speleo_de
Member
Helping a friend to get a non-working Midway Submarine back to life - we are not finished yet 
We had a repair session December 27/28 last year and I want to share some interim results and pictures.
John from flippers.com was so kind sending me the ROM dumps, so I burned two EPROMS in preparation. Our repair session was not too effective, I think I have to get used working together with another person
.
Boards are in an L-shape like many of the early Midway boards. We were able to get the CPU board tested with my Fluke 9010A and working. The original EPROMS were dead, so luckily we had the new ones
. Strapping on the board needed to be changed as the original EPROMS were the TMS2716 kind with three voltages and the new ones good plain 2716 with just 5V. Interesting that the four 6810 RAMs were still ok.
With the I/O board we failed miserably. Someone had worked on the board before, that was clear. So we replaced the missing components and did quick check on the rest. That was the problem: we should have checked more intensively !!
When switching the machine on, 3 transistors got extremely hot and we were not able to fix that during the remaining time.
Finally I took the boardset home and worked on my own place, did a minimum power harnes and simulated motors by some power resistors, required inputs simulated by switches etc ... I found out that several mistakes were made by the guy who had tried to repair the I/O board before. I fixed that plus two of our own introduced mistakes and now CPU and I/O are at least working on my bench. The CPU board had an additional issue with the 'noise' generator and need a new CMOS 4006 - 18 bit shift register.
To be continued to test the I/O and sound with the Fluke on my desk - don't hold your breath
--------------------------------------------------------------
For anyone interested - attached are:
- Pictures
- Fluke 9010A ROM checksums
- ROM - RAM - IO mapping (Inputs missing still)
We had a repair session December 27/28 last year and I want to share some interim results and pictures.
John from flippers.com was so kind sending me the ROM dumps, so I burned two EPROMS in preparation. Our repair session was not too effective, I think I have to get used working together with another person
Boards are in an L-shape like many of the early Midway boards. We were able to get the CPU board tested with my Fluke 9010A and working. The original EPROMS were dead, so luckily we had the new ones
With the I/O board we failed miserably. Someone had worked on the board before, that was clear. So we replaced the missing components and did quick check on the rest. That was the problem: we should have checked more intensively !!
Finally I took the boardset home and worked on my own place, did a minimum power harnes and simulated motors by some power resistors, required inputs simulated by switches etc ... I found out that several mistakes were made by the guy who had tried to repair the I/O board before. I fixed that plus two of our own introduced mistakes and now CPU and I/O are at least working on my bench. The CPU board had an additional issue with the 'noise' generator and need a new CMOS 4006 - 18 bit shift register.
To be continued to test the I/O and sound with the Fluke on my desk - don't hold your breath
--------------------------------------------------------------
For anyone interested - attached are:
- Pictures
- Fluke 9010A ROM checksums
- ROM - RAM - IO mapping (Inputs missing still)
Attachments
Last edited: