Reading HP67 ROMs

[teenix]

Hi all,

I was reading the 1818-0229 ROM chip listed as ROM 5 in the 97 service manual to test some code mods because it is accessed from ROM Bank 1 and is now working ok.

I then thought that this was supposed to be the highest ROM with RAM included on chip and then logically this would be the highest RAM bank, address space $30 - $3F, but it only read back from address space $20 - $2F.

The plot thickens.

cheers

Tony

[teenix]

Hi all,

I used my ROM reader project to read all the ROMs from a HP-97 including the ROM0 display driver, and then also verify the RAM. I found one chip which the RAM seemed consistently faulty.

I've put these in a data base of sorts which the PC program can use for IC access. These are the IC's it has so far, with model and HEX ROM start addresses.

1818-0267 97 0
1818-0228 97 400
1818-0226 97 800
1818-0233 97 C00
1818-0229 97 10400
1818-0230 97 10800

I'll add more as I can. Any user can add more ROMs by entering the reading requirements like start address etc.

During a ROM read if a comparison file is available, the incoming data will be checked while the read is in progress. If the comparison file is not available, then one can be easily generated after the initial read takes place.

The circuit board needs a few mods, so I will tackle that as time permits and then make the project available.


Just a quick note, I've also been playing around with ROM IC replacements and after fiddling with some code and circuit designs, I think I have a setup that will work and It should fit into the same space as the original 8 pin ROMs. The ROM 0 will need extra work because it also controls the display anodes which I haven't quite figured out yet, so it still needs some work, plus I need to create a test jig to simulate the Sync, Is and Data busses and verify its operation before trying it out for real.

cheers

Tony

[Marcovecchio]

This is great news, Tony! If it's already possible to plug it on Woodstock and Spice calculator ROMs, I can help building the library.

And it will help a lot on finally diagnosing and fixing my dead Hp67 machines!

Eduardo

[teenix]

Hi all,

I've uploaded the ROM reader project files to the web page...

http://teenix.org/ROMreader.zip

In the zip file is the circuit and PCB design gerber files plus PIC code, a companion Windows program and PDF help file. I have also included ROM files for HP97 and HP67.

cheers

Tony

[tangdfx]

Hi，Tony

The link seems failed.

[teenix]

(07-21-2022 02:36 AM)tangdfx Wrote:  Hi，Tony

The link seems failed.

Apologies, it seems ok now.

cheers

Tony

[tangdfx]

Thank you Tony.

It works now.

By the way, is there is rom list that still missing on public domain?

As far as I know, there is HP 10 which is a hybrid CPU chip can not use this rom reader.

And also the famous HP95C. Maybe someone own the HP95C could read out the ROM.

Except these two, are there any more? (The domain is limited in classic ,woodstock, spice)

[teenix]

(07-21-2022 03:19 AM)tangdfx Wrote:  Thank you Tony.

It works now.

By the way, is there is rom list that still missing on public domain?

As far as I know, there is HP 10 which is a hybrid CPU chip can not use this rom reader.

And also the famous HP95C. Maybe someone own the HP95C could read out the ROM.

Except these two, are there any more? (The domain is limited in classic ,woodstock, spice)

I don't think you'll find the 95C ROM codes as it was never released.

The ROM for the HP10 is embedded in the large IC so can't be read without destruction of the package. I wrote my own version of the code which is in my HP-10 simulator.

cheers

Tony

[tangdfx]

Hi， Tony

I know.

But I am sure someone in this forum has an physical HP95C even it has never been released.

Of course the owner may not consider desoldering the ROM chip.

[teenix]

Hi all,

I have a design for a ROM replacement now. The circuit board measure 11mm x 8mm and should fit where the original 8 pin ROMs were on the CPU board. I was lucky to find a processor chip to use as there are not many around until it seems sometime mid 2023. I have been working on the code but not fully tested yet. As well as act like a ROM it has to decipher a few instructions and then act on those, mainly for transferring RAM etc. The processor can only execute about 6 instruction at certain times during the PHI clocks so it is a bit of a challenge.

I still have a bit to go because testing will take some time.

cheers

Tony

[Paul Dale]

Amazing work in these trying times.
It would be amazing work even if the times weren't so trying.


Pauli

[Thomas Klemm]

You could use this picture as your avatar:

[teenix]

Perhaps if I get it to work :-)

Just about to power up the bread boarded design. It has a second programmed controller which talks to the PC and also acts a bit like the ACT with Phi1, Phi2 and SYNC output signals, and the bidirectional IS and DATA buses.

cheers

Tony

[teenix]

Hi all,

Struggling to do anything boxed up in a small caravan, but I have a processor hooked onto a HP-67 CPU board monitoring the PHi 1 and 2, Sync and Is buses.

The top half of the image shows the first 3 instructions executed after switch on, the upper and lower traces are from the 67 sync and Is buses and the middle is the Is output from my processor after it decoded the Is information, fetched the ROM code and then output it to a test Is bus.

Code:


Addr  Hex  Disassembled Code
$0000 000  no operation                
$0001 3E3  if no carry go to $0F8      
$00F8 11A  0 -> c[w]

The lower half of the image is a close up of the bus signals. Mine is inverted (center trace) because it interacts with an external open source bus driver.

The next stage is to see if my processor can take over the functions of a 67 ROM by itself. The 67 CPU board I am using has a dead 1818-0232 ROM so hopefully it will spring back into life.

cheers

Tony

[teenix]

Hi all,

I removed the faulty 1818-0232 ROM/RAM from the CPU board and I programmed a microcontroller to operate as a 1818-0232 ROM only. Then I hooked it up to the 67 CPU board and monitored the instruction flow.

At address $0068 the code jumps to a subroutine in this ROM at address $0FC6 for 5 instructions then returns and this is exactly what happened.

The LED display lights up incorrectly, but this is because the RAM in both the 1818-0231 and 1818-0232 chips is faulty. I could not read or write anything in these using my ROM reader. The display mode is defaulting to SCI 0 which is what it should do.

The processor is also programmed to operate as a ROM 0 IC, so the next stage is to make sure it decodes the Is bus and outputs correct LED display information to the anode driver.

cheers

Tony

[wbvreeuwijk]

Hi Tony,

Thanks for creating this awesome project. It is exactly what I was looking for.

As I have 4 HP-67 CPU boards, of which I'm not sure all work as they should, I'm going to build your ROM reader project to at least be able to determine which ROMs are OK and which are faulty. While ordering parts, I encountered an issue sourcing a 649R 1206 resistor. How accurate does this value need to be? Can I replace this with a 680R or 620R?

Thanks in advance,
Bas van Reeuwijk

(07-21-2022 01:34 AM)teenix Wrote:  Hi all,

I've uploaded the ROM reader project files to the web page...

http://teenix.org/ROMreader.zip

In the zip file is the circuit and PCB design gerber files plus PIC code, a companion Windows program and PDF help file. I have also included ROM files for HP97 and HP67.

cheers

Tony

[teenix]

(Yesterday 07:59 AM)wbvreeuwijk Wrote:  I encountered an issue sourcing a 649R 1206 resistor. How accurate does this value need to be? Can I replace this with a 680R or 620R?

I don't think it is critical because there are no LEDs connected to ROM 0 during the read process.

You can probably omit it or try whatever is closest that you can get.

cheers

Tony

[wbvreeuwijk]

Sounds good. Thanks for the quick reply.

(Yesterday 09:11 AM)teenix Wrote:  

(Yesterday 07:59 AM)wbvreeuwijk Wrote:  I encountered an issue sourcing a 649R 1206 resistor. How accurate does this value need to be? Can I replace this with a 680R or 620R?

I don't think it is critical because there are no LEDs connected to ROM 0 during the read process.

You can probably omit it or try whatever is closest that you can get.

cheers

Tony