• Please review our updated Terms and Rules here

A special version of TRS-80 Model I Mainboard. Does anybody know this?

wellswang

wellswang

Member
Joined
Apr 23, 2022
Messages
29
Location
Suzhou, China
I bought a special TRS-80 Model I mainboard in flea market. It's so special, look like integrated with expansion interface.
I can't find any information of this board. I think it was made in Japan (since TEC logo on it) like some Japanese version of M1.
It was marked as "HE11E016530" and "M-157-00", maybe this is the part number.
I just got the board only, no case and keyboard with it. So I didn't know the catalog number of this machine.

I searched on the internet, but didn't find anything. No user guide, no spec info, no technical or service manual.

Does anybody know this machine?

Thanks!

微信图片_20230821182242.jpg

微信图片_20230821182252.jpg

Thanks!

BR,
Wells.
 
Last edited:
The standard Japanese TRS-80 Model I board is the same size as the normal TRS-80 Model I. My first thought was this might be from the TEAC PS-80, but comparing the connectors to the pictures of the PS-80 system shows a difference.

Would it be possible to get a dump of the ROMs?
 
Here's a page that has a photo of the "standard" Japanese Model I's motherboard with TEC markings and, yeah, it just looks like a slightly cleaned up version of the US board. This is a completely different animal which...

MikeModifed said:
Perhaps: https://forum.vcfed.org/index.php?threads/strange-model-i-on-instagram.81006/
Click to expand...

Yeah, I was going to search the board for that, it certainly looks like this is the same motherboard as that machine, which nobody was able to identify. That you found another one suggests it might have been more than a one off prototype, but you'd think if this ever went on sale there's be at least one record of it being sold. There's a picture in that thread that claims the catalog number of this unit was supposed to be "26-7041", but there doesn't seem to be a single relevant hit for that in all the Google-ings so... yeah, no idea? Maybe TEC slapped together a small run of these (or at least the motherboards) and the product was cancelled before it made it out the door?

A ROM dump would definitely be interesting. It would also be cool if you could actually convince this beast to run. Obviously that keyboard ribbon cable is going to be a little bit of a hassle...

Anyway, you sure have some interesting flea markets in your neighborhood. ;)
 
Do you think you could get some bigger/clearer pictures of the board, that would let us read the chip markings? I'm particularly curious about IC 61/62; I assume those are the video memory SRAMs, does it have more than 1K of video memory? (Also, I can't *quite* read them clearly but it looks like the DRAMs are HM4864's? So it has 64K of DRAM onboard? I'd love to know if it has some kind of paging system to access the other 16K...)
 
I took clear photos of this board, under CC BY-NC-SA 4.0 license, you may share and download it from here:
You can see there is a custom ROM board attached on the back of the mainboard. I think this machine is modified to control some manufacture system or printers so they need to use their custom ROM.

I dumped 2 original ROMs (IC6 - ROM A, IC5 - ROM B) and custom ROMs. But I found original ROM B (IC5) is empty, filled with "FF". I asked HwH_Ricky to dump this ROM from his strange M1, his ROM A is as same as mine, ROM B is also like "empty", but filled with meaningless "04 06" or "14 16".
In ROM A we can find "R/S Level 2 BASIC" in it.
ROM HEX.png

You can download these ROM files from here:

Original ROM:
Custom ROM:

I found IC62 which was labeled as "46332P / A50" maybe is the 4KB Character ROM. I will desolder it and try to dump it later.



This machine has 64KB RAM and 2KB video RAM... So strange.

Thanks.

BR,
Wells.
 
Last edited:
Eudimorphodon said:
Do you think you could get some bigger/clearer pictures of the board, that would let us read the chip markings? I'm particularly curious about IC 61/62; I assume those are the video memory SRAMs, does it have more than 1K of video memory? (Also, I can't *quite* read them clearly but it looks like the DRAMs are HM4864's? So it has 64K of DRAM onboard? I'd love to know if it has some kind of paging system to access the other 16K...)
Click to expand...

IC61 is HM6116 2KB SRAM as video RAM. this machine is so strange, 16KB ROM, but 8KB is empty, 64KB RAM and 2KB video RAM, maybe has 4KB character ROM...
 
ROM B showing as empty might be due to the ROM type. I got the same thing on my Japanese Model I until I changed the ROM type.
 
I took a quick look at the ROMs. The IC6 ROM is mostly the same as a Radio Shack 1.3 ROM but there about 30 different bytes in the first $700 bytes. The custom roms look mostly like a BASIC program. This leads me to wildly speculate that the ROM has been patched to load up a BASIC program and run it.

I'll get to a deeper analysis at some point. The patched ROM more-or-less immediately jumps to $301E. This doesn't make sense if I assume the custom ROMs are mapped at $3000. Based on eyeballing the code at the start of custom rom0 I'd expect a jump to $3000.

Spent a few more minutes and extracted the BASIC program living in the custom ROM.
 

Attachments

  • custom.txt
    15.2 KB · Views: 15
wellswang said:
IC61 is HM6116 2KB SRAM as video RAM. this machine is so strange, 16KB ROM, but 8KB is empty, 64KB RAM and 2KB video RAM, maybe has 4KB character ROM...
Click to expand...

I wonder if they used a 2K SRAM simply because it was cheaper/more available than a 1Kx8 part? Is this motherboard more than two layers? Looking at both your front and back pictures it seems like there aren't surface-level tracks on some of the pins, which makes it pretty hard to follow the circuit. (I mean, it would be anyway with the ICs in the way, but...)

The outputs from that ROM next to the video RAM appear to go straight into the '166 shift register? It would make a certain amount of sense if they'd just used a 4K ROM for the character generator; that would be enough memory to let them skip the circuitry that generates the block graphics on a regular TRS-80 and just encode them into the ROM. A dump of the generator ROM would definitely be interesting.
 
Hi Wells,

wellswang said:
I dumped 2 original ROMs (IC6 - ROM A, IC5 - ROM B) and custom ROMs. But I found original ROM B (IC5) is empty, filled with "FF". I asked HwH_Ricky to dump this ROM from his strange M1, his ROM A is as same as mine, ROM B is also like "empty", but filled with meaningless "04 06" or "14 16".
In ROM A we can find "R/S Level 2 BASIC" in it.
Click to expand...

The ROMS are HN61364P , which have selectable polarity on the chip select from the factory.

IC5 is probably OK, but you may need to insert an inverter gate (74LS14 etc.) on the chip select signal to get the data out of it?

I have a standard Japanese Model 1, also manufactured by TEC, and something similar happens there on the A12 signal:

1692883321456.png
1692883518491.png
Regards,
Leslie
 
RedskullDC said:
Hi Wells,



The ROMS are HN61364P , which have selectable polarity on the chip select from the factory.

IC5 is probably OK, but you may need to insert an inverter gate (74LS14 etc.) on the chip select signal to get the data out of it?

I have a standard Japanese Model 1, also manufactured by TEC, and something similar happens there on the A12 signal:



Regards,
Leslie
Click to expand...
Thanks!
In datasheet, it says "The Chip Select input deselects the output and puts the chip in a powerdown mode." Doest it means when I need to read out the data, I should set Pin 20 at HIGH?
Another question, I successfully read out data from IC6, why IC5 has this problem? they are in same model. HwH_Ricky got same result as me.

Thanks.

BR,
Wells.
 
gp2000 said:
I took a quick look at the ROMs. The IC6 ROM is mostly the same as a Radio Shack 1.3 ROM but there about 30 different bytes in the first $700 bytes. The custom roms look mostly like a BASIC program. This leads me to wildly speculate that the ROM has been patched to load up a BASIC program and run it.

I'll get to a deeper analysis at some point. The patched ROM more-or-less immediately jumps to $301E. This doesn't make sense if I assume the custom ROMs are mapped at $3000. Based on eyeballing the code at the start of custom rom0 I'd expect a jump to $3000.

Spent a few more minutes and extracted the BASIC program living in the custom ROM
Click to expand...

perhaps it was attached to a Teletype (or similar) with a paper-tape read/write device?
 
ROM B dump looks good. The original ROM patches seem to be mostly about some extra keyboard scanning and some fiddling with bit 7 and bit 4 of port $FF which normally have no function in the Model 1. A wild guess is that bit 7 of $FF controls which video page is mapped from $3C00 .. $3FFF.

It would appear that the custom ROM is mapped in by default. It copies a small switching routine to $8000, some high-memory code to $F976 and a dump of BASIC memory to $4000 which includes the BASIC program. It then jumps to $8000 which I think switches in the original ROM (well, patched) and jumps to it.

Presumably it does a warm start of the patched BASIC ROM so the program is there. However, I've been unable to prove that out with some emulator fiddling.
 
You must log in or register to reply here.
Back
Top