Cloning a Tandy TRS-80 Model 1

[GK]

Hi,

Not sure what you guys are going for and it does sound very interesting, but just to note the TRS80 was based around the Z80 CPU.  I would think you would want to use that unless you plan to emulate that too.  Z80's were cheap 20 years ago, not sure anymore though.
Planning to expand the memory too?  That would be cool.

The Z84C00XXPEG is the current production, 40-pin DIP, Z80 substitute. Less than $8 AUD. I have 2 in transit: https://www.digikey.com.au/product-detail/en/zilog/Z84C0008PEG/269-3895-ND/929205
The full 48k
No disk drive at the moment. Recreating the Expansion Unit will be a separate project.

[kizmit99]

as there are only 12 lines per character...

Yep, I was going to point out that once you move to having the graphics characters generated by the font rom you might as well let it generate the "blank" lines between rows as well.  This allows you to use a font that has "true" descenders.  Although personally I'm not a fan of that "look" as it doesn't ring true to the original machines (at least to me).

Lots of interesting links added by other folks too!  Thanks for those!  I had no idea there was quite that much of an active TRS80 community out there.

[German_EE]

Lots of interesting links added by other folks too!  Thanks for those!  I had no idea there was quite that much of an active TRS80 community out there.

It's like the first kiss, memories kinda linger 

[biff]

That would be awesome. I don't see a problem with wiring either your adapter unit or the original MCM6670 masked ROM to a 27XX footprint and reading it with your TL866. I use the same programmer.
Here is how you'd wire it. I'd use a machined-pin IC socket in place of the 2716, and plug that directly into the TL866. It's easy to solder wires to the pins from the top.

EDIT: oops, don't connect anything to A10.

OK - this is how I'm going to try wiring it up:

Code: [Select]

Chargen                   2716 (TL866 ZIF)
1 (A6)           ->           22 (A9)
2 (A5)           ->           23 (A8)
3 (A4)           ->            1 (A7)
4 (A3)           ->            2 (A6)
5 (A2)           ->            3 (A5)
6 (A1)           ->            4 (A4)
7 (A0)           ->            5 (A3)
8 (RS3)          ->            6 (A2)
9 (Grnd)         ->           12 (Vss) Grnd
10 (RS2)         ->            7 (A1)
11 (RS1)         ->            8 (A0)
12 (D0)          ->            9 (Q0)
13 (D1)          ->           10 (Q1)
14 (D2)          ->           11 (Q2)
15 (D3)          ->           13 (Q3)
16 (D4)          ->           14 (Q4)
17 (CS) Grnd
18 (Vcc)         ->           24 (Vcc)
                                    15, 16, 17 Grnd

                                     18, 19, 20, 21 N/C


[exrbc]

back in the early 80's I could not afford a trs80 mo1 so made my own from the cct diagrams and remember working out chr gen and hand coding it into an eprom using one of the early eprom programmer,  also added the serial port, high speed tape interface and  floppy disk interface and additional re written roms . then added switching to allow it to become a NASCOM 1 used bank switching to swap out the ram and roms so could switch back and forth between them.
The video section was the problem as the chr set was different as well as the chr's per line ect. still have the original cct diagrams but the hardware has long gone.

[biff]

That would be awesome. I don't see a problem with wiring either your adapter unit or the original MCM6670 masked ROM to a 27XX footprint and reading it with your TL866. I use the same programmer.
Here is how you'd wire it. I'd use a machined-pin IC socket in place of the 2716, and plug that directly into the TL866. It's easy to solder wires to the pins from the top.

EDIT: oops, don't connect anything to A10.

Well,
I tried using every 2716 configuration available in the TL866 database, and every one of them resulted in nothing but 255's after a read.
See attached picture how I rigged it up.
At worst, I was expected jumbled data, not nothing at all.

Jim

[GK]

That would be awesome. I don't see a problem with wiring either your adapter unit or the original MCM6670 masked ROM to a 27XX footprint and reading it with your TL866. I use the same programmer.
Here is how you'd wire it. I'd use a machined-pin IC socket in place of the 2716, and plug that directly into the TL866. It's easy to solder wires to the pins from the top.

EDIT: oops, don't connect anything to A10.

Well,
I tried using every 2716 configuration available in the TL866 database, and every one of them resulted in nothing but 255's after a read.
See attached picture how I rigged it up.
At worst, I was expected jumbled data, not nothing at all.

Jim

That a bit disappointing. Other than double checking the wiring, which I'm sure you've done already, I'm not sure what else to suggest. When I get some time I'll have a play with my own TL866 and the 2716 configurations.

[GK]

Well, my Z84C0006 arrived yesterday afternoon, and holy crap, after some hours of frustration it is now actually alive, although I'm a bit paranoid about even breathing over the breadboard at this stage 

Basic Level I just boots up with a blank screen and a "READY >" prompt. Basic Level II boots up with the question: "MEMORY SIZE?" - not sure ATM what kind of entry it is waiting for. I currently only have the basics of the keyboard matrix interface operating. I can assert any keyboard character by jumpering/poking between the keyboard scan output and scan input matrix lines on the breadboard, but that's not very usable at the moment as trying not to get multiple instances of the wire-poked character is next to impossible. Next up I'll have to adapt the PS2 keyboard interface design that I developed for my PET-2001 clone for the TRS-80.

I also had to add another 74HC74 IC to the video generation circuitry because I originally did not fully emulate the original video blanking function when the CPU takes control of the video memory. When the CPU relinquishes control of the video memory the reactivation of the CRT electron beam needs to be delayed by two parallel-input load cycles of the serial-output shift register to ensure that when the beam turns back on it won't initially scribe several erroneous pixels. This prevents snow-like artifacts being produced on the screen when intensively updating the video memory.

Aside from the data cassette serial interface, which I have not yet implemented, this then pretty much verifies the CMOS clone design.

[oPossum]

Basic Level II boots up with the question: "MEMORY SIZE?" - not sure ATM what kind of entry it is waiting for.

'Enter' for default. It will then ask "CASS?", 'Enter' for default again. Then you will get "READY>".

[kizmit99]

Looking good 
All the "Memory Size" prompt is looking for is where YOU want the top of memory to be (assuming it's not at the top of physical memory) in decimal of course  .  So you can reserve some room for non-Basic stuff.
You can just hit enter and it will set it to the top of physical memory.

[kizmit99]

'Enter' for default. It will then ask "CASS?", 'Enter' for default again. Then you will get "READY>".

Huh???  I've never seen one do that?
I suspect you may be thinking about a Model-III or possibly Model-4 boot sequence?

[oPossum]

Yep, got it confused with Model III/4. Thanks for the correction.

Had to run my emulator to refresh my memory.

[kizmit99]

Oooo...  I notice your Model-1 emulator is running the "upgraded" ROMs (I think they may be called "version 3"?).  The ones where "RadioShack" was abbreviated "R/S" to reduce space and allow the upgraded basic to fit in the 12K.
I've been looking up and down for that ROM image and can't seem to find it anywhere on the 'net.  (I obviously must not be looking in the right places).
Any chance you could either share that rom image, or point me in the direction of where I can find it?
Thanks in advance!

[rrinker]

 I believe Dave is correct, and the lower case mod became standard equipment on later Model 1's. Subsequent systems, Model 3 and Model 4, had it all designed in to the system, no bodge board needed.

 Note that the Model 100 uses an 80C85 CPU and it a completely different animal from what is being discussed in regards to cloning a model 1.

 Now a clone of my old Model 4P - which was NOT just a Model 4 repackaged, there were a few differences in the design.....

[oPossum]

I've been looking up and down for that ROM image and can't seem to find it anywhere on the 'net.  (I obviously must not be looking in the right places).

Already posted in the PET thread  https://www.eevblog.com/forum/projects/otp-eprom-programmer/msg1605259/#msg1605259

This stuff was easy to find 20 years ago when I wrote that emulator!

[kizmit99]

This stuff was easy to find 20 years ago when I wrote that emulator!

Thank you! 

Yeah, it seems like everyone who used to host stuff like that got copyright scared, because a lot of what seems like it used to be out there isn't anymore...   Thanks for the pointer - exactly what I needed.

[GK]

Okay, thanks. I have a copies of both "Level 1 Users Manual" and "Level II BASIC Reference Manual" but have yet to sit down and have a read through.

I just finished laying out the video generator PCB. The hardware configuration will be similar to my PET clone - a separate motherboard and a video generator board. What do you guys think of the crosshatched ground plane as opposed to a solid pour? Gives it that vintage-ish look/feel?

Along the lines of the vintage look/feel theme, on the motherboard, like in the original TRS-80, I've decided to base the +5V rail regulator on a 723 rather than a three-terminal 7805 or equivalent. The pass transistor will be a TO3-packaged 2N3055 PCB-mounted on one of those rectangular, finned heatsinks. It's a pity that the MC3423 OVP/crowbar has gone obsolete. Anyone know of an alternative? Alternatively I can just use a small signal PNP switch and a zener diode to trigger an SCR.

[wilfred]

The crosshatched ground plane does look good. Since you're not reproducing the PCB I don't think it is critical. But, I also don't see a need to change it.

You can get the MC3423 from Rockby.

[GK]

You can get the MC3423 from Rockby.

Yes, but NOS only. I'm not going to use anything that isn't a current production part. I was mostly thinking of the MC3423 from a nostalgic point of view. I don't recall ever seeing a second source for it or any kind of (equally popular) substitute. There are lots of different brown-out/MPU-reset power supply supervisory ICs currently available, but an over voltage protection chip with an output drive intended to fire an SCR not so much. Anyway, I can make a crowbar trigger circuit easy enough with just a few discrete parts. The original TRS80 PSU didn't have any kind of OVP at all by the way. Slip with your DMM probes around the feedback pins of one of the two 723s and you can then go and joyfully tally up all of your fried logic chips.

[GK]

Not terribly exciting, but the preliminary first page of the motherboard schematic. The RAM of the original TRS-80 keyboard unit was expandable to was 16k max. If you had an expansion unit, that could hold an additional 32k for 48k in total. I'm going to put the full 48k onto my motherboard. My clone design will be theoretically compatible with an original expansion unit, so long as the 32k of RAM inside the original expansion unit is disabled to prevent it from conflicting with my clones upper 32k RAM block. The easiest way to disable the expansion unit RAM would be to lift the output enable pin on the RAM's read output data bus buffer and wire it to be fixed HI-Z.

It for this reason that I am not recreating the hardware to generate two specific drive signals originally present at the expansion port connector; these signals being MUX and !CAS, which were originally present exclusively for the purpose of addressing and refreshing the MK4116 dynamic RAM ICs used in the original expansion unit. All of the expansion interface control logic is shown in this schematic.

Page 2 will contain the RAM, ROM and address decoding, the header for connecting to the video generator board and bugger all else. Once you strip from the original design all of the logic for refreshing and multiplexing the addressing of the original DRAMs, simply the address decoding logic to a single AT27C256 programmed with a lookup table, use a single SRAM chip for the full complement of RAM and and a single AT27C256 for the ROM, there's not an awful lot to it any more.

Kinda boring, in a way 

[nick_d]

Not boring

To avoid damaging the expansion unit by lifting pins, could you just allow the expansion unit RAM to do its job and your RAM to do its job in parallel? It won't matter if two devices drive the data bus, providing they're driving the same thing to it. There might be issues with output-enable timing if it's driving the bus before the data has been read, though.

cheers, Nick

[GK]

Not boring

To avoid damaging the expansion unit by lifting pins, could you just allow the expansion unit RAM to do its job and your RAM to do its job in parallel? It won't matter if two devices drive the data bus, providing they're driving the same thing to it. There might be issues with output-enable timing if it's driving the bus before the data has been read, though.

cheers, Nick

Thanks for the interest

Problem with the old DRAM is that it was really slow. I think there would be nasty current spikes and EMI as the RAMs fight each other before they are outputting the same values.
 

[GK]

[German_EE]

I remember the old TRS-80 keyboard where each key was mapped to an address in memory, your scheme using an EPROM and a PIC is very elegant!

Looking forward to the day you run SCRIPSIT.

[GK]

Thanks 

I'm currently laying out a little board for the PS/2 keyboard interface. This will be a universal design compatible with any TRS-80, not just my clone.

SCRIPSIT - know where I can download an audio file? I've been so engrossed with other aspects of this design that I haven't even started looking for software yet. I am about to nip out in 5 to get the paper and another breadboard from Jaycar. The next and final section to get breadboarded is the data cassette interface. Amazingly, despite the rats nest assembly on crappy cheapo breadboard, I seem to be getting the computer reasonably stable now. If it turns out to be stable enough to load and run a game or too, that would make for a cool initial Youtube video, I would think.


EDIT: found in ~30s : https://www.classic-computers.org.nz/system-80/software_cassette_archive.htm