$20 LCR ESR Transistor checker project

[hapless]

Are you flashing both the program file and the EEPROM file or just the program? You need both.

Also, check the MCU speed in the makefile. If it says "FREQ = 16" and your crystal is 8MHz, you will have wrong measurements no matter what else you try.

[TankedThomas]

I just got a T7 with what I believe is an LGT8F328 (QFP32L) and was wondering if someone can please help me with reprogramming it.
Since there's the WIP firmware for this chip, I was going to give that a try, but since it doesn't use ISP, I have no idea how to connect it to a programmer.

I soldered a 5-pin header onto the board but haven't managed to get neither Arduino nor AVRDUDE to read it.
I probably have the pinout wrong but it's hard to find any specifics about what to connect where in order to dump and flash the chip (assuming that's possible).

I'm using a Raspberry Pi because it's the most convenient, though I have a regular USBasp 2x3P connector as well, and some Arduino-compatible boards if necessary (Teensy 2.0++ and ESP8266).
If someone can please point me in the right direction, that'd be really helpful. I couldn't find anyone else trying to do this, though this thread is pretty large so I haven't read every single post (yet).
Maybe replacing it with the Atmega328p is a better idea but I'm not sure if it's worthwhile.

[indman]

Since there's the WIP firmware for this chip

What firmware, which is tested and fully functional, are you going to program into the LGT8F328 chip?
My advice - do not waste your time, but change the chip to a classic ATMega328, for which there are author's working firmware.

[elecdonia]

Since there's the WIP firmware for this chip

What firmware, which is tested and fully functional, are you going to program into the LGT8F328 chip?
My advice - do not waste your time, but change the chip to a classic ATMega328, for which there are author's working firmware.

I agree with indman. At this point in time I think it will be far easier to replace the LGT8F328 with an Atmel ATmega328P. The required PC board modification is relatively simple.

I obtained a quantity of LGT8F328 "Arduino compatible" boards. These boards resemble the Arduino Pro-mini. They come with a bootloader installed and are programmed with a standard USB-TTL adaptor. I was curious to see what they are like. So far I succeeded at programming them after installing the LGT8F328 "boards" package into the Arduino IDE. These boards will run many standard Arduino sketches (such as "blink").

But I think it will still require considerable time and effort to modify existing Transistor Tester software for the LGT8F328.     

[supercap]

Are you flashing both the program file and the EEPROM file or just the program? You need both.

Also, check the MCU speed in the makefile. If it says "FREQ = 16" and your crystal is 8MHz, you will have wrong measurements no matter what else you try.

Hi hapless,
I do not compile the program with the makefile.
I'm using Visual Studio Core and PlatformIO and to be honest is hard for me understand why I have to flash EEPROM too: the program store the data in EEPROM with the files var_italian.h, variables.h.

[Yuriy_K]

I do not compile the program with the makefile.

You are creating problems for yourself. Here are my programs for writing code in arduino mega 2560...

[madires]

I'm using Visual Studio Core and PlatformIO and to be honest is hard for me understand why I have to flash EEPROM too: the program store the data in EEPROM with the files var_italian.h, variables.h.

The text/strings can change between versions (rarely tables and other data). It will also vary with the featues enabled/disabled. And compilers can rearrange data (we don't use fixed addressing).

[supercap]

You are creating problems for yourself. Here are my programs for writing code in arduino mega 2560...

I think not, why PlatformIO and his debug is the best way for me to learn how the program work.

[supercap]

The text/strings can change between versions (rarely tables and other data). It will also vary with the featues enabled/disabled. And compilers can rearrange data (we don't use fixed addressing).

Hmm, yes I start to think about some problem with compilator...
When I have time I will try to debug step by step and check the memory for find the problem.

[supercap]

I resolved my problem with DATA_EEPROM option.
The user hapless give me a input.
Some days ago PlatformIO released a new version.
So I found in Project tasks a function to upload the EEPROM too.
I tried and all was work fine.
I really appreciate your all help!
Thak you.

Last question, is there some way to make the LCD refresh more faster?

[madires]

If you are already using hardware SPI with the 16MHz Mega2560 you could try a display with a parallel bus.

[hapless]

Something slightly different...
https://dragaosemchama.com/en/2019/04/gm328a-reverse-engineering-new-firmware-and-tetris/

I tried the Tetris game, and it works well enough, but the piezo speaker got ripped out of the PWM output early on for sanity's sake.

[Piotrix]

Good evening everybody , This is my first post on EEVblog, though I am  lurker for many years

  Question:
Which tester to buy nowadays and not get a fake Atmel chip? ,  I understand that these 2:
  aliexpress.com/item/1005003606018640.html or
  aliexpress.com/item/1005003678038543.html
  (GM328A the one with terminals on both sides of the PCB) ,  are right, are they?
  By the way, what is the difference between them. As far as I understand the difference is only in LCD screen, but what is exactly the difference?

Or is it better to buy an assembly kit, and then I can chose  resistors and caps with tighter specs   for more accurate measurements. Like this one: aliexpress.com/item/32706710335.html
Basically what I want is a real Atmel chip + the ability modify the tester to get the most reliable and accurate measurements.
Which way should I  go?

[Yuriy_K]

Like this one: aliexpress.com/item/32706710335.html
Basically what I want is a real Atmel chip + the ability modify the tester to get the most reliable and accurate measurements.
Which way should I  go?

The kit uses a display that allows the use of 16 MHz quartz. In the previous examples, you can get a BGR display that works with only 8 MHz quartz ...

[madires]

A treat for the summer break:

v1.50m
- Removed temporary option UI_PROBE_REVERSED_X as it's not needed anymore.
- Option to display battery status in the last line after showing the probing result (UI_BATTERY_LASTLINE).
- Option to display a small battery symbol to indicate the battery status (UI_BATTERY).
- Added battery symbols to most fonts. Also supplemented some fonts with probe numbers with reversed foreground/background.
- Added option for a long reset pulse for an NT7538 LCD controller to the ST7565R driver (LCD_LONG_RESET).
- Added test function for photodiodes (SW_PHOTODIODE).
- Option to scroll menues page-wise instead of item-wise (UI_MENU_PAGEMODE, suggested by indman@EEVblog). Speeds up menu operation with graphics displays.
- Added two new colors to colors.h (COLOR_PALE_GREEN, COLOR_PALE_RED, suggested by .RC.@EEVblog).
- Updated Brazilian Portuguese texts (thanks to wandows@EEVblog).
- Updated Russian texts (thanks to indman@EEVblog).
- Updated Polish texts #2 (thanks to Jacon@EEVblog).
- Updated Spanish texts (thanks to pepe10000@EEVblog).

Please download at:
- https://github.com/madires/Transistortester-Warehouse/tree/master/Firmware/m-firmware
- https://github.com/kubi48/TransistorTester-source/tree/master/Markus

[pepe10000]

Hello.

Thank you madires for the gift for the summer holidays.

Attached the update of the Spanish language file with the latest news.

All the best.

[indman]

Thanks madires!
I have updated the Russian translation of the pdf documentation for version 1.50m as well as the language file.

[Yuriy_K]

Thanks madires!

Checked the v1.50m  on the HilandM644 color variant. There is a warning, do not use

/*
 *  fixed cap for self-adjustment
 *  - see TP_CAP and ADJUST_PORT in config-<MCU>.h for port pins
 *  - uncomment to enable
 */

//#define HW_ADJUST_CAP 

Error! - after adjustment

Haven't found any other comments yet.

[madires]

Is the cap for self-adjustment an MLCC? If so, please try a film cap.

[siealex]

1.50m, GM328A with ATmega in a DIP case. Strange behavior in the Generator mode: rotating the encoder knob exactly by one step changes the frequency in an unpredictable way, e. g. twice, in the opposite direction, by a random step or all the way up to 2 MHz. In other modes the encoder behaves correctly. Do I have wrong encoder settings or is it a bug in the generator mode?

[Yuriy_K]

Is the cap for self-adjustment an MLCC? If so, please try a film cap.

Thanks Madires!    I picked up a capacitor by analogy with a film capacitor MLCC. Everything began to work correctly. Shown in pictures...
The previously standing capacitor worked normally for more than 3 years. I don’t know what happened to him, but the readings differ from those shown in the right picture.

[madires]

1.50m, GM328A with ATmega in a DIP case. Strange behavior in the Generator mode: rotating the encoder knob exactly by one step changes the frequency in an unpredictable way, e. g. twice, in the opposite direction, by a random step or all the way up to 2 MHz. In other modes the encoder behaves correctly. Do I have wrong encoder settings or is it a bug in the generator mode?

Could be the encoder settings or a cheap/dirty/worn encoder. Since the encoder's turning speed impacts the frequency change in the signal gen (faster -> larger change) any issues will be more visible.

[siealex]

No, it was in fact a wrong setting: 2 pulses per notch instead of 4. Now the behavior is way more adequate in all modes.

[siealex]

A couple of questions on color graphic displays.
1. Is it possible to type text with 1 px offset from all edges? If I define a global offset, that line is not cleared at all and displays garbage.
2. Can small fonts (8x8) be upscaled twice vertically without losing flash space?

[siealex]

A small visual bug. Graphic display, 8*8 font (font_8x8_iso8859-2_hf.h). Character "q" is shifted one pixel right, it should be:
 0x00,0x7E,0x63,0x63,0x63,0x73,0x6E,0x60,   /* 0x61 q */