Threaded Mode | Linear Mode

AndiGer · 10-03-2021, 03:57 PM

Tony,

As I understood you still had the design in Protel PCB software?!?

If so - tried to use your layout in this current version?
https://en.freedownloadmanager.org/Windo...-FREE.html

Hope it will work for you
Andi

teenix · 10-04-2021, 02:26 AM

Hi Andi,

Thanks for the help, even though it is a pain, I think I will just have to migrate the project.

cheers

Tony

teenix · 10-09-2021, 03:11 AM

Hi all,

After days of playing around with modern PCB design software I have come to the conclusion that the automation has well overtaken manual methods, and I am sure it is good in many ways, but for an old bloke like me used to doing things a certain way, it will be a hard grind to go down that path. Just so complicated to use even for what I think should be simple tasks.

Anyway, I ended up finding an old laptop in good condition that was really cheap. I have file copies of the old design software and transferred that to the new PC. I tried running the schematic and it groaned about missing DLLs which led me to a few groans of my own. I tracked the listed offender down on the internet and it then groaned about missing another which I could not find any reference to.

I tried the PCB designer and it started ok, but unfortunately, I forgot that it needs multiple passwords to activate. Luckily I have a bitmap screen shot of the original install window (which was legitimate software in its day). I also found the original readme file which has a list of the install files and it listed about 10 DLLs, so I thought finally I've run into the brick wall.

I noticed on the acquired laptop power supply that you can change the output voltage. My old seemingly dead laptop works on 15V, and most others I've seen work on 19V. This supply has both. Ok, I'll try this supply on my old PC. Murphy is still there as the plug was a larger size. Slumped shoulders again. I then found a DC socket in my junk box and the plug fitted, so a bit of soldering and verify the polarity, I plugged it into the old PC and what do you know, it sprang back to life. Great, it seems the old power supply died. Why it never ran on the temporary supply I'll never know. I can now copy the DLLs and back them up.

I turned on my new WIN 10 machine with my backup disc and an unhappy Windows face appeared and said a problem occurred and it shut down. It would then not power up at all. You've got to be kidding. I think Murphy is embedded in my finger tips, everything I touch seems to stop. Cup of tea time.

I copied the required DLLs to the laptop I just got and all the software sprang into life. So now I have 2 machines running the software properly. My older Win7 machine has WinPro running which has a Windows VM for XP mode. I haven't tried but I think it will work there too now that I have all the files restored. The screen died ages ago but works with HDMI.

After I finished my cuppa, the Win10 machine decided to come back to life after holding down the power button for a lengthy time.

A long sorry tale, but in the end, I can now try to finish the 97 design and hopefully it will be the last change.

cheers

Tony

Didier Lachieze · 10-09-2021, 05:52 AM

What a journey ! Congrats for your perseverance.

BobVA · 10-10-2021, 02:50 AM

(10-09-2021 03:11 AM)teenix Wrote: ...I think Murphy is embedded in my finger tips, everything I touch seems to stop. Cup of tea time...

I'm having one of those weeks myself, so your eventual triumph is encouraging! :-)

(I was millimeters away from a successful repair on a 90's vintage audio "treasure" today but managed to snatch defeat from the jaws of victory with one tiny solder bridge. Followed by the drive to the E-waste facility, reflecting on my poor choices.)

Onward! :-)

teenix · 10-10-2021, 07:43 PM

Ouch, that must have hurt :-(

cheers

Tony

hpux735 · 10-17-2021, 07:39 PM

I just finished catching up on this whole thread. What a journey!!

Thanks for all you've been doing Tony. I've inherited an HP-97 from my grandfather-in-law. It requires a fair number of repairs. Your Classic Notes document has proved invaluable.

I've been able to repair the printer, but the LED display and card reader repairs have remained elusive.

I'm pretty certain that the card reader is broken for want of a new Sense Amplifier IC (I've been unable to find replacements anywhere). I've been thinking hard about making a modern replacement of the card reader (WiFi or Bluetooth, similar to your module. The problem, if you're curious, is that the motor doesn't always engage when the !MOT signal is sent by the CPU board. It is sometimes, and the card is pulled through, but often it does nothing. I've attached a trace from my Saleae that is such a case. (MS is the signal from the CPU board to start the motor, and you can see that the motor drive voltage only gets to 1.6 volts. After 1.5 seconds, the CPU throws an error)

Given that you're apparently still working on a new CPU board for the 97, I'm considering just not repairing the card reader and using the new CPU board with the Bluetooth module.

The other issue I have is the LED. It has segment 'f' lit on all the digits, and I believe that implicates the anode driver IC. It seems like this IC is also unobtanium, so I'll probably have to replace it with discrete transistors.

Either way, I've learned that if someone is working on a project such as yours, and you think you might need it some day, buy it immediately. So, if it helps in any way, once your project is ready I'm a certain buyer.

Thanks!
- Will

teenix · (This post was last modified: 10-18-2021 12:05 AM by teenix.)

Does the card motor actually work? Maybe it is loading the sense chip too much if it has a problem. Maybe the capacitor attached to the motor connectors has failed.

Unlikely if it hasn't been moved much, but make sure the ribbon cable traces are not cracked. Maybe wiggle the ribbon in case some high resistance has built up in the connector.

At least with the CPU board, there is storage for 800+ cards so if the card reader is unserviceable, it won't matter.

The [F] segment lit could be a fault in the display, or some corrosion on the display board or an internal short. It is tricky to get the display out but possible. Be careful not to damage the display ribbon cable as it only has single wire strand connections which may break. I would not advise removing the ribbon cable from the connectors as they are held tight and the wires may break. If you have to, be careful.

The driver outputs are only transistors, so if you wanted to you should be able to test the LED display in situ with the calculator switched off, by connecting a (say) 3K3 resistor from the 5V battery and touch the outputs of the anode driver with any cathode output shorted to ground. The upside down construction of the calculator makes things difficult though.

It is hard to isolate the led connections between the circuit boards for testing which is a problem. Please don't be tempted to bend the board to board pins to isolate them, as they "will" break and are also unobtainium. Even with the boards connected, the LED data from ROM0 could be checked to see if [F] is continuously on from there.

I had some small PCBs made that should connect as a replacement for the anode driver in the 97. It is based on MMBT2907 surface mount transistors (which I think are similar to those inside the anode drivers). The only trouble is, I had the transistors upside down :-(

I could get a few new ones made up when I send the (hopefully last) HP97 board design in to the fabricator.

cheers

Tony

(10-17-2021 07:39 PM)hpux735 Wrote: I just finished catching up on this whole thread. What a journey!!

Thanks for all you've been doing Tony. I've inherited an HP-97 from my grandfather-in-law. It requires a fair number of repairs. Your Classic Notes document has proved invaluable.

I've been able to repair the printer, but the LED display and card reader repairs have remained elusive.

I'm pretty certain that the card reader is broken for want of a new Sense Amplifier IC (I've been unable to find replacements anywhere). I've been thinking hard about making a modern replacement of the card reader (WiFi or Bluetooth, similar to your module. The problem, if you're curious, is that the motor doesn't always engage when the !MOT signal is sent by the CPU board. It is sometimes, and the card is pulled through, but often it does nothing. I've attached a trace from my Saleae that is such a case. (MS is the signal from the CPU board to start the motor, and you can see that the motor drive voltage only gets to 1.6 volts. After 1.5 seconds, the CPU throws an error)

Given that you're apparently still working on a new CPU board for the 97, I'm considering just not repairing the card reader and using the new CPU board with the Bluetooth module.

The other issue I have is the LED. It has segment 'f' lit on all the digits, and I believe that implicates the anode driver IC. It seems like this IC is also unobtanium, so I'll probably have to replace it with discrete transistors.

Either way, I've learned that if someone is working on a project such as yours, and you think you might need it some day, buy it immediately. So, if it helps in any way, once your project is ready I'm a certain buyer.

Thanks!
- Will

hpux735 · 10-17-2021, 11:47 PM

(10-17-2021 11:36 PM)teenix Wrote: Does the card motor actually work? Maybe it is loading the sense chip too much if it has a problem. Maybe the capacitor attached to the motor connectors has failed.

It does, which is part of the weird thing. I carefully desoldered the positive lead to the motor, and powered it independently with my lab power supply. It started slowly turning at about 1.9v, and it was easily able to rotate and pull cards through at about 2.5v. In fact, I sent several cards through the reader this way and made sure that the RA and RB signals looked healthy (they did)

(10-17-2021 11:36 PM)teenix Wrote: Unlikely if it hasn't been moved much, but make sure the ribbon cable traces are not cracked. Maybe wiggle the ribbon in case some high resistance has built up in the connector.

At least with the CPU board, there is storage for 800+ cards so if the card reader is unserviceable, it won't matter.

Yes, the ribbon cable is suspect, of course, but I did measure Vbatt, the control signals, etc.. on the card reader side of the ribbon and all those signals looked to be in good shape. If the stars align, and the motor's working, (remember, it's intermittent) I'm able to write cards without an error, but I can't read anything, so it's impossible to verify.

(10-17-2021 11:36 PM)teenix Wrote: The [F] segment lit could be a fault in the display, or some corrosion on the display board or an internal short. It is tricky to get the display out but possible. Be careful not to damage the display ribbon cable as it only has single wire strand connections which may break. I would not advise removing the ribbon cable from the connectors as they are held tight and the wires may break. If you have to, be careful.

It is hard to isolate the led connections between the circuit boards for testing which is a problem. Please don't be tempted to bend the pins to isolate them, as they "will" break and are also unobtainium.

This is very helpful, thank you. I'll look at it more closely than just assuming it's the inductor-based failure case you describe in your notes document.

(10-17-2021 11:36 PM)teenix Wrote: I had some small PCBs made that should connect as a replacement for the anode driver in the 97. It is based on PN2222 surface mount transistors (which I think are similar to those inside the anode drivers). I haven't had a chance to try it yet. I could send one, but the postal service has gone weird here and overcharges big time for tiny packages like this. It stinks because if I send the exact same size and weight in paper the cost is 1/7th. I could try sending one in a 'get well' card ;-)

That's a brilliant idea! I have a PCB mill, so I could go ahead and make some of those myself. Would you mind sharing the Gerbers for it, and the exact part numbers for the transistors (I've been burned by similar parts having different EBC/ECB/etc. layouts.

Thanks!!
- Will

teenix · 10-18-2021, 12:37 AM

(10-17-2021 11:47 PM)hpux735 Wrote: This is very helpful, thank you. I'll look at it more closely than just assuming it's the inductor-based failure case you describe in your notes document.

That's a brilliant idea! I have a PCB mill, so I could go ahead and make some of those myself. Would you mind sharing the Gerbers for it, and the exact part numbers for the transistors (I've been burned by similar parts having different EBC/ECB/etc. layouts.

Thanks!!
- Will

There are no inductors in the 97, just simple transistor switching. I wouldn't have thought a transistor shorting in the driver, but you never know. I would think more likely a display failure, but I hope not.

Hmmm, I modified my post about the transistor but must have hit the wrong button, no matter.

The transistors should be PNP types, I was looking at MMTB2907. My pads were setup as CBE, but should be EBC.

I will redo the board today and pm the gerber files if you like. I'll also get some made up when the 97 board goes in.

cheers

Tony

teenix · 10-18-2021, 02:05 AM

I redesigned the anode driver board, the gerber files are here in zipped format.

http://teenix.org/HP0044.zip

I'll leave them there few a few days.

cheers

Tony

hpux735 · 10-18-2021, 02:32 AM

(10-18-2021 02:05 AM)teenix Wrote: I redesigned the anode driver board, the gerber files are here in zipped format.

http://teenix.org/HP0044.zip

I'll leave them there few a few days.

cheers

Tony

Woohoo!

Thanks! I've downloaded it.

hpux735 · 10-18-2021, 11:03 PM

I've got the PCBs milled out, now I just have to wait for the transistors Smile

hpux735 · 10-19-2021, 12:13 AM

I couldn't find the transistor you specified at digikey, but found this one. I think it's equivalent. Thoughts?

https://4donline.ihs.com/images/VipMaste...353EA87419

teenix · 10-19-2021, 12:20 AM

(10-19-2021 12:13 AM)hpux735 Wrote: I couldn't find the transistor you specified at digikey, but found this one. I think it's equivalent. Thoughts?

https://4donline.ihs.com/images/VipMaste...353EA87419

https://www.digikey.com.au/en/products/d...7/12609579

cheers

Tony

hpux735 · 10-19-2021, 12:29 AM

(10-19-2021 12:20 AM)teenix Wrote:
(10-19-2021 12:13 AM)hpux735 Wrote: I couldn't find the transistor you specified at digikey, but found this one. I think it's equivalent. Thoughts?

https://4donline.ihs.com/images/VipMaste...353EA87419

https://www.digikey.com.au/en/products/d...7/12609579

cheers

Tony

Unfortunately, I don't have a business model set up to justify the purchase of 9,616 transistors at the moment Big Grin

teenix · (This post was last modified: 10-19-2021 01:39 AM by teenix.)

Oops :-)

Element14, Mouser available to you?

https://au.element14.com/diodes-inc/mmbt...t=mmbt2907

https://au.mouser.com/c/semiconductors/?...&instock=y

Postage is always a bummer :-(

Any general purpose PNP, 100hfe, 600mA, SOT-23 package as long as pinout is same.

If you have a local parts supply, maybe just for testing use through hole BC327

cheers

Tony

hpux735 · 10-19-2021, 01:58 AM

Oh, interesting. The datasheet for MMTB2907 shows the graph of Hfe to be closer to 200 (at least at 25℃).

This one seems like another appropriate option...

https://www.diodes.com/assets/Datasheets...060BFH.pdf

I've got other stuff I need to buy at DigiKey, so it really helps to keep to what they have in stock.

hpux735 · 10-19-2021, 06:51 PM

Well, after more analysis... I measured the transistors on all the quad transistor arrays, and they all seemed happy. So, I looked further upstream, and noticed that the trace from the ROM for segment F is bad. In fact, it looks like it has a fried transistor Sad

Well, I guess that's "good" news in that a new CPU board from Teenix will solve this problem Smile

teenix · (This post was last modified: 10-19-2021 10:14 PM by teenix.)

Its a shame if the ROM 0 chip has failed but you might be able to get a spare if you ask around.

The normal signals coming out of ROM 0 are not straight logic levels. They have a common external resistor switched in series with all the anode driver base lines, 220R located just to the left of ROM 0.

cheers

Tony