evil print head instantly kills printer electronics (i865/ip4000/etc)

[benzinerwin]

I've had a couple of defective printers (i865, ip4000, mp750, mp760, mp780) that were basically ok but with bad print heads - which were either clogged or completely burnt to death.

One of the heads did only print colors but no pigmented black on one of the printers.

I found out that this particular print head is responsible for the death of some previously electrically healthy printers: they all were immediately killed after I inserted this head. They all do not print any colors any more. I've tried other, healthy print heads - nothing. Replacing the main pcb did always do the trick and made the printers work again.
So this printer head from hell really killed the printer main PCBs. The signal lines from the print head pass a fuse and then go directly to the MPU. The fuses are all ok. Maybe it's some of the transistors - I don't know, I've replaced some of them just to see if anything changes but with no luck.

Any pointers appreciated...

 

[embguy]

Sorry to learn your problem.
Someone in this forum warned people not to try testing problem printhead in good printer. This confirms the warning.

I think it is the driver circuit that drives the printhead. It will be difficult to repair if the drivers are surface mount ICs.

It may be cheaper to replace it with new one or search your local Kijiji or Craiglist for used printers.

Look for freecycle forum for your location. http://www.freecycle.org/ People posts the offer and wanted ad. All offer must be free to give away. No cash, trade, exchange are allowed in the freecycle forum. This is how I got my Canon MP730 for free.

 

[headphonesman]

Sorry to learn your problem.
Someone in this forum warned people not to try testing problem printhead in good printer. This confirms the warning.

.

Absolutely correct. I have written off 1x4200 and 1x5200 exactly this way, (they were both new out of the boxes).......soon as the head from the "old and failed for no apparent reason machine" was put in they failed in exactly the same way. When a machine fails now (and reason is not obvious) as far as I am concerned the head fails with it !!!!!!

YOU HAVE BEEN WARNED.

 

[websnail]

Sorry to learn your problem.
Someone in this forum warned people not to try testing problem printhead in good printer. This confirms the warning.

Yep... It's happened to me and at least two other people on these forums although I forget who.

I now make it standard practice to replace any printhead AND printer if the printer stops responding as 9 times out of 10 the printhead went with it.

 

[benzinerwin]

Thanks for your replies!

Well, I had a closer look at the hardware and it made me very disappointed in Canon: the expensive parts are left virtually unprotected. The MPU's I/O lines go directly to the printer head contacts. Ok, there are inline fuses, however, these fuses are strong enough to survive voltage overloads and hand them over to the MPU for further processing. Yes, the fuses are there to protect for current overloads. But then again, there's nothing else. Needless to say that all the fuses were healthy in my dead printers.

After isolating the printer head clock line - also directly connected to the MPU - I cannot see any activity there for the dead printers. Well, that's it - the BGA-mounted MPU is defective. In every case.

The burnt killer head shows a weak conducing link between head power ground and logic ground. Also between some of the logic signals and head power supply. This is enough to push some of the 24V head power to the 3,3V signal lines. Obviously the MPU cannot handle it and dies. Well, who would have thought that a printer head may suffer an electric defect and who could have known how such defects will look like.- And who could have known what 24V might do to 3,3V logic... :-/

Now I know about it and I will have a closer look at replacement printer heads. I only wished I haven't learned about it the hard way...

 

[ghwellsjr]

Thanks so much for sharing this information with all of us, benzinerwin. Have you seen any physical damage to your killer printhead or other symptoms that might help the rest of us be alert to the same kind of problem? For example, I posted a picture of a burnt trace on a defective printhead:

http://www.nifty-stuff.com/forum/viewtopic.php?pid=23682#p23682

The printer flashes some error code when turned on with the printhead installed and will not move the printhead but if I turn the printer on without the printhead, the printer seems to be OK except, of course, it can't print without a printhead. I'm still wondering if the printer has escaped damage and if I should buy another printhead.

 

[benzinerwin]

I've got many defective printer heads - most of them were rejected (yellow flashing light). You could easily see the burnt traces on the flex-film-pcb. However, my evil killer head looks perfectly healthy on t he outside. The flex pcb shows no damages. It's a pity.

I guess I have no choice but to check the resistance between the critical signal pads before I put a printer head of unknown status into a healthy printer...

/edit

The only thing the printer does in order to recognize a broken printer head is to check the head's eeprom data and to check the sensor diodes. The eeprom is virtually indestructible while the sensor diodes sometimes survive the damaging process. As a result, the Canon hardware expects to deal with a healthy printer head an turns on all the power switches ... which instantly (or slowly - but irreversibly in either case) kills the affected I/O lines in the main PCB's processor chip (IC101).

 

[nanosec]

Wow! Thanks for the heads up on this, this was bound to happen to me sooner or later, so I think you guys just saved me some bucks!

 

[knuckles_red]

Hello all. benzinerwin; would a BGA refit of a new IC101, combined with a healthy print head cure the problem? This is something i'm willing to try and report back.

 

[Raymond]

I found this out the hard way too, but my print head was already on the way out as well