Hi all,
Some more progress.
I have removed the original printer interface circuit board and have my CPU board driving the print motor and print head directly. I set the motor running and tried to slow it down with finger pressure and it didn't want to, and also I made small changes in battery voltage and that did not have any appreciable effect either so I assume the new motor driver is working ok.
I have removed the idler cog so the carriage doesn't actually move during this early testing as I don't want to damage anything. Unlike the 19C printer, this mechanism can be damaged if the carriage moves faster than the print data is being sent out. The print head is also disconnected until I know for sure that the code will not keep a pixel on too long and risk damage.
If the real 97 goes haywire and the printer motor stays energized either forward or backward then the carriage will jam hard against the printer frame. I cannot see any original mechanism to detect this error, so some serious damage may result. The HP-41 printer has an extra microswitch to detect the carriage limit while printing. This might suggest why the original idler cog is made out of some sort of resin instead of metal. Apart from possibly quieter operation, it might also fail (strip teeth) under improper conditions and avoid other damage. The new motor driver has an overcurrent detect so if the motor stalls for some reason the power will be removed.
At this stage the printer is only doing a line feed from a simulated press of the paper advance switch. This all probably doesn't sound much but when you consider what is actually happening to make this occur it is quite a lot.
The print motor is responding to the HP-97 microcode and the instructions being fed to the simulated PIK chip. The motor moves forward, brakes and then reverses to home, brakes and switches off. I have test points that activate in my code to simulate the home switch operating.
Onward...
cheers
Tony
Hi all,
With the motor running, the first print output to the print head as seen on my scope.
The lower part shown the PWM overlaid on each pixel which keeps the print intensity constant with changes in battery voltage. You can also see the staggering of the pixel lines.
It looks backward because the printer prints right to left.
The key buffer is also working. I can simulate pressing PrintX while the printer is active and it keeps printing until the buffer is empty.
Its somewhat interesting in that it takes the PIC around 2.7uS to process and output the data to the printer as the carriage moves across. The PWM is automatic and needs no code intervention.
Lots to go yet.
cheers
Tony
Hi all,
I just found a potential problem with the 97, but I suspect there is only a very small chance it will actually happen.
There is a possibility that the print head is not in the home position when the calculator is switched on. Reasons for this could be, flat battery, switching off before printing finishes, etc.
If this is the case, then it will try to automatically home itself at switch on so it is ready for printing.
This is fine, but if the print head doesn't get detected as being home, the motor keeps on running continuously. The calculator still runs as normal, although you cannot print anything.
The problem is that if the print head can move, it will butt up against the printer casing and stall the motor. This may result in sheared teeth off the resin printer cog, or as I cannot see any current limiting circuitry, may burn out the motor or driver transistors.
This situation can occur (for example) if the home switch circuit does not work due to a fault or if due to maintenance, a white wire was left disconnected.
If a drive cog was broken then no risk of damage as the print head cannot move.
I haven't heard of this situation arising but you never know.
cheers
Tony
(06-17-2021 06:47 AM)teenix Wrote: [ -> ]The problem is that if the print head can move, it will butt up against the printer casing and stall the motor....
Therefore I used a HP-19C printer module for developing. It cannot stall.
Perhaps install a "normally closed" microswitch at either side to interrupt the motor wire during software development.
Bernhard
Some people are lucky :-)
Actually the original HP-97 has that problem if the home switch circuit wont close when required.
I should be able to connect the print head today. I have the PIC timeout feature enabled while printing so hopefully if the print routine fails, the PIC will reset and make sure the heads are all off.
cheers
Tony
(06-17-2021 05:58 PM)PANAMATIK Wrote: [ -> ] (06-17-2021 06:47 AM)teenix Wrote: [ -> ]The problem is that if the print head can move, it will butt up against the printer casing and stall the motor....
Therefore I used a HP-19C printer module for developing. It cannot stall.
Perhaps install a "normally closed" microswitch at either side to interrupt the motor wire during software development.
Bernhard
(06-17-2021 06:47 AM)teenix Wrote: [ -> ]Hi all,
I just found a potential problem with the 97, but I suspect there is only a very small chance it will actually happen.
There is a possibility that the print head is not in the home position when the calculator is switched on. Reasons for this could be, flat battery, switching off before printing finishes, etc.
If this is the case, then it will try to automatically home itself at switch on so it is ready for printing.
This is fine, but if the print head doesn't get detected as being home, the motor keeps on running continuously. The calculator still runs as normal, although you cannot print anything.
The problem is that if the print head can move, it will butt up against the printer casing and stall the motor. This may result in sheared teeth off the resin printer cog, or as I cannot see any current limiting circuitry, may burn out the motor or driver transistors.
This situation can occur (for example) if the home switch circuit does not work due to a fault or if due to maintenance, a white wire was left disconnected.
If a drive cog was broken then no risk of damage as the print head cannot move.
I haven't heard of this situation arising but you never know.
cheers
Tony
How do other printers handle this situation? Eg 82143A, 82162A, 82240A/B
(06-18-2021 01:43 AM)dmh Wrote: [ -> ]How do other printers handle this situation? Eg 82143A, 82162A, 82240A/B
Unsure, I don't have access to one to play with :-(
cheers
Tony
Hi all,
I turned on the CPU power after connecting the print head and the current meter slowly started to move to maximum. So while having a mild heart attack and mental breakdown I quickly shut it down. All good though, the power supply has current limiting.
The driver chip I chose for the print head is supposed to have resistors to ground inside so I figured I would not need any external ones. It seems I was wrong. I fitted some pull down resistors and gingerly switched back on and the current was normal. It seems the inputs to the driver chip were floating and probably started oscillating.
After my nerves settled a bit, I pressed the test jig Print X button and...Yayyy!!
Here is the first printout from the 97 CPU board with -12345678.12 on display
As you might notice, it is upside down and a bit feint but that is ok, at least it proves all the software seems to work.
The upside down is easy to fix and the intensity can be changed by altering the PWM set point.
The other thing is the motor speed is a bit fast. The print line should have completed about 15mm earlier, but that is just a change in the motor speed set point.
This is encouraging...
cheers
Tony
Hi all,
Still a bit wide but looking much better.
cheers
Tony
(06-18-2021 06:40 AM)teenix Wrote: [ -> ]Hi all,
Still a bit wide but looking much better.
cheers
Tony
Great achievement! Bravo! Perfect! Congratulations.
I'm still waiting for my HP-19C PIK layout to be finished, before having my first print line.
Bernhard
(06-18-2021 08:16 AM)PANAMATIK Wrote: [ -> ]Great achievement! Bravo! Perfect! Congratulations.
I'm still waiting for my HP-19C PIK layout to be finished, before having my first print line.
Bernhard
Many thanks and right back at you.
I have to admit at first I didn't think emulating this would be such a challenge, but it's certainly been that and a bit more. More of those to come no doubt
cheers
Tony
(06-18-2021 06:40 AM)teenix Wrote: [ -> ]Hi all,
Still a bit wide but looking much better.
cheers
Tony
Bravo!!
,,,
!!!!\\\\
'`!_ ||||
` \`-'''|
`\ /
)\ \
/ \ \
\
CLAP
////
!!!! _ \\\\
! -'/ _ ||||
\ / \`-'''|
\ | \ /
) | \ \
/ | \ \
\
,,,
!!!!\\\\
'`!_ ||||
` \`-'''|
`\ /
)\ \
/ \ \
\
CLAP
New toys, how could I resist :-)
cheers
Tony
Not "I'm sorry Dave"??
I could send down an 82240 if it would be useful.
Pauli
(06-18-2021 11:34 AM)Paul Dale Wrote: [ -> ]Not "I'm sorry Dave"??
I could send down an 82240 if it would be useful.
Pauli
Not yet :-)
I was looking to buy one of those but they seem to be expensive.
I saw some advertised on YowCow cheap but never heard of them so don't know how reliable.
I was thinking it would be cool to run the 67 CPU board with a 97 core (which is essentially the same thing) and have the printer run from an IR, or Bluetooth to IR interface.
cheers
Tony
Hi all,
Well, those few forward steps have been overtaken again by some backward ones. Even though the printer is working ok, there is a problem running the printer code with the rest of the tasks the PIC chip has to do. I tried running things in different forms, but there is no way to get it all to work properly with one processor chip.
There are 5 major processing channels to deal with which is proving to be one too many. Although the chip can process everything in the required time, it cannot do so without upsetting the print output, the display and possibly with the card read write. These all operate in real time, so any interference to the data flow by trying to do too much will cause glitches.
The actual HP PIK chip runs the printer as a separate process, so it looks like the CPU board is going to need a separate processor to manage the printer as well :-(
cheers
Tony
(06-20-2021 08:52 AM)teenix Wrote: [ -> ]The actual HP PIK chip runs the printer as a separate process, so it looks like the CPU board is going to need a separate processor to manage the printer as well :-(
Exactly also my experience! Without FPGA, which could handle that, you need two PICs. That is also my concept with the HP-19C repair kit, where an ACT PIC processor and a PIK PIC processor are playing together.
Bernhard
Hello,
Regarding the problem with a printhead end position not being detected, I think the HP-97 solution with a bi-directional motor is somewhat rare. The HP-82240A/B for example, being equipped with the Seiko MTP-201 printer, works like the HP-19c where the motor is uni-directional and the printhead moves from the left to the right and so on while the motor is running in one direction. The only other calculator I have stumbled across that has a thermal printhead with a bi-directional motor is the Canon FP-11P, and just as for the HP-97, this printer only prints the line as long as necessary, where paper feed is the shortest line. I suspect that this does not help, but can be regarded as some superfluous information for the curious.
best regards
Jonas
Hi all,
Here is the new twin processor CPU board for the 97.
I'll leave it sit for a few days, then go back over it. Due to the shape, these boards are costly to make, so hopefully I'll catch any mistakes after a breather.
I'll have to clean out the old Forum attachments, I'm almost full.
cheers
Tony
(06-22-2021 04:35 AM)teenix Wrote: [ -> ]Hi all,
Here is the new twin processor CPU board for the 97.
I'll leave it sit for a few days, then go back over it. Due to the shape, these boards are costly to make, so hopefully I'll catch any mistakes after a breather.
I'll have to clean out the old Forum attachments, I'm almost full.
cheers
Tony
Please do not delete Forum attachments! They are all very valuable.
Can teenix please get some more space?
(hopefully someone in charge reads this as well)
Kees.