Unidentified SGS chip from rare robot

[JehudahDesign]

Greetings.

I am restoring a rather rare robot from 1987, and only have the 16-bit computer module left to repair to make the robot fully functional for the first time in many years.
All the logic gates and smaller components work, so I'm moving on to the larger 40-pin chips, though I'm not exactly sure how or if I can test these, and one of them has been impossible to identify so far.

The full name as it appears on this SGS chip is:
FZ102
28644
57 AC

It seems to be mostly connected to the D70116D CPU (8086 clone), and four 4464 DRAM chips, but I can't find it as anything memory related either.
At least I've checked what all the pins go to, so was hoping someone here might have an idea from that what this chip might be.
I'm relatively new to all this, so the ins & outs might be mislabeled, so please do correct me on this.
First row of numbers are the SGS pins, followed by the pins of what they go to.

01 - 24 - D70116D-5 (interupt acknowledge display - out)
02 - 09 - 74HC08 (3A - in) - & - 32: D70116D-5 (read strobe - out)
03 - 15 - 74HC373 #3 (Q5 - out)
04 - 09 - 74HC373 #2 (Q3 - out)
05 - 02 - 74HC244 (A1 - in) - & - 29: D70116D-5 (write strobe - out)
06 - 03 - 74HC257 #2 (1I1 - in) - & - 02: 74HC373 #1 (Q0 - out)
07 - 10 - 74HC00 (3B - in)
08 - NC
09 - 15 - 74HC257 #1 (output enable - in)
10 - 15 - 74HC244 (B3 - in)
11 - 13 - 74HC244 (B2 - in)
12 - 11 - 74HC244 (B1 - in)
13 - 04 - 4464 DRAM #3 and #4 (write enable - in)
14 - 04 - 4464 DRAM #1 and #2 (write enable - in)
15 - 01 - 74HC257 #1 and #2 (select input - in)
16 - 05 - 4464 DRAM #1 to #4 (row address strobe - in)
17 - 10 - 74HC04 (5Y - out)
18 - NC
19 - 33 Ohm resistor between 19 - D70116D-5 (system clock - in)
20 - VDD
21 - VSS
22 - 05 - 74HC08 (2B - in)
23 - 21 - D70116D-5 (reset - in) - & - 06 - 74HC244 (A3 - in)
24 - 16 - 4464 DRAM #1 to #4 (column address strobe - in)
25 - 09 - 74HC257 #1 (3Y - out) - & - 14 - 4464 DRAM #1 to #4 (address 0 - in)
26 - 07 - 74HC257 #1 (2Y - out) - & - 13 - 4464 DRAM #1 to #4 (address 1 - in)
27 - 12 - 74HC257 #1 (4Y - out) - & - 12 - 4464 DRAM #1 to #4 (address 2 - in)
28 - 04 - 74HC257 #1 (1Y - out) - & - 11 - 4464 DRAM #1 to #4 (address 3 - in)
29 - 09 - 74HC257 #2 (3Y - out) - & - 08 - 4464 DRAM #1 to #4 (address 4 - in)
30 - 04 - 74HC257 #2 (1Y - out) - & - 07 - 4464 DRAM #1 to #4 (address 5 - in)
31 - 12 - 74HC257 #2 (4Y - out) - & - 06 - 4464 DRAM #1 to #4 (address 6 - in)
32 - 07 - 74HC257 #2 (2Y - out) - & - 10 - 4464 DRAM #1 to #4 (address 7 - in)
33 - NC
34 - Resistor array
35 - Resistor array
36 - 11 - 74HC373 #1 to #3 (latch enable - in)
37 - 35 - D70116D-5 (address bus - out) - & - 04 - 74HC373 #3 (D1 - in)
38 - 36 - D70116D-5 (address bus - out) - & - 17 - 74HC373 #3 (D6 - in)
39 - 37 - D70116D-5 (address bus - out) - & - 03 - 74HC373 #3 (D0 - in)
40 - 28 - D70116D-5 (memory selection - out)

[debs3759]

D70116 is an NEC V30. It is compatible with the 8086, but is not a clone, as it has extra instructions plus other minor differences.

Can you post a photo of the SGS chip that you can't identify? You might have to join to do that, or you could upload it to an image hosting site.

[JehudahDesign]

Ah yes, thanks for pointing that out. I noticed it has at least an extra standby function and a lot more transistors, so I'll be replacing it with the exact same if the ones I have don't work anyway.

I uploaded a picture of the SGS chip to my Facebook page for these robots, but can't link to it as a guest. Searching for Modulus Robots should take you there. (I did register, but didn't get the activation email, or found who to ask about it)

[debs3759]

I would have guessed at RAM but you have a 4464 DRAM. Hopefully adding the pic here will help.

[debs3759]

It could be a custom ROM

[Guest]

It was guessed to be a RAM before too, but indeed there's already 4 DRAMs in there, and even a pair of SRAMs on another board in the module.

I've asked a few people who worked on the robots about this module, but they say it didn't have a 16-bit CPU, but did have a 16-bit CPU, so I have no idea what they were on about with this.

Custom was a word I was afraid of... I do have two of these chips, but seeing as 70% of everything was fried inside the robot, I'm worried about their states if I can never replace them or even test them to begin with.

[gshv]

[Jehudah]

Excellent. I'm in. Also sorry I forgot to add my name in the last post while running around between my 80000 projects. I wanted to register since this probably won't be the last chip I need to ask about when owning over 50 robots...

At least fixing this one, my most important robot, would be a major success for me.

[yy2013]

Most likely a DRAM controller chip, which refreshes the 4464 DRAM and generates timing pulses such as RAS, CAS and WE for 4464 and select input for 74HC257 multiplexer.

[gekaufman]

Interesting that the power pins are 20/21. Might help in tracking it down. Not even close to Intel 8203 which would have been my first guess.

[Jehudah]

I have also tried to find it as a DRAM controller since it does just go to some DRAMs, but still finding nothing after a sea of datasheets.

I thought the same about the power pins. I've barely found any chips at all like that, and trying to reverse-search for it by pin functions doesn't seem to work well either.

I'll try to fix up some photographs of this whole module so I can embed them here. There are three cards in total inside one of the robot's module "slices".

[yy2013]

I guess it is a custom DRAM controller, because most DRAM controllers have address multiplexers integrated in them.

[Jehudah]

Here are pictures of everything, in case they make things clearer. (hosted from my website since Photobucket is acting up, so I hope that's okay)

I'm not even sure what goes into the DRAMs without documentation for the robot. I'm guessing the battery backed-up SRAMs hold custom programs from the RS232 port (mic-plug) when the robot is switched off.

The pictures are from before repairs, so dead batteries and old caps have been replaced and shouldn't be the reason for its problems. Everything else is easy to replace.... Except the dreaded FZ102 if that doesn't work.

[aberco]

It seems to be a custom gate array designed for that very purpose. Some form of DRAM control logic would be my guess, but it can also implement a few other simple functions too.

[Jehudah]

That's too bad. I was hoping to at least find a datasheet so I could maybe test the chip. I'm not sure what to do with it now, other than powering it up to check for anything that might harm other chips. At least none of the pins are shorted for one thing...

Like I said, I am relatively new to all this because of my busy schedule, so still have much to read through. But if anyone has any tips on how to check the state of this chip in the meanwhile, it would be much appreciated.