Teardown : Data Proof 160A

[ManateeMafia]

I decided to do a teardown on a couple of Data Proof low thermal scanners. I have owned them for a while but I plan on putting one to work monitoring some references. The one in my test bed has old crumbling insulation on the wires, so I thought a teardown would be appropriate.
The first one was bought for the reference array, the second for parts. The second one was eventually repaired and a third was also acquired. Pictures are listed below.







You can see the different configurations offered from the manufacturer. I prefer the pre-installed cables as they are much easier to connect to a voltage reference.

 





Here you can see where all the magic happens. The isothermal chamber is fairly heavy and is surrounded by foam on all sides.



All point to point wiring on the bottom.



Here we have the rear entry panel of an option 2 scanner. All the binding posts are the Pomona 3770 low thermal connectors.

 

 

   

The boards above are the GPIB interface boards. They are dumb controllers that only receive commands in a XYY format where X=A or B and Y = 00 to 16 or 32 depending on model. It must receive all three numbers and characters.
The bottom board is the newest version of the three. It is supposed to fix timing issues based on comments from the Data Proof website.

 

 

 

These are the relay control and logic boards. There are separate open and close boards.

 

 

 

Here we have a relay and a board where they are mounted. There are 8 relays/board controlling 4 channels. The relays are dual latching and they use the pcb as the stationary contact. No solder connections are made except where the wires attach to the boards.



Data Proof threw in a couple of spares inside each scanner. Data Proof also sells spare relays. No need to send it in for service.

 
 
 

Some assorted photos of the remaining sections in the scanner. The front panel is made up of four separate boards and the power supply is pretty basic with 5V and 14.5V outputs.

[TiN]

It's a joy to see those spare relays located for servicing. Shows the good thinking of engineers who designed these pieces.
Similar like with HP 8642B generator , which have SMA/SMC wrench fixed inside for easier servicing. This not common today to see such attention to details, most of today's gear "just wack it all together, with some secure custom frutty screws, so noone else can take it apart ever".

It also seem modular, as 32-channel model would just have 4 more extra relay cards, 2 more open cards and 1 more close card. And different front panel ofcourse.
I'd expect if you toss all extra boards it will be functional at least over GPIB to control all 32 channels even without correct front panel. 

Thank you for the work and efforts to share  .

[Vgkid]

Thanks for the teardown MM
 That is very cool, neat how they used the pcb as the contact for the relay's.

[quarks]

very nice, thanks a lot for sharing

[zlymex]

Very nice teardown.
I got my 160A several years ago and here is some of my tear down photos  http://bbs.38hot.net/thread-1082-1-1.html

I also modified the scanner so that it consumes less power and hopefully reduce thermal EMF.
In theory, those Printact relays are latching type and need only very little power to drive. I once designed/implemented a scanner with 32 relays but only needs 0.3mA @4V to operate.
In reality, 160A was an old design(use many 74 series) and generate too much heat. Dataproof later modified their new scanners to use 5V supply only.

[ManateeMafia]

Thanks everyone. TiN helped with the editing and usual photo support.

@zlymex
Nice photos. I noticed that the B series scanners have the external power adapter, but it looks like that is only a small part of the problem. I wonder if the schematics for the newer series are the same but with lower power components? I will have to look and see what kind of improvements were made.

[TiN]

Maybe easier just to synth all the needed logic into small CPLD chip and make own single PCB controller for relay boards... Just 5c...

[Vgkid]

Thinking about it, I believe that I have seen pcb traces used as relay contacts before. I just don't remember on what.

[lukier]

Thinking about it, I believe that I have seen pcb traces used as relay contacts before. I just don't remember on what.

Tektronix AM503 (old one) for example. Horrible device, drifty, noisy, very vintage and crusty. Unfortunately, the more modern AM503A/B are silly expensive on eBay :/ I still want to experiment with 1 MOhm termination direct on the probe transformer output (input is fed with bucking current according to the hall sensor readings), without this silly attenuator / amplifier combo. I believe this is how more modern TCP202A or other compact DC current probes work. AM503A/B or TCPA300 still have this attenuator/amplifier arrangement which is a bit of mystery for me, as the oscilloscope input already has these so why duplicate.

[Vgkid]

Thinking about it, I believe that I have seen pcb traces used as relay contacts before. I just don't remember on what.

*snip*

Not sure if that is it, but Tek related.

[ManateeMafia]

Before I replace the cables,I decided to contact Data Proof concerning the solder type they used. The response was that it wasn't anything special. No low thermal solder, especially cadmium based was used. They recommended SN63PB37. Good enough for me.

[Okertime]

When I got my 160A, I noticed the thermal EMF generated by internal power and decided to make some hardware changes as suggested by zlymex in reply #4.
Then I observed, contrary to other scanners, which initialize all relays during power-up, that there is no relay switching during power-up or power-down.
Now I am using a solid state relay to switch on the mains power 0.2 seconds before I send a GPIB command and switch off the power immediately at the end of the command.
The power-on time is reduced to 1 second, while, in my case, the cycle time for relay switching is 72 seconds. This way, the over-all power for my setup is reduced by more than 98%. No EMF problems any more.
The disadvantage, an extra control line is needed to control the SSR. I am using an old XP laptop from 2002 for data-logging. This computer has an ancient LPT-port which controls the Relay.
The 160A is a very old instrument using non standard GPIB commands. I don't know much about the GPIB Port and data protocol, but hardware wise it has some similarities with a LPT-port.
It might be possible to generate a special custom GPIB command, where a signal from the GPIB port is used to control the SSR.

Wulf