SMD fuse blown in Epson WF-7610 Printer

[EHT]

I have a fault which I believe is at least due to an SMD fuse blowing. I'm interested to know if anyone could guess what this fuse is protecting and/or has advice on how to approach replacing it and testing.

The printer is an Epson printer WF-7610, pretty large and complex beast, A3 size with ADF & Scanner. It shows an error code immediately on startup before activating any motors etc. It developed the fault while printing and has had regular but not heavy use. In the service manual the error code is listed with the only possible reason being a fault on the main board: "Circuit error (include blowout of a fuse), possible cause Main Board failure". Sure enough, there is an 0603 part on the main PCB labelled F1 which is open circuit - see pic.

The fuse is really not designed to be replaced. The service manual says for engineers to replace the main PCB, which would be pretty stupid if the cause of the fuse blowing was external, e.g. like the print head!

My plan is to put a pair of leads on the fuse terminals then reassemble the printer. It is a huge PITA to assemble/disassemble with no way of being able to test it disassembled. I will then put a fuse in series with current shunt resistor so I can measure the current and see what happens. 0603 fuses seem to range from about 200mA to 6A. I thought i'd try doubling ratings from 0.25A and seeing what happens. 

I was thinking of using resettable PPTC fuses but i realise these take a few seconds to blow. This would mean I could see if it delays the appearance of the error (<1s), but maybe a bad idea?

Other info - no visible problems on the PCB, not many components I can see that look testable or replaceable, e.g. no power transistors or v regs. A replacement PCB is £30, replacement equivalent printer is £150.

[sean0118]

What are the other components on the PCB? Can you please post some photos of the front and back of the PCB itself? 

[TheMG]

That actually looks a lot like a zero-ohm resistor (jumper) to me...

[attach=1]

I suppose it would work as a very crude inexpensive fuse against gross overloads. Weird though. Maybe they initially intended to populate an actual fuse and later on decided it wasn't necessary?

What is the circuitry before and after the fuse? Any obvious short to circuit ground on either side of the fuse?

[EHT]

Thanks Sean & MG. You gave me a bit more enthusiasm to dig deeper and actually I have a much better picture now. I'm sure that is a fuse. You can see a better pic (following post, '7610 PCB 1_ann_r.jpg'), on the top of the part it looks like it has an 'F' not 'O' and it is open circuit. The area of the circuit which it protects is a MOSFET output stage. You can see this in the first pic. There are two IRF6665 MOSFETs. They seem to be driven by a proprietary Epson chip. The fuse F1 does indeed connect to Q4 Drain. Unfortunately the G, S terminals are underneath these devices so it isn't straightforward to trace and test.

It seems like this pair of MOSFETs drives a resonant circuit of an inductor (with writing R33L on top) and C226. The circuit is near the connectors to the print head (underside, to right) so seems this is for the print head rather than a motor... not sure.

G-D resistance is about 80k. I'm looking to see if there is a D-S short but i dont think so. An important thing is to know if it is a fault on the driver or in the print head. If the latter, i think it writes-off the printer and obviously any time/money spend on the driver also wasted.

Instead of replacing the fuse, i think maybe I'll try powering the PCB with just the display module (not all the 30 or-so other connectors..), see if it gives same error
If it does, I can:
- probe the supply side of F1 to see what the supply voltage is
- see if there are Gate drive pulses
- replace the fuse with a power resistor that would limit the output current quite a bit, depending on supply voltage measured above
If not possible to test outside the printer:
- replace fuse with arbitrary value power resistor

Continuing:
- power up with print head disconnected
- see if the original error goes away
- measure the output developed over the inductor if possible
- measure current through the power resistor
- remove resistor and try with a fuse if above is not showing high current (should be low/zero with no load)
- connect print head
- bang!?

Any better ideas?

[EHT]

See for annotated pic of the MOSFET driver stage, and the rear of the PCB

[EHT]

OK.. I powered up just the main PCB and display. The unit boots, but it is showing a different error obscuring the original one - not too surprising.

Too many assumptions above! The PSU o/p voltage is 42V and this goes straight into one side of the fuse F1. It also connects to Q4D, so F1 is actually NOT in circuit to the MOSFET output stage, just they both use the +42V supply. When I check the output of this mosfet driver, its hovering around +11V DC. There is no gate drive signal, but then the whole thing is in a pretty lame state with nothing connected, so maybe it is deliberately idle.

Anyway... on a +42V supply, I guess F1 is quite low current whatever it supplies, maybe 100-250mA. I can't trace where F1 goes to; it just disappears into the multi-layer PCB

Tomorrow I'll try starting off with a 1k resistor in place of F1 and see if there is any load.

[EHT]

Think I found it.. F1 supplies a couple of lines on the print head. One of these is shorted to ground Eventually tested this by plugging in one of the print head cables and can then measure F1 to GND is short. Print head can't be replaced realistically as I can only see dodgy ones from China for £50. Furthermore, looks like it would need calibrating/adjusting using Epson's 'Adjustment Program', which I don't have, to print the required patterns. Damn.

Baffles me why they made an extensive service manual, ironically one of the most detailed and comprehensive I've seen, when the parts are prohibitively expensive. On their website it has all this 'green' stuff about recycling with pictures of little seedlings and sunshine but if you read through it, all they will do is perhaps assist you to take your old printer to the local tip and buy a new one. What a joke! This company is responsible for creating an AWFUL LOT of waste!

[Koray]

Printers, especially inkjets, are very difficult to troubleshoot and repair due to their construction and ink mess inside.

Instead of fighting a lost match, I carefully tear them down to simple parts and post them on eBay. I salvage screws and motors for future projects. A recent curbside discovered color laser printer brought me over $200 on eBay and there are still unsold parts. This may reduce the environmental impact and also help you financially towards a new one. eBay listing is free and simple.

Just take good pictures, put the numbers on cards or other parts in the listing title, and set the listing to renew automatically. Sometimes it takes 2-3 years to sell, but who cares. I just ship them when I receive notification from eBay. One curbside printer has already saved about 10 printers. Not bad?

K.

[EHT]

Well done! Thing is, I reckon the print head is the part which is most often required and which really needs to be new.  Pity the Chinese pattern parts arent even priced low enough to be worthwhile. Otherwise seems like the lifespan of the unit would be limited just by firmware / driver obsolesce

[fcb]

You're a brave chap. I've got a WF-7620 and it's given great service for the last 3-4 years - hadn't crossed my mind that if/when it dies to fix it though - can you post pics of the insides?

[amyk]

The 42V is the piezo drive voltage, as you have discovered. You might wish to disassemble it carefully to see if the short is caused by a ink leak, and not in the hybrid IC itself.

You may be able to find the adjustment program online; I have ones for older Epsons that I downloaded many years ago.

[TheMG]

Baffles me why they made an extensive service manual, ironically one of the most detailed and comprehensive I've seen, when the parts are prohibitively expensive. On their website it has all this 'green' stuff about recycling with pictures of little seedlings and sunshine but if you read through it, all they will do is perhaps assist you to take your old printer to the local tip and buy a new one. What a joke! This company is responsible for creating an AWFUL LOT of waste!

Service manual is probably intended for techs repairing printers sent in for warranty. They probably never intended for anyone doing out-of-warranty repairs.

Had an HP inkjet printer a few years ago in which one of the print heads was clogged to the point that none of the usual tricks could get it completely unclogged.

HP designed this particular printer with a user-replaceable print head assembly. Just go online and order a new print head assembly, no problem... oh wait, they don't sell them, never have. What the heck is the point of a user replaceable print head if they're not going to sell the replacements?!?!? Truly mind-boggling.

[EHT]

You're a brave chap. I've got a WF-7620 and it's given great service for the last 3-4 years - hadn't crossed my mind that if/when it dies to fix it though

I thought, how could I fail to fix it when I have test equipment and a detailed service manual. Answer: when the mfr won't sell the replacement parts! This error was 0x9A. 0x9A-E seem to be printhead related.

can you post pics of the insides?

Pic shows what is left of it! I wouldn't have been able to disassemble easily without the manual, at least without breaking some plastic catches. Basically, I attempted it because it seemed like it might have been a fault on the PCB which I could see was available for a resonable price if not fixable at component level.

You might wish to disassemble it carefully to see if the short is caused by a ink leak, and not in the hybrid IC itself.

See pic - looks like ink was leaking from the head. As the resistance I measured really is short (<1 Ohm) I expect this leak has led to a short in the head internally, not just leakage through the ink. Even if not, the head seems to be leaking inside so would need to be replaced. There is no visible heat damage anywhere. Ribbon cables are fine. I've given up and ordered a new printer now  At least I am happy that I found the fault without spending any money at all...

Service manual is probably intended for techs repairing printers sent in for warranty.

Yeah, also maybe as this is a small business model, they might have rented them out.

Thanks for your assistance everyone. Hopefully this is a useful warning for others wishing to fix an Epson printer!

[Sk8aaron]

Just use a 1 amp fuse. but I recommend that you put the new fuse outside the printer for easy access incase it burns again you don't have to open it. solder 2 wires to the sides of the F1 fuse and run the wires to the outside then connect new fuse. I had success on the Wf-7610.

[mediaonebataan]

I know this is an old one.. but do you have a picture of putting 1amp fuse I guess its a tube fuse to replace the smd fuse?

[EHT]

On mine, I measured a short in the print head so I didn't try replacing the fuse.

Replacing the print head was not viable - I think almost as expensive as a new printer and also looked like it would be a lot of fiddle to do, especially as alignment is needed. This is shockingly disposable e-waste.

[mediaonebataan]

agree with you, but before I throw away to the garbage at least I tried to salvage this thing also the scanner should have a different module which can bypass... lot of waste

[M4trix]

Epson printheads ? Bad, bad head design.  The inks in the head tend to leak between the chambers due to bad gluing (delamination) of the nozzle plate. This causes the ink cross contamination.
The bad thing is that delamination gets worse over time and printing becomes impossible. I have Epson L800 and I noticed after a nozzle check that Magenta has spots with Cyan ink. I thought, nothing serious but with every other print job the number of spots became larger and after some time it wasn't Magenta anymore but purple ! Then I stumbled on a website where a guy is complaining about the same thing. He got the answer from others... nozzle plate is delaminating. I was shocked especially when I saw the price of the new head ! 
Then, as a person who had poor knowledge about printheads, I bought one from a Chinese ebay seller for 80 bucks. Installed it, powered the printer but nothing happened. Only the power LED blinked once and then all went silent. Took out the mainboard... the SMD fuse was blown, the encoder E09A7218A and two Sanken TO-220 dual transistors, FTD01N/FTD02P. From this point I didn't know what was the culprit, the head or the mainboard. I was so desperate ! So, I decided to take another shot from another seller. This time it worked but all colors were mixed ! What I learned is that they salvage heads from printers thrown in the dumpster and most of these printers have clogged heads. The "smartass" guys over there in order to unclog the nozzles, put the sensitive head under high water pressure which causes a serious delamination. Fortunately, I got my money back. Warning to others, there are no new printheads on ebay. Especially not for 80 bucks. They are all refurbished and most of them are shorted, blown which will cause serious damage on the mainboard ! 

https://inkjetparts.com/blogs/news/what-is-printhead-delamination     

[amyk]

You'd think the Chinese would've come up with a way to repair printhead delamination, given that they've already figured out how to service things like mobiles with glued screen/digitisers...

[M4trix]

I was so naive. Well, look at the bright side. Now I know some facts. 

[mediaonebataan]

Update: soldered the F1 fuse.. did not worked, printer is throwing a printer head error although nothing is connect.. well then to the trash it goes...

[M4trix]

Hmm, interesting theory...

Why do Epson printers die?
Death of PG No. 1

I believe that it is the aggressiveness of the original ink for the L800 that contributes to this defect. With minimal physical impacts, the delamination zone will expand until the channels are completely mixed. Because the physical structure of the print head is exactly the same, we conclude that cyan is the most aggressive ink, it is from this channel that in most cases color mixing occurs.

Well, here it is actually for you and concentrated cyan mixed with black. Black dye you know dear. And the cyan is cheap. So the “epsons” roam that fell into black: (And then the black channel dies, because the same cyan is in it.

Here is the link. Very useful and shocking stuff. The guy is from Russia. There are weird grammar mistakes but you'll get the picture. Anyways, Epson
you...!#@*!!! you won't fool me twice ! 

https://driverstore.ru/en/creatorsupdate/pochemu-sgorayut-pechatayushchie-golovki-na-printerah-epson-pochemu.html