• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Intelligent Curve Tracer 3.0 release

More complex?

The circuit is complex.

What are a few inverters and a dual polarity power supply? Not much compared to the entire range of your project! Look at the PCB, it is complex now!

Why not build one that does the work?

There is a phrase to describe it: "false economy".
This means that while trying to make something cheaper, or save money, you make it more difficult, expensive and save nothing, make it cost more.

Just my opinion.

If your board + software would do the tasks I would buy one. :D

_-_-bear
 
You can try to design a curve tracer circuit use dual power supply ,
and compare with my circuit, who used the opamp and relay more.
If your ADC and DAC is unipolar,you need more opamp.


My circuit to understand is complex, But it is clever.
May the base drive part of the complex, but it realizes the constant current, constant voltage, and the isolation measurement,
 
You load 'PNP Vce-Ic' condition
and plug PNP in slot, Base insert B,Collect into C,Emitter into E,
and press 'GO!' button.

JFET is the same,load 'measure continue' and plug Gird into B,Source into E,Drain into C.

Yes this is what i have done, but pheraps is it important the order?
First i have placed the device into the socket and then i have selected the "PNP Vce-Ic" condition.

But at the last, i not understand well how this happens, now the curve tracer trace also the PNP Bjt and Jfet curves. :rolleyes::spin:
 
The software update
http://www.diyaudio.com/forums/attachment.php?attachmentid=291072&stc=1&d=1341998408

It support script control,
Example use script to measure a cluster of Vce-Ic.

attachment.php

The script is

V0=2.5
mode i
setrc 0
Eout V0
startwatch
color=random(999999)
for i=2 to 10
setib i
notice "IB=",i
for j=0 to 72
cout j/2+V0
waitrelay
getadc 2
ic=1000*vrc/currentRc
ib=1000*vrb/currentrb
curvepoint str(i),vce,ic,ib,ve
next
endcurve str(i),color
cout v0
next
d=stopwatch()
notice "time:",d
end


Ib from 2mA to 10mA,9 curve ,and spend time 6.9s
 

Attachments

  • ads7871_script.zip
    870.7 KB · Views: 439
  • psb.png
    psb.png
    102.6 KB · Views: 2,227
Ex-Moderator
Joined 2011
Most Tube is like N type J-FET in normal voltage, but if work in low voltage,Tube like some BJT.

I had used CT3 to test some Tube in low Cathode-Anode voltage.
but under low Cathode-Anode voltage,Even if the gate voltage is 0, Most tube have no anode current.
you must give gate positive voltage,the tube will have some current,and the Gate current almost equal anode current.



Because High voltage and No P Type ,So this Intelligent Curve Tracer circuit not suit Tube test.
I plan designed another circuit to test Tube,Of course It will be considered to A2 and AB2 state, and measure the Anode current ,gate current and screen grid current.

Any news on the tube curve tracer, looking forward to your new design. :)

Jaz
 
1. I try measurement npn, pnp and n-fet with varios tamplates. And always the same result - one point.
2. DUT pins woks, type of elements is defined, but the measurements are not valid, for example, the drain current for 2SK170 is 663mA when the maximum possible is 16mA, and measured Hfe for 2SC1815GR is 118, when it is in the datasheet 200-400.

Have you checked my tracer before sending it? On the back written number 226#.
 
May be the decimal point in ads7871_v3s.ini file caused,
In some non-English windows, the decimal point is ',' instead of '.'.

You select menu 'Options->config'
Check what is the ‘B/C/E Channel PowerAmp Gain’ display? They all are 4.11.

If not 4.11, it mean your windows system read '.' not as the decimal.
Use notepad to open the ads7871_v3s.ini, and replace all the '.' as ','
 
In INI file on CD that you put in the package, all the gains values are 2.

First, you write that the separator should be ".", And then say to replace all "." to ","... I enter a value through a point, but after opening it changes to a comma.

Ok, so it works! I ran the new version without the old ini file. :)

New question. What values have resistors RB and RC? Values in old ini are correct?

And can you write a manual, how to calibrate tracer? I see, that ADC values on Option/Diagcostic window are not 5V.
 
A year ago, I was working with I/O board manufactured by National Instruments and measured some Soviet FETs. To do this, I was written in LabVIEW similar curve tracer. Unfortunately, your algorithm is in the manual in Chinese, so I can not compare it with mine. But I can compare the results.

PIC19.png

Blue is measured with NI (how much is average, I do not remember)
Red - with tracer with averaging 10 times
Green - with checkbox "Calibration" and averaging 66 times

JFET was n-channel 2P303A

I think that the curve tracer is noisy and suitable only for qualitative screening and selection. To measure the quality of a semiconductor, such as the homogeneity of the slope, this unit in this version is not suitable. Need to work on the noise level. Or am I doing something wrong?

And where are points with the gate voltage -0.09, -0.28 and -0.64V?
 
My instrument is difficult to distinguish within 2~5mV voltage.
So the Vgs-Id/Vbe-ic curve that Vgs/Vbe less 1V is limited accuracy.


For your J-FET measure , you use RC=1K or 12K?
Because your Id is less 1.5mA, If you select RC=1K,
On Id=0.2mA position,The Vrc(Voltage drop of RC) only 0.2V,If the resolution of my curve is 5mV, On this position, Means that the Id error will reach 5/200=2.5%,


If you select RC=12K, then on Id=0.2mA position,The Vrc(Voltage drop of RC) is 2.4V, Means that the Id error only 5/2400=0.2%.

Select big value RC, can get more precision, but you must remember, The Vds(Vce) + Vrc must less 36V. Because 36V is the power supply limite.
 
I used a template, there was selected 75 ohm. I do not remember what resistance I used in my curve tracer. Or 500 ohm or 1K.

The menu for the 2 values ​​1K RC: 1K, 75, 1K, 12K. By default, the template is selected 75.
PIC20.png
I think that resistors values ​​in the menu are not correct. How to fix it?
Using the second value 1K or 12K curve becomes much smoother, you are right :)
PIC21.png
 
In fact, RB / RC value of template can be change, It depends on your measurement range.


The formulation of the measurement conditions can refer to the following schematic.
267816d1329875125-intelligent-curve-tracer-3-0-release-11.png


System only can measure the Vbe/Vce/Vrc/Vrb, and the voltage resolution is 2~5mV.
So if you want more precision, then the Vbe/Vce/Vrc/Vrb much larger than 5mV. Such as more than 0.5V.

The current measurement is based on the I= V / R.
So you have to choose RB/RC large enough so that the volatge drop on resistor is much larger than 5mV.

But RB/RC can not too big,because
1. Vrb(Voltage drop of RB) + Vbe must less than 36V.
2. Vrc(Voltage drop of RC) + Vce must less than 36V.
 
Last edited:
Thank you, I understand. The question was in another. The drop-down list contains two values ​​1K. But in terms of errors, the first value is less than 75 ohms. What resistors are mounted on the board in reality, and how to determine it?

The values ​​of RB: 500, 6K, 91K, 1360K
The values ​​of RC: 1K, 75, 1K, 12K