$20 LCR ESR Transistor checker project

[bitseeker]

That's cool that you can simulate the whole thing including the display.

[broadway]

Hello Everyone,

I'm new here but been lurking for a while.

I received the Fish8840TFT in the mail 3 days ago through aliexp.



I've played with the input voltage and here are the results

Input Voltage / Result
9v = Working (7.8v on display / 320 ma current consumption) 7.8v x 0.320 = 2.88 watt
8.5v = W (7.7v / 300 ma)
8v = W (7.7v / 280 ma)
7.5v = W (7.6v / 260 ma)
7v = W (7.5v / 230 ma)
6.5v = W (7.5v / 190 ma)
6v = W (7.4v / 160 ma)
5.5v = W (7.3v / 110 ma)
5v = W (7.4v / 70 ma)
4.5v = W (7.6v / 50 ma)
4.3v = W (7.6v / 50 ma)
4.2v= W (7.7v / 50 ma)
4.1v = Screen flashes but unit does not switch on. Issue with dropout voltage of zener diode 3.3v (BZX84C3V3). Atmega328 working voltage is 1.8v to 5.5v

Tested on 2200uf capacitor. Zener diodes may probably not work.

Changing input voltage changes the reading data and so needs calibration (join 1+2+3 with small wire and switch on unit. Needs good 100uf capacitor)

Btw, I desoldered the LCD connector flex cable to see whats inside. I would advise readers not to do that since resoldering this back is a pain. IMO, we can desolder/resolder this 3-5 times max before the pitches start disintegrating due to heat. Also set 200-250C heat on solder for this. My resolder attempt was horrible and I had to cover it with tape just to avoid looking at it.



Some components get warm to hot during operation. When powered by 5v Vin, components run cool (ie, at room temperature)

Next mod is to add lithium battery. BZX84C3V3(3.3v) zener is getting hot because it is connected directly to 9v Vin and dropping 5.7v. A stable Vin should be 5v though 3.6v lithium + a tiny booster.




More photos here
https://imgur.com/a/kTM8BPx

[szpila]

The next m-firmware version will include the settings for the TC-1 (in the "Clones" file).

Hi madires.
Could you add to source code translation to polish language? I attached a file with translation.
Good luck with further development

[perieanuo]

hello,
I just tried v 1.33m on ay-at clone with 7735 LCD, cryqtal 20 Mhz.
BD237 is detected as NFET, after 3 calibrations I still got big errors on 2200uF capacitor, 22pF seen as 30 pF, some resistors also bad measurements (30K in place of 33K).

IMHO the calibration needs a real procedure from A to Z also for new users.

Reflashed 1.29m, back to correct values(I made always backup for .hex and .eep after calibrating and validating the measurements with verified R/C/Q).

I'm not critic, I just try to make the world a better place
Best regards to all,
Ovidiu

[madires]

I've checked a few large caps with a LCR meter, an 8MHz 328 and 16MHz 644 transistor tester running the latest m-firmware, but wasn't able to reproduce the issue. Could you please PM me your config.h and config_328.h for 1.29m and 1.33m?

[madires]

Here's the complete modification package for the TC-1:

[jellytot]

Here's the complete modification package for the TC-1:

Nice work madires  At least now when I kill my current atmega I will have a way to reuse the TC-1     and maybe with further development I might just scrap the original software for
this user friendly upgradable one. 

[joeyjoejoe]

Is there some sort of modification to the Fish8840TFT to prevent the 3v3 regulator from blowing up when flashing it?

[perieanuo]

try to add a 3.7V (or 4.1) zener between 3v3 reg output and GND-anode to GND, and a schottky diode between input and output of 3v3 regulator with anode on out and cathode on input.maybe a 5v6 zener on input, then you're immortel.I think the spikes kill the 3v3 reg.
best regards,ovidiu

[madires]

While checking the configuration for the TC-1 I've noticed that the power supply seems to need additional or larger decoupling caps. I also had to lower the voltage of the brown-out detection (e-fuse).

[timelessbeing]

Just for curiosity's sake, I decided to photograph the ATMEGA that came with my component tester alongside a brand new one I just ordered from Microchip. They look very similar, except the laser markings on the new chip are much better quality.

Are there any tests for identifying counterfeits?

[timelessbeing]

(previous post size was exceeded)

[madires]

Hasn't the one with the higher quality markings the later production date stamp? Maybe they got new lasers. The markings on my Atmel MCUs with on older production date are also a bit weak.

[timelessbeing]

You could be right.

Top on says '1725' ... 25th week of 2017 I'm guessing.

Bottom one (came with tester): 35th week of 2015.
The chip factory to eBay seller time would have to be relatively quick. I ordered my tester on the 50th week of 2015.

[edavid]

Just for curiosity's sake, I decided to photograph the ATMEGA that came with my component tester alongside a brand new one I just ordered from Microchip. They look very similar, except the laser markings on the new chip are much better quality.

Are there any tests for identifying counterfeits?

No, because one has ever seen a counterfeit AVR chip 

[robca]

Just for curiosity's sake, I decided to photograph the ATMEGA that came with my component tester alongside a brand new one I just ordered from Microchip. They look very similar, except the laser markings on the new chip are much better quality.

Are there any tests for identifying counterfeits?

No, because one has ever seen a counterfeit AVR chip 

There were cases of companies being sold duds instead of AVR chips (https://www.sparkfun.com/news/350), which is what some people might be referring to. But you are right, there are no known cases of working (or semi-working) AVR chips (unlike, say, many buck regulators chips, where the fakes work enough like the original to pass a cursory test; pretty much any buck regulator from eBay China is built on a counterfeit chip)

Chip rebadging is also common: take a chip, erase the markings, replace the markings with the ones for the came chip with higher specs (i.e. an Intel processor rebadged to look like a faster version, or higher temp tolerances)

[ziplock9000]

I just bought this one (M328P LCR-T4, MTester) from Amazon UK for £5.72 GBP
https://www.amazon.co.uk/SODIAL-Quality-Transistor-Capacitance-Inductance/dp/B015X8Q1DS/ref=sr_1_fkmr1_1?s=industrial&ie=UTF8&qid=1529772441&sr=1-1-fkmr1&keywords=SODIAL%28R%29+High+Quality+Brand+New+LCR-T4+ESR+Meter+Transistor+Tester+Diode+Triode+Capacitance+SCR+Inductanc

It seems to be identical to the one in Dave's video. It identifies itself as an MTESTER on the LCD and LCR-T4 on the Amazon page, but with no firmware version numbers at all. I've just ran some very cursory tests and it seems to fail on a BF245A JFET, which it thinks is a BJT (shown in the attached images below). I need to do more extensive testing and post them here if anyone cares?

I'd like to identify and update the firmware if it helps with component identification and accuracy. I'll need to dig through this monster thread. I've seen the firmware listing (possibly) on www.mikrocontroller.net, but I'd need more help I think

UPDATE: I found the post https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg659867/#msg659867

[madires]

k-firmware: mega328_T3_T4_st7565 or mega328_T4_v2_st7565
m-firmware: see Clones file for the settings for T3/T4

[Azure]

The next m-firmware version will include the settings for the TC-1 (in the "Clones" file).

I  may be looking in the wrong place, can someone provide a link to the file Madires mentions.

I am trying to upgrade one of my TC1 units.  I have attached a schematic I put together, it may not be 100% but should be close, let me know if you think there are errors and I can check.

I have upgraded the processor to a ATMEGA644PA-AU.  Will probably ask questions as I try and figure out how to build config files and choose correct options.

[madires]

The k and m firmware won't run on an unmodified TC-1 because of the additional MCU U4. But it's possible to replace U4 with a small circuit which I've posted on this page. Also I've run into problems with the TC-1's power supply, it's not stable. The "Clones" file is in the m-firmware's tgz archive and lists settings for a few clones and some hints. And here's the wiring of the LCD for your diagram:

PB7 (SCK)  - LCD SCK
PB6 (MISO) - LCD SDA
PB5 (MOSI) - LCD A0
PB4             - LCD /RESET

If the designer of the clone would have payed more attention to the pin assignment the TC-1 would be able to use the hardware SPI for driving the LCD.

[KD0CAC John]

This is another time & place where mentioning some sellers to buy from , or not , to help avoid buyers from ending up with problems ?

[madires]

The AY-AT clone is still the best choice at the moment, IMHO. I wonder why none of the Chinese clone manufacturers has contacted us for collaboration yet. They could sell a nice clone with an ATmega644 and a few hardware options, and we would take care about the firmware as we already do.

[KD0CAC John]

Thanks madires for the replies .
The 1st one I bought is deep in my ebay to find , the one I got last week is this one https://www.ebay.com/itm/2017-DIY-KITS-GM328-Transistor-Tester-Diode-Cap-ESR-Volt-Freq-Meter-PWM/182586325643?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649
I must have mixed up , the too many I was looking at , I was expecting one with the clear case , oh well .
And now you mention another hope-full coming out with the ATmega644 , maybe the 3rd .
On the one I bought , hopefully to help someone , they had a couple of electrolytic caps - that were to tall - they held the display from fully seating into the socket , also could not reach the stand-offs for hold down .
Then the screws for the stand-offs heads could reach solder points on the other side of board .
The kinda stuff that if a builder knew in advance , you could mount the caps horizontal & glue down - rather than like I did after the build - try to seat display and then found issues , dismount caps add wire to mount horizontal vs. vertical .
Now have to find a way to calibrate - this has internal calibrate and is off , do not remember how good the matching values can be .
 

[b_force]

I'd like to identify and update the firmware if it helps with component identification and accuracy. I'll need to dig through this monster thread. I
UPDATE: I found the post https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg659867/#msg659867

Maybe it's an idea that someone actually can make a summary, make a new topic and make that as a sticky?
I have the idea that a few people here know exactly what's going on and what/where to find everything.
But to me it's also a huge maze to go through.

[ziplock9000]

I'd like to identify and update the firmware if it helps with component identification and accuracy. I'll need to dig through this monster thread. I
UPDATE: I found the post https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg659867/#msg659867

Maybe it's an idea that someone actually can make a summary, make a new topic and make that as a sticky?
I have the idea that a few people here know exactly what's going on and what/where to find everything.
But to me it's also a huge maze to go through.

Yeah it's a bit confusing for me too although I'm starting to get my head around it. Especially since I have to set up a compile environment and edit make files. I'm surprised nobody has set up a self contained wizard that asks you a few questions, has a self contained compiler and uploads the end product to the MCU. All nicely wrapped up in a single download and .exe