HP97 The journey begins

[BobVA]

(06-22-2021 09:18 PM)Kees Bouw Wrote:  Can teenix please get some more space?

(hopefully someone in charge reads this as well)

+1

Easy for me to say, since I'm not doing the programming, moderation or paying the bills, but.... :-)

I think it would be a benefit to the community to select some of our more in-depth tech contributors for an upgraded storage limit. Or, alternatively, perhaps not count attachments in highly read/referenced threads against the OP's limit?

[anetzer]

(06-22-2021 09:18 PM)Kees Bouw Wrote:  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)

+1!

[teenix]

Hi all,

I've been busy behind the scenes. The circuit board is getting a bit complex now but getting there. I had a bit of difficulty sorting out the new motor drive. Normally not much of a hassle but I wanted to have programmable print width and print intensity and this required some fine tuning of the circuit. This was the main reason I could not complete the project using a single processor.

Both of these functions seem to work ok, but I won't know 100% until the new circuit board is made and running the software.

I have also installed a chip for the Constant Memory feature which can be disabled if required.

I'm still hanging off getting the board made due to little glitches I see and it's a lot to throw away if it's broken. It's amazing how everything looks fine until you look at it a few days later.

I've modified the main CPU program and written the (PIK) CPU program. These test ok during simulation and on a test jig, but because they have to interact with each other I won't know 100% until the board becomes operational.

What a ride :-)

cheers

Tony

[teenix]

Hi all,

I thought while I have the board almost ready, (always the case lately) that I would complicate it further by considering the HP-97S IO functionality. I am not going to provide an interface for the original HP IO module as that had direct connections to the 97 serial busses and a complicated way of providing digit data for the C register. I thought a simple RS-232 type of interface will be ok for the task as it doesn't have to be blisteringly fast.

There is a Bluetooth module on the board which might be able to offer this function but that will require an external board with a master Bluetooth module set up to connect with the 97 CPU.

I have added a chip to the board which will provide an isolated RS232 connection to the outside world via a four wire interface. This won't require any major setup, just a 5V supply and a host to talk to the CPU. I will publish a communication protocol for this when I get to testing it all. This setup will allow platforms like Arduino etc to interface to it and therefore have functions like analog to digital, digital input output, PC interface, GPS, pretty much anything. All it would have to do is format and send the data as 14 x 4 bit nibbles so it can be placed in the C register. The other functions that the original IO module had, like run from [A], will be enabled.

I've actually designed the board with 2 Bluetooth modules. I want to test a more modern one and see if it will have communications that are good enough to re-flash the CPU chip. Whichever is better will be the one used on the board.

cheers

Tony

[teenix]

Hi all,

The new 67 Continuous Memory circuit boards finally arrived. I built one and it seems to do what I expected. The card reader can be working through the continuous memory so I'll have to test that and if all is well should be right to go.

cheers

Tony

[teenix]

Hi all,

The HP-67 Continuous Memory CPU board and card read/write seem ok.

When enabled, the memory will become non volatile and will still be there after power is switched off and back on. As this process requires no power at all, this includes indefinite removal of the batteries.

The Continuous Memory will preserve all memory, C Register, LastX, DEG RAD GRD setting, program pointer, and the Default Key state. (A-E default functions)

The COMM switch has been removed as this function is now automatic.

The price for the new 67 CPU board is the same as the original one.

http://www.teenix.org

cheers

Tony

[teenix]

Hi all,

I took a deep breath and the new 97 CPU PCB design has gone to the fabricators. I hope it is right as these, being so large, are reasonably expensive to get made. Third time lucky :-)

cheers

Tony

[Massimo Gnerucci]

(07-14-2021 02:51 AM)teenix Wrote:  Hi all,

I took a deep breath and the new 97 CPU PCB design has gone to the fabricators. I hope it is right as these, being so large, are reasonably expensive to get made. Third time lucky :-)

cheers

Tony

Well done Tony!
Ready for a combo order with the new 67C board.

[teenix]

Hi all,

Well, that didn't take long, the new 97 CPU boards are here already.

There is a lot of new stuff on the board including Continuous Memory so I'll have to build and test sections at a time and then open up the modified dual processor software in stages.

Here we go... :-)

cheers

Tony

[Massimo Gnerucci]

(07-19-2021 05:25 AM)teenix Wrote:  Well, that didn't take long, the new 97 CPU boards are here already.

There is a lot of new stuff on the board including Continuous Memory so I'll have to build and test sections at a time and then open up the modified dual processor software in stages.

My wallet is already open. :)
Gotta have one! And a 67C, too.

[PANAMATIK]

(07-19-2021 07:23 AM)Massimo Gnerucci Wrote:  

(07-19-2021 05:25 AM)teenix Wrote:  .... the new 97 CPU boards are here already.....

My wallet is already open.
Gotta have one! And a 67C, too.

Mine too! Gotta have this wonderful HP-97 teenix board! I have a defective HP-97 with three missing components, ROM ACT PIK, which should be alive again.

A HP-67C does already exist in form of the HP-67E ACT chip for the HP-67 calculator, which can store 21 x 224 step programs in flash memory and includes the preprogrammed "Standard Pac" solutions, however it cannot read/write from/to the card reader.

Bernhard

[Paul Dale]

Agreed, very tempting.

[teenix]

Hi all,

The power supply is up and running.

The next stage is the main processor and check out the software operation without the printer.

Onward...:-)

cheers

Tony

[teenix]

Hi all,

A rather frustrating day today after trying to populate the board.

I don't know if it's me but it seems to be getting more difficult to solder some of the chips on the circuit boards I have been getting made lately - even after cleaning. It used to be that using flux gel and just whisking the iron tip over the multiple close spaced pins resulted in a nice clean solder job. Now, the soldering takes a few goes to get the job done and even then some pins won't solder properly resulting in trying to re-solder again and having to clean up inevitable solder bridges. Even some of the actual joints look a bit sad.

It is always a bit of a "squint the eyes" moment when applying power to a new board, and in this case, no smoke, but the PIC programmer would not recognize the PIC. I verified that the MCLR pin was not soldered to the board with a multimeter and I eventually got the programmer to recognize the PIC chip after about 10 soldering attempts of the MCLR pin using thin solder and a very fine solder tip.

Having damaged miniscule IC pins on delivery doesn't help either. If you miss a small bend on one of these pins, it can create a solder short to the adjacent one and is somewhat difficult to repair, and removing the chip from the board is a pain as well. The circuit board tracks are so small that they can peel off the substrate quite easily if one or more pins decide to stick.

I also added a picture of the 97 print head and ribbon cable. I don't know if it is normally attached to the flat side of the actual metal part of the print head or just sits loose like in the image. I thought if it has started to peel away I will have to put a small clip on it to stop it eventually tearing at the bottom.

I think I might take the rest of the day off

Fun with programming next... :-)

cheers

Tony

[schoudericrs29]

Hello everybody,

a long process, but everything goes on. All key functions and the printer are working. Except for some new formatting for the display of notes, timer, and so on, all code is still owned by HP.

[mfleming]

The soldering problem you describe is partly due to board quality. The typical low-quantity PCB prototype board needs brisk precleaning and drenching in flux when soldering. Gold plate pads OTOH need no flux. You should be able to specify that option (at an eye watering bump in price) to save time and damaged parts & boards.

I eventually went from hand soldering to reflow soldering using a homemade reflow oven to deal with fine pitch IC's and small SMDs. Another route is to drop a major bundle on a high end soldering station, something a friend swears by. Give the PCB upgrade a try first though, perhaps on a smaller project board. You might be surprised by the difference.

~Mark

[Hans Brueggemann]

(07-23-2021 05:10 AM)teenix Wrote:  ...
I verified that the MCLR pin was not soldered to the board with a multimeter and I eventually got the programmer to recognize the PIC chip after about 10 soldering attempts of the MCLR pin using thin solder and a very fine solder tip.
...

cheers

Tony

rather than using an electron microscope and a bee's knee as a soldering tip, try to go with either a 5mm (yes, 0.2") "hoof" or "spade" style solder tip and a very generous application of no-clean solder flux. worked like a charm on my 100-pin FPGA projects: the magic of using enough flux. it's really as simple as that.

[mfleming]

(07-24-2021 11:30 AM)Hans Brueggemann Wrote:  try to go with either a 5mm (yes, 0.2") "hoof" or "spade" style solder tip and a very generous application of no-clean solder flux. worked like a charm on my 100-pin FPGA projects: the magic of using enough flux. it's really as simple as that.

These are some magical videos! Important to note the author is using a quality soldering station and the proper tip as mentioned. The best surface finish is ENIG but specifying that bumps you out of the low cost prototyping category and into the regular commercial PCB category.

[teenix]

Thanks guys :-)

Maybe the flux I am using has degraded a bit. I bought it some time ago but has always been in the fridge when not being used. I'll get some more to try out on a next parts order

cheers

Tony

[teenix]

Hi all,

It's alive :-)

Lots more to go however, but encouraging.

I've only got the PC controlling the keyboard and remote programming to minimize HP-97 fiddling so the next step is the real keyboard and then on to the card reader. Hopefully the CRC code written for the 67 CPU board works here as well. If that all works then I'll check the Continuous Memory.

cheers

Tony