$20 LCR ESR Transistor checker project

[madires]

1. edit Makefile
2. edit config.h
3. edit config_328.h for an ATmega328 or config_644.h for 324/644/1284
4. run "make"
5. run "make upload"

[eti]

1. edit Makefile
2. edit config.h
3. edit config_328.h for an ATmega328 or config_644.h for 324/644/1284
4. run "make"
5. run "make upload"

So just to clarify, there exists both "config.h" *and* "config_<mcu_part>.h", or is your "3. " step a mere extension of step "2. ", and there's only ONE config .h file?

Thanks

[madires]

config.h and config_<MCU>.h are two different files. In config_<MCU>.h you'll find all MCU specific settings like pin assignments (including the display), and config.h is for features and all other settings. BTW, this is explained in the README file (see "Building the firmware").

[eti]

config.h and config_<MCU>.h are two different files. In config_<MCU>.h you'll find all MCU specific settings like pin assignments (including the display), and config.h is for features and all other settings. BTW, this is explained in the README file (see "Building the firmware").

Ah  okay, I'll have a look, thanks again so much. )

[ansic]

The replacement LCR-T4 arrived today and it works!
From flashing the previous one I noticed neither the K or M firmware looked as nice as the from-factory programmed firmware.  Is there any way to configure the open source firmware to have the same appearance?
Did the manufacturer create an in-house firmware just for this clone?  The open source ones did not even have the 180 degree rotation feature for the display,  which must mean either there is a third firmware out there that I've missed, or it's manufacturer modified.

Also, are there any PCB faults known with the LCR-T4 that can be improved upon?

[madires]

The manufacturer's firmware is a modified k-firmware which has a nice user interface but less functionality. More kBs for fancy graphics means less flash memory for features, it's a trade-off. The m-firmware supports a flipped display output for ST7565, but it's a compile time setting. I think, the k-firmware supports that also. AFAIK, the T4 has no issues besides a missing level shifter for the 3.3V display.

[carrascoso]

Hi, Madires

I am classifying SMD components and have noticed a failure in your firmware.
Observed in several components, a problem that does not occur with the K firmware, nor with the Chinese firmware that incorporates the
GM328 or LCR-T4 meters.

An example in photos with a transistor C5707 for the test, in the datasheet says that it is a NPN Bipolar transistor.

The original Chinese GM328 and LCR-T4 meters, and the GM328 with the K firmware, the transistor put it in position
that you put it tells you that it is a bipolar NPN transistor, and it appears as the datasheet says, everything is fine here.

Now the same transistor with the GM328 and with the firmware M, in one position it appears as a NPN Bipolar transistor, and in the other position it appears as a JFET, leading to error by not knowing positively what the component really is.

Look at the photos and look at the position of the transistor that is only rotated in the meter socket.

 

The same applies to the double SOT23 three-pin encapsulated diodes, connected to the original Chinese GM328 and LCR-T4 or to the GM328 with the K firmware and appears as two diodes, either in series, opposed or as they are inside the component but always
It detects it as two diodes.



With the firmware of yours, you connect the three pins and it appears as a single diode or as a JFET transistor and it costs you
much measurement and in some cases it gives error.

I tell you because I do not know if it is my problem and if it happens with other components, and if not, if this problem could be solved and corrected.

Thank you very much.
Cheers

[hpmaxim]

Given that there seems to be an infinite number of variations on these units, is there a good guide as to which one to buy and/or where to buy it?

[madires]

The original Chinese GM328 and LCR-T4 meters, and the GM328 with the K firmware, the transistor put it in position
that you put it tells you that it is a bipolar NPN transistor, and it appears as the datasheet says, everything is fine here.

Now the same transistor with the GM328 and with the firmware M, in one position it appears as a NPN Bipolar transistor, and in the other position it appears as a JFET, leading to error by not knowing positively what the component really is.

That's strange. I haven't seen that before.

The same applies to the double SOT23 three-pin encapsulated diodes, connected to the original Chinese GM328 and LCR-T4 or to the GM328 with the K firmware and appears as two diodes, either in series, opposed or as they are inside the component but always
It detects it as two diodes.

With the firmware of yours, you connect the three pins and it appears as a single diode or as a JFET transistor and it costs you
much measurement and in some cases it gives error.

I tried two diodes in all possible ways and can't reproduce that issue. Do you have different GM328s for k-firmware and m-firmware? If yes, maybe there's something wrong with the one running the m-firmware.

[madires]

Given that there seems to be an infinite number of variations on these units, is there a good guide as to which one to buy and/or where to buy it?

Two pages back there's a link to an overview of different models. You can get the testers from all well known market places.

[carrascoso]

I will continue to select components, if it happens again I will put photos for you to see.

Yes, I have a GM328 with firm K, another with firm M and another with Chinese firm.
I tried several times and it gave me that problem.
Tomorrow I do tests again and I also tell you if it can be with photos.

[madires]

Please flash the GM328 running m-firmware with k-firmware and try the same components to check if that GM328 has a hardware problem.

[hpmaxim]

Given that there seems to be an infinite number of variations on these units, is there a good guide as to which one to buy and/or where to buy it?

Two pages back there's a link to an overview of different models. You can get the testers from all well known market places.

Thanks... other than the features is there a benefit to one versus another? 8 vs 16 MHz?  Accuracy? Etc?

[timelessbeing]

the overview table shows the benefits you want. Don't bother with ones without a rotary encoder.

clock speed is configurable.

accuracy is determined by the sense resistors (0.1% is better) and the voltage reference, which is typically bad on the clones. upgrade to a 0.1% voltage regulator if you want accuracy.

[carrascoso]

Please flash the GM328 running m-firmware with k-firmware and try the same components to check if that GM328 has a hardware problem.

Hi Madires.
Fixed issue, just had to reinstall firmware.
Sorry for the inconvenience.
Thank you.

[hpmaxim]

the overview table shows the benefits you want. Don't bother with ones without a rotary encoder.

clock speed is configurable.

accuracy is determined by the sense resistors (0.1% is better) and the voltage reference, which is typically bad on the clones. upgrade to a 0.1% voltage regulator if you want accuracy.

Looks like the one to get is either the GM328A or the LCR-TC1/T7 (can't tell what the difference is between the TC1 and T7 are).  The GM328A+ seems worse.  I don't really understand how they are using a 7805 on a 3.7/4.2V battery...

I like the extra voltage on the zener measurements on the TC1 and the internal battery, but it does seem to be lacking the rotary encoder and all the extra measurements.

[timelessbeing]

I don't really understand how they are using a 7805 on a 3.7/4.2V battery...

Step up converter I'd imagine.

[madires]

I like the extra voltage on the zener measurements on the TC1 and the internal battery, but it does seem to be lacking the rotary encoder and all the extra measurements.

The TC1 (actually the whole TC family) has an additional MCU controlling power and the test button. To run k/m-firmware that additional MCU needs to be reprogrammed (https://github.com/atar-axis/tc1-u4) or replaced with a simple circuitry (TC1-Mod, see transistor tester repository). And the TC1 is also known for shorted MLCCs in the power supply section.

[madires]

I've received several requests to add support for the DHT11 and DHT22 (DHT21, DHT33, DHT44) sensors. The code is nearly done but I don't have any of those sensors for testing. So I'm looking for someone with DHTxx sensors who has some time and coding skills to help with testing and debugging. Alternatively, DHTxx samples would be also appreciated.

[capitainekurck]

accuracy is determined by the sense resistors (0.1% is better) and the voltage reference, which is typically bad on the clones. upgrade to a 0.1% voltage regulator if you want accuracy.

Hi Madire

I got a LCR T7 and i wait for an hiland M644 because the T7 boot only when he want.
i think he voltage regulator is the tl431 who is 1%. What better regulator who can be use in place of the tl431, and what are the resistors to change for 0.1 %.
an other question are they more accuracy  with your firmware than the Chinese firmware ?
thanks

[madires]

The TL431 is used as voltage reference by several clones and a better one would be LM4040 for example. The critical resistors are the test resistors, i.e. 680 and 470k Ohms. Nearly all of the modified firmwares are based on the k-firmware and there isn't much difference in accuracy between k and m-firmware. What helps is a better voltage regulator (e.g. MCP1702) and low tolerance test resistors.

[timelessbeing]

What better regulator who can be use in place of the tl431...

The answers to 90% of the questions in this thread are answered in the manual, which is mentioned hundreds of times. RTFM.
https://github.com/svn2github/transistortester/blob/master/Doku/trunk/pdftex/english/ttester.pdf

[indman]

The answers to 90% of the questions in this thread are answered in the manual, which is mentioned hundreds of times. RTFM.
https://github.com/svn2github/transistortester/blob/master/Doku/trunk/pdftex/english/ttester.pdf

I would change your correct point somewhat - not 90% but 99%!

[capitainekurck]

hi

yes i have read the manual. but in fact i want to know if there is better components to make the tc1 more accurate with the same pin out. the mcp1702 and the 7l05 don't have the same pinout and it not easy to to change it. idem for the tl341  and the lt1004. I m a beginner hobbyist 
it s probably easier to use TO92 part than SMD to change them so it possible to bend pins to solder them in the right place.
if i change this 2 components is there other resistors or capacitors to change around them ?

[timelessbeing]

don't have the same pinout and it not easy to to change it.

That's what wires are for. You can always just glue the part in somewhere.

I don't know what board you have, but the mcp1702 is a drop in replacement for a to92 78L05

You can throw away the TL431. Voltage reference is not needed with a precision regulator.

All of this has already been discussed. I recommend opening this thread in single page mode ('Print' button at the top) and using the find function with some strategic keywords. Or use google with the "site:https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project" parameter.