The Ware for February 2018 is shown below.
Ware courtesy of Hernandi Krammes!
Every board designer leaves a thumbprint on their ware — and this one is from a region I had never previously seen a ware from before. So while probably easy to guess the function, I still appreciated it for the small, unique details.
Function is easy indeed: some ISA BNC network card, probably doing 10 Mbit/s. No clue about either brand or model or where to look for that special thumbprint.
Other possibility is an Irma 3270 terminal emulator. I see ? three ? z80 processors, a bit much for an ethernet. 8900 ethernets were about 30 ttl ssi gates.
It’s not three Z80 CPUs, it’s just a chipset of a single CPU, CTC and SIO each.
Looks like a Brazilian card to me. You can tell from the various SD.. cards, some of them only returning Brazilian hits on Google, from Brazil printed on many chips, and the use of Brascoil components.
The use of BNC makes me think of a video output, the maquest on the bottom left firmware label could be a short for maquina de escritorios.
Let’s add the z80 controller, the serial driver for communication with the host pc, and I’d say it’s a video compositing card to output text in a video for consumption in a TV/video studio ?
I have nothing to add, but I absolutely love those capacitors soldered to the IC U15. Look at how beautifully the leads are bent in to a rectangle!
Indeed it is an early graphics card for an isa slot. The z80 ctc chip is the give away.
Is it an I/O board for a Maquet Servo 300?
Naw, looks more like a 900.
If the card connector was ‘down’ and silkscreening read top to bottom left to right, then all the ICs are upside down from what I’d expect. Maybe it made layout easier.
It was made in the Free Economic Zone of Manaus — in the middle of the Amazonas!
Ahhh, that’s what the prod. Z.f.m stands for. Was wondering..
Another hint that points towards Brasil is the “PROD.Z.F.M” marking near the analog section of the board, which could refer to the Manaus Free Trade Zone.
As to the purpose of the board, I have no idea.
It might be video, but only monochrome: The only crystal on the board is 16MHz, and it’s nearly impossible to derive the 3.58MHz color carrier from that. So it’s likely not a TV studio part.
It might be networking, but then it’s likely not standard 10Base2 Ethernet, as that would require a 20MHz clock. (Sure, that could be synthesized from 16MHz, but why should they do that?)
Token Ring’s timebase was 16MHz. Yet the connector on the back is more typical of ARCnet or 10base2, with no 2.5MHz or 10MHz timebase or multiple thereof in sight.
Interestingly, Brazil is the only user of the PAL-M system, which mixes NTSC number of lines. Still apparently using 3.58MHz for its Color carrier.
Still unsure how the 16Mhz could play with that. Maybe with the Z80 CTC ?
Just from a glance over it I’m going to guess a card for generating those big blocky text overlays on video that used to be used for stuff like sports timing in the late 80s.
I doubt that, because it would need a video input port for that.
Not necessarily. This doesn’t do the compositing, it’s just teh chargen. It outputs characters on a solid color and a chromakey machine will drop the BG color and do the compositing.
Some weird video card with PAL/NTSC output. Or maybe PAL/NTSC is an input, I don’t know, for pirating video. Just guessing :)
This is most likely a Brasilian “clone” running CP/M. From what I’ve found several domestic companies were producing such systems based on Zilogs Z80. These CP/M machines often ran IBM mainframe emulators.
Whats odd about thus ware is that it looks to be an ISA-card. Could be that they optet to reuse those connectors or it is a Z80 plugin for an IBM PC XT…
This is most likely a Brasilian Z80 clone running CP/M. These were apparently popular in Brasil in the 80s and several domestic companies produced them. In the banking sector they often run some IBM mainframe terminal emulation software.
What is odd about this ware is that it comes in the form of an ISA-card. Either they re-used that form-factor or it was used as a plugin in a IBM PC XT. My bet is that they re-used the ISA-spec.
using bus slots for everything including the “main” board was common (and still is in some modern hardware designs).
I would say “where’s the ram” but that would just on one or more other cards populated with DIPs. we’re talking z80 here.
Regardless, my first reaction on seeing this board is still “very early networking”, not even 10mbit. The unshielded wires from the BNC connector bother me, but at low speeds it wouldn’t matter.
The dual A-F jumpers are some form of address selection (for the bus? network? both?).
Note the EPROMs: One is soldered down, the other is socketed for reprogramming. Why are there two?
I’ve just scrapped three very similar-looking boards (long, 8-bit ISA, Z-80 plus (dynamic in my case) RAM and support, BNC connector, no EPROM though) and they were early Novell G-Net 10Base2 boards.
http://retro.co.za/ccc/Novell/
Similar but not the same.
Hmmm… What strikes me is the fairly long connection from the analog-looking part of the board to the BNC connector. If it was video or network, they would probably have used a shielded coaxial, or at least an impedance-controlled cable. Also, early network cards tended to have an AUI (attachment unit interface) on a 2-row, 15-pin SUB-D connector and left the work with the physical layer (i.e. the network medium which would be the coax cable) to the box you would attach. And: The fairly large analog-looking corner doesn’t look like a design for signals in the range above 1 MHz at all. These are the reasons why I tend to rule out anything “network” or “video”.
The two DIP-8 ICs could be simple logic with transistor outputs (SN75454…), able to drive up to 400 mA.
The large inductors are paired with a lot of capacitors and resistors, so these could be filters in the audio-frequency range (or maybe up to a few 100 kHz).
The board is too complicated for something as simple as a trigger output or anything else that would just be something like switched DC or a slow PWM signal.
Thinking about the whole combination of blocks – fairly complicated digital stuff, and a large analog part with a lot of filtering – and asking myself what type of system all this stuff would be needed in, my guess goes towards something like the output for the write head in a tape data storage device. You know… Cartridges or even the good ol’ stuff that used 1/2 inch tape on large 10.5″ reel-to-reel systems with linear tracks. If this is the case, it must be the board for the modulator to the write (record) head, because the read (playback) head would need shielding and a front-end for conditioning the weak signal from the tape. The processor and RAM might fit the purpose of taking some bits, adding error correction or a CRC and Manchester Code, and modulating the signal onto a few carriers suitable for tape storage, i.e. roughly within the audio frequency range that can be recorded onto tape.
If it’s a tape system, I would love to see the pictures of the additional boards that would be needed for reading the data from the tape head and for the control of the machine’s motors and solenoids…
After I had shown this ware-to-be-named to a friend, he agreed this could really be some sort of a circuit for generating a bitstream. He said this is backed up by the presence of the Z80 SIO chip – so there is a serial interface on a board with no connectors except the BNC. This chip generates the serial bitstream, and next to this chip, there is a 74LS123 with monoflops, and ‘174 flip flops. These can be used to get the coded data stream ready for something that fits onto a tape or into a wire… MFM, NRZI, …
Well. If it’s not for a tape, maybe the “MOD” part of a MODEM using an audio channel? However, a MODEM would have a good synchronization between its transmitter and receiver, so it would probably be good to fit both onto one PCB, and this appears to be the transmitter only. On a tape storage device, this would probably work: You either write to the tape, or you read from it. Having the write and read circuits on two separate boards would probably work just as fine. Also, the parts, especially R28, R29 and some transistors like Q5, look “strong” enough to directly drive a tape head’s impedance.
Could be some sort of satellite modem or satellite communications card, or its used to generate some kind of signal clock.
Some low speed (33k, 56k) modem ? The analog circuitry looks a bit like the one found in the Minitel.
Or maybe an early DOCSIS modem, because of the BNC connector.
A TV + Radio card with the antenna as BNC?
Doesn’t explain the Z80..
This is definitely a network card, but the exact standard is unclear. The front end analog looks just like some early ARCNET stuff, and it would be possible to implement something quite like ARCNET with the Zilog SIO/2. The analog front end has the SN75454B peripheral driver on one end, and a LM319 dual comparator on the other. Looks an awful lot like the ARCnet transceiver on this old Zenith NIC (page 12-9)
http://www.s100computers.com/Hardware%20Manuals/HeathZenith/Z-100%20NET100%20Ethernet%20Interface%20card.pdf
The 16MHz crystal wouldn’t be compatible with ARCnet though. Maybe a local brazilian variant that used a slightly lower bitrate to work on the SIO (which maxed out at 2Mbps, compared to ARCNET’s 2.5) so that they could make it without the much more expensive integrated ARCNET controllers
I’d be curious if it was a 3270 emulator, but I’ve never seen anyone do that with an SIO, and I’m not familiar enough with the 3270 layers 1 and 2 to know if it could be shoehorned into an SIO.
Wow. Thanks for posting the link to that amazingly detailed service manual. The schematics there are indeed very similar to what can be seen on the picture of the board. Searching for terms ‘like token ring “2 mbps”‘ does yield some results. There may have been an OS called LANtastic network operating system on MS DOS (matching the 8-bit ISA interface of its time) running at 2 Mbps or 0.7 Mbps and either ARCNET, ethernet, or IBM token ring, which could explain the clock frequency (one search result pointed to a magazine called Network World, issue of 11. Sept. 1989. And Artisoft, who made LANtastic (?), did use the combination of the Zilog Z80 with the SIO chip, as this picture suggests: http://www.mcamafia.de/mcapage0/jpg/6343_1.jpg. It’s still a bit early (and of course, I’m just one of the participants), but I guess I can say “congratulations to you” for the closest guess so far.
Its a Network card. Probably Ethernet
*SiD was a IC manufacturer from Minas Gerais, BR
*The IC logo probably stands for Itautec Componentes from Sao Paulo, BR
*Brazil is graphed in portuguese as Brasil (is not for export; they should mark Brazil in this case)
*ABC Xtal was a crystal manufacturer from Rio de Janeiro, BR
*MAQUEST should stand for MAQuina de ESTado (state machine)
I bet BOTREM stands for BOot REMoto (remote boot) and, finaly, in the early 90´s the most popular standard in Brazil was Ethernet (with coax cables, 50R terminators, etc).
I was questioning why there are two EPROMs… but as a network card, this makes some. Plus I believe there are _three_…
One would be the firmware for the card’s z80 itself.
Another could be for a possible boot / driver rom for the system it is put in?
I’m not sure why the third one would be needed.
Well… multiple layers PCB was uncommon in Brazil that time. So, I’m making some assumptions based in a 2 layers PCB and PTH board densities.
The U40 EPROM (FW 3040) is placed near a RAM chip and the z80 CPU. A nice place for a firmware EPROM for the z80. The U35 EPROM (FW BOTREM) is placed near the address lines of ISA bus. Make sense for a remote boot ROM. And the 3rd EPROM (a ceramic chip from SiD with the label MAQUEST) is placed near 2 74LS174 (hex D flip flops) an a z80 SIO.
At that time, imported components was expensive, so is not unreasonable to use cheaper brazilian components to make a state machine (as I think it is). Even if it looks weird. These cost constraints has already produced bizar results.
Hah, yes! An EPROM plus some flip flops is effectively a tiny FPGA. :)
When I initially commented I clicked the -Notify me when new comments are added- checkbox and now every time a remark is added I get four emails with the same comment. Is there any way you possibly can take away me from that service? Thanks!
Network card of some sort due to activity light. Definately not a video card.
My bet would be some sort of Eth 10base2 (thinnet) mainframe terminal emulator card. 3 EEPROMs is unusual for any network card. One must be boot sequence (DOS + Drivers ) and the others must be IBM connectivity layer and then emulator of some sort ( 3270 emulator/terminal ) most likely.
Lol. I see now that it is revealed already. Although I think that “an Ethernet card” is not entirely the whole story. There are too many proms for it to be “just” a thinnet card. I stand by my emulator card theory.
F*ckin?amazing things here. I very glad to see your article. Thanks a lot and i am looking forward to contact you. Will you kindly drop me a e-mail?
Aciu uz straipsni!