$20 LCR ESR Transistor checker project

[pizzigri]

Thank you for this information. How did you find out that it has an APT32F172K8T6 microcontroller? I recently bought an LCR-T4 and, unfortunately, I received the "V4.1.3_2022/5/18" hardware version of this tester. I immediately saw something unusual about the microcontroller, and I checked how the (+5V) supply connects to the MCU: GND to pin 18 and VDD to pin 19!

No, this cannot be an Atmel ATmega328P, and the supply pins do match the APT32F172K8T6 pinout! Device marking: it does have an Atmel logo (that looks a bit weird), and "MEGA328P / U-TH / 35473D / 220642H" printed/etched on it. No 8 MHz crystal is mounted, and judging from the empty footprints there is a beginning of a serial-to-USB interface (for a CH340N) on the PCB. (I have seen pictures of this PCB version with the CH340N and USB connector mounted.)

Yes, be careful when ordering a new LCR-T4. If you are expecting to receive a component tester that has a genuine Atmel ATmega328P (which you could then reprogram... ) then a little chat with the seller is recommended. However, resellers may not be aware of this fake clone, or resellers may not be interested in customers demanding a product that has genuine components. My seller stated that the reason for this LCR-T4 having different components (with fake device marking on the MCU) is a "chip upgrade", and no comments on the microcontroller device markings.

So, if you have a chance to see the LCR-T4 before purchasing, it is advised to check the bottom side of the PCB to verify whether there is a crystal present. Also the top side of the "V4.1.3_2022/5/18" version PCB has a distinct pattern of traces running straight from ZIF socket to the 1, 2, & 3 test pads. Many customers have bought this product without knowing or complaining about the MCU. Some customers ask (themselves) what this USB connector is all about, but most customers just report that the device is working fine. I just wanted an LCR-T4 to experiment with and to reprogram it, but I now have to search for an original LCR-T4 clone.

Yeah, well, I got a T7 and it also has the SAME chip in it, with Atmel markings  identical to the last digit. Only silver lining, I paid 13 Euro for it, so it was not a huge loss; and maybe I can recycle some component (namely, LiIon battery, lcd screen, case)...

ETA: and I forgot to say that performance is worse - not hugely, but noticeably - to the other "antique" versions of the tester I already have, all based on the 328.

[pizzigri]

Hi guys,
I've been looking for a Hiland M644 everywhere for some time - including here, with a post on the Buy/Sell/Wanted board.
On aliexpress the Hiland when available goes for almost 40 euro, with the risk that what is received is not at all what it's advertised. Having already lost 35 Euro, including the 13 Euro to the last T7 debacle, the whole point of "cheap" is moot - now it is actually preferable to spend 50-60 Euro in parts and a JLCPCB sourced board to be at least able to have a perfectly working higher end device, rather than throwing literally away money out the window... like i just did with the two dud testers I got chasing the white herring of a M644.
At this point, the option of actually drafting on diptrace a PCB becomes a sort of necessity if I want a real, certified Atmega644 inside.

.... Problem is I tried to search info on the Atmega644 based transistor tester and there is LOTS. In reality, even overwhelming.
Karl-Heinz's 1.13K manual covers a Atmega644 schematic, with many add-ons, including a relay based protection - But, no definite PCB design.
In the repository, the only PCB available is a discrete, thru hole developer kit design without LCD, built around the DIP version of the MCU - and I have a few M644 but in the smt TQFP 44 package.

The most approachable and sensible thing to do could be to use as a basis the Hiland TT-M644 schematic, and either try to loosely redraft (reverse engineer) the PCB following the pictures posted, or try to design a completely new layout...
I can do the grunt work - I can use Diptrace, but a few things are a bit out of my league. Can anyone help me, at least providing an updated schematic based on the Hiland M644 that includes all improvements and corrections discussed here since the introduction of the Hiland a few years back (including the apparently missing Measuring ports protection), and maybe once I manage to actually produce something, check on the drafted design I did not do noob errors?
Once I have a definite design, I will make both diptrace and gerber files available to everyone here! It could become sort of a "reference" design.
Please help...

[indman]

Karl-Heinz's 1.13K manual covers a Atmega644 schematic, with many add-ons, including a relay based protection - But, no definite PCB design.
Please help...

In fact, this work has already been implemented for you by the author of the extended ATmega644-1284 tester scheme - dear Nick Lanchak. It is his great contribution to the development of this project that Karl-Heinz mentions in his manual.You can read the full documentation on his device here
https://drive.google.com/folderview?id=0B4yqyYy2Daakfl9DeTBkMEJVTVNwajkzNGtjVi0wcjdMUG5ZODJXa216WFZEZG1ZZE1ud3M&usp=sharing&tid=0B4yqyYy2DaakZzBVU3JDSXB6S1E
Alternative link
http://vrtp.ru/screenshots/3588_Rev_A0.02C_m1284.zip

[pizzigri]

OK. So, that's FANTASTIC.
In a way, I was aware that someone called NickL. did something - I read the whole mikrocontroller.net/topic/transistortester-avr thread...  But, I could never get access - and as a matter of fact, cannot open the link even now. I login my Gmail but it says I need to request access...
ETA and now, I also tried to ask for access. using the Google form.

[indman]

New modification English k-firmware for GM328A(AY-AT) based on the current 899th revision 1.13K.Quartz 8 or 16 MHz.Font 8x15.

The author's code has been supplemented and corrected in many places to speed up the work and best display the test results:
1. Improved font 8x15, added new symbols of inductance,zener diode,triac,thyristor, "eta" as well as large symbols UJT and PUT.
2. Test results (value data) are highlighted in a separate color.
3. A full pinout for 3 contact parts is always placed at the bottom of the display for convenience when sorting.
4. A new color icon has been added to assess the battery status. If an external power supply(>10V) is used, the display shows the inscription "DCPwr".
5. A small (simple) modification of the circuit allows you to measure the zener diodes at the terminals of the voltmeter "GND" - "V+" up to the voltage of the power supply (9-15V) by holding the Test button for a long time at start. The zener diode symbol will appear on the display if it is connected correctly, and on the 2nd line its breakdown voltage is Vz. If the zener diode is connected as a diode, only the forward voltage drop on it will be displayed on the display Vext.
6. Improved auto-detection of UJT and fixed the display of the transmission coefficient "eta".
7. The tester's algorithm has been changed for cyclic modes [RL] and [C] - now exit from these modes without turning off the tester. Long press of the Test button - exit the cyclic mode and restart the test.
8. Fixed calculation of the capacitance of diodes at a reverse voltage of 5V (negative capacitance values are no longer displayed for germanium diodes).
9. Added calculation and display of base-emitter capacitance (Cbe) for BJT.
10. Added auto-detection of silicon and Schottky/Germanium diodes with the display of the corresponding designations in the 1st line.
11. For a more accurate assessment of the quality of the potentiometers(trimmer), a display of its total resistance between contacts 1-3 has been added, as well as the percentage of resistances R1<=>R2.
12. There are now 8 full active lines in the menu without an unnecessary "Mode-Select" line. The page mode of the menu display is enabled, which increases the speed and convenience when moving through the items.The "Turn off" item is highlighted in a separate color.
13. To save memory space, the ESR in-circuit measurement mode has been removed.A similar mode can be obtained using the cyclic capacitance measurement point on terminals 1-3.
The screenshots below compare the standard firmware and the modified one
 I will answer questions about receiving this modified firmware in my personal message!

[Yuriy_K]

A year has passed since the publication of the changes for ST7735 TTesters, posted on page 277. During this time, many changes have accumulated for the software from Karl-Heinz. The images of the elements and the program code for their display have been redesigned. Here are the new firmware, they were tested on my homemade products with ILI9341. I have no way to test on ST7735. There may be comments, write in a personal mail.
Here are pictures of some of the changes.

Added Mega64_T7 to ST7735 and similar changes for my version of Hiland M644

[Pukker]

New modification English k-firmware for GM328A(AY-AT) based on the current 899th revision 1.13K.Quartz 8 or 16 MHz.Font 8x15.

Thank you, but where to find it for download?

[Yuriy_K]

I tried to add for BJT transistors measurement of junction capacitance be - Cbe. There are problems with darlington transistors with internal resistors and epitaxial transistors. With the latter, the problem was solved, but there may be other problems.
I offer the RGB version for testing. Comments and bugs are welcome...

P.S. there is not enough memory for the ATmega328, I had to refuse to show component icons on the screen. All other functions of the TTester are preserved.

[techie]

Hi,

I was able to configure ComponentTester-1.47m firmware to work properly on a LCR-T5 meter, based on the STRIP_GRID_BOARD = 5 LCD port mapping found in Karl-Heinz firmware. In order for the LCD contrast to work properly within the LCD_CONTRAST range of 1-63, a minor edit to ST7565R.c was required:

The lines:
/* set contrast: resistor ratio 6.5 */
LCD_Cmd(CMD_V0_RATIO | FLAG_RATIO_65);

should be changed to:
/* set contrast: resistor ratio 4.5 */
LCD_Cmd(CMD_V0_RATIO | FLAG_RATIO_45);

I've attached Markus v1.47m firmware binaries for the LCR-T5 tester , along with the modified source files (Makefile, config.h, config_328.h, and ST7565R.c) needed for re-compiling  :

Hope this helps!

[madires]

Has your LCR-T5 a case similar to the TC-1 and includes a LiIon battery? Or is it just a red PCB with two bush buttons at the bottom right?

[Danyit]

Hello I have the DIY Kit "‘AY-AT"’ or GM328A, with your help I Updated it with the wonderfoul M firmware (I also updated with the latest revision).
Now I am here looking for info about cap checking.
I use this 3 function for cap checking:
1) general function for component check (Capacity, ESR and Leakeage)
2) ESR (capacity and ESR NO leakeage information)
3) Cap Leakage (charge and discharge)

For Esr checking, I usually use Function 1 and 2 (they give same result). For reference I use Typical Esr values tables, founded online.
With function 1 I can aslo check for leakeage, comparing the result with the following table, founded in the manual.
10 - 220UF 1-3UA
330 - 470UF 4-5UA
470 - 820UF 4-7UA
>1000UF 5-7UA

But now for function 3 what can I do? What I have to check?

I think that if a Cap pass all the test, it should be ok, Am I wrong?
Thank you, for your support

[madires]

There's no simple answer. Based on the capacitor type the leakage current can vary a lot. An idea would be to compare the leakage with a cap of the same type and known to be good. Some dataheets might give a hint. Also be aware that it's a low voltage test. For a proper leakage check you would need a tester able to provide the voltage the cap is rated for.

[techie]

@madires   Yes, my LCR-T5 has a case similar to the TC1 and a Li-ion battery.

[Yuriy_K]

Added measurement of Cbe for BJT and Cg for JFET transistors for Hyland M644. The display is clearly not suitable for such changes. I made two firmware for fonts 7x12 and 6x8. If anyone needs it please post your comments.
In the measurements, I highlighted the Darlington transistor in red and showed the measurements of this transistor on the "m" firmware and the datasheet of this transistor.

[madires]

Would the auto-hold mode help (next probing cycle is triggered by a button press)?

[madires]

Then maybe the 'monitor' functions?

[madires]

A diode/LED monitor shouldn't be too hard to add. The other features would screw up the UI concept and might be an idea for an alternative firmware.

[gipetto]

I know the porting to a 3.3v mcu question gets asked a lot and shot down because 5v is superior, but i was thinking of how split transformers have a negative, 0v and positive output. some of the transistor testers use two 3.7v cells in series.
Could a pair of cells be tapped in the middle for 0v and run a pair of 3.3v mcu in a split rail setup. that way you could get 6.6v of range, superior to the 5v mcu.

[madires]

The reason for 5V is that more components can be detected than with 3.3V. Placing two 3.3V MCUs in series won't work because they will be exposed to voltages out of spec. One solution for using a 3.3V MCU is to add an external switching matrix plus ADC which make things more complex and expensive.

[gipetto]

Ah I see.

[jwperry]

Hey guys, quite the thread!

So like many of you, I purchased a random tester off of ebay several years ago and noticed some limitations that might be rectified by better firmware. Going off of the clones folder, I believe I have a MK-328. I attempted to flash Karl-Heinz Kübbeler's firmware based off some instructions from a Russian blog, but sadly I don't speak Russian and might have messed up some fuse settings. In addition, I was unable to dump the original firmware prior to flashing as I believe the chip was set to not allow dumping. (I'm not super knowledgeable when it comes to this stuff). Anyway, I am attaching images showing the unit, the board, the original fuses and lock bit settings, and the output from avrdudess. As it stands, a single button press will turn the screen on and then eventually fade to black after about 10 seconds.
Any input would be appreciated
-John

[indman]

jwperry,try to fill in the firmware that I attached below.
Download the archive, unzip it to the root of the disk from your computer. There should be no long paths to the firmware source files, as you have shown in your screenshots.
Now the fuses - they should be like this
fuses_lo = 0xf7
fuses_hi = 0xd9
fuses_ext = 0xfc
lock_byte = 0xff

[jwperry]

indman,
Thanks for getting back to me. I have updated and set the fuses and lock bit. I attempted flashing the firmware you shared with me and now I am getting strange characters on screen. This only happens once. If I pull the battery from the unit and reinstall then I get the same "fade to black" behavior. I have attached a photo of the unit and a screenshot of avrdudess when using your firmware. This was flashed from a short path. Thanks again
-John

[indman]

Try 2 more firmware versions from the archive that I attached.The fuses are exactly the same as I wrote before. What will be on the display in this case?

[Yuriy_K]

Question to users of my firmware. Do I need to include in the firmware The memory of the ATmega 328 is at its limit.
Improved the measurement of single diodes. Expanded the measurement of the forward voltage drop with a measuring resistor of 470k (in the original texts at 680 Ohms). Here are measurement examples for various types of diodes ...