Advantest TR6143 Teardown/Attempted Mains Voltage Modification

[42Khz]

Hi all, I recently acquired an Advantest TR6143 SMU from Japan for cheap on the 'bay. It has six banana binding posts on the front and three triax connectors on the back.

Some quick specs from the datasheet:

• Sources up to 110V (0.5A max) and up to 2A (32V max);
• Source (and measure) resolution of 10μV (320mV range) and 1nA (32μA range);
• Stable as a voltage source into 1μF for all current ranges in SLOW mode.

I find it quite interesting that the unit is so tolerant of capacitive loads (relative to its western counterparts such as e.g. Keithley, HP/Agilent), even in FAST mode (100nF from the 3.2mA range upward), moreso considering that it's fairly old. I found a bulletin from 1992 so it has to be at least that old. If I'm not mistaken, that dates it around the time that Keithley released their 236/237/238 SMUs, which only tolerate up to 20nF across the board.
I'm not sure when exactly Keithley and Agilent started adding a "high capacitance mode" to their SMUs, perhaps someone else here knows. The relatively new Keithley 2450 for example still supports up only to 20nF in normal mode, and up to 50μF in "High-C mode". As far as I can remember, this mode roughly boils down to lowering the gain bandwidth of the integrator by a factor 100 or so, and increasing the time constant of the current sense resistors by another factor 25 or so. Those rough numbers can be found from the response times cited in the datasheets, I might be misremembering how the factor 2500 (50μF/20nF) is distributed. The concept is described in this Keithley patent from 2007(!). Advantest must have done something similar in their TR6143.

Anyhow, I was hoping that I would be able to rewire it for European mains like I did for my Advantest R6246. Unfortunately, no dice. The unit contains two (welded?-shut) black cans of quite different size as can be seen below (top left), with only two wires going into each. Judging by their size, the transformers experience unequal loading in operation so for that reason as well as some others wiring their input windings in series seems like a recipe for disaster. I guess I'll have to buy an external conversion transformer...

Since I had to rip the unit apart to find this out anyway, I took some pictures of the interesting stuff along the way. I'll update things as I get around to figuring out more about the unit, but no promises.

[42Khz]

Here is the power stage. It contains 3 separate heatsinks, and the middle one holds components on both sides. The side view is taken from the right side of the board. Apparently the design is from 1987.

I have no idea why it's so complicated, perhaps some very elaborate cascode to handle the worse-case dissipation of >55W. The PMI OP-07 is also quite interesting, no idea what it is doing there.
Also note the (partly) guarded trace on the left of the board, running to the relay. Probably something to do with the most sensitive current range.

[42Khz]

Moving on to the DA board, which holds a DAC703 16-bit DAC and an LM399, along with some other precision components.

The circuitry around the crystal in the bottom left looks like it might be some discrete crystal oscillator. Quite a lot of components to make an oscillator, though, maybe it's doing something interesting.

[42Khz]

Then finally the current sense board, which is covered by a shield on both the front and back. It looks like the current sense resistors are spread all over the place. There is a fancy reed relay near the bottom left, presumably for the most sensitive current range (1nA resolution). I'm starting to think the purple ceramic ICs might be custom ASICs of some kind.

The first closeup photo shows two high-current resistors, one being 1 Ohm (2-wire) and the other 0.1 Ohm (4-wire). These must be for the 320mA and 2A range, respectively.

A second closeup shows the fancy hermetically-sealed resistors for some of the more sensitive current ranges. I wonder why Advantest felt the need to add trim capacitors here. That together with the large amount of TL072 suggests that something interesting is happening here.

The little red thing in the top middle is a trimmer that is accessible when the card is inserted in the unit. It looks to be the only trimmer, presumably the majority of the calibration is done in digital (i.e. predistorting the DAC code to perform offset and perhaps gain correction). The guarded trace around U16 (PMI1008) and the big ceramic package also stands out. If anyone knows what the ceramic ICs are, please let me know!
I bet the control circuitry (IV clamp and integrator) are also on here, but I can't instantly make them out.

[DavidKo]

How it looks with the transformer? I have found out, that my R6143 seem to have only 100/110 or 200/220 transformer (based on the color coding of the wires, one wire is missing compared to my another Advantest stuff). I'm not sure if I have measured the voltage across the other winding to be sure or if I have used only the info from voltage sticker.

[42Khz]

How it looks with the transformer? I have found out, that my R6143 seem to have only 100/110 or 200/220 transformer (based on the color coding of the wires, one wire is missing compared to my another Advantest stuff). I'm not sure if I have measured the voltage across the other winding to be sure or if I have used only the info from voltage sticker.

I'm not sure I understand exactly what you mean, but perhaps the following is helpful.

In my unit there are three grey-sleeved cable pairs going from the transformers down into the unit (see top left of the image included in the first post). Each contains a red and a white wire, and one of the sleeved pairs goes into a bridge rectifier on the bottom of the unit. Probably feeds the front panel or something, I didn't care to check. The other two (one from each transformer can) go to the mains switch. From this I concluded that there is only one transformer tap externally available per transformer can, so no proper way to wire it for 220V. I don't want to break the cans open, so I don't know what's inside them.