Skip to main content
Topic: YAUP: Dell Printer teardown for PCB making (56K warning) (Read 25190 times) previous topic - next topic

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #15
Still waiting on that look bearmos! :P

I did not do much lately, at least I am starting with the electronics. As you all know, I wanted to use the stepper motor driver that came with the scanner, so I soldered on wires to the chip package so I can put it into my breadboard. Turns out the cables that I used in the pictures above (Solid core wire, from ethernet cables) is too thick and I can't bend it, so getting it into the breadboard was a horrible pain. I de-soldererd those wires and used wires from a IDE cable (who uses IDE these days !?) and had a much easier time getting it to connect to the bread board.

So, it went from this:

http://http://archive.hak8or.com/pictures/HP_Printer_Teardown/P1250653.JPG

to this:

http://http://archive.hak8or.com/pictures/HP_Printer_Teardown/P1250669.JPG

Now, I put it into the breadboard and working on setting it up. After finally getting it all wired up, nothing happened. :( I rechecked my connections, checked power, checked if Vin_high (2v) was under the voltage I was using to toggle (2.8ish V), but still nothing. I quickly let go of the stepper driver because I wanted more control of the stepper, I was not interested in the other "features" of the stepper driver, and it was probably not working since I did not use a current sense resistor nor any Vref's.

http://http://archive.hak8or.com/pictures/HP_Printer_Teardown/P1250661.JPG

So I tried to find some transistors or mosfets, and I quickly found a few mosfets I scavenged from other items a long time ago.
2x IRLI520G
1x IRLI530G
1x IRLI540G
All of them have a VGS(th) of 2 volts, and at 2.8ish volts I get 2 enough current flowing, but of course, at such a low VGS these suckers have a huge RDS(on), causing them to heat up very fast. But, you are probably wondering, where am I getting 2.8 volts from, and why is it "ish" too? I can't find any 3.3 voltage regulators anywhere for some reason, I guess I ran out of them a while back and ended up forgetting about them.  But, I had a LM317! And I have a large resistor assortment! So I wired it up, and the closest I can get with my resistors is 2.8 volts, and I don't want to bother to parallel/series them to get close to 3.3v, so that is why I have 2.8 volts. "Ish", because the multimeter I have is starting to go bad so it jumps a bit, but not really a big loss since it was a very cheap multimeter. I think the probes cost more than the multimeter itself, heh. :P Also, the PIC24FJ runs from 2 volts, so I should be fine.

Here is what the current setup looks like!

http://http://archive.hak8or.com/pictures/HP_Printer_Teardown/P1250668.JPG

Turns out the pickit2 I have on hand is a dead on. :( Long story short, I bought a pickit2 many years ago and since I was so new to electronics I fried it. Software detects it fine, but it does not read any pic's, and visually there is no difference. I bought a new pickit2 after that, and ever since then I constantly get mixed up which is which. For some reason I never thought about marking it with a marker or something to tell them apart. :S

Checking out the stepper today made me extremely surprised! These steppers are INSANELY strong! I have never witnessed such a strong motor! Meaning, if one of the phases in the stepper is energized to just 12 volts (rated for 24), it has a very noticeable resistance to manually turning it.  I am inspired to make a few boards with a PIC32 that would be similar to the nedoCPU-32 but surface mount, and some additions just for me, like high voltage mosfets and a USB connector.

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #16
[quote author="hak8or"]Still waiting on that look bearmos! :P[/quote]
It's a good think you reminded me - I had completely forgotten.

Here's a quick tear-down of a HP deskjet 895Cse:

An overview of the printer, assembled (sort of):
[attachment=5]

Without the cover:
[attachment=4]

Closeup of the sled with the print heads:
[attachment=3]

the single linear bearing that the whole thing rides on.  The sled is belt driven with an encoder strip above the belt, there is a piece of formed sheet metal that loosely guides as well to keep things in place.
[attachment=2]

the brushed DC motor driving the sled:
[attachment=1]

the same brushed DC motor, the mount to the toothed belt can be seen here, as well as the linear bearing beneath it.  The formed sheet metal that the sled guides on is seen in the top center of this photo.  The spring is actually holding the linear encoder strip taunt:
[attachment=0]

there seems to be a six attachment limit? (TBC)

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #17
a closeup of the encoder strip for the sled:
[attachment=5]

Now for a look at the roller encoder (I think that's why I'm posting all this after all).  It has 200 LPI (presumably lines per inch) and 1800CT written on it:
[attachment=4]

A look at the roller drive assembly - this is another DC motor - with a whole lotta gearing between it and the roller.
[attachment=3]

Finally, the ink cleaner/slop/drip well thing.  This actually has a bipolar stepper (with some gearing) driving it:
[attachment=2]

And a look at the whole tray:
[attachment=1]

As well as the PCBA responsible for taking care of business:
[attachment=0]

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #18
Woh, that is a huge amount of pictures! :D

The printer is very very similar to the printer I took apart, it has the same:
stepper motor (I think)
idea to hold the linear encoder
location of DC motor
Very similar paper hold down mechanism (The flaps in first post last picture)
and a bunch of other things

I think these printers are all manufactured pretty much the same, give or take a few non critical part locations. Can the cartridges wobble a bit, like, rotate on the linear rod? The metal at the top should have a small tab which prevents the cartridges from rotating further, it is weird that they didn't design it so they don't wobble at all.

I think 1800CT means 1,800 counts if you rotate the rotary encoder a full revolution. I am jealous yours is 200 LPI though, heh! :P So, these encoders can go above 150, I wonder if they put a lower LPI encoder strip to save money in the printer. >:( Also, your rotary encoder has numbers going all the way around, mine only goes maybe half way around.

Hp makes their own ASIC's !? I had no clue!

Thanks a huge bunch for the awesome pictures bearmos! :D

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #19
[quote author="hak8or"]The printer is very very similar to the printer I took apart, it has the same:[/quote]
Yeah, these seem like they haven't change much through the years

[quote author="hak8or"]
I think these printers are all manufactured pretty much the same, give or take a few non critical part locations. Can the cartridges wobble a bit, like, rotate on the linear rod? The metal at the top should have a small tab which prevents the cartridges from rotating further, it is weird that they didn't design it so they don't wobble at all.[/quote]
Yes, the cartridges wobble on this one as well.  The trouble with no wobble would be manufacturing to those tolerances (it gets pretty tricky (i.e. expensive) with less "slop").  Since these are made as inexpensively as possible, I'm thinking the design is simply relying on gravity to load down the print head assembly - after all, there is no "Y" movement to worry about here, just "X", which is quite solid.

[quote author="hak8or"]I think 1800CT means 1,800 counts if you rotate the rotary encoder a full revolution.[/quote]
Agreed.

Quote
I wonder if they put a lower LPI encoder strip to save money in the printer.
Don't forget, the resolution realized is dependent on the gearing between the encoder and paper roller, which can be different.

[quote author="hak8or"]So, these encoders can go above 150[/quote]
Yep, here's an example of one that's > 670 LPI:
http://http://www.avagotech.com/pages/en/motion_control_encoder_products/incremental_encoders/incremental_encoders_and_code_wheels/aedb-9340-w13c/

[quote author="hak8or"]
Hp makes their own ASIC's !? I had no clue! [/quote]
It always surprises me too.

Hopefully the pictures are helpful!

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #20
Bearmos, thanks for the pictures and information! :P A huge majority of help from you!

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #21
[quote author="hak8or"]Bearmos, thanks for the pictures and information![/quote]
Sure thing, glad to help.

Also, beware of my last comment on the LPI of the avago sensor - it was calculated based on the assumption that one count would equal one line.  After looking into the datasheet a bit, I'm not really sure what their sensor setup is, to the actual number of lines could be fewer, but still provide greater resolution, due to quadrature encoding.

Re: YAUP: Dell Printer teardown for PCB making (56K warning)

Reply #22
Yeah,I am not exactly sure either on if a count is a line, I am seeing extra channels on the sensor. :(

Again, thanks :D