BBTC

The Alien probe was useful for checking the wiring and memories. Memories are working fine. Next step is to check the bit bucket with both the Z8001 and PC XT Alien Probes.

Last year, I had a lot of free time until late autumn. I returned to work in November, so my projects will slow down.

The possible next steps:

implement the Trump card with discrete logic to see if it works with the software

making a simulator to run Z8000 code on Linux

port this simulator to run in the Z8001 alien probe

add a UART in the CPLD to help debugging or

add an interceptor socket with which I can add peripherals to the current BBTC

there is already a disassembler for the Z8000 running on Linux

I figured which chip goes where. I combined the 2 binary files and disassembled the result. I tried a disassembly of the swapped file but it was obviously not good so I scrapped it. I added the files in this project.

Command to combine the 2 binary files:

srec_cat -o BOOT.bin -binary BOOTODD -binary -unsplit 2 0 BOOTEVEN -binary -unsplit 2 1 

 Location of boot chips:

D8-15: BOOTEVEN

U7 on BBTC, U21 on TC

D0-7: BOOTODD

U8 on BBTC, U22 on TC

The listing in the original article seems to be an old version. The 2 files on the software disk are a derivative.

I tried swapping the 2 boot chips, same result. Removed both, same result.
At this point, I need to know:
Do the files BOOTEVEN and BOOTODD match the original Trump Card boot chips and which goes where (D0-7/D8-15)?
I reproduced the circuit from the article except for the dynamic RAM. I know there was an error in the DRAM logic but that didn't affect my version. I don't know if there are difference between the schematic and the official Trump Card.
If someone have a working Trump Card, I would like to see the result of the DIAG.COM program.

The initialization sequence in the article seems to be incomplete. I used the files BOOTEVEN and BOOTODD provided on the Trump Card's floppy disk.

LDZSYS doesn't load either, but using diag.exe, I can test the bucket circuit further this time.

I used my PC XT alien probe to check every control lines, data and address bus.
The CPLD was mysteriously keeping address A0 stuck at 0. 

There was no short with ground, so I added an inverter on A0 in the CPLD with the output going to an unused pin. There was no need to make any patch on the PCB (yet).

With the pin A0 configured as an input, it was no longer hanging the bus and I could plug the BBTC on the PC XT motherboard. 

However, the BBTC still didn't work. LDZSYS was not able to connect to the Z8001, and the DIAG program remained stuck.

The bucket simulation seems to work OK.

The listing of the bootstrap initialization program doesn't match the files BOOTODD and BOOTEVEN.

So far, the BBTC doesn't work. I removed all the chips on sockets. Only chip remaining on board is the RAM and CPLD.

At this point, the BBTC's CPLD is programmed.

When I plug the BBTC with the Nano 8088, the Nano 8088 doesn't boot. No video, no message, nothing at all. It boots normally if I remove the BBTC.

I soldered an unprogrammed CPLD on an empty BBTC and I have the same result. When I put an empty BBTC, the Nano 8088 boots normally.

Huge milestone reached!
I also made a version 2 of the Nano 8088 board because the compact flash connector wasn't available anywhere. Fortunately, once assembled, the board worked! The compact flash doesn't seem to work yet. Once this works, I'll try to push my files on github to the same project.

Meanwhile, I assembled 2 PCBs from Microcore Labs: the MCLZ8 and the MCL86+. That give alternatives to my alien probes for testing. At least I don't need to debug the 8088 anymore.

The next test was to plug the BBTC on the passive motherboard along the Nano 8088 v2. Unfortunately, the card blocks the bus and nothing works. When I removed it, the Nano 8088 works normally.

Next step will be to whip out my Alien Probes. Time to analyse this more deeply!

My assembled Trump card isn't tested yet, I need a PC to test it. I ordered a Nano 8088 kit with a motherboard but it doesn't work so I need to test both. I make some tools to be able to test the hardware at the lowest level. I call that Alien probes, the name says it all. 

I made one for the PC XT bus, one for the Z8001, one for the 8088 and a generic one to monitor bits and generate digital signals. These are based on the STM32F405 chip which is similar to the MCU on the black board I used on the Z80 ICE.

I plan to make Alien Probes for other CPUs used in the retrocomputing community. This will help locate hardware faults.

I just made a simulation model for the bucket logic.

In the original Trump Card schematic, I noticed (what seems) a redundancy with the signal /BIOW at IC30 pin 13 and IC33 pin 12. That may explain the 470pF capacitor at the output of IC28 pin 9.

While the original circuit modif may not be trivial, the CPLD version would be much easier to correct, and no need for capacitor.