HP Forums

Full Version: ACT available
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
ACT Manual new chapter about HP-01

You can download the actual version of the ACT manual, which containes 15 new pages, describing the HP-01 version.

ACT manual

Thanks for reading.

Bernhard
Christmas ACT!

During December 2016 I will offer a special repair service for any defective woodstock calculator from model HP-21 to HP-29C by sending the unit to me for repair for the price of the standard DIY repair kit 80,- Euro + shipping.

Meanwhile I got some more experience and also a defective internal power supply can be repaired, which sometimes has damaged transistors and diodes. The chance of a successful repair has now increased from 98.5% to 99.8% Smile

Bernhard
HP-25E with Continuous Stack!

For everybody, who owns the ACT Update Kit, I just released two new versions 1.09 for the HP-25, a full version without GPS function, and a GPS version with reduced program area. Their only new feature is, that they preserve the user stack registers when switched off, as well as the display mode and trigonometric mode and actual program step. It is highly useful and I love it.

HP-25E 1.09

This is an improvement even above the HP-29C, which originally only preserved the X register in Continuous Memory.

Thanks to Mark Henderson, who has made this suggestion to me.

Bernhard
A very special repair in the new year!

The story:

In my collection of woodstock calculators, there was residing one item, an HP-25, that I bought last year as defective, an attribute, which I like for nearly all of my aquisitions. But this one refused to be repaired by the new ACT and all other means. Its display driver chips, both the cathode driver and anode driver, were damaged. Also the internal power supply didn't work. It was a hopeless case and was laying around for nearly one year, waiting for what to come.

But meanwhile, thanks to another HP-25 repair, which I had promised to a member here, I had to learn the hard way how to repair the woodstock power supply, by replacing active components like transistors, diodes and condensators by their modern equivalents. And I got its power supply repaired. Accordingly I could repair the power supply of my calculator as well.

But how to repair the display driver chips? I needed a replacement. I unsoldered both chips from another calculator of mine, a defective HP-19C, which was beyond repair for other reasons, but its display drivers seemed to be still working. I gave it a try. Yes, I got a running display, but not completely running, unfortunately some LED segments always were lit dimly, only the cathode driver was ok, and this was not a successful repair. The precious anode driver spare part was not completely healthy. I realized, that I was short in original display drivers.

[Image: uc?export=view&id=0Bwx8KUfOUL_RbVdrN2xNNkVYekk] [Image: uc?export=view&id=0Bwx8KUfOUL_RNllPTDdwQ0Y3OFE] Left the display with dot and segment e always on, right the display driver chips on sockets.

With the idea in mind, that I never should give up and finally have to repair all of my calculators, I was thinking about a display driver replacement, based on the knowledge about the timing betweeen the chips I had. I modified an newACT circuit and reversed its input and outputs, replaced many components and wrote a completely new program, which was able to read the display digits serially from the bus and to set the LED 7-segment patterns with exact timing. I found, that the 16MHz PIC processor was way too slow for the task listening to all of the 200kHz signals. I couldn't use interrupts because they were not fast enough, only polling of the clock signal was a chance and I had to count each machine instruction to get a result. Every embedded programmer knows, what it means to relinquish on any interrupts.

After several days working with the oscilloscope, I finally succeeded. The prototype was working. And it was no problem to integrate the small circuit into the calculator. I repaired my poor calculator, which was waiting so long for getting alive again. Smile

Here is the source code for this approach:

https://drive.google.com/open?id=0Bwx8KU...0prbHFXTlU

The following images show the prototype board attached to the pins of the anode driver chip and the running display. The brightness can be adjusted by the resistor values for each segment, I choose 470 Ohm to get a bright display for my calculator. This my HP-25 is fully functioning again.

[Image: uc?export=view&id=0Bwx8KUfOUL_RbUZ1QjFGbDdoaEk] [Image: uc?export=view&id=0Bwx8KUfOUL_RVm9XU2pWTi1WTHM] [Image: uc?export=view&id=0Bwx8KUfOUL_Rei1ZcWs3dE9XSVk] On the right the final placement

Don't be disappointed, but this anode driver replacement concept will not become available for public, because it replaces only the anode driver and therefore is not good enough. But I'm planning a design, now I have more knowledge, which will replace both display driver chips and has even more advantages. It will be available in late spring, as I can hope, and you will be really surprised about its features.

Happy new year!
Bernhard
Amazing as you've accustomed us to - congrats and happy new year!
Well done Bernhard, very nice!
All that's left now is to create a new LED display :-)
(01-08-2017 05:08 AM)teenix Wrote: [ -> ]All that's left now is to create a new LED display :-)

Right down the bottom: http://www.panamatik.de/html/hp_calculat...r_kit.html
Replacement LEDs.


Pauli
(01-08-2017 05:27 AM)Paul Dale Wrote: [ -> ]
(01-08-2017 05:08 AM)teenix Wrote: [ -> ]All that's left now is to create a new LED display :-)

Right down the bottom: http://www.panamatik.de/html/hp_calculat...r_kit.html
Replacement LEDs.


Pauli

As Homer says ... Doh! :-))
Bernhard Emese, aka 'Woodstock'!

Check your email Bernhard!
The prices for broken/used Woodstock Calculators is bound to go way up if it hasn't already, now that Bernhard has made so many useful products to completely refurbish these classic favorites!
(01-09-2017 12:14 AM)BarryMead Wrote: [ -> ]The prices for broken/used Woodstock Calculators is bound to go way up if it hasn't already, now that Bernhard has made so many useful products to completely refurbish these classic favorites!

A 3D printed case and a keyboard assembly and new ones will be possible!


Pauli
(01-09-2017 12:39 AM)Paul Dale Wrote: [ -> ]
(01-09-2017 12:14 AM)BarryMead Wrote: [ -> ]The prices for broken/used Woodstock Calculators is bound to go way up if it hasn't already, now that Bernhard has made so many useful products to completely refurbish these classic favorites!

A 3D printed case and a keyboard assembly and new ones will be possible!


Pauli

We can always hope the price for broken woodstocks will come down as we will no longer need broken calculators for spare parts. Of the dozen woodstocks I have, 3 are nearly bare boards cannibalized to get the others running.

I do have a case or two (or three) if someone needs one.
(01-09-2017 12:39 AM)Paul Dale Wrote: [ -> ]A 3D printed case and a keyboard assembly and new ones will be possible!

Somehow this falls in line with my darker musings : About any part of those cherished machines can and is now being tinkered with: Emulators, firmware, logic boards, IO. Some of the already achieved results are radical beyond belief...

The only thing I have yet to see is a new keyboard as elegant, as sturdy and as aesthetically pleasing as those of the golden era of HP.

Honestly: What is the main reason, one still graps for a real HP calc if not the touch and feel of those keys?

Still dreaming of pairing a bluetooth-enabled basic calc with a good rotate and click keyboard to, say, an emulator on my smartphone, which could then take over all cpu functions and offer the whole display area for extended display, printing or plotting...
(01-09-2017 03:19 AM)Nate Martin Wrote: [ -> ]We can always hope the price for broken woodstocks will come down as we will no longer need broken calculators for spare parts. Of the dozen woodstocks I have, 3 are nearly bare boards cannibalized to get the others running.

I do have a case or two (or three) if someone needs one.

Your bare boards will come to life again, if you wait for my HP-25E Low Power circuit, which will be available end of spring. The schematic is in its final stage. Main features are 20-40mA when display is on, Microamps when in standby mode, full Continuous memory like the LCD models, USB connector for program upload, download, 512k Flash memory, enough for all models in one, Real Time Clock, Buzzer, Infrared printing, easy to install by inserting it below the display area.

I'm not responsible for higher prices in ebay! If you calculate the ratio of avarage increase of other models 42S, 33C, 34C etc. in the last year, I think it will be the same as for the woodstocks. Nevertheless I admit, there could be less broken woodstocks on the market, because many of them are repaired now.

I got so much positive feedback and thanks for making these machines running again. Thanks to all of you, which are taking care of these old calculators.

IMO the woodstocks are in size and quality the best LED pocket calculators ever made by HP or in general.

Bernhard
HP-25E without 49 program step limit!

For everybody, who owns the ACT Update Kit, there is a new ACT version available. I just released version 1.10 for the HP-25E with dynamic program overlay capability.

It overcomes the HP-25 49 program step barrier. You now can virtually write HP-25 programs with more than 5000 program steps. Please read the new chapter about dynamic program overlay loading in the ACT manual 1.10 for more information. It is a milestone and highly useful and I hope you love it.

HP-25E 1.10

All new HP-25E ACT orders will be shipped with new version 1.10.

Thanks to Geir Isene, who has demanded this feature and made this suggestion to me.

Bernhard
Very nice addition!


Pauli
(01-27-2017 10:11 PM)PANAMATIK Wrote: [ -> ]HP-25E without 49 program step limit!

For everybody, who owns the ACT Update Kit, there is a new ACT version available. I just released version 1.10 for the HP-25E with dynamic program overlay capability.

It overcomes the HP-25 49 program step barrier. You now can virtually write HP-25 programs with more than 5000 program steps. Please read the new chapter about dynamic program overlay loading in the ACT manual 1.10 for more information. It is a milestone and highly useful and I hope you love it.

HP-25E 1.10

All new HP-25E ACT orders will be shipped with new version 1.10.

Thanks to Geir Isene, who has demanded this feature and made this suggestion to me.

Bernhard

This is the sort of thing that makes me want to type an excessive number of exclamation marks in a row. I'll restrain myself - !!!!!!!!!!!!!!!!!!!!!!!!! - 25 only! (plus one).

Nigel (UK)
(01-28-2017 10:12 AM)Nigel (UK) Wrote: [ -> ]This is the sort of thing that makes me want to type an excessive number of exclamation marks in a row. I'll restrain myself - !!!!!!!!!!!!!!!!!!!!!!!!! - 25 only! (plus one).

Nigel (UK)

No, no, only one exclamation mark would have been enough - ! Thank you. Smile

If the HP-25E ACT would just be a functional simulation like in some HP calculator apps, this would have been a more easy task. But it emulates the original firmware.

Here some inside views:

First I had to find the code pieces, where the f ENG 7-9 instructions were executed. After several tries with a self written debugger I found address 01427 will be called by an indirect jump for either FIX SCI and ENG 0-9 instruction. It will be called also when entering one of these as program step in PRGM mode, so I had to filter out this condition. I had to verify, that it will be called only when program is running, when SST (and not when BST) is executed, and when executed manually by typing f GTO 9.

In register a[0..1] could be found the code of the instruction, which distinguishes between the display format instructions. ENG 9 code is 0x79. When emulating the HP-25 firmware, each instruction is checked whether it is executing at address 01427. If ENG 9 is executed, then execution is interrupted and the 49 byte contents of the program memory is replaced by the new program by reading from the library. Fortunately address 01427 is called after the program step counter is already incremented, so I could replace it just by 01 to resume from step 01, which is coded in register m[1..2]. If I would had to manipulate the counter before increment, I would have set its value to 00, which causes program execution to stop.

Next the ENG 9 instruction should not change the display format, so I had to manipulate the address to skip the change of the display format by setting pc to address 01743 and execute the next machine code from there. You see I still emulate the original HP-25 code and didn't change the firmware.

For overcoming also the HP-29E 98 program step barrier, I will have to analyze its behaviour in similar way. Although this is not critical, because it is already double the program space of the HP-25, I think this will follow soon.

At the end it sounds logical and not at all complicated, but to find the right path is not always easy.

Thanks again
Bernhard
(01-28-2017 01:41 PM)PANAMATIK Wrote: [ -> ]For overcoming also the HP-29E 98 program step barrier, I will have to analyze its behaviour in similar way. Although this is not critical, because it is already double the program space of the HP-25, I think this will follow soon.

HP-29E Version 1.10 with more than 3000 program steps for a contiguous program is released.

HP-29E 1.10

This was faster than I thought. Smile

Bernhard
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Reference URL's