• Please review our updated Terms and Rules here

S-100 PSU & Chassis (Delta Products)

JonB

JonB

Veteran Member
Joined
Jan 26, 2014
Messages
1,652
Location
South Herefordshire, UK
Hi

I am now starting to look at an S-100 box I acquired recently. First check is the PSU - I measure +18, -18, +9.5 volts on the three power rails, unloaded. I think these are OK. The PSU is a typical early linear supply having 1/2 wave rectification (the +-19v rails are supplied by a single piece bridge rectifier, whereas the 9.5v rail looks to be supplied by two separate enormous diodes of obviously high current rating. There are three very big electrolytic caps on the board - 95,000uf 15v and two 5500uf 45v. I've removed the cards and plugged it in, it's humming along nicely giving the unloaded readings above.

I'm thinking of leaving it like this for a while - a sort of soak test. Agreed there is no load, but I'm not ready to plug boards into it just yet..

Cheers
JonB
 
A load while testing an unregulated linear supply isn't particularly necessary.

Later, you may want a load when checking for AC ripple (i.e. how are the capacitors?).
 
Well, I haven't tried with the 'scope yet but I have connected the CPU board (seemingly ultra rare Teletek FDC-1) and smoked a tant.

Replaced that and one other and it is running with my Z80-ICE. I can see the content of the monitor ROM and the 1K static memory (64k dynamic RAM board not connected yet!).

I am having problems getting it to communicate over its serial console port. Have some electrical checks to do - it needs hardware handshaking but looping back CTS/RTS and DSR/DTR has no effect. Maybe one or more of the driver ICs is toast (or, worse, the Z80-SIO).
 
Parallel I/O board
64K DRAM
2xEPROM programmers (one of which can handle the TMS2716)
A prototyping board, unpopulated.

The processor board is a Teletek FDC-1 which is really an SBC, having 2 serial potrs, 2 parallel ports, a 2K monitor, 1K SRAM, and FDC. So all it needs to boot CP/M is the 64K card and the MEGASTOR 8" drive unit.

Since my last post I got into the monitor. just a matter of tracing the wires - carefully - and knocking up a serial port adapter.

Suddenly:

Code: 
FDC-I 3.0 Monitor (a©
>

That last character is odd. It's a ) with the 8th bit set, I think. I see the same effect in other prompts from the monitor. I suppose if my terminal stripped out the 8th bit (per bog standard ASCII) I'd be none the wiser. It's stored in the ROM correctly.
 
I wonder if the 8th bit isn't being used as an end-of-string deliminter--I'll bet that it is.

In which case, you can ignore the parity bit.
 
Yeah, I have seen that before. But like I said there is no 8th bit in the strings in the ROM. I've looked and it says "FDC_I 3.0 Monitor (a)" as expected. The monitor must be adding it - but it doesn't do this consistently (I mean, for every string it prints).

I'm pretty pleased it appears to be working. Have yet to connect the disks and additional RAM for a full on CP/M boot.
 
Only on the final character of certain strings. I reckon it's a bug in the monitor; maybe they tested with terminals that stripped bit 8.

So we have the sign on message, the "Completed" message (it prints this when you exit certain commands like dump). But the error messages (ending in "ERR!") don't do it.

We can see the strings in memory using the monitor's DA command:

Code: 
E780 1F 1E 1F 1E 1F 0D 0A 46  . . . . . . . F
E788 44 43 2D 49 20 33 2E 30  D C - I   3 . 0
E790 20 4D 6F 6E 69 74 6F 72    M o n i t o r
E798 20 28 61 A9 0D 0A 3E A0    ( a ) . . >
E7A0 20 45 73 63 61 70 E5 20    E s c a p e
E7A8 43 6F 6D 70 6C 65 74 E5  C o m p l e t e
E7B0 56 20 45 72 72 21 A0 41  V   E r r !   A
E7B8 64 64 72 A0 26 20 6E 6F  d d r   &   n o
E7C0 2E 20 6F 66 20 73 63 74  .   o f   s c t
E7C8 72 27 73 A0 D8 EA E1 C9  r ' s   X j a I

No corruption there...

Oh well, it's not the end of the world. I plan on being in CP/M all the time with this box anyway.
 
I've now connected the 64K RAM board and tested it using the monitor. Looks OK, but as there's no RAM test routine I've had to do block copies of the ROM into the RAM; fortunately the copy performs a verify.

Then, another tantalum capacitor gave way (on the memory board) and I replaced it.

I've connected the 8" drive unit and I'm attempting to read data off a disk using the monitor's RD command but it is coming up with errors. Typically address mark not found. So I'd like to ask what this really means and how the address marks are put onto the disk (I assume by formatting using the FDC's format command)? The FDC is our old friend the NEC uPD765 but the monitor has no command to format a disk. SO, writing to / reading from a blank or unformatted disk doesn't work, and neither does reading from a formatted disk (albeit a disk of dubious functionality, but one with a label that indicates it's a Teletek FDD-1 CP/M disk).

Naturally I have tried to boot using these disks but there is no response - it either locks up, or returns to the monitor with no error. So, it looks like I'm down to diagnosing a disk problem and I guess it is hardware related.

I should say here that the drive unit is known to be working; it will boot up a TRS-80 Model II into CP/M, reliably.
 
Minor update - I looked in the source code for my monitor v3(a) and fond that, yes, bit 7 in a string indicates end of string, and yes, it sends the last character without masking the bit, so it comes out as an extended character. Ha!
 
You must log in or register to reply here.
Back
Top