$20 LCR ESR Transistor checker project

[indman]

i am only wondering because a 2200uf cap reads it as 2500ish while the DIY kit reads it as 2400ish
Although reading the K firmware manual here it seems that between different units of the same MCU there can be differences ?
the showdata screen says for the pre-built
Ri_Hi= 21.8                 
Ri_Lo= 20.5           
while DIY kit says
Ri_Hi= 15.6                 
Ri_Lo= 12.8           

If you do not have a more serious tool for measuring RLC parameters than these testers, then it will be difficult for you to determine the actual value of your capacitance 2200mkF or another similar one and correctly configure the reading correction as Yuriy_K wrote above.
You really have different MCUs in these testers, this is clearly visible by the different resistance of the Ri_Hi measuring ports and Ri_Lo. I have highlighted these values ​​in a thicker font in your post. You should pay special attention to the quality of the MCU +5V power stabilizer!!! If possible, you should install a truly QUALITY voltage stabilizer, because the measurement results and their stability will depend on this. This is my advice.

[Feliciano]

Most Electrolitic Capacitors have 20% tolerance (and they age). That's why they follow the E6-series values.

Nevertheless, besides what indman and Yuriy_k just told you, if not only a few but many capacitors are reading too high values on your k-firmware unit, you can compensate modifying the makefile and compiling/flashing again:

Code: [Select]

# The WITH_AUTO_REF option enables reading of internal REF-voltage to get factors for the Capacity measuring.
CFLAGS += -DWITH_AUTO_REF
# REF_C_KORR corrects the reference Voltage for capacity measurement (<40uF) and has mV units.
# Greater values gives lower capacity results.
CFLAGS += -DREF_C_KORR=12
# REF_L_KORR corrects the reference Voltage for inductance measurement and has mV units.
CFLAGS += -DREF_L_KORR=40
# C_H_KORR defines a correction of 0.1% units for big capacitor measurement.
# Positive values will reduce measurement results.
CFLAGS += -DC_H_KORR=0

[indman]

you can compensate modifying the makefile and compiling/flashing again:

These makefile settings are intended for more experienced users who have had experience working with this project for more than one year and know what adjustments are best to make and what settings are best left untouched, because they can make the tester work unstable and unpredictable.
For Cosmic_Starlight, these settings are still premature and can only do harm and not benefit. In addition, as I said above, in order to correctly adjust the readings, it is very desirable to have a more serious measuring instrument as an exemplary standard.

[Feliciano]

I totally agree with you indman, but he keeps insisting, so I'm telling him RTFM and have fun.

[Cosmic_Starlight]

First of all, I want to sincerely thank everyone for their detailed responses and for sharing your knowledge. I realize this forum is highly technical, and I’ve done my best to do my homework before asking questions here (though it’s clear however much i did isnt enough). I’ve mentioned in earlier posts that I’ve read parts of the manual I thought were relevant to the question i was asking and somewhat understand the general concepts, but as a complete beginner, I still find myself struggling with some of the finer details. For an electronics engineer or someone with moderate to high knowledge in electronics, the 100+ pages might be straightforward, but for someone like me, it’s overwhelming. An analogy i can think of is like it would be that it’s easier to make a plane fly when you’ve already seen one fly or flown one before—not so much when you’ve done neither.

My main goal with this question was to seek simple guidance. I fully understand that this Chinese AY/AT clone is off-spec compared to the reference design in its stock form, even more so to professional tools, and I’m not expecting miracles from it. I also know that swapping voltage regulators or using 0.1% resistors can improve performance, but I simply dont have them right now

I was perhaps naively hoping for a straightforward answer, like: 'Yes, this kind of 100uF discrepancy is normal for this variant in stock form,' or, 'No, that seems unusual—try X or Y with what you have.' I understand this might be oversimplified, but I was hoping for guidance at that level. (though its completely understandable if its not possible i know that sometimes oversimplifying things isnt that easy)

For now, I’m working with what I have and am genuinely happy with these devices—I’m not trying to compare them to $1000 professional equipment. I’m also aware the manual mentions a 4% error for capacitance readings around 1uF, and I’m okay with that range. My specific question was whether the ~100uF difference I observed is due to MCU variance or the calibration steps and capacitors used.

I hope this doesn’t come across as me being annoyed or offended—I truly am not. I genuinely appreciate the time and effort everyone has put into responding. I’m just trying to better understand what’s realistic for this device and for my current skill level. Like for example if something could be done about the 100f discrepancy with the devices as is thats good if not that's ok too not a problem

i realized this was the best place on the internet to ask knowledgeable people about this device so thats why i came here

But i realize I might be overstaying my welcome here, and I apologize if I’ve been bothersome. Hopefully, the next time I post, I’ll have the proper voltage regulator and correct tolerance resistors and the M firmware (though I’m not sure when that will be).

Thank you again for your patience and help—it really means a lot to someone just starting out. Wishing you all the best!

[indman]

I was perhaps naively hoping for a straightforward answer, like: 'Yes, this kind of 100uF discrepancy is normal for this variant in stock form,' or, 'No, that seems unusual—try X or Y with what you have.
My specific question was whether the ~100uF difference I observed is due to MCU variance or the calibration steps and capacitors used.

I already reminded you above that you have installed MCUs of different quality! This is just one of the important reasons why measurement results may differ, even if you did the calibration correctly and used the correct capacitors for it.
On the other hand, the difference of 100mkF for a capacitance of 2200mkF is not significant, since, subject to a 20% deviation from the nominal value, it can range from 1760mkF to 2640mkF and this will be considered an acceptable capacitance value.
Therefore, you don't need to make any more adjustments now until you gain more experience with this project. Good luck! 

[Feliciano]

I was perhaps naively hoping for a straightforward answer, like: 'Yes, this kind of 100uF discrepancy is normal for this variant in stock form,'

I think my previous answer covered that part:

Most Electrolitic Capacitors have 20% tolerance (and they age). That's why they follow the E6-series values.

[snapper]

@Yuriy_K
the PE6 is not brought out, does it now have to be bridged directly on the 2560?

[Yuriy_K]

@Yuriy_K
the PE6 is not brought out, does it now have to be bridged directly on the 2560?

The PE5, PE6 pins are can be jumpered  on the Mega 2560, this is shown in the diagram and indicated in the author’s description.

Unfortunately the PE6 pin is not connected
to a pin of the Arduino female connector strip. The PE5 pin_7 is connected to the connector 3 of
the PWM socket strip and can be jumpered with the PE6 pin_8 of the ATmega2560.

[snapper]

Sorry, but my board has no jumper !

[Yuriy_K]

If you enlarge the photo from page 318, then in the upper left corner there is a photo of the changes on the Arduino board itself. These are added capacitors on the power pins and a jumper on the pins themselves, as described by the author. Check the correspondence between the pin numbers and the soldered capacitors, I initially had an error.

[hiei27]

hi everyone, i just got one of these testers, it came with the usbc port and an atmega328p (quad flat pack) with the 1.12 firmware. my question is is it possible to flash an upgraded firmware without desoldering it? i do not have the schematics, so i cannot rule the possibility of having mods that are unsupported by the firmware. do you have any suggestion in this regard? or i should directly try to get a different board, maybe a kit?

[Feliciano]

Welcome to the forum.

Normally these testers doesn't include a port for flashing. Nevertheless, for some models you can add (solder) such port without needed to desolder the SMD ATmega.

In any case, in order to be able to provide you better guidance please check whether your unit is one of the OSHW major clones, or kindly provide a photo of it.

[UltimateX]

I have this one from ebay....
It has  1.07 FW
Could I flash it with new FW and what version please??

Specifically, in the case of using the k-version you can use directly the contents (files "TransistorTester.hex" and "TransistorTester.eep") of these folders:
http://www.mikrocontroller.net/svnbrowser/transistortester/Software/trunk/mega328/?view=tar
or
http://www.mikrocontroller.net/svnbrowser/transistortester/Software/trunk/mega328_2X16_menu/?view=tar

Setting of fuses for ATmega328:
lfuse:0xf7 hfuse:0xd9 efuse:0x04 (0xfc)

Hi,

I know the above answer is many years old but I would like to (finally) upgrade my component tester that exactly matches the one shown in the photo preceding tom666 and madires' answers to "9a4wy", i.e. "EZM Electronics Studio M8/M168/M328_9V_V3.0" printed on top left of the PCB.  I tried finding the "classic editions" of the 1.12k and 1.19m firmware on GitHub, but I cannot find them.  Does anyone know if there are links to a repository(s) where the firmware files for these versions can be downloaded?

Please ignore the missing transistors.  They were removed to test at the time of the photo but were fine and reinstalled.   The tester works fine but I believe later firmware versions have improved accuracy, especially on ESR.

Thanks!

[indman]

Does anyone know if there are links to a repository(s) where the firmware files for these versions can be downloaded?

Try upgrading to this version, which I've attached. It is not the latest, but it works stably and has been tested on many similar clones.
I have compiled it before for this particular model.

[hiei27]

thanks for your reply, my board doesn't seem like any clones from the list you linked, it's similar to the GM328a (bgr) 1 but it uses  a usb C port (with no other pins connected beside vcc and gnd, pic attached).

Here are some pics made with a phone, i will snap some later with a microscope that i hope will be better, but maybe these will already do?

[indman]

hiei 27 , What doesn’t suit you about the factory firmware? I also advise you to first wash the board thoroughly to remove any flux and dirt!

[hiei27]

i don't know, i just assumed that newer firmware improved things like speed and precision, am i wrong? Oo

[indman]

i don't know, i just assumed that newer firmware improved things like speed and precision, am i wrong? Oo

No, the speed and accuracy will remain the same on the new firmware. My advice is not to touch the factory firmware, but rather wash the board and use a health tester. There is no need to interfere with the operation of a working device again.

[hiei27]

what do you mean for "health tester"? something like a multimeter?

[indman]

This is a poor translation into English. I meant your tester, use it with factory firmware.

[hiei27]

oohh i understand, thanks man 

[Yuriy_K]

i don't know, i just assumed that newer firmware improved things like speed and precision, am i wrong? Oo

There are some of my firmwares for the color display on page 277. Your board has 6 contacts for connecting the programmer (located vertically, below MEGA328)

When setting up the ESR readings in the k-firmware for Mega328 with a low R0 value (below .10 Ohm), I encountered damage to some film capacitors several times smaller than normal ones. Such capacitors can be classified as Low Esr.
Here is an example of measurements on different firmware...

[hiei27]

mmhh i will probably end up building one from scratch, using this one as it is and maybe arranging the ports to better suit my taste and my poor eyesight lol. do you have any recommendation for wich schematics/file/chip i should go for?

[Yuriy_K]

Two pages earlier, an example of assembling a TTester on Arduino Mega2560 with a 2.8" TFT SPI 240x320 display is shown. This resolution is maximum for k-firmware, for m-firmware the display resolution is higher.