Homebrew PCBs

The Epson R200 Inkjet printer that also prints CD's is now on sale at
the Epson store online. I just ordered one It is factory refurbished.
The price $57 including free shipping. Not bad for a fully functional
printer for pcb's.

The list price is $71 but using the code 6WEJ41CC gets you 20% off!

TANSTAAFL!

Myc


PS; Stefan, any luck with the direct to pcb process? We are all waitng
to hear your results.

Just a quick note about this printer...

I bought one, it's a good printer and I like it. If you decide to 
try and run a PCB through on the CD tray, please accept this warning 
for what it's worth....

you may jam up the printer pretty bad.  I did this and still have 
not gotten it back working again.  The printer is cheap enough that 
I shouldn't be upset about it, but for some reason this really got 
me down.  Had I taken the time to clear a "path" and run a PCB 
through without the CD tray etc., all would be okay.  So, my tests 
are on hold till I get over it, and get back to experimenting with 
it again :-(



--- In Homebrew_PCBs@yahoogroups.com, "mycroft2152" 
<mycroft2152y@...> wrote:
>
> The Epson R200 Inkjet printer that also prints CD's is now on sale 
at
> the Epson store online. I just ordered one It is factory 
refurbished.
> The price $57 including free shipping. Not bad for a fully 
functional
> printer for pcb's.
> 
> The list price is $71 but using the code 6WEJ41CC gets you 20% off!
> 
> TANSTAAFL!
> 
> Myc
> 
> 
> PS; Stefan, any luck with the direct to pcb process? We are all 
waitng
> to hear your results.
>

On Fri, 30 Jun 2006 13:56:57 +0200, mycroft2152 <mycroft2152y@...>  
wrote:

>
> PS; Stefan, any luck with the direct to pcb process? We are all waitng
> to hear your results.


I'm minutes before finishing the flatbed conversion.
I've worked out a clever (i think!) way to allow a carrier with PCB to  
pass the C84 while being fully aligned to always print in exactly the same  
spot.
This should allow more tests, although the chip issue and the ink  
refilling issue need sorting soon.
Anyway, this weekend i will finally be able to make test boards wider than  
1cm.

Is anyone willing to work together on making a chip replacement with a  
microcontroller? I won't pay $50 for a set of stupid auoto reset chips. If  
i were a "normal" user and the chip was emptied that fast i would have  
thrown the printer out the window a dozen times already. I mean it really  
is 100% 80% 50% - and then the red LED starts already. And that's only  
powering on in between and printing nothing.

ST

Stefan, 
If it is possible send me your inkjet cartridge data. I can help you if you wish to do your SW. I didn't finalize my circuit it is still using RAM instead of E2.  I have just received E2 chips from Digikey most probably end of next week I could be able to add E2 interface.
  
Volkan 
.... making a chip replacement with a  microcontroller?

> -----Message d'origine-----
> De : Homebrew_PCBs@yahoogroups.com 
> [mailto:Homebrew_PCBs@yahoogroups.com] De la part de Stefan Trethan
> Envoyé : juin 30 2006 09:43
> À : Homebrew_PCBs@yahoogroups.com
> Objet : Re: [Homebrew_PCBs] Epson R220 on sale
> 
...
> 
> Is anyone willing to work together on making a chip 
> replacement with a  
> microcontroller? ...
> 
> ST


Stefan, this is one aspect of the project that I'd LOVE to help.  But Volkan
has been quite clear that there are important timing issues in this circuit.
I'm limitted in what microcontrollers I can program so I'm not much use here
(I only have a PIC programmer and related software/compiler).

Robert
:)

>If i were a "normal" user and the chip was emptied that fast i would 
>have thrown the printer out the window a dozen times already. 

For what it's worth there are 119 reviews by "normal" users for the C84 
at Amazon.com. 

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@...> wrote:
>

On Fri, 30 Jun 2006 16:53:34 +0200, Volkan Sahin <vsahin@...>  
wrote:

>  Stefan,
> If it is possible send me your inkjet cartridge data. I can help you if  
> you wish to do your SW. I didn't finalize my circuit it is still using  
> RAM instead of E2.  I have just received E2 chips from Digikey most  
> probably end of next week I could be able to add E2 interface.
> Volkan
> .... making a chip replacement with a  microcontroller?


You said it must be a fast micro, i'm not sure that's the case. That one  
webpage you know already the guy used a atmel.
He told me the problem was the startup time, but if you power the  
controller from the printer power and not the chip power that should not  
be a problem (he put the circuit atop the CIS).

Have you tried always sending the same ink level back, or does the printer  
notice if the value doesn't change?

ST

The printer that I am using has a serial clk of 4usec it is fast for a micro and I could not able to write a fast enough interrupt handling routine that can run on 8Mhz MSP430 micro. It can be different for your printer model.   
 I think in Eddie's solution micro works in parallel with cartridge chips, either he is sending complete zero for first 2 byte or reprogramming the chips (but I am not sure). In my solution I am completely emulating chips and not using the power source from the cartridge connection so I don't have any startup or power on reset problem. Sure new printers check everything, after writing it verifies the written data. If it fails, it gives error messages. I have tried to send same ink level; it accepts it after doing recharging (waste of ink). BTW Atmel has also very fast AVR series. 
   
Volkan  
....Have you tried always sending the same ink level back, or does the printer   notice if the value doesn't change?

The C84 has about 15us reading clock.
Writing is much slower, but that is only done when it is powered off.

I think eddie just pulled the data line low at the right moment.


What do you suggest how i proceed? I'm not sure how to go about making a  
software for this.

Do i have to get the logic analyzer out to decode the pattern? I don't  
think the memory of my scope is deep enough to do that. Or is the pattern  
known already?

ST


On Fri, 30 Jun 2006 23:33:48 +0200, Volkan Sahin <vsahin@...>  
wrote:

> The printer that I am using has a serial clk of 4usec it is fast for a  
> micro and I could not able to write a fast enough interrupt handling  
> routine that can run on 8Mhz MSP430 micro. It can be different for your  
> printer model.
>  I think in Eddie's solution micro works in parallel with cartridge  
> chips, either he is sending complete zero for first 2 byte or  
> reprogramming the chips (but I am not sure). In my solution I am  
> completely emulating chips and not using the power source from the  
> cartridge connection so I don't have any startup or power on reset  
> problem. Sure new printers check everything, after writing it verifies  
> the written data. If it fails, it gives error messages. I have tried to  
> send same ink level; it accepts it after doing recharging (waste of  
> ink). BTW Atmel has also very fast AVR series.
> Volkan

If it is 15us then you don't need to worry about. You are right writing is much slower. You don't need a logic analyzer. Just built a micro connected to clk, data, resetb lines and start storing data after rising edge of reset signal, sampling should be done positive edge of the clk.
 Their format is indeed so easy first 4 bit hold address and write bit (MSB) and the 32 bytes of data comes.
 In T60x series cartridges first 2 bytes hold ink counter 3rd byte holds cartridge replacement counter, that's what I am calling, and 4th byte holds specific to printer ID. In empty cartridge first 4 bytes are always zero and after recharging they're initialized. Before start of every printing, printer checks all of the 4 bytes if they're different than the ones in battery backuped internal counters, it initiates recharging cycle. 
 Don't forget that I have only experience with my CX4200 and these are my observations.
  Volkan

... Do i have to get the logic analyzer out to decode the pattern? I don't  think the memory of my scope is deep enough to do that. Or is the pattern  
known already?

I did some more testing.
It seems the stupid printer really notices if the value in the chips is  
different from the value in it's own memory.
Because i noticed the writing is only done when the power is removed. So i  
simply pulled the chips before switching off. But at next power up the  
printer did a cleaning cycle.

I'm still not sure how the chips work. The readout is clear, there are 4  
transmissions, i think one for each chip. It's always clock and data, very  
normal. But for writing there seems also to be some fast clock, and then  
some slow clock. Any idea what is transmitted there? It is only a single  
block fast, then slow, not 4 blocks fast/slow fast/slow fast/slow  
fast/slow as i would have expected. How does it write to the 4 chips then?


There is a small EPROM in the C84. I believe this must be the place where  
it hides away the copy of the values the chips should have. I wonder if  
one were able to disable writing to this the printer would not notice the  
constant chip value. But that doesn't seem easy either since this is a  
serial eprom and one can not simply interrupt the write enable wire.


So how is this going to work? Make a software for the micro that fakes the  
chip, including the writing, and each time it is empty automatically reset  
it to full? doesn't this waste ink?


My PC just reset itself for no apparent reason at all. I hope it's not  
developing something, i do not need that right now ;-)

I will look into the eprom issue, ideall i would like to keep the value  
constant everywhere forever with the printer being none the wiser.

ST



On Sat, 01 Jul 2006 00:58:16 +0200, Volkan Sahin <vsahin@...>  
wrote:

> If it is 15us then you don't need to worry about. You are right writing  
> is much slower. You don't need a logic analyzer. Just built a micro  
> connected to clk, data, resetb lines and start storing data after rising  
> edge of reset signal, sampling should be done positive edge of the clk.
>  Their format is indeed so easy first 4 bit hold address and write bit  
> (MSB) and the 32 bytes of data comes.
>  In T60x series cartridges first 2 bytes hold ink counter 3rd byte holds  
> cartridge replacement counter, that's what I am calling, and 4th byte  
> holds specific to printer ID. In empty cartridge first 4 bytes are  
> always zero and after recharging they're initialized. Before start of  
> every printing, printer checks all of the 4 bytes if they're different  
> than the ones in battery backuped internal counters, it initiates  
> recharging cycle.
>  Don't forget that I have only experience with my CX4200 and these are  
> my observations.
>   Volkan

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan" 
<stefan_trethan@...> wrote:
>
> I did some more testing.
> It seems the stupid printer really notices if the value in the chips 
> is  
> different from the value in it's own memory.
> Because i noticed the writing is only done when the power is 
> removed. So i  
> simply pulled the chips before switching off. But at next power up 
> the  
> printer did a cleaning cycle.
> 
> I'm still not sure how the chips work. The readout is clear, there 
> are 4  
> transmissions, i think one for each chip. It's always clock and 
> data, very  
> normal. But for writing there seems also to be some fast clock, and 
> then  
> some slow clock. Any idea what is transmitted there? It is only a 
> single  
> block fast, then slow, not 4 blocks fast/slow fast/slow fast/slow  
> fast/slow as i would have expected. How does it write to the 4 chips 
> then?
> 

Epson printers store the ink level information on the chips, not in 
the printer. If you take one set of cartridges out of the printer, and 
put it in another, it will still show the ink depleted. The printer 
checks for this data, and also checks to see if they are Epson branded 
chips (aftermarket regular chips and auto-reset chips were built from 
scratch to avoid patent infringement, while they work, they are not 
the same, and the printer can tell). The printer should not have a 
problem with the chips being re-set, otherwise the $10 resetters from 
Staples wouldn't work for people who refill their cartridges. Unless 
you mean you are resetting while the printer is still on and not in 
cartridge-change mode, which might confuse it.

If a chip is reset to 100%, the printer will do a very full, intense 
cleaning cycle, as it thinks that a new cartridge has been installed 
and wants to start the ink flow properly. Newer generation auto-reset 
chips only reset to 90%, so the printer never does the full purge 
again. However, Epson printers also tend to do an ink cleaning 
whenever you turn the printer on. This seems like a huge waste of ink, 
but on the C-series printers with OEM Durabrite ink, you do NOT want 
the ink in the cart to run out, and you DO want the printer to run 
massive amounts of cleaning cycles, otherwise the quick dry acrylic 
enamel they call ink will dry up in the printer within days and wreck 
it.

On Sat, 01 Jul 2006 17:24:37 +0200, fenrir_co <fenrir@...>  
wrote:

>
> Epson printers store the ink level information on the chips, not in
> the printer. If you take one set of cartridges out of the printer, and
> put it in another, it will still show the ink depleted. The printer
> checks for this data, and also checks to see if they are Epson branded
> chips (aftermarket regular chips and auto-reset chips were built from
> scratch to avoid patent infringement, while they work, they are not
> the same, and the printer can tell). The printer should not have a
> problem with the chips being re-set, otherwise the $10 resetters from
> Staples wouldn't work for people who refill their cartridges. Unless
> you mean you are resetting while the printer is still on and not in
> cartridge-change mode, which might confuse it.


I thought the same as you do.

BUT, the printer _does_ store a copy of the ink level in that little eprom  
on the controller PCB.
If the chips show a different value than they should it does a cleaning  
cycle at power on. I don't think this is a long priming cycle, only a  
normal cleaning. When i just switch it on without and changes to the chips  
it does not do that.

Anyway, i have found reasonably priced spongeless carts with auto reset  
chips at <http://www.4-u-all.com>.
The spongeless cart with auto reset chip is about 3.5eur per cart. I think  
this is acceptable and have placed an order.
I'm not sure it will go through though because they say one needs a  
verified paypal address (meaning you need a credit card), which i don't  
have.
This outfit is in ireland so that would be great for me customs-wise.

Anyone know another place where the sell spongeless carts in case this  
doesn't work out? It would need to be reasonably priced, not $15 per auto  
reset chip and $5 per chipless spongeless cart like inksupply.com. I think  
with those prices they are just nasty parasites latching onto the epson  
chip fraud. Nobody can make me believe a chip is worth $15.

ST

Hi List
Do you know a solution to invert the polarity (from positive to 
negative) of a Gerber generated file of a  PCB layer,
independently of any PCB program?
Cristian

-Please- do -not- start a new thread by responding to an old one. Even
after you change the subject line and delete all visible traces of the
original message, Yahoogroups links it into the old thread.

Thanks,
Steve Greenfield
listowner Homebrew_PCBs

--- In Homebrew_PCBs@yahoogroups.com, Cristian <bip@...> wrote:

>
> Hi List
> Do you know a solution to invert the polarity (from positive to 
> negative) of a Gerber generated file of a  PCB layer,
> independently of any PCB program?
> Cristian
>

----- Original Message ----- 

From: "Cristian" <bip@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Sunday, July 02, 2006 6:41 PM
Subject: [Homebrew_PCBs] Gerber Layer Polarity Inversion


> Hi List
> Do you know a solution to invert the polarity (from positive to
> negative) of a Gerber generated file of a  PCB layer,
> independently of any PCB program?

GC-Prevue might do it for you. The Pulsonix software I use will generate 
either positive or negative Gerber files.

Leon