Hacking the HP45

Powder and inkjet printing
User avatar
dragonator
Site Admin
Posts: 595
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

Both CLK and CLR are actively used in the program. Clear is made high (active low), clock is made low and the data is entered into the driver. The clock pulses high, followed by the clear going low about 2us later. I have not yet looked at a way of eliminating one of these. Maybe I can, but I will have to look at the datasheet before I try. I first wanted to make it work.

If all pins are on one port (as they are on the Teensy 3.6) you can do without the latch. If you do not have all pins on one port (like the Teensy 3.2) you will either need a latch or trigger each port separately.
daywalker2006
Posts: 3
Joined: Sat Apr 08, 2017 12:19 pm

Re: Hacking the HP45

Post by daywalker2006 »

thats help alot, thank you
dlamet
Posts: 1
Joined: Sat Apr 15, 2017 5:11 pm

Re: Hacking the HP45

Post by dlamet »

Hello, group. Much thanks to everyone, especially dragonator, for all the pioneering effort. I'm part of a team of three developing a powder printer. We've just switch direction to adopt inject pens in our design, so just getting up to speed still. However, we're on deck to provide any assistance we can.

Coincidentally, I worked on the PostScript printer driver for the HP 1600CM, which was the first printer to use the HP 45 way back in '95. Unfortunately, that was hell & gone from the pen driving firmware. Wish I still had some contacts from those days.
david
Posts: 73
Joined: Wed Sep 21, 2016 8:43 pm

Re: Hacking the HP45

Post by david »

Thank you all for the great job. I'm wondering if there is a built-in method of the HP45 to detect if the ink has run out, for a) replacing/refilling the cartridge; b) avoid damaging the HP45 by dry-run.
User avatar
dragonator
Site Admin
Posts: 595
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

AS far as I know there is not a built in way to detect it directly (like with most cartridges, that simply say they are empty, while they are not). There should be an indirect way of measuring it. If the head temperature suddenly spikes without a good cause, the chances are that the head is empty. This is because there is no longer any fluid that takes the heat of the nozzles.
david
Posts: 73
Joined: Wed Sep 21, 2016 8:43 pm

Re: Hacking the HP45

Post by david »

Image

I'm right in thinking this pad footprint is the "bottom view", i.e. as the following picture,

Image

Many thanks
User avatar
dragonator
Site Admin
Posts: 595
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

You are right in thinking that yes.
david
Posts: 73
Joined: Wed Sep 21, 2016 8:43 pm

Re: Hacking the HP45

Post by david »

thanks. I'm currently building a driver board with,

- LVDS serial interface for long distance connection;
- on board buffer and trigger and sync;
- on board DC-DC converter to control the head voltage;
- ESD protection;
- temperature control;
- nozzle diagnose

Is there anything I should be careful with this printhead? According to the timing, 2us setup, 2us fire, and 2us hold, it would take 132us to scan all 22 addresses, which means 7.5khz firing rate. IIRC, HP45 can do up to 12kHz, how could that be possibly achieved?
david
Posts: 73
Joined: Wed Sep 21, 2016 8:43 pm

Re: Hacking the HP45

Post by david »

also in order to ease the PCB routing, the scanning order is not A1, A2, A3 ..., but something like A1, A5, A2 ... I think as long as all 22 addresses are scanned it would complete the firing. Please let me know if I missed anything.
User avatar
dragonator
Site Admin
Posts: 595
Joined: Fri Aug 14, 2015 4:48 pm
Location: The Nethelands
Contact:

Re: Hacking the HP45

Post by dragonator »

I have ordered a batch of 10 of the HP45 connectors (http://www.cfconn.com/product-show_14712/). This is only a sample. I will need to order in the 100's to make the price reasonable, but I hope to receive them somewhere next week. I will share photo's and data when I receive them. I will be able to send samples to other people interested (under the idea of getting a production run later in a designed product), but it will be one per person and they are quite spendy (order of magnitude, €50 with shipping). With bigger production runs price drops to $6-10 each.

@ David: I am also slowly working towards a new controller. In my case something with it's own microcontroller that can work stand alone. I can give more details of what I had in mind later, but perhaps we can share resources. I am sure that we both have similar things required.

I have sent you the datasheet of the C8855m, the original HP45 controller, by mail. It holds a lot more information than I can share through a few posts. A few things to consider. The 2us comes from a time before the documentation, based on what someone I know said. It is crude but works. It is actually 1.875us. Also setup and hold times can be optimized. The address fets act like a mosfet, so if you have faster charge and discharge times, you can probably get away with less. 1us on both is the time needed to get to 12kHz.

I am currently rebuilding the current V1.01 controller on a breadboard to find an issue where I cannot fire more than 5-6 nozzles at a time (in stead of the 14 required). This effectively divides the fire rate by 3. When I have the cause, I will share as well.

Address scanning in order is also something I will drop, but it is not completely without payoff. The physical offset for each nozzle under the head in the printing direction is the same as the address of each nozzle. if address 0 is 0, address 1 is 1.75um, 2 is 3.5um. The distance between each address is roughly 1.5-2um each. This is not much, but is most definitely there.
Post Reply