$20 LCR ESR Transistor checker project

[indman]

Phil1977, you need to watch for short-term dips in the supply voltage, which are not so easy to detect, although this can be done on your device. In addition, you need to look at the noise characteristics of the power supply, as done in the photo that I will show below. For example, oscillograms are shown if a cheap Chinese 7550 and a high-quality MCP1702 are installed. Look carefully at these photos, the inscriptions are in Russian, but I think you’ll figure it out.  7550 stabilizers also come in different quality, this is also shown in the photo.

[indman]

Any other ideas? I think about adding programming pins to the device, reprogram the AVR, (re-)set the fuses and as last resort exchanging the AVR against an original one.

Yes, these are the things that helped solve the problem. You are not the first to encounter this effect of the Chinese industry. Take it calmly.

[Phil1977]

I don't think it´s the voltage regulator. I did some more measurements with AC mode and zoomed into the worst voltage drop before the power off. It´s slightly more than 20mV over 100us.

With another device of the same type I added a few high quality decoupling caps directly at the AVR, and it didn't change a thing. If the voltage drop shown in the measurement would be responsible, it should easily have been solved by 100nF COG and 4.7uF X7R.

I´m not angry, these little gems are wonderful even if they have a few bugs. I just want to report it because it seems to be the right place here :-)

[indman]

I don't think it´s the voltage regulator.

Quite agree, if you have not quite a bad 7550 installed. So now connect the programmer and set the fuse to ExtentedFuse=FD.

[Yuriy_K]

Davo013
I have good and bad news for you. The screenshots are normal, but the information in your T-tester is not always displayed correctly. I highlighted the unnecessary information in red.
My conclusion is that the ATmega is defective.

I can advise to temporarily replace U1 with any 5V stabilizer - on wires. Solder parallel to C16 10 uF. Solder the power leads of the ATMega itself.

[Phil1977]

My conclusion is that the ATmega is defective.

How do you came to this conclusion? By the "1"-numbers in the "Show-Data"-display at the 3 resistances?

[Yuriy_K]

How do you came to this conclusion? By the "1"-numbers in the "Show-Data"-display at the 3 resistances?

I didn't understand the question. I'll explain my conclusion in more detail on the screenshot below. The text constant "RH/RL" (line 381) should end with the code "00" - the end sign. But on the screen we see a continuation in the form of "left half of the resistor - code 04" and a continuation of the text constant of line 380. This can only happen if the ATMega328 is not working correctly...

[indman]

My conclusion is that the ATmega is defective.

This is not a fact. This effect can occur if the firmware was written with an error. Look at the photo of Davo013 with firmware Rev.804 from my archive - there is no such effect there
https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg5580231/#msg5580231

[Yuriy_K]

In his screenshots it's not constant either, but it does happen. So I'm right...

[Phil1977]

Can you tell typical good values for the internal resistance and capacitance?

And what I also can not understand is that the continuos C+ESR measurement works flawlessly with exactly the same hardware that is crashing in the standard universal test:

[indman]

I've explained before - automeasuring a capacitor is one of the longest processes in terms of time and load power supply, because it also searches for and detects other components in front of it!
And in C/ESR mode, the algorithm is immediately set to measure only capacitance.
The original ATMEL AVRs had an internal resistance in the 20-22 ohm range. Chinese knockoffs and new generation controllers can have this range from 12 to 18 ohms. 

[Yuriy_K]

Davo013
To eliminate any errors during compilation, I checked it on my color display beforehand. The original font is terribly straining for the eyes, so I remade it for my own. Here are the results of checking various capacitors. All tests passed normally. There are no extraneous symbols. I rebuilt the firmware for your display, check the results...

[Davo013]

Thnx Bro!
I will check in the morning, i had a terrible tuff day…

[Davo013]

Davo013
To eliminate any errors during compilation, I checked it on my color display beforehand. The original font is terribly straining for the eyes, so I remade it for my own. Here are the results of checking various capacitors. All tests passed normally. There are no extraneous symbols. I rebuilt the firmware for your display, check the results...

I Just burned the firmware and at the same places the strange symbols came up...
It's not random
But i found something which might be something, or maybe not...

When i copied from the manual a line to a .txt file i saw this... check picz
Could it not be something like that? fi is shown as an arrow up
Maybe something using different platforms like mac, linux?? I'm not a real computer guy haha

I found because tried another attempt to compile, because the contrast is very dim, after burning the firmware i can hardly read the screen...
When adjusting the contrast it starts at 40, and you have to click many times.
Then it counts down from 128 when it's at "116" it's perfect, if click further on, the screen dims again...
At 63 it's at full brightness again and at "56" you will reach again the perfect balance
I Checked the Makeme file and i think it's this line:

# The contrast value can be predefined with the constant VOLUME_VALUE
# for ST7565 controller the value can be between 0 and 63, for the SSD1306 0 to 255 can be selected.
#CFLAGS += -DVOLUME_VALUE=25

Should it be like this to change?
# The contrast value can be predefined with the constant VOLUME_VALUE
# for ST7565 controller the value can be between 0 and 63, for the SSD1306 0 to 255 can be selected.
CFLAGS += -DVOLUME_VALUE=56

I still don't know how to compile haha i'm trying to understand how it all works

In the manual there was a tiny thing about compiling the Makeme file with AVR, and this lol

Chapter 9 To Do List and new ideas

5. Write User's guide for configuring the tester with the Makefie options and description of the
build chain.
This would make the manual really good though! the Makefile is like the engine computer from a car

[indman]

Davo013, You don't need to compile and invent anything right now. You need to get a working tester right now, right?
So, my advice - connect the programmer, make sure to perform the procedure of erasing the entire crystal in the program(I showed the option in red in your screenshot), then write the firmware Rev.804 from my archive, and set the fuses so (low: 0xF7, high 0xD9, ext: 0xFD).
If your device works fine, then you can further experiment with compyling your own firmware. If the tester works with errors, then you need to look for the reason further.

[Davo013]

That worked!

I can test now all the caps without the tester shuts down...
With the calibration i saw the first time 1 error with the characters
after the 2nd calibration it was gone.

[Davo013]

Within a few days some Chinese 328 will arrive...
Hopefully they are not fake... But you never know with damn Chinese crooks... I had some spare one's, but after cleaning with iso i had a black cotton swab :X
The seller where i bought them last week hag quite good reviews, but even that you cannot always trust... They somehow know almost know every dirty trick to rip you off for a few cents haha
Even parts that are originally 35cents, they sand it & rebrand 'm.... that's totally nuts though...

I was thinking to do an order at digikey, but can all the 328's go on the bside? i guess so, as long as it's a TQFP32...
The prices are quite different, and the specs look quite the same... only that FuSa series has more in & outputs which can maybe causes issues?
I guess you guys know more about it lol

ATMEGA328PB-AU €1.52 AVR® ATmega, Functional Safety (FuSa)
ATMEGA328PB-AN €1.54 AVR® ATmega, Functional Safety (FuSa)
ATMEGA328PB-AUR €1.59 AVR® ATmega, Functional Safety (FuSa)
ATMEGA328PB-ANR €1.64 AVR® ATmega, Functional Safety (FuSa)
ATMEGA328-AU €2.26 AVR® ATmega
ATMEGA328-AUR €2.34 AVR® ATmega
ATMEGA328P-AU €2.46 AVR® ATmega
ATMEGA328P-AUR €2.56 AVR® ATmega
ATMEGA328P-AN €2.61 AVR® ATmega
ATMEGA328P-ANR €2.69 AVR® ATmega

Awesome to see that there's a whole worldwide community behind such a device!
To bad those chines make a huge profit of it...
Ah well at least they sell it cheap! (most of them)
In my country all the sellers complain about the Chinese, but hey over here the flip the price's 3x or 4x for the same garbage out of China pffft and that's excl. shipping
If they weren't that greedy people will buy less directly from China

[indman]

To Davo013,
Atmega328PB completely replaces Atmega328P!!!

[Davo013]

To Davo013,
Atmega328PB completely replaces Atmega328P!!!

Ah okay, thnx! I thought I’d better ask  I couldn’t find much info about on this forum, and they have more connections on the same footprint…
Most info was about people having problems with the PB’s
You guy’s are the real specialists at these testers i’m just a noob with stuff haha
I’m more into tube amp gear and that kinda stuff 😂
The next coming week there will arrive some some parts from AliEpress the most stuff from Ali is fake, but I’m trying out my new strategy buying parts on Ali….
If I have to buy Parts on Digikey every it will cost me to much on shipping costs, or I have to buy more then 50 euro’s

[indman]

Davo013, This topic discusses the transistor tester project! Discussion of other devices and projects is inappropriate here so as not to clutter up an already extensive topic!

[madires]

I found a while ago a B-Side mod with a relay protection, but somehow I can’t find it anymore, and I didn’t make a book marker also…’ But I want to protect myself against destroying the tester and my tester with a charged capacitor through!

The discharge/protection relay is explained in Karl-Heinz' documentation (https://github.com/kubi48/TransistorTester-documentation) and supported by both OSHW firmwares.

[Davo013]

Sorry for that, i shall remove the post…

I found a while ago a B-Side mod with a relay protection, but somehow I can’t find it anymore, and I didn’t make a book marker also…’ But I want to protect myself against destroying the tester and my tester with a charged capacitor through!

The discharge/protection relay is explained in Karl-Heinz' documentation (https://github.com/kubi48/TransistorTester-documentation) and supported by both OSHW firmwares.

Thnx!