ADAM a 25 years old digital transient recorder

[maelli]

I do have ADAM in my lab!
Adam stands for Analog Data Acquisition Memory.
It is a 22kg beast with a with a tiny green 5.5 inch monitor.
Datecode on the ICs, it is from 1989 or 1990.

I have fond memories on these things.
As a young engineer I used to work with them, back then they were
very powerful machines, even though extremely expensive.

They could see what happened before the trigger!
So much better than the analog storage scope of the time.
The amount of memory was massive, 64K samples per channel. Crazy!
And the memory could be segmented, from 256x256 to all 64k in one shot.
All inputs are differential, back then we used them with TEK 6015 1:1000 probes.
Sample rate is 20 Megasamples per second.

Mine is 4 Channels, however, 12 Channel versions also existed.
Every channel comes with 6 extra digital signal inputs, so 24 digital channels
in mine.

not many were built, the company does not exist any longer

Should I open it up and see what is inside?

[T3sl4co1l]

Sure, give it a crack!

I'd expect a fairly standard ACQ board, nothing special.  20MHz bandwidth in those days wasn't a huge challenge, nor was 64k of memory, so it shouldn't be too bad.  The seemingly special purpose nature, custom design, and above all the software running inside, would dominate the cost I think.

Tim

[Bored@Work]

not many were built, the company does not exist any longer

Looks like the owner sold the company in 1995 and later on moved to Thailand http://th.linkedin.com/pub/rene-maurer/6/969/113

Should I open it up and see what is inside?

Do you really have to ask?

[HighVoltage]

All inputs are differential, back then we used them with TEK 6015 1:1000 probes.

Around in that that time frame, we used the Tek P6015.
It was filled with Fluorocarbon 114 (C2CL2F4) as insulation dielectric and  worked really well.
We used a Tektronix scope to measure high voltage and high speed transients. Our goal was
to measure down to the low ns range but neither the P6015 nor the scope was fast enough.
And the pictures of the scope we took with a Polaroid camera.

I did not know ADAM at the time, I am sure we would have tried one, money was not an option
at the time and place where I did this research.

Yes, I would like to see a tear down of this unit.
Thanks for sharing.

[maelli]

Starting with the analog front end, one of the four channels.
These are easily accessible, remove 2 screws, then pull the module out.

Nice modular design, but quite complicated.
Two boards + Display board. Most recent datecode is 8940.

Ten relays, 11 trimcaps, 13 trimpots...

(the attenuator is not 1-2-5-10, but 0.1V, 0.12V, 0.15V,0.2V all the way to
100V, in total 31 settings)

Googled some of the components:
The 2 DIL40 are simple LED drivers
A1-5195 and A1-2540 in the ceramic DIL14 are high speed op-amps.
The UAA170 must be responsible for the LED bar display, similar to a LM3915.

I expected the A/D and RAM to be here, but no, nothing, must be somewhere else.

[Refrigerator]

Ooh, that's just beautiful! You can clearly see where all that expense came from. 

[maelli]

open the hood and see what is inside.
Not quite a single board construction...

[T3sl4co1l]

Thinking, 68k processor?

[maelli]

A/D Board, RAM, and Trigger board  (one of the four)

A/D Board on the left
The A/D is a UVC3101-10 from ITT Germany. Could not find a datasheet on this one.

According to the manual, there were different A/D options. This one is 20Msamples, 8Bit.
There were also 10 and 12 bit options, at various speeds and prices.
A 10Msamples 12bit version was 9200 bucks PER CHANNEL ;-)

RAM board in the middle
with 16 x 8kx8 static RAM. D4361C

Trigger board to the right.
Triggering was quite sophisticated on these machines, it was possible to link the
trigger source on different channels, with logical OR / AND.

Then the Processor board: a humble Z80!

All boards together: 550 odd ICs on 30 odd PCBs.
Taking this into account, the power consumption of 130W is quite modest.

So, I put it back together again, still works, I did not break anything :-)

[PA0PBZ]

RAM board in the middle
with 16 x 8kx8 static RAM. D4361C

Are you sure? I can't read the chips but they look like DRAM to me, and that is usual for a Z80
D4361 has 28 pins, these have 22.

[maelli]

no, I am not sure!
googling  "nec d4361c datasheet"

brought me there:

http://www.alldatasheet.net/view.jsp?Searchword=D4361C

and then I was guessing.

here is a better foto of the mystery ram.

[bigsky]

What an amazing piece of technology. I would strongly advise saving the EPROM contents - otherwise you will be in trouble if one ever dies. Same goes for any PALs or GALs (often used in this era for address decoding) and any other programmable devices.

It's interesting that the input modules use DIN41612 connectors but the processor backplanes are on edge connectors. My experience is that edge connectors are less reliable, especially after a couple of decades. Other people may disagree.

Do you have any manuals for it?

[edavid]

NEC D4361 = 16K x 1 SRAM OK, I guess not!  Sorry.

[maelli]

NEC D4361 = 16K x 1 SRAM

mmhh...  still does not make sense, since the machine has 64kwords (16bit) of total
storage per channel and only 16 chips.
Finally I found this in the net (in German):

http://de.sci.electronics.narkive.com/VsWVllgL/was-ist-ein-d4361k-40

apparently 64k x 1 SRAM, 45ns (!)

Do you have any manuals for it?

I do have the operating manual and calibration instructions for the
analog front end, but no schematics.
And the price list from December 1989, for laughs..