Group Actions
Engagement
  • Author Author: Jan Cumps
  • Date Created: Date Created
  • Views 4968 views
  • Likes 10 likes
  • Comments 16 comments
Related
Recommended

DIY Source Meter Unit - initial thoughts

Jan Cumps

I'm thinking about building a low power SMU.

We're playing with OpAmps and discrete semiconductors regularly over here, and there's a Keithley 2450 SMU with I-V Tracer Software  road test going on.

Learnings from those activities can be applied here.

 

image

image: over simplified concept

 

Goals:

Only two;

  • Make a Source Measure Unit.
  • Make it programmable or scriptable

 

Possible capabilities:

This would be a low power instrument, µA up to maybe a few mA. Low voltage. A utility that can be used with analog low power devices.

Source and sink voltage and current. positive and negative.

Constant current, constant voltage.

(undefined) precision and speed

 

It should be possible to make this modular:

  • the driver part
  • the measure part
  • controller and converters
  • human and digital interfaces

 

I'm going to read up on the subject now. Ideas and support are welcome.

  • in reply to shabaz

    Just noticed that blog post has punctuation stripped out, probably during a Jive upgrade : (

    I'll go through and fix that sometime..

  • Hi Jan,

     

    There's a couple of open source SMU-like devices that might be worth exploring, in case it gives some ideas for your project.

    One is ADALM1000 ADALM1000 see here for more info Getting Started with the Active Learning Module ADALM1000  it has limits of 5V and 200mA, so that is quite restrictive. It cannot handle negative voltages, although there is an on-board 2.5V output which could be used instead of ground for some measurements.

    It is only £25+tax, so well worth it just to experiment with.

    Also, there's some benefits maybe to using the same protocol/API for your project or extending it. When I wrote up about it, people just (I'm sad to say) whinged about the software because they couldn't install it on Linux without making some effort, but nowadays there's Windows software for it too. The open source software is called PixelPulse2, maybe it could be adapted to suit your SMU design, to save some effort. I don't know if you intend on-board LCD, or attaching to a PC. The same interface protocol is used by a software suite called ALICE which provides additional functionality although then it not really functioning as an SMU but as other instruments, like impedance meter, logger, etc.

     

    There's a more advanced model called ADALM2000, but that's unfortunately ten times the price, but looks impressive (again within its limits) - just too much beyond my budget for experimentation though, although I think it is still value-for-money as a lab-in-a-box.. I just wish it was lower cost. It's a much more complex design than ADALM1000 though.. uses an FPGA, and I think USB 3.0..since it supports 100Msps. The ADALM1000 has a limit of (I believe) 100ksps.

  • in reply to michaelkellett

      wrote:

     

    I don't think that TI are telling the whole story with their diagram of the current output of the DAC8775.

    As they draw it (page 35 of the data sheet)

    considering just the lower or sinking half:

    Iout = -(DACvolts - VNEG_INx)/Rsense, so the sink current is a function of the difference between the negative supply and the DAC voltage.

     

     

     

    ...

    MK

    Yes, just shopping for ideas at the moment, browsing, reading and collecting. I'm fully aware that the drawings and text in the TI application note are a vast oversimplification.

    I'm also looking at your circuit and the way you approach building for performance and precision.

     

    And worse, at zero DAC volts both amps will be trying to turn their MOSFETS on enough  to drop the full rail voltage across their sense resistors.

    They don't show it in the drawing, but in the operating narrative they explain that they never drive both outputs: "Only one of the current stages are active at any given time".

  • I was thinking of separate modules for driving current and voltage, though I can see how they might be combined

    This is an interesting idea.  Going a little bit further, it might me nice if the voltage/current source could be modular enough that you might be able to a high and low current version.  In addition it might be nice if you could have the option to have single and dual channel versions.  I was really interested in the Keithley SMU that was roadtested until I realized that a single channel unit could not be used for testing DC-DC converters, which would be handy.  Just blue sky thinking at this point as this is a large undertaking and the last thing you need at the start is a lot of feature creeping.

     

    Nice to have so many experts and extremely thorough thinkers in one room!

  • in reply to Jan Cumps

    I don't think that TI are telling the whole story with their diagram of the current output of the DAC8775.

    As they draw it (page 35 of the data sheet)

    considering just the lower or sinking half:

    Iout = -(DACvolts - VNEG_INx)/Rsense, so the sink current is a function of the difference between the negative supply and the DAC voltage.

     

    And worse, at zero DAC volts both amps will be trying to turn their MOSFETS on enough  to drop the full rail voltage across their sense resistors.

     

    You could fix this by using a diff amp to look at the voltage across Rsense.

     

    If   is reading this is right up his street for simulation.

     

    The problem is that with two diff amps this circuit needs 10 precision resistors to work.

     

    Here's a very quick and messy LTSpice of it, with diff amps, can't make it work without.

     

    image

     

    The classic Howland current source (TI app note snoa474a) only needs 4 precision resistors.

     

    MK