[Thice] had himself a problem. As luck would have it his HP laser printer died shortly after the warranty period expired, and HP was ready to charge him €350 to repair it. Since that would pretty much buy [Thice] a new one, he decided to try fixing the problem himself. He scoured the Internet for a solution to his problem, and luckily discovered that his printer might be recoverable.
The entire LaserJet M1522 series is apparently pretty prone to breaking, with the formatter board being the usual point of failure. To fix his printer, he disassembled the outer shell, removing the formatter board from the unit. Once the onboard battery was removed, he constructed a set of standoffs using aluminum foil, and set the board in his oven at 180°C (~356°F) for about eight minutes.
After cooling, he reinstalled the board, and his printer behaved as good as new. [Thice] says that the only problem with his fix is that he needs to bake the board every 6 months or so, making this a great hack but not the most ideal solution in the long term.
That looks like a combo convection / microwave. It’s good to know that the procedures for repairing and destroying the board differ by only one button press.
its the solder than develops over time miniscule breaks. Its the so called HP environmental soln. We use Lead solder. Its old School, & it works. Cook the bord at 356F for 8…356/8..it works. to be on the safe side: Let it sit overnight.
The entire HP product lineup is prone to breaking.
It’s optimized for their largest customer base: offices that replace all the hardware every 3-4 years as the warranties run out.
I did this with my graphic card in my Dell XPS M1710:
http://www.youtube.com/watch?v=YsMDnHIS8vg
Great post! I am on the 3rd baking session with my HP 1522 now.
The first two bakes lasted 5 months or so, but it’s beginning to act up already 2 months after the most recent one. Hope that is not a sign of the end.
I’m gonna keep baking it until the current toner cartridge needs to be replaced. Then, off to buy a different brand.
very interesting. Great hack, but why does this work?
Cheap, lead free solder. That and a basic misunderstanding of Tin.
http://www.empf.org/empfasis/july05/whiskers705.htm
over time. the lead free solder develops, because of heat..very small crack and the current arcs. Lead free is supposed to be good for the environment. We still use lead!
its OLD school, and it LASTS!!!
lol sounds like what i have to do to my ps3 every 5-8 months….damn lead free solder!!!
So, are the solder joints cracking? Is there physical stress on the board that keeps causing this?
Yeah, it’s the joints on the main BGA chip that are cracking. There’s no external physical stress, it’s just a combination of crappy RoHS solder and heating/cooling cycles.
I’ve repaired several of the P2015 formatter boards, which are prone to the same problem. I’ve done them both ways, using the oven method at first, and then after I acquired a hot-air station I started doing them the proper way. Of course doing it the right way is preferable, but even after a couple years in service I haven’t seen any failures on the ones I fixed the ‘ghetto’ way.
I *do* go a step further, though, and add heatsinks to the affected BGA chips… I just use random spares salvaged off old motherboard southbridges and such, whatever I have that’ll fit, and attach ’em with a daub of thermal epoxy. That might be why mine have lasted so long.
I would love to see the ‘right way’ depicted in a tutorial… :) I tire of re-baking my board every few months.
Is it possible to home-brew a hot air station?
It will work for a while, but you’ll have to pull it back out and re-bake it every couple of months or so. I threw mine out and bought a cheap Brother wireless laser printer and have never looked back. Every HP I’ve owned has crapped out.
I did this with a Geforce 8800GTS and an old HP netbook.
While it isn’t a permanent reflow, it’s a quick hack that can save you quite a bit of cash.
I’ve always felt that the newer and smaller HP Laserjet series printers were designed by a different group than the larger Laserjets. I’ve supported various 1xxx and 2xxx series printers that were pure junk and I’ve also owned 4000, 8000 and 9000 series printers that refuse to die. I think the main thing is to avoid personal and desktop type printers in favor of workgroup and departmental models.
That’s just what i did to my rrod xbox, worked for both xboxes i own.
I’ve installed custom clamps to keep the boards totally flat and Upgraded the cooling on one, the other xbox just got rrod again 2 weeks ago.
Nice job btw!
He might be better served redoing or adding solder to suspect joints. Most components are only made to withstand so many reflows.
It’s not even reflowing at those temperatures. It’s got something to do with crystal/tendril growth, which temporarily bridges the gaps until the flaw sets in again.
It’s pretty difficult to get a double sided board to reflow – you can only do it on one side because if you try to reflow both sides, the components on the other side will drop off – and it ages the components about ten years per pop.
Its not the solder joints cracking per se. What must be happening is that there is a mismatched coefficient of thermal expansion between one or more parts on the board and the design of the solder pads and / or the pc board itself. Remember that FR-4 is not just FR-4, I believe if my memory serves me correctly, according to IPC, there are 31 or 33 registered FR-4 types. Nothing to to with the overall thickness, but plenty to do with the individual layer thickness of the glass cloth and the weave direction. I am talking about the mechanical construction of the FR-4, not the total number of copper layers. There are 31 or 33 registered FR-4 materials with the IPC. Choose the wrong one, or worse, as I have seen on plenty of production PCB drawings, just specify FR-4 and let the Chinese use whatever they want…poor design work all around.
I used an HP Laserjet 4MV for many years, until I could no longer find toner for it. I just picked up a Laserjet 5p off craig’s list last week for $20, and $20 worth of toner and I am set for another 15 years.
Why anyone would buy a new printer is beyond me.
Maybe because they don’t want an extremely slow printer that takes forever to spool a document being transfered through a parallel port?
My $5 Laserjet 4M+ has a 100Mbps NIC in it :)
I did this with my HP laser too. I was fairly skeptical, but since the printer wasn’t working anyway I figured I had nothing to loose.
Baked the formatter in the oven and it’s been trouble free for almost a year now. I though it was pretty crazy, seeing how I got the printer for free from a surplus bin and it was basically brand new otherwise.
Nice reminder there are thrifty hacks like this out there.
I’ve done the same thing with one of our Laserjet M1522nf’s.
It seemed to fix the problem as well.
This doesn’t make sense for a reflow operation. I don’t know of any formulation of solder that flows at 180C, and the board is probably RoHS, which means the solder should flow at about 265C. To reflow, the oven needs to be at least 280C, which gets into territory that most household ovens have difficulty achieving.
My thoughts exactly. So what is actually going on here? Does some sort of metal oxide form in the cracked joints over time which gets reduced by the oven heat alone?
Or could it just be that radiated heat from the oven’s element gets absorbed more by black chip packages causing them to get hotter than the surrounding air?
So I should stop telling people that I reflowed my printer board, or risk appearing a fool? Oops. Too late.
Seriously though, it makes perfect sense to me why I keep having to repeat this process if I am not actually causing the solder to reflow. Ugh.
I am about to break out the heat gun and do some selective ‘reflowing.’ Sound like a bad idea??
I guess that the heating cycle in the oven just expands the solder connections just enough that they touch again, and the board works for some time.
Generally, baking boards in the oven is not a possibility, because if you do hit reflow temps, you’ll melt the hell out of any connectors on the board that isn’t rated for reflow temps (only a small fraction of connectors can handle this) – Usually any connectors on the PCB are selectively soldered by hand or robot after reflow of the main components.
And you’ll cause any solder to rise inside the connectors that are rated for reflow temperatures.
I’ve had to swap some components from boards by reflowing them on a hotplate, and it’s a matter of seconds as the solder creeps up the gold plating. After that’s happened, the device won’t work for long because the connectors start to fail.
Hmm I have a really nice Linux-compatible Brother MFP that was hit by a power surge, I wonder if this would be worth a try…
A power surge does not crack the solder joints, it usually fries stuff, so no, won’t work….
Have done this to plenty of HP formatter boards, I like to use a cheap hot-air gun and a cheap IR thermometer. Not bad for $20 of equipment.
Amazing, just did the same yesterday. Same Printer same Problem.
https://plus.google.com/113261528535143564780/posts/VZt9b9WT4L2
People are suggesting Brother printers? On Hackaday? Really? Aren’t those known for using a higher temperature toner that doesn’t work well for transfering to PCBs for etching? What are you all using your printers for? Documents?
I like Thrashbarg’s recommendation of adding heat sinks. Would it be possible to put a sheet of mica or mylar or something even between the board and the heat source in the printer to help keep this from happening?
My thoughts exactly. Although I love my Brother’s output quality much better compared to HP, it sucks badly in toner transfer.
I’ve been doing toner transfer on a Brother laser (HL-2140) with stock toner on shiny newspaper ad/flyer paper and have been getting very good results down to 20mil (0.5mm) traces. Granted it took a while to get the process down, but once you get a feel for the iron (yes, hand iron) the results are consistently good.
…while I was researching for my formatter board cooking recipe I became under the impression that this is a manufacturing flaw that HP is aware of, but won’t admit is present. Even more concering the apparently braod range of models it coveres. My printer is a LaserJet 2015dn.While a good cheap printer can be had for people willing to tinker (perfect for the people one this blog), the best thing would be for consumers not to purchase these printers because there’s obviously a defect that HP is willing to pass on to their customers. Basically, good for hackers, but crappy they exist in the first place. /soapbox
It’s not a flaw. It’s planned obsolescence.
The whole BGA/RoHS fiasco sounds suspiciously similar to the great capacitor plague of the 2000s. Manufacturers are aware of the problem, but actually like it since it usually only strikes once the warranty is up.
Sadly, we have very little recourse. Manufacturers deny the problem as long as possible, then switch over to blaming someone else. Eventually the public becomes aware, and they switch over to a new problem. I can’t wait to see what the next form of planned obsolescence is.
All because BGAs just suck. What a horrible package that needs to be abandoned. My work uses them on aerospace electronics, mostly because that is the only packages they can get for certain ICs, but we have had nothing but problems with them ever since.
If I were to design anything life-critical or mission-critical, I would avoid BGAs. Thi package needs to die off and pretend it never existed.
Cellphones would be a lot bigger without it, though.
I’ve used this trick to revive an Nvidia 7800GTX and was successful. I’m still unsure of how it was actually fixed though. If you look at the melting point of leadfree its normally way above the max temperature of a household oven.
What I think may be happening is that the heat expansion of the board shifts some of the cracks and makes the connection again, making the board work for a while longer.
I guess you could always just buy a hot air station, I bought an Atten one for $50 on ebay
Er, leadfree melting ABOVE 500 degrees F?
No way. I have some here that melts at 226F. Most of the common tin/lead alloys melt at 361, and some lead-free melts at little more than 5 degrees over that.
The CHEAP lead-free solder melts way under that. It also gets “soft” despite being under its melting point when hot. Parts shift with gravity or lift off the board. I swear, chinese subcontractors use ANYTHING that melts as solder.
I’ll join the “Been there done that” crowd. I have a couple of 2015d in different offices and after thinking I’d have to replace them I hit on several web pages describing how to cook them back to life.
My wife came home to me carefully watching a known-accurate oven thermometer and timer as the imaging board on standoffs toasted away in the toaster-oven. She’s used to such things.
8 minutes at 350°F in the “convection oven” setting. Take all the stickers off first. So far I haven’t had to do it twice, but I like @old thrashbargs heatsink idea.
lol @ bothering to repeatedly do (shitty) reflow
should just re-ball the bga(s) once (preferably with non-bullshit, lead-bearing solder) and get it over with.
also done this nifty trick with my 8800gts vid card, was getting graphical errors across the screen, works like new now
Please don’t do this in an oven you intend to use for cooking….
IT’S CALLED A REFLOW, PEOPLE.
We’ve been doing it for decades. I’ve been cooking PCB’s from failed consumer equipment in my kitchen oven since the 90’s… It makes me sad when electronics hobbyists (like I assume hang out here) say they have never heard of this method. Why buy a new one when you can fix what you have!!!
need themo control, then i use it ;)
Is there any solution to repair other models of HP printer formatter boards? Such as ho 1220/1280
I’ve successfully done this twice now to fix my MF1522nf printer. It’s a great resource, Thank you!!