Micro controller based curve tracer

Thread Starter

talhaali

Joined Nov 24, 2011
19
The aim of the project is to design and develop a Curve Tracer. A curve tracer is a useful test & diagnostic equipment that facilitates plotting the characteristic curves of various electronic components like Resistor, Diode, Bipolar Junction Transistor (BJT), Field Effect Transistor (FET) etc. These curves offer insight into the properties of the electronic component and is helpful to a design engineer. The design will involve a microcontroller (or FPGA) based system for generation of precise test signals (stimuli) for probing an electronic component. The test signals get converted to analog form by Digital to Analog conversion circuitry. The sensors will capture the desired signals from the electronic components, and convert them to Digital form using Analog to Digital (A/D) conversion circuitry. This will be processed further by microcontroller (or FPGA). An interface to an input device (e.g. keyboard, mouse) and display device (e.g. a monitor) will also be developed. The results (plots) get displayed on the display device.
 

t06afre

Joined May 11, 2009
5,934
So what's the question?
I think it is more some sort of assignment description
I think it is way to complex for any member here on AAC. But feel free to prove me wrong. Post a complete project within 4 weeks on the "The Completed Projects Collection" forum. With all schematics and code. And I will agree to have humble pie for dinner that day. Are you up to the challenge Mrchips
 

SgtWookie

Joined Jul 17, 2007
22,230
There was a project similar to this published in EDN a decade or so ago.
http://www.edn.com/article/491912-Trace_voltage_current_curves_on_your_PC.php
Unfortunately, they dumped the PDF files from the server; but the .zip file is still available, and you can view the complete article in this .pdf:
http://www.edn.com/contents/images/101101di.pdf
It starts on page 112 (the 4th page in this file).

Of course, this would need to be converted to uC code, and use USB communication rather than a LPT port. However, it can still be useful as an idea for something that once worked.
 

MrChips

Joined Oct 2, 2009
30,707
I think it is more some sort of assignment description
I think it is way to complex for any member here on AAC. But feel free to prove me wrong. Post a complete project within 4 weeks on the "The Completed Projects Collection" forum. With all schematics and code. And I will agree to have humble pie for dinner that day. Are you up to the challenge Mrchips
Sure. I'm in the process of building one right now. The challenge is getting the speed up so that you do not have to wait all day for a decent scan.
 

t06afre

Joined May 11, 2009
5,934
@talhaali, joking aside. I do not think anybody will just hand over a complete project with code and schematics. But you are free to ask any specific question regarding your project. If you put in some effort. You will be reward on this forum. No doubt about that
 

CDRIVE

Joined Jul 1, 2008
2,219
My wife complains that I never finish anything I start and most of the time she's right. This topic is a good example of my shortcoming. I started a PC based curve tracer last year and got side tracked,.. as usual! Anyway, it is based around a Picaxe UC and outputted to the PC via RS232, so I could use VB's MSComm as the hardware interface. Someday, I'll finish it ....... after I finish all the other ones! :rolleyes:
 

Thread Starter

talhaali

Joined Nov 24, 2011
19
tnx guys actuy m having problem with micro controller part , its intrfacinf n code :( i want to do it with pic micro controller
 

MrChips

Joined Oct 2, 2009
30,707
I have an analog curve tracer that sweeps through the supply and bias voltages and displays the characteristic curves on an analog X-Y scope.

All you would need are two DACs and one ADC interfaced to a computer. As I said you will run into speed constraints and it will take awhile to acquire sufficient data points in order to get some decent curves.
 

Yako

Joined Nov 24, 2011
245
I couldn't do this in a fit (at least I am honest)

Requirements: (all MCU-based)

  • Display to RGB monitor
  • Mouse input ...
  • Keyboard input too!
You're asking for like an operating system man. And in 4-weeks!
 

Yako

Joined Nov 24, 2011
245
I have an analog curve tracer that sweeps through the supply and bias voltages and displays the characteristic curves on an analog X-Y scope.

All you would need are two DACs and one ADC interfaced to a computer. As I said you will run into speed constraints and it will take awhile to acquire sufficient data points in order to get some decent curves.
Well that's a start, you have what is called the 'front end'
 

Thread Starter

talhaali

Joined Nov 24, 2011
19
I couldn't do this in a fit (at least I am honest)

Requirements: (all MCU-based)

  • Display to RGB monitor
  • Mouse input ...
  • Keyboard input too!
You're asking for like an operating system man. And in 4-weeks!
there isnt much time , i need it till march 2012
 

Thread Starter

talhaali

Joined Nov 24, 2011
19
I have an analog curve tracer that sweeps through the supply and bias voltages and displays the characteristic curves on an analog X-Y scope.

All you would need are two DACs and one ADC interfaced to a computer. As I said you will run into speed constraints and it will take awhile to acquire sufficient data points in order to get some decent curves.
can u provide me with the data for this analog trace , so that i cud start off,
 

Yako

Joined Nov 24, 2011
245
there isnt much time , i need it till march 2012
You're joking right?

Schematics, PCBs, a prototype and product documentation for a project of this scale done by one person in this time frame?

Cannot see it being much of a project. Something like this a couple of engineers would likely quote around USD 50K to do man.
 

SgtWookie

Joined Jul 17, 2007
22,230
Yako,
It's not all that bad. There are lots of projects out there that have interfaced with peripheral equipment.

Rather than discouraging talhaali from the project, it would be better to point out that the human interface could be left up to a personal computer, and then the problem is reduced to communication with the PC via an existing protocol, with USB being high on the list of preferred protocols.

That way, much of the pressure is taken off, and talhaali can concentrate on the basic functionality of receiving commands from the PC, performing the trace on the semiconductor in question, and returning the data to the computer for processing.

Once that is working, then bits could be added on, like a USB interface to a mouse and a keyboard; even RGB video if they plan for those things in the original hardware concept.

If he needs data from a trace, LTSpice (available as a free download from Linear Technology) has an educational simulation called curvetrace.asc supplied with it. The simulation could be used to write out data from simulated curve traces.
 

MrChips

Joined Oct 2, 2009
30,707
As I said, I am already working on this. I had my system up and running in one afternoon. What you need are two DACs and one ADC.
 

Yako

Joined Nov 24, 2011
245
Depends on how determined you are I guess. I gave myself 6 migrianes doing some Java programming assignments. USB, even harder again -- the protocol is documented in like about 200 pgs of text. Some of these MCUs with on-chip USB ports that are coupled with drivers for your PC making it much easier. I have seen it done without a USB specific port on the host side using a cheap PIC.

There is also the option of a serial-to-USB converter too. Doubt that fast transfer speeds will be required for the PC software to read the data from the front end.

What PC language do you plan on using to write the software?
 

MrChips

Joined Oct 2, 2009
30,707
Here is the B&K Precision Model 501A Curve Tracer




and the curves being displayed on an old Tektronix Type 503 Oscilloscope:

 
Last edited:

ErnieM

Joined Apr 24, 2011
8,377
I've used some curve tracers such as this one:



While ancient they are incredibly useful devices as they cover such a wide range of voltages and currents. For example, to test a diode for leakage and breakdown one may be putting 100V across the diode and reading uA or even nA, then switch to a mA scale to find where the diode breaks down at some higher voltage. So the horizontal scale is calibrated in volts while the vertical is in amps.

So it's not quite just some D2A and A2D units. You also need some current to voltage conversion, scaling, voltage sources to perhaps 1500 volts, programmable current sources (base drive if nothing else), quite a lot of things.

I'm not discouraging the project, just trying to begin to define the extent of the blocks necessary. One good starting point is to take an existing curve tracer, get it's list of functions and use that to define the necessary blocks.

Modern curve tracers are modular in design to accommodate (expensive) functions only some users mey desire, such an approach may be used here also.
 
Top