$20 LCR ESR Transistor checker project

[slybunda]

I have purchased an LCR-TC2 which appears identical to that in Swainster's pics(Reply #8574 on: May 09, 2023, 05:00:09 am). My pcb is marked T7-PLUS V2.0 also and f/w is V3.1E.

My tester exhibits odd behaviour when measuring resistors of 15 Ohms and 18 Ohms in that it identifies them as Inductors with some arbitrary value of inductance and with the nominal resistance shown. If I use 12 Ohms or 22 Ohms these values are correctly identified as Resistor with the nominal values displayed.

The ability of this tester to produce an accurate value of resistance appears to be quite poor. I measured a nominal 56KOhm resistor with two different multimeters and both produce a reading of 56k6 Ohms. The TC2 produces a value of 58k07 - not what you would call close to the result of both multimeters.

thats odd, your tester seems to be a bit like my T7 from the inside but your missing the crystal oscillator: https://www.eevblog.com/forum/testgear/t7-tester-410146/msg5276989/

[indman]

thats odd, your tester seems to be a bit like my T7 from the inside but your missing the crystal oscillator: https://www.eevblog.com/forum/testgear/t7-tester-410146/msg5276989/

There is an X2 quartz crystal on the board, but it is small. What this tester lacks is this normal ATMEL chip and the original authors firmware!

[NF6X]

I posted photos of the "AITRIP" brand T7 tester I just got from Amazon in slybunda's T7 thread. Mine has an APT32F172K8T6 microcontroller, and I don't see any crystal or resonator on it. I presume they're using an internal RC oscillator on the microcontroller. When I short 1-2-3, it identifies itself as "FNIRSI-TC1", and does not show a firmware version number.

[indman]

NF6X,you bought the most unsuccessful model of a Chinese clone from FNIRSI.
If for clones with LGT8F328 installed, you can replace it with the classic ATmega328 and with a little effort get a completely working version with authors firmware, then your controller in case of failure cannot be replaced with anything.

[jdev99]

jdev99


For m-firmware, similar divider settings are in the config.h file
#define ZENER_DIVIDER_CUSTOM
#define ZENER_R1         100000        // R17     
#define ZENER_R2         12000          // R18

After measuring the resistors out of circuit, and adjusting above mentioned values, it is now pretty good.
Values below 5v is a bit low, but everything up to 39v is 99% accurate.
I have a TC1, with ATmega644 and "ComponentTester-1.51m" firmware.
Thank you very much for all excellent advice and effort you guys are putting into this little tester. 

[rherber1]

Bummer! Unfortunately some of the tester clones with an alternative MCU come without the self-adjustment feature.

It certainly seems that this version doesn't have a self test function.

I also tried the factory reset switch on the bottom and it doesn't seem to do anything.

[rherber1]

I have purchased an LCR-TC2 which appears identical to that in Swainster's pics(Reply #8574 on: May 09, 2023, 05:00:09 am). My pcb is marked T7-PLUS V2.0 also and f/w is V3.1E.

My tester exhibits odd behaviour when measuring resistors of 15 Ohms and 18 Ohms in that it identifies them as Inductors with some arbitrary value of inductance and with the nominal resistance shown. If I use 12 Ohms or 22 Ohms these values are correctly identified as Resistor with the nominal values displayed.

The ability of this tester to produce an accurate value of resistance appears to be quite poor. I measured a nominal 56KOhm resistor with two different multimeters and both produce a reading of 56k6 Ohms. The TC2 produces a value of 58k07 - not what you would call close to the result of both multimeters.

thats odd, your tester seems to be a bit like my T7 from the inside but your missing the crystal oscillator: https://www.eevblog.com/forum/testgear/t7-tester-410146/msg5276989/

This board accepts either the larger xtal in a can or the tiny ceramic version. Mine appears to use the latter.

[rherber1]

thats odd, your tester seems to be a bit like my T7 from the inside but your missing the crystal oscillator: https://www.eevblog.com/forum/testgear/t7-tester-410146/msg5276989/

There is an X2 quartz crystal on the board, but it is small. What this tester lacks is this normal ATMEL chip and the original authors firmware!

You might be right about the ATMEL chip and original firmware.

I asked another guy who owns a similar unit and his definitely correctly identifies a 15 Ohm resistor as "Resistor". So perhaps my unit is an imperfect clone of the other "clone". see pic of his unit.

[wasedadoc]

.. see pic of his unit.

Tell the other guy that some of those testers can be switched between Chinese and English text by holding down the Start button  for an extended period.

[slybunda]

My T7 measured a 10ohm thermistor as 11.7 ohm, but thats gonna vary with room temps but was detected as a resistor

[NF6X]

It's unfortunate that the clone I got doesn't support the open source firmware, but I think it'll still be suitable for my "is this still a transistor?" testing purposes. Maybe I'll buy or build an Atmel-based one later. This was just the first one I tried buying from Amazon, and I knew there was a pretty high chance that I'd be getting a non-Atmel-based one in January, 2024. So, I'm not too disappointed about it.

Does the open source firmware implement the IR remote control decoder feature? I don't specifically need that feature, but it's kind of neat.

I hope there will eventually be an open source toolchain and English documentation for the APT32F family, so we can easily hack the things it shows up in.

[madires]

The m-firmware supports the detection/decoding of about 15 IR RC protocols (or 22 with the  SW_IR_RX_EXTRA option). You can also transmit about 15 protocols. The firmwares of the Chinese tester clones are all based on the OSHW firmwares (mostly k-firmware), but the manufacturers ignore the obligation to publish their modified versions.

[NF6X]

Thank you for your explanation. Now I feel more motivated to buy or build hardware which works with the open source code, and maybe even contribute to the open source tester ecosystem in some small way.

[mojoe]

This thread is much too long, so I hesitate to post my comparison here. If the mods think this would be better as a separate post, please "Make It So".

I had a MK-328 tester that I bought years ago. It has a real 328P uC and a 8 MHz xtal. I posted a picture of the internals here a few weeks ago (https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg5277592/#msg5277592). It is pretty standard for its age. Quite some time ago, I did flash 1.12k firmware on it.

I just bought a TC1 tester, and a MTT-V1 tester (picture below). Both are smaller and use rechargeable lithium batteries. The TC1 has a micro-USB charging jack. The MTT-V1 has a USB-C jack, which surprisingly actually charges via USB-C.

No documentation comes with the TC1, but I did find a PDF online. The MTT-V1 came with a folded color sheet with basic instructions. One thing not documented for the MTT-V1 is how to do a self test. Even though the tester came with three pins shorted together, I couldn't find any way to do a self test.

The MTT-V1 is much thinner, and has swappable adapters - one for SMD, one for a ZIF socket, and one for two buttons to send IR commands.

Both the TC1 and MTT-V1 receive IR codes. the MTT-V1 handles several different manufacturer's codes. I'm not sure about the TC1. The document says Hitachi, but my two IR remotes both speak NEC, which was decoded. Interestingly, the MTT-V1 reports a two digit code, the TC1 reports a four digit code, the two MSD are the same as the MTT-V1.

The MTT-V1 also sends IR codes. To access the send/receive functions, you need to short press the button to go into normal testing mode, wait for testing to finish, then double-press the button to bring up a menu.

There are several YT videos on the TC1. Only one person has published videos on the MTT-V1. If you want to see the inside of the MTT-V1, watch his first video. I tried to take mine apart, but the display is glued in too well, and I didn't want to risk damage.

I did take a picture of the TC1 PCB, which is attached. The uC has the markings removed.

I did some testing of various components with all three testers, using the ZIF sockets. For resistors, capacitors and inductors, I also used my DE-5000 LCR meter, with the built-in slots.

For LCR measurements, none of the four instruments gave the same values. I'll consider the DE-5000 to be the most accurate.

For 1R to 1M all were close, but the old MK-328  agreed more closely with the DE-5000.

For 4.7pf to 470uF, the MK-328 and TC1 were both very close to the DE-5000., although below 33pF, the testers either gave no reading, or an inaccurate one.

For 0.47uH to 820uH, the three testers didn't register an inductor until somewhere between 10uH to 56uH (I didn't test every inductor I had). The MK-328 was closest to the DE-5000.

Next, I tested a variety of semiconductors (you know, those guys who only conduct an orchestra part time). Obvioulsy, I couldn't use the DE-5000 for this.

For the basic values on transistors and diodes, all three were fairly close. However, the MTT-V1 showed some odd values for transistor Ice. It also wouldn't read the only JFET I had (J310G), identifying only the gate. It also showed much higher values for diode reverse current than the other two testers. I'll have to dig up some specsheets to see which is correct. The TC1 showed some extra measurements vs the MK-328, so that may be of use.

As for zener diodes, both the TC1 and MTT-V1 test up to 20V. Looking at the numbers, I'd trust the TC1 more.

I don't need the IR capability, and for LCR measurements, I'd use the DE-5000. I mostly would need the semiconductor testing to identify unknown parts, and sometimes to match parts. For that, either the MK-328 or TC1 would do, with the TC1 having the edge. With some of the odd readings from the MTT-V1, I don't think I'd trust it.

If you don't have a better LCR meter, then the old MK-328 agreed best with the DE-5000.

I really wanted to like the MTT-V1 because of the size, the USB-C charging, and the swappable adapters. The measurements just didn't make it, compared to the two other testers and the DE-5000.

EDIT: One thing I forgot to mention when measuring caps. The DE-5000 used 1kHz for measuring ESR. I'm not sure how the testers measure ESR. All three testers reported much more optomistic ESR numbers, when they did show ESR.

EDIT again: Although the second picture is captioned MTT-V1, it is actually the TC1. I'm not sure how I messed up the name.

[indman]

mojoe, As has already been repeatedly said and proven in this topic, it is not advisable to trust testers who are not built on original AVR controllers and do not have the author’s normal firmware to check parts, as there may be unpredictable results. Two respected authors our project have been developing and polishing code for more than 10 years, and this is immediately visible when compared with devices on another platform.

[mojoe]

Yes, you don't need to tell me the obvious.  I know full well from reading most of this thread over the years that the newer testers are likely to have non Atmel uC's in them and different firmware. My reason for posting was to let others know how these two newer testers performed, compared to my old one.

Since many people don't want to build one from scratch, we have to buy one. This means buying a Chinese tester. The MK-328 I bought years ago has a real Atmel uC, as I mentioned. And, as numerous posts show, many people are buying various models, all made in China.

Despite all of this, the TC1 tester does a decent job. As I mentioned, for measuring LCR, I have better equipment.

If you can tell me where to buy an assembled tester (or a kit) with a case that uses the correct parts, and the authentic firmware, please do. If this information is in this very long thread, then I must have missed it.

[madires]

If you want to buy a tester at the moment the best bet is to look for a DIP ATmega. The problem with the alternative MCUs is that they come with a poorly adapted firmware. All tester clone firmwares are based on the OSHW firmwares (mostly k-firmware) which are tailored to the ATmega MCU family. Some measurements can't be simply ported 1:1 to other MCUs, leading to subpar results or complete screw-ups. Another issue (also for clones with a genuine ATmega with 32kB flash) is the fancy UI. That fancy UI increases the firmware size considerably so that features, cross-checks and other things need to be dropped, resulting in less reliable component detection and less component related values.

[indman]

madires, Of course, I’m not 100% sure, since my Chinese colleagues do not provide their source files, but most likely they write the code entirely in the ASM assembly language, which of course reduces the size of the firmware and increases the speed of its operation, but significantly complicates the possibility of customization and configuration many necessary parameters.

[alex_D]

Yes, you don't need to tell me the obvious.  I know full well from reading most of this thread over the years that the newer testers are likely to have non Atmel uC's in them and different firmware. My reason for posting was to let others know how these two newer testers performed, compared to my old one.

Since many people don't want to build one from scratch, we have to buy one. This means buying a Chinese tester. The MK-328 I bought years ago has a real Atmel uC, as I mentioned. And, as numerous posts show, many people are buying various models, all made in China.

Despite all of this, the TC1 tester does a decent job. As I mentioned, for measuring LCR, I have better equipment.

If you can tell me where to buy an assembled tester (or a kit) with a case that uses the correct parts, and the authentic firmware, please do. If this information is in this very long thread, then I must have missed it.

This kit is the safest bet to use with OSFW and you get an Atmega with a DIP socket.
https://www.ebay.com/itm/142849416622

[NF6X]

Despite its shortcomings, I like my new APT32F172K8T6 based tester. It has gotten me interested in designing my own ATmega-based variant. I think I'll order one of those DIP socketed kits to start playing with the m-tester code while I begin designing my own hardware.

[mojoe]

This kit is the safest bet to use with OSFW and you get an Atmega with a DIP socket.
https://www.ebay.com/itm/142849416622

Thanks for the link. However, there is no description of capabilities. Do you know more about it?

If it doesn't do zeners, then it's pretty much the same as my old MK-328. The only extras I can see is the rotary encoder and the DIP socketed uC.

No internal battery (9V or otherwise), so not really portable. Also, these acrylic open frame "cases" aren't much protection if you want to toss it in with other tools to take with you.

[mojoe]

I just checked some diode datasheets. As I suspected, the MTT-V1 is wrong for diode reverse current. It's reporting uA values, which should be in the nA range. It's not just a matter of the wrong units, the numbers are just plain wrong.

Both the MK-328 and TC1 values are similar to each other. The specsheet uses a higher voltage to test this, so it lists a higher reverse current, but still in the nA range.

[wasedadoc]

This kit is the safest bet to use with OSFW and you get an Atmega with a DIP socket.
https://www.ebay.com/itm/142849416622

Thanks for the link. However, there is no description of capabilities. Do you know more about it?

If it doesn't do zeners, then it's pretty much the same as my old MK-328. The only extras I can see is the rotary encoder and the DIP socketed uC.

No internal battery (9V or otherwise), so not really portable. Also, these acrylic open frame "cases" aren't much protection if you want to toss it in with other tools to take with you.

I got one of those kits about 5 years ago.  With the same case I think it was close to £10.

The case does hold the 9 Volt battery inside.  It does not do zener diodes but has some extras.

A frequency counter.

A square wave generator of adjustable frequency.

A 10 bit PWM generator

A DS18B20 tester

A DHT11 tester

An IR_Decoder.  Needs a 3 pin IR sensor plugged into the ZIF socket

A Voltmeter

Section 2.9, Fig 2.27 of the attached pdf.

[madires]

This kit is the safest bet to use with OSFW and you get an Atmega with a DIP socket.
https://www.ebay.com/itm/142849416622

That's the classic AY-AT (aka GM328A) and still a good choice. And it's supported by both OSHW firmwares.

[NF6X]

For my own component tester variant, I want to use a 3M Textool transistor test socket (or one of the less expensive ones from Sullins) like this one from my Tektronix 576 curve tracer. Sadly, they apparently aren't made with this lovely green plastic any more. I haven't seen a TFXDTOL brand knockoff of this socket on Aliexpress, so it won't be a cheap approach.



If I include the Zener diode measurement function, one simple approach would be to have a 3PDT switch to connect the test socket to either the normal probe pins or the Zener test circuit. It would be nicer if there was just one button, though. I'm thinking about including relays to select the normal or Zener test modes, and modifying the m-firmware to automatically perform the Zener test when it detects a single diode with its cathode on pin 1 and its anode on either pin 2 or 3.

Do y'all think it is likely to be problematic for my tester to always try applying a reverse bias of around 30V with 5mA current limit whenever it sees a diode junction? Is that likely to burn out random components?