$20 LCR ESR Transistor checker project

[Rooster Cogburn]

That would be great! I'm busy with the m-firmware.

(please don't take this personal)

This general attitude is absolutely the norm among software & EE. Most people refuse to even spend 1min on documentation for every 100min hacking. GitHub is littered with projects that smart people spend combined lifetimes on that'll never be seen or be of use to anybody because they could not be bothered to spend 5min on adding a screenshot and a paragraph of text on 'how to build & install'. How may times have you seen some README on GitHub where you literally can't even figure out what this is supposed to do because just writing ONE SENTENCE on what this is was asking for to much.

If you Google for 'transistor tester m-firmware' I don't find your canonical repository. And how would I even know to search for that if I didn't dig through a giant forum thread telling me that's the current branch of code for this project (I guess it is?). Looking at the janky SVN browser (for most people it's anyway 'if its not on GitHub it might as well not exist') and trying to click on the PDF docs all I get is 'Display of files larger than 512 KB disallowed by configuration'. I guess I need to install SVN.

I just see this kind of stuff over, and over again. It's the engineer's disease. Brilliant people dedicate thousands of hours of their time to build truly impressive and useful things, only to stop two steps in front of the finishing line. All it would take is brief, simple page on GitHub or a community wiki with a few words on what is the state of this project, where to get the current docs/code/schematics, a paragraph or two for the 'Buyers FAQ', done. It's like you spend a year of your life building me this beautyful car, but when I ask you for the keys all I get is a shrug and 'read through this forum thread, somewhere in there is a link to a gopher server containing a Rot13 encoded text file with the coordinates of where I buried it in your backyard. Good luck, I need to build another car'.

[madires]

Have you seen the project's page at https://www.mikrocontroller.net/articles/AVR_Transistortester? And Karl-Heinz' documentation with over 130 pages? The m-firmware's README with about 1800 lines? We can't please everyone, but there's no lack of documentation.

[timelessbeing]

the only way to know what the differences between the hardware and software releases are is either...

The best way is to just dive in. The beauty of this tester is that the cost of entry is so low. You could even breadboard it yourself. Easy to modify and upgrade too. It's not a life changing decision.

How may times have you seen some README on GitHub where you literally can't even figure out what this is supposed to do

It's usually because they wrote the code for themselves, and you should be thankful that they decided to share it.

[Rooster Cogburn]

Have you seen the project's page at https://www.mikrocontroller.net/articles/AVR_Transistortester? And Karl-Heinz' documentation with over 130 pages? The m-firmware's README with about 1800 lines? We can't please everyone, but there's no lack of documentation.

Yep, I've seen those and have even explicitly mentioned all three as things I checked out. That wasn't really what I was looking for and still left me rather confused, probably the same for the other two people that immediately chimed in asking for advice as well. Way too in-depth, confusing and mostly helpful if you're already hacking on the project and have the general idea down. Maybe I'm being a bit stupid here, but it's always a good assumption to make about your users I can guarantee you that if I actually took the time to digest all of this I could write a few paragraphs of text that would answer 90% of the questions of 90% of the people approaching this for the first time and being lost. And so could you. From experience, this kind of feedback is about as likely to be accepted as explaining to a Linux proponent why Linux on the desktop is unlikely to take off ("I don't understand, it's really easy and works well for me, all you have to do is...").

How may times have you seen some README on GitHub where you literally can't even figure out what this is supposed to do

It's usually because they wrote the code for themselves, and you should be thankful that they decided to share it.

You don't understand where I'm coming from. I'm not saying this because I think I'm entitled to receiving documentation and a nice website from the author. They don't owe me anything. I don't get it from their point of view. Writing an introduction text, tutorial, making a friendly landing page and taking a few pictures should feel like a victory lap, not a chore. You get to show of your amazing creation to the world and get the pleasure of people actually using it. Plus it's like 15min of work and most likely the most effective 15min of the entire project in terms of real world results for your time. I'm not thankful for code thrown over the fence, it has no value to me.

[timelessbeing]

Yes but that takes time away from actual development.

They are not interested in your thanks either.

[Rooster Cogburn]

Yes but that takes time away from actual development.

Yes, it does.

They are not interested in your thanks either.

But you said I should be thankful?

[timelessbeing]

ok, happy, if you like.
And if it doesn't make you happy, then you can get a refund.

[Fuzzy Star]

I often find that long-running projects are quite impenetrable for outsiders who have not been following along from the beginning...
after >1h of Googling I'm not really getting anywhere...

Apparently the README should explain things. It does, if you already know the history of the project and who Karl-Heinz is and what he's been up to...

...it has 2012 in the title, so no idea if that still has any relevance to the current state of the code and whatever I can buy on AliExpress today.

There's some info on https://www.mikrocontroller.net, but even that doesn't really help in telling the different hardware platforms apart, getting some kind of idea what the differences between the software releases are and what to actually buy...

I think it would be super awesome if somebody in the know could just spend 5min writing a basic 'Buyers FAQ' with a last-updated date on it that just tells people what to buy...

This would be what you're looking for:
https://github.com/svn2github/transistortester/blob/master/Doku/trunk/pdftex/english/ttester.pdf

[bitseeker]

You don't understand where I'm coming from. I'm not saying this because I think I'm entitled to receiving documentation and a nice website from the author. They don't owe me anything. I don't get it from their point of view. Writing an introduction text, tutorial, making a friendly landing page and taking a few pictures should feel like a victory lap, not a chore. You get to show of your amazing creation to the world and get the pleasure of people actually using it. Plus it's like 15min of work and most likely the most effective 15min of the entire project in terms of real world results for your time. I'm not thankful for code thrown over the fence, it has no value to me.

I get where you're coming from and you're right that it would be beneficial to have user-friendly documentation and a buyer's guide that covered the many testers available (or maybe just the most popular ones).

However, writing good end-user documentation, just like writing good technical documentation or good code, isn't easy and takes more than a couple of pics and 15 minutes of work. These tasks require different skillsets and few people are good at them all.

Are you good at seeing technology from a user's perspective, anticipating how best to answer their questions before they ask them? Do you have the skill to reinterpret dense technical information into an easily digestible form? If so, then this project (and many upcoming users) can benefit from a quality user-friendly manual that you could make.

[texaspyro]

Another problem with testers is that they are like snowflakes... no two seem to be alike.  With clones and clones of clones everywhere (even though they look / are labeled identical), what you received yesterday may have no relation to what came in today... different pin assignments and displays are a particular problem.

[bitseeker]

Yeah, that's a whole other problem. Buying from China, you never know what you're going to get (and they often steal the firmware, mod it, and don't let you see what changed, let alone document it).

[madires]

The svn2github service was shut down a few months ago. So the repo mirror at GitHub is obsolete now and therefor we've decided to set up a "warehouse" at GitHub with the most interesting files, like firmware packages, PCB files and docs for example. This provides a convenient way for you to download the latest firmware since the SVN repo is considered cumbersome by most users. You'll find the warehouse at https://github.com/madires/Transistortester-Warehouse.

And some news about the m-firmware. The next version will support STE2007/HX1230 based displays (aka Nokia 1202 display). Possibly I'll also add a driver for the PCF8814 (aka Nokia 1100 display) which is quite similar to the STE2007.

[indman]

Hi, men!
I want to offer you the comparative table created by me on the most popular clones.
Look attentively whether I made a mistake or passed some moments?
Sorry, if my English translation is not absolutely good

[madires]

The LCR-TC1 requires a different firmware for the control MCU (U4) or to replace U4 by a simple circuit (TC1-Mod) to become compatible with k and m-firmware. 

[indman]

The LCR-TC1 requires a different firmware for the control MCU (U4) or to replace U4 by a simple circuit (TC1-Mod) to become compatible with k and m-firmware.

These are already technical details. Nothing prevents to update the main firmware of ATmega324(644) 

[sdancer75]

I have the same problem with my GM328. What steps did you follow to resolve the problem ?

[madires]

The 5kg hammer from the garden shed will fix that. Have you tried to reprogram the ATmega? Or what did go wrong?

[sdancer75]

The 5kg hammer from the garden shed will fix that. Have you tried to reprogram the ATmega? Or what did go wrong?

Yes, I tried the mega328_GM328 version I used from https://www.mikrocontroller.net/svnbrowser/transistortester/

I always get a white blank screen.  The fuses I used are: lfuse=0xF7 hfuse=0xD9 efuse=0xFF

I even tried the versions below without success.

GM328_EZM_v1.12k_r464_EN.zip
GM328_EZM_v1.12k_r508_EN.zip
GM328_EZM_v1.12k_r541.zip

 
I suspect that the firmware does not support this display (I don't know what kind of display it has), but I am not sure.

The seller description reports that "The display unit using a 160x128 pixel color display, full-screen characters 8x20". The GM328 was working when I received it, but it had a calibration problem so I decided to reprogram the MCU. Also I can for sure that it wasn't color.



Any recommendations?


Regards

[timelessbeing]

did you try changing the contrast setting?

[madires]

Yes, I tried the mega328_GM328 version I used from https://www.mikrocontroller.net/svnbrowser/transistortester/

It's an AY-AT clone. Please try mega328_color_kit. And don't forget to reprogram the EEPROM also.

The seller description reports that "The display unit using a 160x128 pixel color display, full-screen characters 8x20". The GM328 was working when I received it, but it had a calibration problem so I decided to reprogram the MCU. Also I can for sure that it wasn't color.

The AY-AT clones come with an ST7735 based color display usually. If your clone has a monochrome display you could try ST7565 for example.

[sdancer75]

Yes, I tried the mega328_GM328 version I used from https://www.mikrocontroller.net/svnbrowser/transistortester/

It's an AY-AT clone. Please try mega328_color_kit. And don't forget to reprogram the EEPROM also.

The seller description reports that "The display unit using a 160x128 pixel color display, full-screen characters 8x20". The GM328 was working when I received it, but it had a calibration problem so I decided to reprogram the MCU. Also I can for sure that it wasn't color.

The AY-AT clones come with an ST7735 based color display usually. If your clone has a monochrome display you could try ST7565 for example.

When you say don't forget to reprogram the EEPROM what do you mean? When I re-program the MCU with the hex file that's exactly what I am not doing?

The display does not have any PCB connection marks. It just says J1 with 8 pins. I tried to connect the display with my breadboard according to the following table

https://cdn.instructables.com/FD6/5BW7/IJ3763UJ/FD65BW7IJ3763UJ.LARGE.jpg?auto=webp&width=580

and linking with the Adafruit_ST7735.h and Adafruit_GFX.h libraries.

The final project looks like

https://cdn.instructables.com/FEF/2S4D/IIYWSHR9/FEF2S4DIIYWSHR9.LARGE.jpg

It lights up but I get the same result, blank screen.

Regards,

[madires]

When you reprogram the flash (hex file) you also have to reprogram the EEPROM (eep file). The EEPROM contains most text strings and other important data.

[sdancer75]

When you reprogram the flash (hex file) you also have to reprogram the EEPROM (eep file). The EEPROM contains most text strings and other important data.

Shit no I didn't this !!! I will try it, but I am curious why it did not work in my breadboard!


Anyway, the display has the following ICs. A) TI Hex Inverting Buffer and Converter (7CA676M) CD4050BM G4 and B) AMS1117 DROPOUT VOLTAGE REGULATOR and an SMD resistor. Nothing else.

 

[sdancer75]

Yes, I tried the mega328_GM328 version I used from https://www.mikrocontroller.net/svnbrowser/transistortester/

It's an AY-AT clone. Please try mega328_color_kit. And don't forget to reprogram the EEPROM also.

The seller description reports that "The display unit using a 160x128 pixel color display, full-screen characters 8x20". The GM328 was working when I received it, but it had a calibration problem so I decided to reprogram the MCU. Also I can for sure that it wasn't color.

The AY-AT clones come with an ST7735 based color display usually. If your clone has a monochrome display you could try ST7565 for example.

When you say don't forget to reprogram the EEPROM what do you mean? When I re-program the MCU with the hex file that's exactly what I am not doing?

The display does not have any PCB connection marks. It just says J1 with 8 pins. I tried to connect the display with my breadboard according to the following table

https://cdn.instructables.com/FD6/5BW7/IJ3763UJ/FD65BW7IJ3763UJ.LARGE.jpg?auto=webp&width=580

and linking with the Adafruit_ST7735.h and Adafruit_GFX.h libraries.

The final project looks like

https://cdn.instructables.com/FEF/2S4D/IIYWSHR9/FEF2S4DIIYWSHR9.LARGE.jpg

It lights up but I get the same result, blank screen.

Regards,

The mega328_color_kit firmware worked !!! Thank you madires you are the best !

Is this display ST7735 compatible? This pinout 1-8 is the same as the
following table (TFT display pin)?

[madires]

Most SPI driven displays got the same signal lines plus power and backlight. So you can't deduce the controller from the pinout.