$20 LCR ESR Transistor checker project

[LateLesley]

It looks like they are SOT23 versions of the 2SC9014, and 2SC9015. If you look for MMBT9014 and MMBT9015, you find them in SOT23 format.


https://www.aliexpress.com/item/32626794124.html

http://www.aliexpress.com/item/32853116372.html

https://alltransistors.com/pdfview.php?doc=mmbt9014lt1.pdf&dire=_upd

https://alltransistors.com/pdfdatasheet_upd/mmbt9015lt1.pdf

[ansic]

Replaced the transistors but the problem is a short from 5V rail to ground, 1.4 ohm.  The trace disappears under the atmega328p so without de-soldering the chip its difficult to guess where it goes next.  LM7805 was not the fault.  Could it be a short inside the atmega chip or the LCD?

[timelessbeing]

some people have reported shorts on their PCBs

[eti]

I purchased a T4 which arrived broken.  There was no backlight at first, there was a "Testing.." text when the button was held down.  Nothing would happen past that.  If the button was released the tester shut down instantly.  Seemingly randomly the tester would function for a couple seconds.

After flashing the chip, both the M and K version, with various configurations, sometimes it worked for a while, a full calibration went through but after the next boot, nothing would show up on the display.  It *did* identify components correctly a few times, both resistors and capacitors.

The tester is totaled now.  After applying 5V to the LM7805 regulator output, 1.5A of current was rushing through a SOT23 device and thus burning my finger. 

Does anyone know what the three SOT23 devices close to the 9V input are?  The silkscreen says 9014 and 9015, the print on the device is J6 and M6.

Maybe these Chinese manufacturers could make and sell a transistor tester, tester.

[ansic]

The short was in the atmega328p chip itself. 1.8 ohm from VCC to GND measured after taking the chip off with a hot air gun.  To replace it I only have atmega3208 chips which have a different pinout (and architecture..!)

[capitainekurck]

Hello all

I new on this forum.
I bought one of there tester on ebay.
I receive it yesterday it work 5 or 10 minutes and it fail. i was trying to check some resistor and led to test it (not capacitors) and suddenly it stop to work. it don't turn on any more. The battery is full load. when i push the button to power it on, the  screen become white ( it normally black when it power off) and it turn black few seconds after.

Is it the same problem as yours or for you the display stay black ?

regards
L.S.

[madires]

If just the backlight turns on the MCU isn't running. The TC clones are known for shorted MLCCs in the power supply circuitry. And some have a Schottky diode with a too low reverse voltage for the Zener test voltage.

[capitainekurck]

hello thanks for you reply. is there a way to fix it or is finitely dead.

you mean C11 et C12 capacitors. the one i receive seem to be T7 on the box but the pcb seem yo be the TC1

[madires]

Yep, same PCB. C11, C12, D2 and the MLCCs around the 5V regulator. It shouldn't be hard to find the culprit.

[capitainekurck]

i test C8 C9 C10 c11 c12 c14 c15 without de soldering them with my multimeter. they don't seem to be shorted. i plug the bat an i test the 5 V on the AVCC from the MEGA and the 5 volt is correct. while i was making the measures it start normally on time and fail each time after.

I'm not an electronician, just a hobbyist)

[madires]

Maybe a bad solder joint?

[capitainekurck]

i don't know, hard to find . all components are surface mount. yesterday it work for 3 or 4 minutes and stop again. this morning it work since 2 or 3 hours. I cant find what's going wrong.

perhaps it the MCU who got a problem

[Chris02]

If the tester powers off after the first run something else wrong. Any warning about a low battery? Are you sure that U4 is programmed with the alternative firmware?

Sorry for my  late reply.
I'm sure that is U4 programmed and even correctly.

And in my opinion it is also no low battery because I see also "ciao" message. Measurement result is shown for about 0.5s and also "ciao" screen for 0.5s. As I told in my prviews postings, my tester has no Mega324 inside, like TC1 has normaly. It has 644 inside.
Any proposal?

[madires]

There could be something wrong with the setting of TEST_BUTTON or the button signal, i.e. the firmware thinks the button is pressed.

[Chris02]

There could be something wrong with the setting of TEST_BUTTON or the button signal, i.e. the firmware thinks the button is pressed.

Yeah U R right On PD0 is constantly +5V and switch isn't connected to PD0 as in clones config. I'll try to figure out what has been changed next week.

Edit: PD0 isn't for power on. It was my fault

[capitainekurck]

Now mine sometime work sometime don't work. sometime the mcu seem boot and an other time it seem don't boot.
Dot you think it a problem of the mega that can be damage or something in the power. it correctly power with the 5 V. when it don't  boot the 5V seem to be correct
I will change C11 and C12 because there is a lot of post that said there a often the culprit but I'm not sure that is the problem. 

an other thing. if i remember well, the fist time i power it on when it arrive there was a sreen with the version of the firmware. now when it boot up there is a blank white screen a very quick "garbage" screen and the test screen

[madires]

Now mine sometime work sometime don't work. sometime the mcu seem boot and an other time it seem don't boot.
Dot you think it a problem of the mega that can be damage or something in the power. it correctly power with the 5 V. when it don't  boot the 5V seem to be correct

Another cause could be the brown out detection (see e-fuse). Try a lower threshold. I had to adjust the BOD for one TC-1 to make it work.

I will change C11 and C12 because there is a lot of post that said there a often the culprit but I'm not sure that is the problem. 

As long as they aren't shorted they won't cause any trouble for the 5V rail. The flyback transformer powers the 5V rail and the Zener test (up to 50V). But it's a good idea to replace C11 and C12 because of the known DC bias problem of MLCCs. You won't get MLCCs spec'd for higher voltages in the same package size.

an other thing. if i remember well, the fist time i power it on when it arrive there was a sreen with the version of the firmware. now when it boot up there is a blank white screen a very quick "garbage" screen and the test screen

I can't say much about the modified firmware the TC-1 comes with. As soon as the 5V rail is available the LEDs for the backlight are powered (blank screen) and after the MCU started the firmware initializes the display. When the display's RAM isn't cleared before enabling the display's output you'll see random pixels briefly. It's an annoying feature of the ST7735.

[capitainekurck]

"Another cause could be the brown out detection (see e-fuse). Try a lower threshold. I had to adjust the BOD for one TC-1 to make it work."

Sorry i don't understand. Where is the e-fuse and how to lower is threshold.

for C11 and C12 i will not use SMD i find on an other post that someone use an electrolytic 10µf 100V and a ceramic 100nf 100v.

In a other way if i can fix it and i decide to buy an other what is the best clone i can take to flash the last m-firmware with all option (changing the mega if it need to have enough memory)

[madires]

The ATmega has a few special configuration bytes, like the extended fuse byte. Check the documentation of your ISP programmer for changing the fuse bytes. The ATmega's datasheet explains the fuse bytes in detail.

The best choice for clones is the AY-AT (ATmega328) or the Hiland M644 (ATmega644, monochrome display) at the moment. They aren't perfect, but ok.

[eti]

Hi.

When I got my LCR-T4 (generic, beautiful peppermint green backlit mono LCD, orange PCB, the one you see everywhere) it came with "MTester" firmware - I backed it up using avrdude. The I flashed in another build, and when I did so, efuse value of 0xfc was complained about, and it asked for 0x04, so I set it to 0x04; I understand that is something to do with differing revisions of the atmega 328, and the efuse would still work ok?

[madires]

0xfc is fine. The 0x04 is caused by a known issue in avrdude's configuration file. Simply ignore it.

[eti]

0xfc is fine. The 0x04 is caused by a known issue in avrdude's configuration file. Simply ignore it.

Thanks every so much, I hope you're enjoying your weekend 

[capitainekurck]

The best choice for clones is the AY-AT (ATmega328) or the Hiland M644 (ATmega644, monochrome display) at the moment. They aren't perfect, but ok.

do you mean this one ?
https://fr.banggood.com/Original-Hiland-M644-Transistor-Tester-LCR-Diode-Transistor-Zener-Quartz-Measurement-p-1279042.html?rmmds=search&cur_warehouse=CN

[madires]

Yep, it has a boost converter for the the Zener test, the extended frequency counter without input buffer and a built-in adjustment cap.

[eti]

Okay, since this thread is SO MASSIVE, and since things aren't always explained clearly and in steps (or maybe it's me?) I am struggling to find a clear tutorial as to what needs editing (Makefile? config_328.h?) and how, how one goes about compiling, etc... little fragments of the whole bigger picture are scattered everywhere through the thread, and I value all the hard work, but I am struggling to find an "Step 1, Step 2, Step 3..." kind of beginner's guide for firmware compilation (and also we need to know the FULL avr-gcc commands and arguments supplied to it).

If someone could kindly help, thank you.