hand.. those caps are expensive... J/K) and its all come up well. Unit works quite well.
I suspect the power supply is a little undersized on the logic rail as the LCD backlight flickers quite badly during boot suggesting some sort of current draw issue but other than that its working a charm
I'd like to thank the guys here that invested their time to help me with something that was a bit beyond my skill set and I couldnt afford to blow it up to learn from my mistakes (or worse.. injure myself)
So Thanks.
I'm headed up to QLD later this year and I'll make sure to pack some interesting teardown stuff to drop of @dave's local post office as I didnt catch him down here in Melbourne and I think its a bit big to take on a plane.
Also its a shame there is no "this guy is awesome" or "helpful post" thing here otherwise Johansen would be getting some from me.
-
Are we sure the cap was a cap and not a mov?
https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch (https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch)
-
As far as im aware a Mov is usually thinner and fails open instead of shorted. Definitely a cap as its right on the Fet as a snubber.
-
Are we sure the cap was a cap and not a mov?
https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch (https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch)
no one ever puts MOV's directly across a high frequency flyback primary. they put snubbers there instead.
-
Are we sure the cap was a cap and not a mov?
https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch (https://www.google.com.au/search?q=mov&source=lnms&tbm=isch&dpr=1#q=mov+varistor&tbm=isch)
no one ever puts MOV's directly across a high frequency flyback primary. they put snubbers there instead.
Hmmmm, I really shouldn't post when I'm tired. I had missed the link to the pics, and was reading (or half reading) the thread through eyes that didn't want to remain open. I agree, it would be a silly place for a MOV.
-
Also. Ive got a Melbourne made Scope from the 70's that I think would be great for Dave to do a teardown.. but he hasn't replied to my enquiry email.. I'm guessing because hes sick.
http://www.radiomuseum.org/r/bwd_oscilloscope_509b.html (http://www.radiomuseum.org/r/bwd_oscilloscope_509b.html)
Just in case you need the full 509b manual, see attachment :)
I have a the earlier BWD 506 sitting here in a corner.
-
for not getting killed you need:
an isolation transformer
fuse and 100w series conected bulb
luck and knowledge
beware of suprises
generally smps with failed smd components never work again
measure all electrolitics change if in doubt