Show Us Your Curve Tracer

[Electro Fan]

It is very interesting to see the various approaches to "Curve Tracers":  analog, digital, hardware, hardware plus software, standalone, with an oscilloscope, with a computer, with a SMU - in each case the design reflects the objectives and other thinking of the designer(s).  Some curve tracers are for transistors, some for tubes, some for other active devices and/or passive components - lots of possibilities.

I think there might be a bit of a resurgence coming in curve tracers, for various reasons and purposes, not the least of which is education. 

Please feel free to post your favorite curve tracers (including any images or links) that you use, built, or designed.  Tell us a little bit about your use case(s) and what you like about your curve tracer and what if anything you might like it to do differently.  Or feel free to post other related info that helps show the various uses and designs for assembled, kit, or DIY curve tracers. 

Some people say that the world's first curve tracer was a Tektronix 570 but maybe someone knows of earlier curve tracers.

Here are several links for people wanting to get oriented toward curve tracers (intended to show some of the principles of curve tracers, not to advocate any particular product):

https://vintagetek.org/curve-tracers/

https://en.wikipedia.org/wiki/Semiconductor_curve_tracer

https://www.evaluationengineering.com/instrumentation/article/21126325/whats-the-difference-classic-curve-tracer-vs-smu-with-curve-tracer-software

https://www.rs-online.com/designspark/how-to-use-an-oscilloscope-as-a-curve-tracer

https://www.keysight.com/main/editorial.jspx?cc=US&lc=eng&ckey=2838609&nid=-32110.1203274.02&id=2838609

https://www.allaboutcircuits.com/technical-articles/understanding-i-v-curves-part-i/

https://www.qsl.net/kd7rem/pdf/octopus.pdf

https://www.microwaves101.com/encyclopedias/curve-tracer-measurements

https://www.ossila.com/pages/iv-curves-measurement#

https://www.electronics-tutorials.ws/blog/i-v-characteristic-curves.html

[Electro Fan]

A few more links in case it stimulates any posts.

Excellent fundamental video on curves by W2AEW:


A very cool curve tracer:  ELV KS7000


A pretty practical/useful tracer:  DCA75 - Atlas DCA Pro Advanced Semiconductor Analyser
https://www.peakelec.co.uk/acatalog/dca75-dca-pro-semiconductor-analyser.html

An ideal entry/intermediate curve tracer (imo) might combine some of the functionality of the ELV KS7000 and the Peak DCA75 

Edit  Added comments and link:

An extension of the integration of ELV-like and PEAK-like functionality might be something like these Huntron products:

https://youtu.be/DHoEQivIdQQ

https://youtu.be/s-mRtOIHqqM

The realm beyond Huntron seems to go in the directions of perhaps still more functionality and flexibility and no doubt higher voltage thresholds as pioneered by the vintage Tektronix curve tracers.

Edit  Added:



Edit  Added:

https://www.eevblog.com/forum/testgear/transistor-curve-tracer-kit-on-ebay/?all

[graybeard]

Here is a 3D red-cyan anaglyph of my Tektronix 575 curve tracer.  It requires red (left) - cyan (right) anaglyph 3D glasses to view.  Click on it for a full size version.



Chris

[capt bullshot]

Recently I've used my Tek 577 to test a simple circuit (PWM to Resistance converter for a particular purpose):





The curve tracer is set to 0.1V/div, 50uA/div and 0.1mA per step with some offset.
Not too linear, but I believe it'll do the job as it will be part of a closed loop. The circuit is intended to control the output voltage of a COTS SMPS by paralleling its output side to the adjustment pot. Voltage across that pot is low enough for this circuit to work. Of course, the curve tracer doesn't apply an PWM signal to the photocoupler, but PWM will be fast enough it gets smoothed out by the systems frequency response.

[bd139]

I used a 33120A and TDS210 as a curve tracer here to test a tunnel diode:

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg1974956/#msg1974956

33120A has a floating output which makes this easier to do.

[flash2b]

See my topic in the Dutch blog: https://translate.google.com/translate?hl=&sl=nl&tl=en&u=https%3A%2F%2Fwww.circuitsonline.net%2Fforum%2Fview%2F141079

[jan28]

BB3 running micropython curve tracer script: https://youtu.be/I0y6fgv8G00

Screenshot from YouTube video:

[Martin.M]

575  with the 400v mod

[Cubdriver]

Gotta love those old beasts!





-Pat

[jjoonathan]

SMU gang represent!

(glances to the left, then to the right -- coast is clear, SPA gang isn't around yet)

Custom 3D printed feet and "panel" (bar) turn this SMU into a convenient bench tool. Actual work is done in jupyterlab / holoviews / pyvisa-py.

[Johnny10]

575! No, that is my curve tracer!

[bd139]

SMU gang represent!

(glances to the left, then to the right -- coast is clear, SPA gang isn't around yet)

Custom 3D printed feet and "panel" (bar) turn this SMU into a convenient bench tool. Actual work is done in jupyterlab / holoviews / pyvisa-py.

This is quite funny. I started designing a generic Tektronix grade curve tracer replacement last year. By the time I’d broken it down into logical blocks, made everything as generic as possible and attached a USB interface to it on paper I stepped back and squinted and it turned out I’d designed an SMU. I lost interest then for some reason.

[Electro Fan]

BB3 running micropython curve tracer script: [url=https://youtu.be/I0y6fgv8G00]https://youtu.be/I0y6fgv8G00[/url]

Screenshot from YouTube video:
(Attachment Link)

Pretty Nifty!

https://www.crowdsupply.com/envox/eez-bb3/updates/micropython-arrives-on-bb3

EEZ Bench Box 3:
https://www.crowdsupply.com/envox/eez-bb3

Edit:  Looks like a thread on the BB3:
https://www.eevblog.com/forum/projects/building-log-for-eez-bb3-enclosure-bare-boards/

[Electro Fan]

(Attachment Link)

575  with the 400v mod

Beautiful!

[Electro Fan]

Gotta love those old beasts!





-Pat

Beautiful outside and inside!!

[Electro Fan]

See my topic in the Dutch blog: https://translate.google.com/translate?hl=&sl=nl&tl=en&u=https%3A%2F%2Fwww.circuitsonline.net%2Fforum%2Fview%2F141079

Nice scope curve trace display.

[Cubdriver]

Found the one I took of my youngest lab assistant with the 'newest' curve tracer, too.  He is observing a color-cycling LED as it transitions between hues, thus the weird instability in the trace (and he's now a third again bigger than he was in that picture):



-Pat

[jxjbsd]

Found the one I took of my youngest lab assistant with the 'newest' curve tracer, too.  He is observing a color-cycling LED as it transitions between hues, thus the weird instability in the trace (and he's now a third again bigger than he was in that picture):



-Pat

Cat: You can increase the vertical gear one more gear.

[rcjoy]

Here's HP's version of the curve tracer, the 4145A Semiconductor Parameter Analyzer!

[simyaci]

this device is also my design

[wolfy007]

Heres an old one from me, using the classic 7603 + 7CT1N (sorry taken in low light with phone)

[Johnny10]

The Tektronix 577

[Shock]

My Tektronix 576 curve tracer requires a power supply restoration but aside from that is a fairly clean example.

Here is a video that shows one approach to filter capacitor replacement. I may do mine a little differently so it's in the queue till I can devote the time. 49m40s shows their placement.

[rsjsouza]

I have been monitoring prices for years and would love to have a higher voltage unit such as the 575/576/577, but unfortunately the disposable income and space in my lab are very limited.

So, this past Christmas I got a reasonably spec'd USB based curve tracer from a Chinese developer nicknamed locky_z. The unit is well made, the software is rock solid and well featured (despite having a small translation issue here and there) and it has good midrange specifications of 0-36V up to 2.5A.

In the pictures the unit is testing an old 2N6907 and bejng driven by my Power Designs 5015A.

I also did a video about it for my channel (in Portuguese only, sorry)
https://youtu.be/NjhEYeM-FWM

[Electro Fan]

I have been monitoring prices for years and would love to have a higher voltage unit such as the 575/576/577, but unfortunately the disposable income and space in my lab are very limited.

So, this past Christmas I got a reasonably spec'd USB based curve tracer from a Chinese developer nicknamed locky_z. The unit is well made, the software is rock solid and well featured (despite having a small translation issue here and there) and it has good midrange specifications of 0-36V up to 2.5A.

In the pictures the unit is testing an old 2N6907 and bejng driven by my Power Designs 5015A.

I also did a video about it for my channel (in Portuguese only, sorry)
https://youtu.be/NjhEYeM-FWM

Cool

Looks like locky_z sells kits and assembled versions on eBay:

https://www.ebay.com/itm/Locky-z-2019-Curve-tracer-/114233327020

https://www.ebay.com/itm/Locky-z-2019-Curve-tracer-with-new-component/114234796434?hash=item1a98eca992:g:91cAAOSwncJeyNdX

Also saw your unboxing video - it helps show what comes with the kit.  Thx!

[Calvin]

Hi,

Having waited for years, I ordered my locky_z on ebay as soon as he offered the latest batch last year.
Unfortunately does it not work as I hoped for 
It mostly draws just weird fuzzy lines on the display.
So I just put it aside since I had no time to dig into the issue deeper.
I heard though that another user had similar problems and got a replacement.
Unfortunately is locky_z  hard to contact and doesn´t respond very well (starting from not giving a rating on ebay to not responding on mails).
My old Tech 576 is a much more trustworthy guy in comparison. 

regards
Calvin

[Johnny10]

What about the B&K Model 501A ?

My first Curve Tracer.
Now where did I put it?

Remember also that all of the Tektronix and B&K models came with terrific manuals,

[rsjsouza]

I am glad you liked the content, Electro Fan.

The kit is really good as the schematics and the microcontroller code are freely available. Also, one additional aspect is the ability to create your own scripts for testing, as well as save the data in tabular text form for post processing. I haven't done that yet, but it is quite promising.

I am having a blast with it.

[rsjsouza]

Having waited for years, I ordered my locky_z on ebay as soon as he offered the latest batch last year.
Unfortunately does it not work as I hoped for 
It mostly draws just weird fuzzy lines on the display.

Calvin, I don't know what extent of testing you did but, depending on the set conditions, the DUT will only show really squiggly lines as it is not properly biased. I did that at the beginning and was really miffed at the results.

For example, the software by default sets the base resistor to 1M, which is quite high for most of the general purpose bipolar transistors. Also, if the collector resistor is too high, the Ic will be too low and therefore potentially cause the transistor to be simply in cutoff.

At any rate, I am sad to hear that you had a bad experience. His user's guide has a very thorough debug section, though. Yes, not an ideal solution but still some hope. Good luck!

[Messtechniker]

My semi-finished  semiconductor curve tracer. Completed and verified D/A converter and USB link. Output stages need to be built one day.
More on this here:
https://www.dl4jal.de/kls/kls.html

[FrancisM]

Here's my own one, designed for small signal transistors. I got the idea from an old Elektor article.
It can show output characteristics and transconductance curve of jfets or mosfets.

The sweep is an adjustable triangular wave of +-14V max, the current limit can be set to 25mA or 250mA. Done with a TL074.
The base (gate) is voltage driven by a pic18F1320 and a double buffered dac MCP4822. I select the base resistor with jumpers.
Having a single bipolar supply, the current is monitored across a 0.1 Ohm shunt in the emitter (source). It's then amplified 100x by an AD797 to the Y output. I can change triangular's wave frequency to get rid of the looping effect with highly capacitive transistors.

The result is not that bad on newer DSOs, no flickering at all with the right setup, but still not such a nice trace as with my TEK 2465B.
Things may probably change in the next coming years with affordable high definition scopes.

The first screenshot is from an NPN ZTX851, little transistor with big die. It behaves nicely at low Vce. I don't like the colour but can't change it.
The second one is the transconductance curve from an N chanel PF5102. It was first in the fridge, then on the setup and heated with an air gun. With persistence, one can see the sweet spot of it's zero tempco.

I also own a USB Locky_Z one. I got it many years ago at the first run for cheap. It's fully made of used parts and was first released for the chinese market. But it still works.

Francis

[Martin Hodge]

A Tek 575 connected to an Eico 667 tube tester through a custom curve tracer circuit.
Curves are from a 6AU6 pentode.

[graybeard]

A Tek 575 connected to an Eico 667 tube tester through a custom curve tracer circuit.
Curves are from a 6AU6 pentode.

I have been considering how to adapt tubes to my 575.  Thank you for this example!

[Jay_Diddy_B]

Hi group,

I have a few. I have a Huntron 2000 for troubleshooting.

Tektronix 576 for looking at transistors and diodes.

Tektronix 571

My best curve tracer is an HP 4155B Semiconductor Parameter Analyzer:




This picture is Id and GM versus Vgs for 2N7002 MOSFET.

The vertical scale (Id) is logarithmic and goes from 1fA (1E-15) to 100mA (1E-1) 14 decades 

Regards,
Jay_Diddy_B

[Electro Fan]

Hi group,

I have a few. I have a Huntron 2000 for troubleshooting.

Tektronix 576 for looking at transistors and diodes.

Tektronix 571

My best curve tracer is an HP 4155B Semiconductor Parameter Analyzer:

(Attachment Link)


This picture is Id and GM versus Vgs for 2N7002 MOSFET.

The vertical scale (Id) is logarithmic and goes from 1fA (1E-15) to 100mA (1E-1) 14 decades 

Regards,
Jay_Diddy_B

14 decades to a femtoamp?!  Very neat!

[jh15]

Where is pic?
I'll show you mine (Tek  575 with pussy on top enjoying the heat, , if you show yours. (Might just show 2 photos offset by eyelength for 3d.)

[sorenkir]

This one is quite an oscure model (Br 01 from Leybold-Heraeus, made in Germany):



Front panel is mislabeled : read Uc instead of Ub on the right!
I think I will have to do a lot of reverse engineering to get it back to work, as it died the other day...



Michel.

[Calambres]

I made one based upon the infamous "Thai Kit" (https://www.ebay.com/itm/141138017533).
Already talked about it some time ago in this very forum.



This kit is also heavily based upon an ELEKTOR magazine article (Dic. 1989). I made a small modification to the kit to let it switch between 8 and 4 curves and subsequently the total µA/mA. This is a photo of the guts while I was still prototyping it:




It needs a good ole CRO (preferred) with XY mode to show the curves.



The most interesting feature of my build is that it can compare (match) up to 6 transistors/FETs at the same time at the flip of a rotary switch. Here you can see it in action testing some transistors, first with four small transistors (2SC1815) and one power transistor:



The second is showing the details of four small transistor (2SC1815) curves:

[Ground_Loop]

This is my vintage Heathkit. It has served me well. The one thing I would change about it is the single ended sweep to bipolar. It does sweep positive and negative, but not continuously. You need change the control settings to sweep the opposite direction. I gave some thought to modifying it, but gave up after realizing I would need to hack on the enclosure to accomplish it.

[Electro Fan]

This is my vintage Heathkit. It has served me well. The one thing I would change about it is the single ended sweep to bipolar. It does sweep positive and negative, but not continuously. You need change the control settings to sweep the opposite direction. I gave some thought to modifying it, but gave up after realizing I would need to hack on the enclosure to accomplish it.

(Attachment Link)

(Attachment Link)

That's a nice combination of curve tracer and oscilloscope.  Definitely good idea to retain their vintage dignity.   

[Electro Fan]

I made one based upon the infamous "Thai Kit" (https://www.ebay.com/itm/141138017533).
Already talked about it some time ago in this very forum.



This kit is also heavily based upon an ELEKTOR magazine article (Dic. 1989). I made a small modification to the kit to let it switch between 8 and 4 curves and subsequently the total µA/mA. This is a photo of the guts while I was still prototyping it:




It needs a good ole CRO (preferred) with XY mode to show the curves.



The most interesting feature of my build is that it can compare (match) up to 6 transistors/FETs at the same time at the flip of a rotary switch. Here you can see it in action testing some transistors, first with four small transistors (2SC1815) and one power transistor:



The second is showing the details of four small transistor (2SC1815) curves:

Really nice post and congrats on your curve tracer.  I was going to ask how you did this:
"The most interesting feature of my build is that it can compare (match) up to 6 transistors/FETs at the same time at the flip of a rotary switch" but then I saw your link which explains your approach as you steadily worked it all out.  Your posts in that the thread and the whole thread had a lot of good info so I edited one of the posts at the front of this thread to make it easy to find the thread.

[Calambres]

If someone is interested, I have a pdf of the original ELEKTOR article the Thai kit is based upon. Not that it is very diffiult to find but if someone wants it without much hassle, drop me a note.

[edavid]

This is my vintage Heathkit. It has served me well. The one thing I would change about it is the single ended sweep to bipolar. It does sweep positive and negative, but not continuously. You need change the control settings to sweep the opposite direction. I gave some thought to modifying it, but gave up after realizing I would need to hack on the enclosure to accomplish it.

(Attachment Link)

(Attachment Link)

That's a nice combination of curve tracer and oscilloscope.  Definitely good idea to retain their vintage dignity.   

I disagree - I would never connect a curve tracer to an MCP scope like the Tek 2467B 

[Electro Fan]

This is my vintage Heathkit. It has served me well. The one thing I would change about it is the single ended sweep to bipolar. It does sweep positive and negative, but not continuously. You need change the control settings to sweep the opposite direction. I gave some thought to modifying it, but gave up after realizing I would need to hack on the enclosure to accomplish it.

(Attachment Link)

(Attachment Link)

That's a nice combination of curve tracer and oscilloscope.  Definitely good idea to retain their vintage dignity.   

I disagree - I would never connect a curve tracer to an MCP scope like the Tek 2467B 

Your concern is using up the useful life on the MCP or something else?  (I'm guessing Ground_Loop doesn't trace curves all day every day, and his MCP likely has it's timeout feature operating.)

[Ground_Loop]

I also have 7 scopes on my bench and a 555 in the corner.   On the rare occasion that I use the curve tracer I usually go to the 67 because i have a preset for the curve tracer.  And the timeout works just fine.

[edavid]

I disagree - I would never connect a curve tracer to an MCP scope like the Tek 2467B 

Your concern is using up the useful life on the MCP or something else?  (I'm guessing Ground_Loop doesn't trace curves all day every day, and his MCP likely has it's timeout feature operating.)

The timeout is a convenience, but it won't protect the CRT from burns if you make a determined effort.

(MCP = microchannel plate image intensifier)

[Martin Hodge]

A Tek 575 connected to an Eico 667 tube tester through a custom curve tracer circuit.
Curves are from a 6AU6 pentode.

I have been considering how to adapt tubes to my 575.  Thank you for this example!

The circuit is designed by Dennis Tillman. Attempting to attach PDF... nope, too big.
Instead search for "An Inexpensive Vacuum Tube Curve Tracer Adapter for All Tektronix Semiconductor Curve Tracers"

[Martin Hodge]

Here's the 575 testing the tunnel diode in the trigger of the 'b' timebase of a Tek 535A.

Vertical: 0.5 mA/div
Horiz: 50 mV/div

[Electro Fan]

Here's the 575 testing the tunnel diode in the trigger of the 'b' timebase of a Tek 535A.

Vertical: 0.5 mA/div
Horiz: 50 mV/div

Vintage Tektronix, tunnel diode, beautiful photography... doesn’t get much better.

Super Cool! 

[TERRA Operative]

I built one of the Thai Kits tracers, it works well.

https://youtu.be/uU_KjH-jJnA

I also have a Tek 577-177 that I have to repair one day (blown tants etc) and a 4145A that I have to track down a bunch of triax cables to connect it to the test fixture (and do the SD card floppy drive swap and maybe a colour LCD swap too if I feel like spending money).

[Anthocyanina]

If someone is interested, I have a pdf of the original ELEKTOR article the Thai kit is based upon. Not that it is very diffiult to find but if someone wants it without much hassle, drop me a note.

I would like that, please, i think it would be fun to build a curve tracer. Thank you!

[Calambres]

Curve_Tracer_(ELEKTOR_12-89).pdf

[QuadESL63]

My Philips PM6507 curve tracer:

Jac

[Calvin]

Hi,

regarding the locky-z intelligent curve tracer I had a bit of time to test it again.
Seemingly his own software rev.3.1 generated the problems on my system.
The now installed rev 3.5 which user erikovsky put on github seems better.
At least I could get some MOSFET measurements to look almost like MOSFET curves.

regards
Calvin

[jastreb]

Where did you buy that transistor holder. Love it

[mjs]

Some years ago I bought a boatanchor HP 4142B with 3 SMU modules. Some Python in Jupyter Notebook created a nice programmable curve tracer. Unfortunately it's way too big to keep in my home lab at the moment..

Here's an example of results:
https://chart-studio.plotly.com/~msyrjala/13

[Anthocyanina]

Curve_Tracer_(ELEKTOR_12-89).pdf

Thank you! I have a question about the transformer used for the power supply. how do you get +14v and -14v from a 9v transformer? (they list a 9v transformer in the BOM) Thank you!

[tubularnut]

Thank you! I have a question about the transformer used for the power supply. how do you get +14v and -14v from a 9v transformer? (they list a 9v transformer in the BOM) Thank you!

The notes refer to 14V as the 'unloaded' voltage, which will drop under load.

The secondary voltage rating will be the RMS voltage.

For a half rectified voltage with capacitor, the peak 'unloaded' voltage be 1.414 x secondary voltage, so for 9 volts, this will be just short of 13V. Given that a transformer will often give out a higher than rated voltage, its possible to see 14V.

When under load it will drop to around 0.9 x secondary voltage.

The +14V and -14V is because there are 2 half wave rectifiers, one for +ve and one for -ve.

[Larryc001]

Another project for another day I’m afraid.

[Anthocyanina]

Thank you! I have a question about the transformer used for the power supply. how do you get +14v and -14v from a 9v transformer? (they list a 9v transformer in the BOM) Thank you!

The notes refer to 14V as the 'unloaded' voltage, which will drop under load.

The secondary voltage rating will be the RMS voltage.

For a half rectified voltage with capacitor, the peak 'unloaded' voltage be 1.414 x secondary voltage, so for 9 volts, this will be just short of 13V. Given that a transformer will often give out a higher than rated voltage, its possible to see 14V.

When under load it will drop to around 0.9 x secondary voltage.

The +14V and -14V is because there are 2 half wave rectifiers, one for +ve and one for -ve.

Oooh, that makes sense. I have a 12-0-12 transformer that i would like to use for this. should i then use +9v and -9v regulators? Sorry if this shouldn't be the place to ask these questions, thought to ask since it's related to a curve tracer project! Thank you!

[tubularnut]

Oooh, that makes sense. I have a 12-0-12 transformer that i would like to use for this. should i then use +9v and -9v regulators? Sorry if this shouldn't be the place to ask these questions, thought to ask since it's related to a curve tracer project! Thank you!

If you use one half of the 12v-0-12v(i.e 0-12v) I think you should be able to swap it directly in place of the 0-9v transformer. The documentation for the tracer refers regularly to use of the unregulated supply.

[khutch004]

A uTracer kit plus an old tube tester:

(picture will follow when I figure out how to do that!)

[Electro Fan]

A uTracer kit plus an old tube tester:

(picture will follow when I figure out how to do that!)

Thanks for posting.  That caused me to look up the uTracer.  The comments from the designer are pretty neat as he walks through his thinking on evolving the design to higher voltage.

https://www.dos4ever.com/uTracerlog6/tubetester6.html

[THDplusN_bad]

Good Day,

@Ground loop, I totally agree that the Heathkit IT-1121/3121 is a joy to work with. Both types are virtual identical, but have different colours and front panel knobs.

I have repaired my unit back in 2014, and modified it with BNC-output connectors for the outputs to the oscilloscope (to replace the 4mm/banana connectors), added a fuse holder for base step currents with a "fuse fail" indicator lamp and a custom scale.
I have also made a simple adapter with a ZIF-socket for the A/B testing of components, which has proven its value. Matching transistors is a joy with this. 
Photos attached.

I also enjoy using a Tektronix 5CT1N (basically the Tektronix 7CT1N's little brother for the Tektronix 51xx/54xx oscilloscopes). And the latest addition to my labs is an Agilent N3280A DC component tester. More of the latter soon...

Cheers,

THDplusN_bad

[jh15]

I have one (575) in top shape, what is the problem with yours?

In college, there was a brand new one, and I really liked the manual that was under that top compartment 0n the cabinet. Can't find one anywhere.

[THDplusN_bad]

@jh15: This might be well known to you, but electronic copies of all Tektronix manuals for the type 575 curve tracer can be found here: https://w140.com/tekwiki/wiki/575

Cheers,

THDplusN_bad

[jh15]

Thank you, I may have missed it. It is a small manual that fits in the pop-up storage area on the top of the right shell. I'll look again. Next best thing I liked were the pull-out quick guides on the bottom of most of my Sencore equipment.

[jh15]

Couldn't find the pop-up chamber manual. It must have a p/n, but I went down a rabbit hole of interesting things like info on my scope cart, etc.

[khutch004]

A uTracer kit plus an old tube tester:

(picture will follow when I figure out how to do that!)

Thanks for posting.  That caused me to look up the uTracer.  The comments from the designer are pretty neat as he walks through his thinking on evolving the design to higher voltage.

https://www.dos4ever.com/uTracerlog6/tubetester6.html

Yeah, he is one of us and he and his wife are a great team with occasional packing and shipping help from the kids, I think. The kit is very nicely done with a manual that does Heathkit one better. It was a great experience to work with them and build the kit. I don't need the new version, I don't know that I will be able to resist it!

[Electro Fan]

A uTracer kit plus an old tube tester:

(picture will follow when I figure out how to do that!)

Thanks for posting.  That caused me to look up the uTracer.  The comments from the designer are pretty neat as he walks through his thinking on evolving the design to higher voltage.

https://www.dos4ever.com/uTracerlog6/tubetester6.html

Yeah, he is one of us and he and his wife are a great team with occasional packing and shipping help from the kids, I think. The kit is very nicely done with a manual that does Heathkit one better. It was a great experience to work with them and build the kit.

Cool! 

I don't need the new version, I don't know that I will be able to resist it!

If you go for it, let us know.

[THDplusN_bad]

Good Day,

here are some details about the latest addition to my labs, an Agilent N3280A "component test DC source". Which I have turned into a curve tracer with some Python magic 

• There is surprisingly little information on this unit available on the internet. The official data sheet specs, user guide and an App Note on "Optimizing Power Product Usage to Speed Design Validation Testing" are available on the Keysight website https://www.keysight.com/us/en/product/N3280A/component-test-dc-source-5-w-4-outputs.html
  
• The unit basically provides four isolated, independent voltage/current sources, up to +/-10.25V and +/-512mA each. It sports four-wire outputs with current and voltage sensing and programming of both with 0.1% basic accuracy.
• It features four quadrant bipolar outputs, and it has exactly *one* front panel knob. Which is the POWER ON/OFF switch. 
  Which means that you have to program the device via GPIB to use it.
• The output connectors are a bit unusual, these fit 6 pole connectors, Würth Elektronik type 3673 male for custom cabling.
• I suspect these are the main reasons why this unit was offered at a relatively low price by a US-based surplus dealer. I have paid about as much for shipment and import duties as for the item itself...
• I have inspected the unit upon arrival, changed the line voltage to 230VAC and some basic cleaning of the housing was all it needed. I have replaced the noisy Pabst type 612 fan with a Noctua NF-A6x25 FLX 60 mm fan, which has slightly less air flow. This has reduced the noise emissions from a "screamin', I do not want this"-noise to a still noticeable, but completely acceptable level.
• The attached photos provide a few details, i.e. a XILINX Spartan XC2S50 as the digital brain, AD677 16-Bit 100 kSPS ADC from Analog Devices. The output power amplifiers are from APEX, type PA46.
  Most datecodes are from 1999. Note the massive toroid transformer.
• A closer inspection revealed that the input section of one power supply has been repaired in the past, apparently with original-spec'd components; but the solder joints do not like this was completed by Agilent/Keysight. 
  Anyway, all power supply units were found to be well within specs.
• I have also replaced the Schaffer line filter, which is another story on its own.
• The mounting screws for the GPIB-connector always need some moderate re-tightening. It pays back to use some quality 9/32 inch-/imperial tools for this.
• It is nice to see that the old HP plastic feet still fit this.

I have started to familiarize with the user guide, then to program the unit using Python and my NI USB GPIB-adapter about four weeks ago. I am quite pleased that the program creates some pretty nice plots such as the one included below.

That is more than 2,100 readings, from testing a standard BC547 NPN transistor. 

Cheers,

THDplusN_bad

[THDplusN_bad]

Added screenshot in PNG format

[Ground_Loop]

Good Day,

@Ground loop, I totally agree that the Heathkit IT-1121/3121 is a joy to work with. Both types are virtual identical, but have different colours and front panel knobs.

I have repaired my unit back in 2014, and modified it with BNC-output connectors for the outputs to the oscilloscope (to replace the 4mm/banana connectors), added a fuse holder for base step currents with a "fuse fail" indicator lamp and a custom scale.
I have also made a simple adapter with a ZIF-socket for the A/B testing of components, which has proven its value. Matching transistors is a joy with this. 
Photos attached.

I also enjoy using a Tektronix 5CT1N (basically the Tektronix 7CT1N's little brother for the Tektronix 51xx/54xx oscilloscopes). And the latest addition to my labs is an Agilent N3280A DC component tester. More of the latter soon...

Cheers,

THDplusN_bad

Nice mods. Got me thinking now.

[THDplusN_bad]

Curve tracing - or at least looking at those lovely curves --- it surely can turn into an obsession.

I have been using Python for programming an (inexpensive) USB-6008 DAQ box from National Instruments.

Guess what my hands grabbed for the creation of two input signals for testing two inputs...
An "Octopus" tester (a crude form of a component tester, well known, but still useful). That is not a curve tracer, but I just had to see a diode's I/V curve like in the attached screenshot... 
I cannot help it! 

Pls. ignore the voltage scaling; as it's irrelevant in this case.

Cheers,

THDplusN_bad

[Johnny10]

B&K Model 501A

[Electro Fan]

B&K Model 501A

Thanks for posting the B&K 501A.  I looked it up and found this, looks like a good product and educational demo.

[jh15]

Makes me want to sell my 575, and get one of those. Probably not anywhere near the capabilities of the 575.

And the tubes warm the basement.

[Markus2801A]

Deas anyone have the schematics for the ELV Curve Tracer KS7000?

Thanks in advance :-)

[Messtechniker]

My semi-finished  semiconductor curve tracer. Completed and verified D/A converter and USB link. Output stages need to be built one day.
More on this here:
https://www.dl4jal.de/kls/kls.html

(Attachment Link)

Completely scrapped the pcb. for space reasons and rebuilt this:







More here:
https://www.wellenkino.de/557/viewtopic.php?f=7&p=3205&sid=141deca1ed14b466610df61b065a9b98#p3205

and here of course:

https://www.dl4jal.de/kls/kls.html

For something more advanced see here:

http://diy.ucborgmann.de/index.php/de/audio-messtechnik/kennlinienschreiber

[0culus]

I have three curve tracers...I have one Type 575 Mod 122C that is in perfect working order, another nearly identical one that isn't in good working order (repair queue, eventually), and a 577M1 with working storage. Pictured here is my experiment at rejuvenating a set of tired Burroughs Nixie tubes that came out of an -hp- 3440A voltmeter (also in repair queue, mainly for the plug in now). It actually worked very well, and the curve tracer provides excellent visual feedback of when the rejuvenation is getting there...the negative resistance region goes from being absolutely crazy to rock stable (and of course the poisoned bit of cathode is now lighting properly). I need to designed a test jig with proper sockets for Nixies to make the process less manual.

Also, I *highly* recommend anyone who owns a curve tracer find a copy of the Tektronix Measurement Concepts book on semiconductor devices.

[rodpp]

I have two curve tracers, a 567 and a 370A.

Both are great instruments, responsive and with a great range of voltage and current.

The 370A have some nice additions compared to the 576:

- Cursors to measure parameters of the curve.
- GPIB interface that allows operate all functions of the instrument remotely, and acquisition of data.
- Memory to save/restore different configurations of the instrument. You can set an experiment configuring all parameters of the test, then save it in a memory position. Configure another one test setup, and save in another memory position, and so on. After that, you can restore as needed each one from the respective memory position.
- Save/restore curves. It's handy to compare/match components. You trace one (various) component(s) curve(s) and save it. After that you can test another component and show, at the same time on screen, actual and saved curves to compare/match.

[edavid]

I have a Tek 371, complete with bubble memory cartridge, but I've never been able to find a test fixture for it.  Anyone have a spare?

(the second photo shows me defeating the fixture interlock, the hard way)

[Johnny10]

Oh, I want one of those !!!!!

[Cubdriver]

Oh, I want one of those !!!!!

Join the club!  The 576 is sweet, but that thing - wow!


-Pat

[rodpp]

I have a Tek 371, complete with bubble memory cartridge, but I've never been able to find a test fixture for it.  Anyone have a spare?

(the second photo shows me defeating the fixture interlock, the hard way)

Wow! What a nice equipment.

The 371 is a beast, it can supply 3000V or 400A (pulsed) to the DUT, up to 3000W!

It does a pair with the 370. The 370 is the "high-resolution" curve tracer with a resolution of 1pA and 50uV (but can only supply 2000V or 20A), while the 371 is the "high-power" brother.

The test fixture for the 371 is not easy to find, but sometimes some shows up on ebay.

[edavid]

Oh, I want one of those !!!!!

Well, feel free to PM me 

[Wolfgang]

Hi,

I do it using SMUs. My SMUs are Keysight B2902A and B2962A. Why do I do this:
- I need precision for the base voltage (e.g., for Gummel plots)
- The measurements must be pulseable to mitigate device heating at higher powers
- I also need wide range current measurements (0.x uA to several 100mA)
- The data is imported to a PC and all the curves are made there using Python.
  (I have output curves, like a normal curve tracer, Gummel plots, beta curves, ...)
- The imported data is used to create SPICE models.

An analog curve tracer is nice to make quick measurements or go/no go tests. I have not built, but I like a nice
project from Changpuak (complex curve tracer) or the Curve Tracer Arduino shield.

[Shock]

I do it using SMUs... (proceeds to explain whole procedure)

Looks at Hameg button, looks back at post, reaches for Hameg button.

[david77]

Deas anyone have the schematics for the ELV Curve Tracer KS7000?

Thanks in advance :-)

You can still get it from ELV

https://de.elv.com/elv-serie-7000-kennlinienschreiber-ks-7000-teil-12-204310

https://de.elv.com/elv-serie-7000-kennlinienschreiber-ks-7000-teil-22-204319

I've built one this year, its quite capable. Although I do still have some issues that I need to resolve.

[kripton2035]

elv don't sell anything outside of germany/austria/swiss.
I bought from them years ago, now it's impossible.
they have now a french store elv.fr, but there are quite nothing in this store, only their domotic line.
it was a great shop, it's even no more european !

[Jacon]

You can still get it from ELV
----

Iron Curtain apparently still exists...
Unable to buy them from Poland

[0culus]

Man, I don't think I've seen a 370 or 371 in the wild! Neat! While we are on the subject of "modern" curve tracers, I give you my -hp- 4145B system I have put together. We are showing Ic-Vbe and Ib-Vbe characteristic curves for a 2N2222A, on a log-log plot. This makes it easy to compute AC beta for any values of Ic using cursor math (shown at the bottom of the graphics display). The flat lines at the end are the SMUs going into compliance at 100 mA. The real strength of the 4145B (and it's later and far more expensive brethren, the 4155 and 4156) is very low current measurements. In fact, most of these were used in wafer probe stations and not as bench instruments, even though you totally can with the 16058A test fixture.

[0culus]

I have a Tek 371, complete with bubble memory cartridge, but I've never been able to find a test fixture for it.  Anyone have a spare?

(the second photo shows me defeating the fixture interlock, the hard way)

I seem to recall seeing a complete 371 for sale on the bay of fleas once, for a suitably obscene amount of money. 

[TERRA Operative]

Nice, I just picked up a 4145B a couple days ago for about a hundred and fifty bucks, will look nice sitting next to my 4145A. Couldn't pass it up....
I just need a few test pins for the test fixture, and 4 triax cables and I have a full 16058A test fixture.

A 371 just sold here in Japan for around a thousand bucks, on the lower end of the going rate for them these days, depending on exact model, they sell for one to two grand, approaching three if it comes with all the accessories etc.

[Johnny10]

That was a really good buy!

HP 4145B

[0culus]

Nice, I just picked up a 4145B a couple days ago for about a hundred and fifty bucks, will look nice sitting next to my 4145A. Couldn't pass it up....
I just need a few test pins for the test fixture, and 4 triax cables and I have a full 16058A test fixture.

A 371 just sold here in Japan for around a thousand bucks, on the lower end of the going rate for them these days, depending on exact model, they sell for one to two grand, approaching three if it comes with all the accessories etc.

Very good price. Personally, I really think the 4145B complements an analog curve tracer nicely in terms of capabilities. It does the low current stuff the Type 575 just doesn't do (no Kelvin sense connections at all) and the 577 honestly isn't that great at (though I have not yet found the Kelvin sensing transistor adapters for a price I'm willing to pay, so there is that).

[TERRA Operative]

I have a 577 with a handful of adapters waiting to get its turn on the healing bench (I know it has some of the usual shorted tants to begin with), but I've not used it yet so I can't comment on it.
I wouldn't mind a 576 but they are either stupid expensive here, or pop up at a more reasonable price only after I just made another big purchase already.

[0culus]

The 576 is quite complicated too...the 577 has the huge advantage of sharing the upper chassis with the 5000 series scopes, so working on that stuff is as easy as finding a suitable 5000 series to cannibalize if needed. I know of a fellow who repaired a 577D1 with a bad storage tube by getting one from a dead 5000 series scope with a storage tube.

[TERRA Operative]

Oh, that's good to know if I need parts.

Next the trick is to find a reasonably priced broken 5000 series scope, I have seen one being sold so far, but because they are 'rare' here, the price goes up beyond what they are really worth..

[AaronLee]

I was reading about this guy's (named Paul) project:
http://www.paulvdiyblogs.net/2017/
And his latest version 3 of it:
http://www.paulvdiyblogs.net/2021/03/building-curve-tracer-version-3.html

Looks interesting to me, but I don't have a curve tracer yet. I'm quite interested in acquiring a curve tracer, but not in a big hurry, so maybe I'll wait until he's finished version 3. Anyone have any opinions on his approach?

[Electro Fan]

I was reading about this guy's (named Paul) project:
http://www.paulvdiyblogs.net/2017/
And his latest version 3 of it:
http://www.paulvdiyblogs.net/2021/03/building-curve-tracer-version-3.html

Looks interesting to me, but I don't have a curve tracer yet. I'm quite interested in acquiring a curve tracer, but not in a big hurry, so maybe I'll wait until he's finished version 3. Anyone have any opinions on his approach?

Prior to this thread while looking at curve tracers from the high end to the entry level, and especially at the ELV and successor designs, I came across Paul's work which has evolved into the VBA curve tracer.  Fwiw, the work done by Paul and his colleagues was something that helped stimulate this thread.  It struck me at the time that Paul's project was a potentially very good or maybe excellent approach to designing a curve tracer that could address a lot of needs and interests, possibly at the professional/commercial level and definitely at the academic/teaching level and at the enthusiast level.  IMO, Paul (V) and his colleagues (B and A) have been on a very thoughtful, open, tenacious, and admirable path and I'd bet that they will continue until they are confident they are "there."  And my guess is they might keeping going with further enhancements and ideas beyond "there."  I'd love to see V, B, and A participate here and have the opportunity for forum members to become a sounding board, adopters, and potentially contributors to the advancement of the VBA curve tracer.

[Shock]

Get them to post about it in the projects forum, it's there to discuss projects.

I'll also add it's not so much a resurgence in curve tracers but electronics more accessible and collaborative, especially with video streaming. Dave himself mentioned about lurking in the electronics newsgroup back in the day, it was fairly primitive.

Online auction/classified sites also give us access to cheaper gear and at some point past the multimeter, scope, psu stage people may want to characterize components. So it's not like curve tracers are the next biggest thing, they have always been around, we just don't have wait months to discover a cheap design in a magazine anymore. Now it's go on youtube or ebay and a few bucks later you can get a basic IV curve. Or many curves if you throw more money at it.

[Dwaine]

Here is my setup....

[antenna]

I went as cheap as I could. An old DC adapter cut open to convert to AC (5.82vRMS), a 400ohm resistor, a BNC jumper cut in half and the probe from my LCR meter. Picoscope 3206D XY mode with custom probes programmed to get the scales correct.

[Messtechniker]

Poor man's curve tracer on the Hameg HMO 1022 scope:



Same green LED connected to DL4JAL curve tracer:

[david77]

Here's my copy of the ELV KS7000. It works quite well and has turned out to be a useful tool for the workshop when fixing stuff.

Hooking up devices with clip leads was not only very messy and annoying, it also resulted in nasty oscillations. So I decided to go the
Tektronix route and made up a testfixture. So far I've got this one that works for small & medium power transistors. I've got parts on order to make another one for TO3 packages. This works very well so far.

Certainly no Tek but ok for me.

[Neomys Sapiens]

I posted some of my initial experiences with the Peak Electronics DCA75
here:
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3530794/#msg3530794
,here:
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3671740/?topicseen#msg3671740
and here:
https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3503112/?topicseen#msg3503112

Besides this, I have the model 3110 from Vu-Data corp., which seems to be a (licensed?) copy of an earlier Huntron model, maybe tre Tracker 1000.

[AaronLee]

Here's my copy of the ELV KS7000. It works quite well and has turned out to be a useful tool for the workshop when fixing stuff.

Hooking up devices with clip leads was not only very messy and annoying, it also resulted in nasty oscillations. So I decided to go the
Tektronix route and made up a testfixture. So far I've got this one that works for small & medium power transistors. I've got parts on order to make another one for TO3 packages. This works very well so far.

Certainly no Tek but ok for me.

Wow, impressive!

[Messtechniker]

Hooking up devices with clip leads was not only very messy and annoying, it also resulted in nasty oscillations.

Had the same problem with my DL4JAL curve tracer.
Fitting a small ferrite bead in the collector line fixed that.

[david77]

Had the same problem with my DL4JAL curve tracer.
Fitting a small ferrite bead in the collector line fixed that.

Yes, I did the same inside my fixture. Killed all the nasties instantly.

[Electro Fan]

As discussed earlier in this thread, this project is showing a lot of excellent potential:

https://www.eevblog.com/forum/testgear/diy-curve-tracer-(analog)/

[KungFuJosh]

I had a terrible experience with a popular tube curve tracer. As a result of that experience, I designed better hardware, and hired somebody to create new firmware. The result is what I'm calling the FUtracer (for obvious reasons if you're familiar with the one that pissed me off).



It's setup to test (from left to right):
12AX7 etc...
EF800 etc...
EF86
EL84
EL34/6L6/6V6 etc...

It can be setup for whatever tube types/sockets you want, but those are the ones I cared about, so that's what's on the tube pcb part.

Here's the Windows GUI:


Here's my ridiculous / overkill test I did. 301 points from Va 2 to 400, with 31 sweeps of Vg from 0 to -30:

[Johnny10]

All the popular guitar amp tubes.
Cool! 

Can you switch between triodes?

[Cubdriver]

That thing looks sweet!!  Very nicely done!

-Pat

[KungFuJosh]

All the popular guitar amp tubes.
Cool! 

Can you switch between triodes?

There's no need to switch, it tests both triodes in dual triodes at the same time. Though technically you can switch between them on the plot tab in the GUI. You can show both Ia1 and Ia2, or show either. I'll take a new test and post it later, but this is what the plot screen looks like:

That thing looks sweet!!  Very nicely done!

-Pat

Thanks! I'm probably going to sell the PCBs along with the programmed PIC chip if anybody is interested. Though there are a few SMD items on there, so it's not a project for everybody. If there is enough interest, I might find an SMDA service that carries the parts I use, but I don't want to do it as a service myself. I do it well enough, but not well enough to charge for it. 🤣

Overall, it's not that expensive to build. The BOM on Mouser is about $180 or so with all premium parts, and I was thinking about selling the boards (main PCB, tube socket PCB, and a external PSU connection PCB) and programmed firmware chip for like $150 shipped in the US.

Thanks,
Josh

[Johnny10]

Is this full voltage or like uTracer short pulse. Is the Gm from database or calculated?

[KungFuJosh]

Is this full voltage or like uTracer short pulse. Is the Gm from database or calculated?

It's pulsed, most tube testers are pulsed (there are a couple that aren't that have their own limitations). Gm is calculated. The GUI is a special newer version of utMax (which is open source), and it has modeling (modeling currently needs to be fixed though) for better Gm readings based on operating point.

[Cubdriver]

I would definitely be interested in building one of them!

-Pat

[KungFuJosh]

I would definitely be interested in building one of them!

-Pat

Awesome, I'm glad to know there's more interest out there. My tube amp forum obviously has interest too. I'll need to buy more PIC chips before I can sell more than...2 lol. But also, unless you have a stockpile of convenient parts, a lot of the parts on the BOM are backordered until next year. Yay for covid supply chains. 🙄

I'll eventually build a website with build info and share the link here.

How are you with SMD stuff? There's a few tiny things on there. 😉

Thanks,
Josh

[Cubdriver]

Awesome, I'm glad to know there's more interest out there. My tube amp forum obviously has interest too. I'll need to buy more PIC chips before I can sell more than...2 lol. But also, unless you have a stockpile of convenient parts, a lot of the parts on the BOM are backordered until next year. Yay for covid supply chains. 🙄

I'll eventually build a website with build info and share the link here.

How are you with SMD stuff? There's a few tiny things on there. 😉

Thanks,
Josh

Haven't played too much with SMD stuff thus far, but do have a preheater, a hot air tool, and fine tips for my regular iron.  And yeah, the supply chain is a total clusterfuck.  Been dealing with it trying to get parts for work - it's crazy.  Stock expected six or more months down the road.  On things like Molex Connector housings.  WTF?   

-Pat

[KungFuJosh]

Haven't played too much with SMD stuff thus far, but do have a preheater, a hot air tool, and fine tips for my regular iron.  And yeah, the supply chain is a total clusterfuck.  Been dealing with it trying to get parts for work - it's crazy.  Stock expected six or more months down the road.  On things like Molex Connector housings.  WTF?   

-Pat

I recommend trying out some of the SMD practice kits on Amazon. I built a few before I attempted this build. I have a few more to build for practice/fun.

On my tracer PCB:
Here's one difficult one (only one of them obviously): https://www.mouser.com/ProductDetail/895-FT232RL
And there's two of these: https://www.mouser.com/ProductDetail/595-PGA113AIDGST

I use solder paste for both of those, but for some reason the PGA113s are trickier than the FT232RL for me. The pin pitch seems almost the same. Maybe it's cause there's two. lol There's a few other easier items, only basic SMD skill is really needed for.

I've had some luck going back and forth between Mouser, Arrow, and Digikey or Newark for supply stuff. But it's obviously gotten much worse.

Thanks,
Josh

[Johnny10]

At least your not trying to remove SMD.
I find it very trying. Its a different world now. Parts that can fly!
I keep looking at the JCB NANO kit but big $$$$$$

[KungFuJosh]

At least your not trying to remove SMD.
I find it very trying. Its a different world now. Parts that can fly!
I keep looking at the JCB NANO kit but big $$$$$$

Removing SMD is much easier. I have one of those "fast chip" magic goo kits. SRA sent me this free with some other stuff once upon a time: https://www.amazon.com/Fast-Chip-Removal-Temperature-Alloy/dp/B015RV4NBS

It takes seconds to remove annoying SMD stuff with that. I really think Magic Goo would be a better product name. I suppose it would be harder if the pins aren't visible though.

[rsjsouza]

Haven't played too much with SMD stuff thus far, but do have a preheater, a hot air tool, and fine tips for my regular iron.  And yeah, the supply chain is a total clusterfuck.  Been dealing with it trying to get parts for work - it's crazy.  Stock expected six or more months down the road.  On things like Molex Connector housings.  WTF?   

-Pat

I recommend trying out some of the SMD practice kits on Amazon. I built a few before I attempted this build. I have a few more to build for practice/fun.

On my tracer PCB:
Here's one difficult one (only one of them obviously): https://www.mouser.com/ProductDetail/895-FT232RL
And there's two of these: https://www.mouser.com/ProductDetail/595-PGA113AIDGST

I use solder paste for both of those, but for some reason the PGA113s are trickier than the FT232RL for me. The pin pitch seems almost the same. Maybe it's cause there's two. lol There's a few other easier items, only basic SMD skill is really needed for.

I've had some luck going back and forth between Mouser, Arrow, and Digikey or Newark for supply stuff. But it's obviously gotten much worse.

Thanks,
Josh

Josh, if you are talking about supply troubles, the FT232RL can be sourced in many USB-to-serial converters (you may even use a serial connection and plug an external adapter). The PGA113AIDGST seems to be in stock in TI's page. 

If you are talking about soldering alone, then yeah, it is worth practicing on some older boards or kits. Also, you need to see things - get a powerful loupe with good lighting, a stable soldering iron with a thin and clean tip and good solder paste and solder.

BTW, your curve tracer looks quite interesting; congratulations on making it.

I might not have enough capex to commit at this moment, but I will keep this in mind.

[KungFuJosh]

Josh, if you are talking about supply troubles, the FT232RL can be sourced in many USB-to-serial converters (you may even use a serial connection and plug an external adapter). The PGA113AIDGST seems to be in stock in TI's page. 

If you are talking about soldering alone, then yeah, it is worth practicing on some older boards or kits. Also, you need to see things - get a powerful loupe with good lighting, a stable soldering iron with a thin and clean tip and good solder paste and solder.

BTW, your curve tracer looks quite interesting; congratulations on making it.

I might not have enough capex to commit at this moment, but I will keep this in mind.

Thank you! It was quite the journey getting it to where it is.

I have stock of everything. 😉 But I'm not selling the parts; when I do, it will just be the boards and the programmed chip. SPA07N60C3, LM7805CT/NOPB, DR127-331-R are all hard or impossible to get right now, and there's a bunch of other stuff out of the ~92 items on the BOM for the main PCB that have supply issues.

For SMD work, I use my Andonstar AD208 digital microscope. It does a pretty darn good job. No lag issues if you don't max the resolution.

My soldering station is an Aoyue 2703A+ (I've been abusing this one for over 4 years- the last 3 stations I had from other brands didn't last that long). The hot air works great for the SMD stuff. I use Kester EP256 solder paste- it works well but sometimes is a pain in the ass to stick at first.

Thanks,
Josh

[KungFuJosh]

All the popular guitar amp tubes.
Cool! 

Can you switch between triodes?

I ran a trace on a new production Tung Sol 12AX7. Here's the plot:



The GUI window is resizable, so you can make the plots full screen. The other 3 screenshots attached below are my monitor's resolution. They have both or each individual triode.

Thanks,
Josh

[Cubdriver]

Yeah, I definitely want to build one of these!

-Pat

[Johnny10]

Very Nice!

[KungFuJosh]

For anybody interested, I'm going to order some more PIC chips to program, and then I'll get 5 to 10 kits ready. Initially they won't have anything on the boards, so I suggest anybody interested has or develops some SMT skills before proceeding with this set. For the future, I'm getting quotes on PCBA service to cover the 5 smaller SMT parts (there's also a couple larger ones setup for easy hand soldering that don't need any special skills).

Thanks,
Josh

[Shock]

Please start a thread in the projects forum.

[Cubdriver]

For anybody interested, I'm going to order some more PIC chips to program, and then I'll get 5 to 10 kits ready. Initially they won't have anything on the boards, so I suggest anybody interested has or develops some SMT skills before proceeding with this set. For the future, I'm getting quotes on PCBA service to cover the 5 smaller SMT parts (there's also a couple larger ones setup for easy hand soldering that don't need any special skills).

Thanks,
Josh

I'm definitely interested.

-Pat

[KungFuJosh]

Please start a thread in the projects forum.

Done. Here's the new thread: https://www.eevblog.com/forum/projects/project-vacuum-tube-curve-tracer-(the-futracer)/new/#new

Thanks,
Josh

[lugaw]

SMU gang represent!

(glances to the left, then to the right -- coast is clear, SPA gang isn't around yet)

Custom 3D printed feet and "panel" (bar) turn this SMU into a convenient bench tool. Actual work is done in jupyterlab / holoviews / pyvisa-py.

Please share the STL of your custom feet, I need it on my picoammeter. Thanks!

[Roehrenonkel]

Hi,
 
At https://jacmusic.com/Tube-testers/index.html you can find under No.10 most of the digital/PC-controlled curve-tracers for tubes.
Have tried some but nothing is perfect. What i wanted is more accuracy and flexibility.

More than the kit that calls itself "professional tube testing system" in germany.
Not even grid-current measurement possible there. Nice software but Windows-only.

So i put some VXI-modules (16 Bit DAC +-40V, seven voltmeters and a PSU) together
with a selfmade relay-matrix 4-In 16-Out and some GPIB-Powersupplys.
For now controlled by Agilent-HP-VEE, would like to do that from my Raspberry.

Have also checked transistors and LEDs with it,
just depends on the controll-software and power-sources.

[jjoonathan]

SMU gang represent!

(glances to the left, then to the right -- coast is clear, SPA gang isn't around yet)

Custom 3D printed feet and "panel" (bar) turn this SMU into a convenient bench tool. Actual work is done in jupyterlab / holoviews / pyvisa-py.

Please share the STL of your custom feet, I need it on my picoammeter. Thanks!

Whoops I must have missed this last time. Here they are!

[THDplusN_bad]

Good Day,

this is not about a self-made curve tracer, but those interested in PC-based measurement equipment might be entertained by this.

I use a "Electronic Explorer" device from Digilent/US for a few years now. This is basically a simple USB-connected, FPGA-based multi-instrument that integrates an 4-channel oscilloscope, 2-channel arbitrary waveform generator, simple logic analyzer and a power supply in combination with a solderless breadboard. I primarily use it for the quick evaluation of simple circuits whenever needed. It also supports Bode plots etc. The "Electronic Explorer" hardware platform is outdated by now, but it still supported by Digilent's "Waveforms" software. If any Digilent employee reads this: Thank you for doing that.

Digilent (which is a NI company, by the way) has released a "transistor test adapter" some time ago. That is an add-on for their more recent USB-based instruments such as the "Analog Discovery 2" device. https://digilent.com/reference/test-and-measurement/transistor-tester-adapter/reference-manual
And so they have also added the required settings in form of a virtual curve tracer to the mentioned "Waveforms" software. I have created my own version of a curve tracer in Waveforms some time ago (this is primarily a configuration of the two-channel waveform generator, and not too hard to do). However, I have tried this new "tracer" configuration and attached are a few plots I have generated from a standard BC337-40 BJT today. (Side note: The noise level results from the wiring and the two curves on the right are artefacts).

It's quite nice to see how the curves are generated, a large data set gets acquired very quickly and how the data can be exported to a database, MS Excel etc.

Also, the software offers a 3D-view of the collected data that is quite fascinating. https://digilent.com/reference/test-and-measurement/guides/waveforms-curve-tracer

This is not a real curve tracer in terms that the voltage and current supplies are limited (+/-10V, 35 Ohms output impedance each channel for the "Electronic Explorer"). But it's an excellent starting point into the fascinating world of component testing. It's also certainly good enough to match some small signal transistors /BJTs or FETs/ as well as diodes.

Have a great weekend, all.

Cheers,

THDplusN_bad

[6SN7WGTB]

I built the Paul Carlson curve tracer, but modified to be (a) SMD and (b) have a power op-amp to provide a split rail supply from a single external PSU.

I use it for go/no-go testing in circuits.

Shown below, inside and out, and with a germanium diode under test.

[Martin72]

Nice work ! 

[Electro Fan]

Thanks to rstofer's posting in this thread:
https://www.eevblog.com/forum/beginners/oscilloscope-triggering-from-function-generator/msg4058536/#msg4058536

I came across this link to a thread on a curve tracer project:
https://circuitcellar.com/research-design-hub/create-your-own-i-v-curve-tracer/
- another nice inexpensive combination of hardware and software

[deemarkay]

Here's my copy of the ELV KS7000. It works quite well and has turned out to be a useful tool for the workshop when fixing stuff.

Hooking up devices with clip leads was not only very messy and annoying, it also resulted in nasty oscillations. So I decided to go the
Tektronix route and made up a testfixture. So far I've got this one that works for small & medium power transistors. I've got parts on order to make another one for TO3 packages. This works very well so far.

Certainly no Tek but ok for me.

Looks really nice! The testsocket adapter, frontpanel and all is very shmick!
I'm currently in the process of collecting all the parts for a build of my own and having trouble sourcing some components. Especially the relais are a mystery and it turns out even replacements for the transformer  are no easy feat to aquire. Could you share how you solved parts sourcing problems on this build?

[david77]

Yes sure I'll help.

I use two transformers. One is 2x15V 0,5A for the OP's and CMOS stuff and a big toroid that used to give 2x44V output and I've added a few windings to get it up to 2x60V.

Ah yes, those Siemens Relais are long obsolete but not really necessary. Just use regular Relais, there's nothing special about them.

[basinstreetdesign]

Digital, smidgital...I've probably posted about this before but here is my home-built tube curve tracer.  I used it to answer a question about the biasing of some 6AU6's in a stereo amp I was making.

This is the first attempt at making it, which proved successful enough.
https://www.instructables.com/Tube-Curve-Tracer/

Then some added functionality came a bit later:
https://www.instructables.com/Tube-Curve-Tracer-Ver-11/

A word about my build style.  I don't like to spend a lot of money.  So parts are usually drawn from my glory boxes and whatever I have on hand.  If I can find enough workable stuff then it may happen that I spend exactly $0 on a project.  And so it was with this one.

Included is a shot of it in use and a shot of the HP 1741A showing the curves for one triode in a 12AX7 tube.

[deemarkay]

Yes sure I'll help.

Thanks! 😎

[RoGeorge]

This is from a Python script sending SCPI commands to a Rigol DP832 power supply, in order to browse/measure Vgs, Vds and Ids of a PHP45N03LT TrenchMOS.


https://www.eevblog.com/forum/testgear/rigol-dp832-power-supply-as-automated-curve-tracer/

Does Python qualifies as a curve tracer? 

[caulktel]

I bet not many have seen my curve tracer. https://www.nutsvolts.com/magazine/article/build-an-oscilloscope-octopus

Joel
N6ALT

[branadic]

I have the updated version KLS 2.0 of DL4JAL's curve tracer, that recently got the upgrade to PIC18F46K80. The board rev2.0 with the improvements is discribed here:

https://www.bartelsos.de/messtechnik/untersuchungen-an-bauteilen/kennlinienschreiber

Attached a measurement of a 2SK170.

-branadic-

[David Aurora]

Anybody else got one of these? I picked it up in a deceased estate sale a while back, it seemed to be mid-repair (half the screws missing, cord cut off etc). Haven't been able to find any record of it ever existing, let alone schematics.

I cleaned it up a bit at the time and then kind of forgot about it because it needed all new switches and I couldn't be assed fully disassembling it, but I finally changed them over tonight after work and the thing is now 99% working (the switches were so bad that you kind of had to hold them in place to get any display and even then it was a crapshoot). All that's left to do now is troubleshoot the step generator, it doesn't put out anything on voltage mode.

I've got a Tek 576 arriving soon so I don't imagine I'll bother going much further with this one, but it seems way too nice inside to waste so it'll likely get passed on to my assistant or something if I can work the last bugs out of it

[Electro Fan]

New video from W2AEW on curve tracing.  As usual from W2AEW his teaching is outstanding; his videos on a scale of 1 to 10 are typically about a 12 to 20 or better for clarity and conciseness.

[RoGeorge]

This yellow alligator showed up once, while identifying the type of photodetector from a former punched-card reader.  The curve tracer was improvised from a Rigol DG4102 DDS generator, a scrapped transformer and a Rigol DS1054z oscilloscope in XY mode:



The bar of 12 optical sensors from 1975 proved to be made out of NPN phototransistors.

[Electro Fan]

RoGeorge - very nice project!

https://www.eevblog.com/forum/testgear/rigol-dp832-power-supply-as-automated-curve-tracer/

[EinErik]

With this new tracer, compared to my old lockyZ, this one is definitely better with no doubt
https://www.diyaudio.com/community/threads/new-tracer-is-coming.379759/page-5

[RichardM]

I would love to make this version of the curve tracer. I don't suppose you are willing to share the PCB/Scematics ? I have the ELV articles.

Regards

Richard

[Electron123]

What's the difference between a "transistor curve tracer" and a more basic octopus version (like this https://www.qsl.net/kd7rem/pdf/octopus.pdf)?  I know the fancy versions have more features, but are they both able to test the same components? Or is the basic one not able to test transistors?

[Cubdriver]

What's the difference between a "transistor curve tracer" and a more basic octopus version (like this https://www.qsl.net/kd7rem/pdf/octopus.pdf)?  I know the fancy versions have more features, but are they both able to test the same components? Or is the basic one not able to test transistors?

A transistor curve tracer allows you to characterize the transistor or other device - see how it behaves under different drive conditions.  The octopus is more a go/no-go thing that tells you it works, but not much more.

In this image of a curve tracer, each slanted line represents a different base current; the sweeps are the collector current as Vce is increased and decreased at each base current step.  In this image, the vertical scale is 5 mA/division, and the horizontal is 5 V/division.  The base current steps are 0.01 mA each.  Note that it has two sockets and the big toggle switch, it is set up to permit two transistors to be connected and by flipping the switch back and forth you can see how closely they are matched.



-Pat

[Cubdriver]

To elaborate on how the curve tracer works, it applies a base current (or gate voltage if testing a FET), then sweeps the C-E (or S-D) voltage from minimum up to whatever max you've set.  It then increments the base/gate drive by the step interval and sweeps the C-E/S-D voltage back down to zero.  Increments the drive again, repeats the low-to-high sweep, increment again and high-to-low.  Lather, rinse, repeat until the desired number of steps have been generated.  The beam on the screen thus traces out the current through or voltage across the transistor under test.

The curves for a PNP device are similar, but start in the opposite corner of the display as the polarities are reversed from those of an NPN (top right for PNP, bottom left for NPN):


Curve tracers can also characterize things like diodes and the strike voltage of neon lamps (sometimes used as voltage limiters/regulators).  In this instance, the device in the curve tracer is a color-changing LED, and the photo was taken as it was lighting two of its emitters to get a blended color.  Each of the colored LEDs inside has a slightly different forward voltage, thus the dual curves with slightly different voltages and currents, and the jumping between the two curves is caused by the internal IC in the LED switching the two different LED dies on and off alternately:


-Pat

[david77]

I would love to make this version of the curve tracer. I don't suppose you are willing to share the PCB/Scematics ? I have the ELV articles.

Regards

Richard

It's not that I don't want to share my PCB's but they're in one file married to the schematics and I'm not willing to put that online, copyright reasons, you know?

Maybe I can unmarry the layout and the schematics... I'll see what I can do.

[G0HZU]

I recently upgraded my curve tracer jig so it can produce accurate IV curves for diodes. The jig is homemade but I use a programmable PSU and a couple of decent bench DMMs to do the measurements. It's all controlled with a VB app in Windows.

The output data is then loaded into Excel as below and this system can test the diode down to a few nanoamps and compare the result to a theoretical curve as in the image below.

[croma641]

Hi, G0HZU,

what is Your curve tracer setup for other semiconducors ?

thanks

I recently upgraded my curve tracer jig so it can produce accurate IV curves for diodes. The jig is homemade but I use a programmable PSU and a couple of decent bench DMMs to do the measurements. It's all controlled with a VB app in Windows.

The output data is then loaded into Excel as below and this system can test the diode down to a few nanoamps and compare the result to a theoretical curve as in the image below.

[jord4231]

Here's my curve tracer 
https://youtu.be/Jb6N5H1kS2Y

[armandine2]

I have the Elektor magazine linked below - so no excuse

https://worldradiohistory.com/UK/Elektor/80s/Elektor-1988-10.pdf

for not having a simple curve tracer?

[tautech]

Not mine but member mawyatt had some time for experimentations of what he could do within the limits of a AWG and DSO.
https://www.eevblog.com/forum/testgear/fooln-around-with-dso-awg/

Reply #2 has his hand drawn setup and layout as an attachment.

[Calambres]

I have the Elektor magazine linked below - so no excuse

https://worldradiohistory.com/UK/Elektor/80s/Elektor-1988-10.pdf

for not having a simple curve tracer?

One year later (1989-12) Elektor showed a new instance of that curve tracer that was later cloned in the (in)famous Thai Kit. See my take on that kit in this very topic:

https://www.eevblog.com/forum/testgear/show-us-your-curve-tracer/msg3503702/#msg3503702

[TopQuark]

I'll put in my entry to the topic with probably the cheapest curve tracer mentioned in this thread.

I was once a teaching assistant for a mechatronics course in university during the pandemic lock down. It was not possible for students to access the lab gear on campus, all everyone had was an Arduino, some R and C and a couple of diodes from a basic component kit.

I was put in charge of designing a lab project for basic electronics, and I designed a fun "curve tracer" / "SMU" with the Arduino board. I have attached the lab manual I wrote for the project and a couple of test result plots. Mind you the students I taught were majoring in Integrated Systems and Design (not EE) and they were first or second year students, so the language of the lab manual was purposefully oversimplified.

What I hoped the lab would do was to instill in student's mind, the importance (and fun) of doing test and measurement, even if we were on a shoestring budget and locked down. I don't think many students understood or appreciated the lab project, but I think a few students got something useful out of the experience.

[mbrennwa]

Just stumbled over this thread. Cool stuff!

Here's my curve trace project: https://pypsucurvetrace.readthedocs.io

PyPSUcurvetrace takes a somewhat different approach than most other curve-tracer projects that require some specific hardware / board to work. PyPSUcurvetrace is a software toolbox that controls programmable power supplies to set and read the voltages and currents at the DUT. This means you can make a curve tracer from PSUs that are available off the shelf, and you choose the PSU units that suit your application: small transistors, power transistors, vacuum tubes, etc.

The PSUs don't need to be expensive "high end". For example, two RIDEN 6006P units can be had for about $300 or less, and will make for a very decent curve tracer setup! 

PyPSUcurvetrace also allows controlling the temperature of the DUT during curve tracing. I attached an example of an IRFP150 power FET measured at 30°C, 50°C and 70°C.

PyPSUcurvetrace not only allows measurement of the curves, but also has post-processing tools for high-quality curve plotting and to determine the DUT parameters from the raw data files (e.g., bias voltage/current for a given operating point, gain parameter, output conductance). More tools like curve matching/pairing of parts are in the works...

[THDplusN_bad]

Good Day,

Nice work, mbrennwa!

I have paused my activity here for some time - life happens...

At least, I have found some time to adopt my simple Python scripts for my Agilent N3280A component tester so it can also test MOSFETs now...
Plot taken from a BS170 N-ch. MOSFET attached. 

Cheers,

THDplusN_bad

[jonpaul]

tektronix 576, std, pulsed 100A front porch, many Tektronix and DIY sockets and adapters.
jon

[Edison]

Hi, I started building my tracker, what I could buy, I bought and ordered the rest of the parts, but the delivery company lost my shipment, so I got angry and got this - unfortunately the tracker arrived with a faulty function, managed to fix it, but not much - documentation no, all ICs have laser defaced markings. A repaired small TEK serves as a display.
A few photos - something from the repair of the oscilloscope and the resulting effect

[Neepa]

That's kinda cute. Using that tiny mobile scope as a screen.

Below my Pride and Joy with it's new to me NOS rack handles. Test Fixture on the top left and still missing the connection cables for it. But those are under procurement.
Always a treat pushing the powerbutton and seeing the monster coming to life with its flashing LEDs and that jet turbine mascerading as a fan in the back.

[Edison]

A beautiful piece of technology
I love these classic testers, but the size won't fit in my study, I'm unlucky, I have to miniaturize, I bought the Tek damaged, the connectors oxidized, the timer switch glued with hot glue and most importantly no probes - I couldn't let it die like this, it got a new power supply ( batteries are not needed) I restored it as best I could and it has been working like this for two years.

[0culus]

That's kinda cute. Using that tiny mobile scope as a screen.

Below my Pride and Joy with it's new to me NOS rack handles. Test Fixture on the top left and still missing the connection cables for it. But those are under procurement.
Always a treat pushing the powerbutton and seeing the monster coming to life with its flashing LEDs and that jet turbine mascerading as a fan in the back.

Very good. A very sophisticated bit of kit, with a lot of uses. I use my 4145B quite a bit. Make sure you make plenty of back-up boot disks though, since the instruments depends on them to be useful.

[alm]

Both the 4145A and B can be fitted with a Gotek or HXC floppy emulator. That would reduce the boot disk problem to making sure you keep an usb stick or sd card around with the right file. Here are some links. Apparently the main challenge is getting the cable pinout right, but 4145A and B are quite similar in that regard.

https://hxc2001.com/showroom/HxC_Floppy_Emulator_Semiconductor_Parameter_Analyzers_HP4145A.html
https://torlus.com/floppy/forum/viewtopic.php?t=1595
https://torlus.com/floppy/forum/viewtopic.php?p=24043&sid=90a698af4dbda5c4568c26813d8da371#p24043
https://torlus.com/floppy/forum/viewtopic.php?t=3878

I'd think you could also get it to work with a Gotek drive and the open source FlashFloppy firmware (cheaper), but I only found a brief mention of that.

[RoGeorge]

Not mine, it's from the 'xofunkox-scientific experiments':

Characteristics varistor


 

[gjames50]

I would love to make this version of the curve tracer. I don't suppose you are willing to share the PCB/Scematics ? I have the ELV articles.

Regards

Richard

It's not that I don't want to share my PCB's but they're in one file married to the schematics and I'm not willing to put that online, copyright reasons, you know?

Maybe I can unmarry the layout and the schematics... I'll see what I can do.

I'm curious about the overall costs of your version of the ELV KS 7000. Paul's DIY has an updated version of it, but the build cost for his is just under 1,000.00 USD. I posted a reply to the forum post located here: https://www.eevblog.com/forum/projects/1986-transistor-curve-tracer-worth-while-building/

I converted the Gratis documents from ELV to English (although I need someone who speaks/reads German/English to help verify it).

Your take on this is very interesting. What are the copyright issues on yours? Since ELV is giving this information away for free now, it must be for some other reason, correct?

Thanks!

[watchmaker]

Here is my Doc Brown (Back to the Future) version of the Thaikit. Curve Tracer, octopus for zeners and TCRL tester on one unit.  Makes for easy sorting.