HP-19C Repair Odysee

[PANAMATIK]

Hi Tony,

Indeed the HP-19C PIK also generates a PWM modulated signal on all 7 print pixels for controlling their intensity. This is done by gating all pixel signals R1-R7 with the STB signal. The STB (PIK/Pin 28) signal is a continuous 10kHz signal, generated by an external oscillator, with normally 50:50 Duty cycle, if battery is fully charged. It increases the duty cycle when battery gets low. The signal level of R1-R7 is 1V, which fully switches the driver transistors without base resistor. In the newPIC I will add base resistors to get the same base current with 5V levels. The gating will be done by external NAND gates and I will generate a STB signal internally on the CCP1 PWM output. The PIC16F1519 has capability to measure the actual battery voltage and adjust the duty cycle of the STB signal. Then intensity control will still work even when the HP-19C STB signal is faulty or not generated at all.


screenshot of print pixel R1 and STB intensity control (original HP-19C PIK while printing).

Getting closer and closer to the final concept.

Bernhard

[teenix]

Yes, the pixel length will change if just changing the actual pulse width for them.

I used the mathematical formula that HP used for the motor/head driver. The controller uses this to determine the PWM signal.

I wasn't sure of the gain of the driver transistors as I cannot find data on them, (I thought something like PN2222) but if it is quite low I wasn't sure if the PIC could drive all of them directly. I figured the PIK chip might have some internal resistance to limit base current.

cheers

Tony

[[kby]]

Tony,
On your schematic for the HP-10 it is the 20k pot in the upper left corner labeled “motor speed.” I know this works as I’ve done the adjustment.

(05-09-2021 02:16 PM)teenix Wrote:  Looking at Tony Duel's circuit, the 50K pot should adjust the motor speed, and the 200K pot should adjust the intensity.

cheers

Tony

[PANAMATIK]

Hi all,

As described in post #32 my HP-19C printer motor was way too fast, whatever poti position selected.

I checked the printer board transistors and components a thousand times, replaced ICs and compared the components and wires with a known good board: nothing to find. Then finally confused, I thought that the big motor capacitor did not have a value in Tony Duells schematics, I unsoldered it to measure its capacitance (1.5uF). Then helplessly I soldered it back to the motor, assembled the two boards and tried again.

And it works!



The capacitor wasn't soldered properly and had lost its contact to the motor, but it is essential for generating the VM signal, which controls the motor speed.



I printed on the same paper from the days before with the reverse printed number Pi. But now I could change the motor speed with the poti. "0.00" is printed with too slow speed, "3.14" with the right speed.

But this was not the "totally corroded" HP-19C, it was my second HP-19C. The challenge for the PIK replacement continues, and signs are good, because I know much more about the HP-19C printer now.

Bernhard

[PANAMATIK]

A next milestone for the newPIK chip is achieved!

Today for the first time I placed the newPIK prototype board on the HP-19C printer board. With keyboard buttons, which are only test buttons for now, I was able to drive the motor with PWM controlled different speeds. The FWD, REV and BREAK signals are working as expected. With the PRX button I could start the line printing sequence, the motor moves 1s with slow speed forward, then with full speed backward until it reaches the HOME switch position, where the motor windings are shorted 1/2 s for fast stop. This works very well.



HP-19C main board with newPIK prototype stacked on "printer/power supply board"

One problem is left: The VM signal (Velocity Motor?) goes Low during the PWM Low phase if the speed is too slow, but the High level is only 1.8V, I cannot read it on TTL digital input. Therefore I cannot regulate the motor yet.

There next tests will be: regulating the motor speed by amplifying the VM signal
and, last not least, driving the printhead pixels.

I hope this will be done soon, then I can draw the schematic for the newPIK. I have also the idea to include infrared printing for the unfortunate cases, where the printhead is damaged, then the newPIK will be able to make these HP-19Cs printing again by sending the identical lines to the HP82240B printer, when an Infrared diode is installed.

[Edit: One day later I can regulate the motor speed and adjust the speed with the 50k Poti. I achieved to read the VM signal on an analog input every ms and now use it for regulation of the motor speed, also the 7 pixels printhead pattern is output every ms]

Bernhard

[PANAMATIK]

The newPIK Prototype is complete and controls the printer motor!

Motor Video (16s 5.5MB) This is my first mp4 video in this forum. I hope you can play it in your browser. If not you can download the file.

I programmed the motor regulation for the HP-19C printer motor. The motor moves faster or slower when I turn the 50k Poti, clockwise is slower, counter clockwise is faster. I'm totally happy with it. Also the printhead pixels are driven and generate the Italic typeface like the original. I did not yet print on paper, because I need a hardware modification, because the STB signal must be anded to all 7 pixel signals. The duty cycle of the STB PWM signal is dependant on the battery voltage and the 200k Poti, and adjusts the printer intensity. I checked the printhead signals on oscilloscope. Timing is not critical. Keyboard buttons can be detected while printing, like in the original HP-19C.

This PIK firmware, when finished, should also work in the HP-97, but also needs the new ACT inserted. The challenge remains to make it compatible with the HP-97 and HP-19C with an original ACT inserted.

Next steps: Drawing the schematics, then designing the layout.

Bernhard

[Harald]

That looks great! Put me on the waiting list, I still have the 19C in the box that needs a new PIK!

[PANAMATIK]

(05-18-2021 10:35 AM)Harald Wrote:  That looks great! Put me on the waiting list, I still have the 19C in the box that needs a new PIK!

Hi Harald,

Your are number 1 on the waiting list (after number 0 which is me).

I hope there will be more HP-19C machines to resurrect. All HP-19C calculators can be repaired with the new PIK if only the printhead heating elements are OK. All other components can be replaced by modern equivalents. In the rare cases of defective display drivers I will also offer a repair solution.

Bernhard

[jjohnson873]

(05-17-2021 06:25 PM)PANAMATIK Wrote:  The newPIK Prototype is complete and controls the printer motor!

Motor Video (16s 750kB) This is my first mp4 video in this forum. I hope you can play it in your browser. If not you can download the file.

Hi Bernhard....I can hear the audio of the printer, but I don't see any video, just a black screen. I can't seem to download the MP4 file either.

Happy to see the HP-19C repair moving forward. I've got a few HP-19C's to work on, sitting on my workbench.
~ Jim J. ~

[AndiGer]

No problem here to watch the video - even on (nearly vintage) iphone 4s

OK - on windows 10 computer same problem - only the sound :-)

[PANAMATIK]

(05-18-2021 05:09 PM)jjohnson873 Wrote:  Hi Bernhard....I can hear the audio of the printer, but I don't see any video, just a black screen. I can't seem to download the MP4 file either.

I also get only audio on my Windows Firefox, but I get video on my macOS Safari. With Firefox I could at least download the file with right mouse button and play.

Does anybody know why a .mp4 file sometimes plays only audio?

Bernhard

[rprosperi]

On Windows, FireFox, Edge and Chrome all bring up only an embedded audio player, but it's easy to right-click, then save-as...

It has to do with the codec that's used when the video was made. Even though it's ".mp4" there are many codecs that are lumped into .mp4, and modern browsers don't have the matching codec for the video stream used when that was made. I can only guess, but I'd assume the video was made with either a rather old device, or perhaps a very new one, and that specific video codec support ended, or has not been widely adopted yet.

Most likely case, the video was made on an iPhone, so it naturally plays fine on Safari.

-- Well sure, those walls are very, very high, but isn't the garden lovely today?

[PANAMATIK]

I used another video codec and now it should work. Use Link in post #46

Bernhard

[rprosperi]

(05-18-2021 06:14 PM)PANAMATIK Wrote:  I used another video codec and now it should work. Use Link in post #46

Confirmed, the new link opens up video/audio no problem w/FireFox on Windows.

[[kby]]

(05-18-2021 05:57 PM)rprosperi Wrote:  On Windows, FireFox, Edge and Chrome all bring up only an embedded audio player, but it's easy to right-click, then save-as...

It has to do with the codec that's used when the video was made. Even though it's ".mp4" there are many codecs that are lumped into .mp4, and modern browsers don't have the matching codec for the video stream used when that was made. I can only guess, but I'd assume the video was made with either a rather old device, or perhaps a very new one, and that specific video codec support ended, or has not been widely adopted yet.

Most likely case, the video was made on an iPhone, so it naturally plays fine on Safari.

-- Well sure, those walls are very, very high, but isn't the garden lovely today?

That means the likely codec was HEVC and the video was made on a recent iOS flavored device. It’s possible to get the recording format to default to H.264 which is more widespread, but I don’t think there’s a “change it just thus once” method other than manually switching the default.

[PANAMATIK]

(05-17-2021 06:25 PM)PANAMATIK Wrote:  The challenge remains to make the newPIK compatible with the HP-97 and HP-19C with an original ACT inserted.

I found a new PIC processor PIC16F18877 for this project, which might be able to achieve this. It should be able to communicate via ISA and DATA line concurrently, which its necessary for communicating with the original ACT. It is also 2x faster than my previously used processor PIC16F1519 and has some other advantages.

But the combination of the newACT (Arithmetik Clock Timing circuit) and newPIK (Printer Interface Keyboard circuit) will be my preferred solution, because it allows to add new features like more memory, real time clock, alarm time beeper, sleep Mode etc.

Still I see the Microchip PIC processors in advantage against a FPGA solution (like tinyFPGA), because FPGA chip sizes and voltage levels does not match the requirements.

Bernhard

[Jonas Sandstedt]

Hello,

I do not know if this will help the Odysee, but when I read about using the printhead of the HP82240 printer, I remember that this printer uses a Seiko printer that is still available (but is about to be discontinued), see

https://www.keesvandersanden.nl/calculat...inside.php

The printer is MTB201-166, and at the page above there is a link to a technical specification (press "MTP201-166") where the printhead is described in detail, and in the end it is also described how to remove the printhead.

The smaller printer MTB102 is adapted for 38mm paper width.

Kees, I hope it is OK to refer to your fine homepage.

best regards
Jonas

[PANAMATIK]

(05-28-2021 09:32 AM)Jonas Sandstedt Wrote:  I do not know if this will help the Odysee, but when I read about using the printhead of the HP82240 printer, I remember that this printer uses a Seiko printer that is still available (but is about to be discontinued), ...

Very interesting! I just ordered one.

Thanks Bernhard

[Jonas Sandstedt]

Hope it works, Bernhard.

I just want to correct my misspelled printer reference, it is MTP, not MTB.

best regards
Jonas

[PANAMATIK]

My HP-19C PIK prototype was not longer to handle. The ratsnest of wires caused a lot of problems, broken wires, unreliable connections, it was difficult to attach probes. At the end nothing worked, I was deperate and could not track down any software and hardware errors. Finally I threw it away and made a new prototype board.



terrible to work with



this works much better!

Bernhard