$20 LCR ESR Transistor checker project

[hoangtran]

In my pics, there is a big difference in ESR measuring between the original firmware (8.19 ohm) and the m firmware (0.14 ohm). And the ESR results of m firmware and DE-5000 are nearly equal.

[Feliciano]

Have you calibrated the component tester before and after flashing the OSHW firmware? and again when alternating between the ZIF socket and the test leads? This is an important step to do every time to change the environment.

[elecdonia]

In my pics, there is a big difference in ESR measuring between the original firmware (8.19 ohm) and the m firmware (0.14 ohm). And the ESR results of m firmware and DE-5000 are nearly equal.

Regarding Transistor Testers with their original Chinese factory firmware:

It is always necessary to "short the probes" and then press "Start" to perform the "self-test" before expecting accurate test results. Chinese firmware performs "self-test" for 15 to 60 seconds and then displays a message instructing the user to "isolate probes" (remove the short) and then the self-test continues for another 20 to 60 seconds before it finishes. Unlike open-source Transistor Tester firmware the Chinese firmware does not display intermediate "step-by-step" data screens during self-test. It may be necessary to perform "self-test" several times before getting accurate test results.

Many Transistor Testers with Chinese firmware are capable of selecting either English or Mandarin for the displayed language. This is (of course) desirable for Transistor Testers which are sold both inside Chinese-speaking countries and exported. For example I have an LCR-TC1 unit (MCU is LGT8F328P) where holding down "Start" button for 5 to 10 seconds changes the displayed language between English and Chinese characters.

Because Chinese display fonts occupy a substantial amount of flash, the Chinese software developers must remove large portions of the original Transistor tester code and/or greatly simplify it.

This most certainly impacts their accuracy and reduces their feature sets.

[hoangtran]

In my pics, there is a big difference in ESR measuring between the original firmware (8.19 ohm) and the m firmware (0.14 ohm). And the ESR results of m firmware and DE-5000 are nearly equal.

Regarding Transistor Testers with their original Chinese factory firmware:

It is always necessary to "short the probes" and then press "Start" to perform the "self-test" before expecting accurate test results. Chinese firmware performs "self-test" for 15 to 60 seconds and then displays a message instructing the user to "isolate probes" (remove the short) and then the self-test continues for another 20 to 60 seconds before it finishes. Unlike open-source Transistor Tester firmware the Chinese firmware does not display intermediate "step-by-step" data screens during self-test. It may be necessary to perform "self-test" several times before getting accurate test results.

Many Transistor Testers with Chinese firmware are capable of selecting either English or Mandarin for the displayed language. This is (of course) desirable for Transistor Testers which are sold both inside Chinese-speaking countries and exported. For example I have an LCR-TC1 unit (MCU is LGT8F328P) where holding down "Start" button for 5 to 10 seconds changes the displayed language between English and Chinese characters.

Because Chinese display fonts occupy a substantial amount of flash, the Chinese software developers must remove large portions of the original Transistor tester code and/or greatly simplify it.

This most certainly impacts their accuracy and reduces their feature sets.

In this Chinese firmware, it looks like self-calibration has been cut off. I have tried many times, but only this result appears:

[hoangtran]

Have you calibrated the component tester before and after flashing the OSHW firmware? and again when alternating between the ZIF socket and the test leads? This is an important step to do every time to change the environment.

Yes, I know, but 8.19 ohm vs 0.14 ohm is not acceptable. And most importantly, I cannot run self-calibration with this Chinese firmware; it looks like they cut off that function.

[elecdonia]

In this Chinese firmware, it looks like self-calibration has been cut off. I have tried many times, but only this result appears

Please measure resistance of each 680 ohm probe resistor. There are 3 of these, each connected between a MCU pin and a test socket pin.

These 3 resistors must match each other within +/- 1 ohm.

I had a similar issue with a “kit” style Transistor Tester where the 3 resistors supplied with the kit measured as follows:  675, 676, 691 ohms. The resistor which measured 691 ohms was far enough out of range to  prevent my tester from entering “self test” mode. Instead, when powered up with 1-2-3 shorted it would only display the following test result:    “1 - [resistor] - 2 - [resistor] - 3”

Details in my earlier post:
   https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg4726172/#msg4726172

[Aldo22]

Please measure resistance of each 680 ohm probe resistor. There are 3 of these, each connected between a MCU pin and a test socket pin.

Hi!
Because of your posting, I looked at the circuit board of the TC1 with the APT32F172K8T6 again.
It looks like the middle "1" connectors go somewhere different than the first "1".
The first goes to R4 (474 Resistor), while the middle ones go to U6 ("V05").
I can't see what's happening on the other side of the board because that's where the connector is soldered, but to me it looks like there are different "1"s.
Does anyone understand what's going on here?
Thank you.

[madires]

Have you checked for continuity between all the #1 test pins? U6 plus D1 is the overvoltage protection. Each test pin should be connected to an ADC pin, a 680 Ohms resistor, a 470 kOhms resistor, and one I/O pin of the SRV05-4 (U6).

[Aldo22]

Have you checked for continuity between all the #1 test pins?

No, I have not. I was just looking at a photo of the inside of my device that I took earlier.
I've opened it up again now and checked continuity.
You're right, all #1's are connected.
On this occasion I also measured all resistors.
They are very close together (<1%).
Everything's ok.

Thank you!

[vk3em]

I have the same item like this. Mod to Atmega328 with 1.48m firmware. Here is the result:

Can you provide details on how you modded to replace this mcu with the ATMEGA328?

Luke

[hoangtran]

In this Chinese firmware, it looks like self-calibration has been cut off. I have tried many times, but only this result appears

Please measure resistance of each 680 ohm probe resistor. There are 3 of these, each connected between a MCU pin and a test socket pin.

These 3 resistors must match each other within +/- 1 ohm.

I had a similar issue with a “kit” style Transistor Tester where the 3 resistors supplied with the kit measured as follows:  675, 676, 691 ohms. The resistor which measured 691 ohms was far enough out of range to  prevent my tester from entering “self test” mode. Instead, when powered up with 1-2-3 shorted it would only display the following test result:    “1 - [resistor] - 2 - [resistor] - 3”

Details in my earlier post:
   https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg4726172/#msg4726172

I have read your post before. Those resistors in my kit are good: 680, 680, 680.01.

[hoangtran]

I have the same item like this. Mod to Atmega328 with 1.48m firmware. Here is the result:

Can you provide details on how you modded to replace this mcu with the ATMEGA328?

Luke

Yes, I can.

In the hardware: Replace LGT8F328P with ATmega328P. Wiring pin 3 and 21 to GND; pin 6 and 18 to +5V.

Here is a change in makefile:

Code: [Select]

FREQ = 16

In config.h

Code: [Select]

#define HW_REF25
#define ZENER_R1         100000
#define ZENER_R2         10000
#define ZENER_UNSWITCHED
#define ZENER_BOOST_HIGH                  /* high active */
#define HW_PROBE_ZENER
#define ZENER_VOLTAGE_MAX     20000     /* max. voltage in mV */
#define HW_IR_RECEIVER
#define UI_ZENER_DIODE
#define BAT_DIRECT
#define BAT_OFFSET       0
#define BAT_WEAK         3600
#define BAT_LOW          3400
#define ADC_LARGE_BUFFER_CAP

In config_328.h

Code: [Select]

/*
 *  ST7735
 *  - 4 wire SPI interface using bit-bang SPI
 */
#define LCD_RES          PD1            /* port pin used for /RESX (optional) */
//#define LCD_CS           PD5            /* port pin used for /CSX (optional) */
#define LCD_DC           PD2            /* port pin used for D/CX */
#define LCD_SCL          PD4            /* port pin used for SCL */
#define LCD_SDA          PD3            /* port pin used for SDA */
#define LCD_FLIP_X                      /* enable horizontal flip */
//#define LCD_FLIP_Y                      /* enable vertical flip */
#define LCD_LATE_ON                     /* turn on LCD after clearing it */

#define TP_REF           PC5       /* test pin for 2.5V reference and relay */
#define TP_BAT           PC4       /* test pin for battery */
#define TEST_BUTTON      PD5       /* test/start push button (low active) */
#define IR_DATA          PD0       /* data signal */
#define BOOST_CTRL       PD7       /* control pin */

You can play around with other options as you want.

The schematic is based on MrSqueaky's work.

[Vitaly_Ne]

There was a need to measure megaohm resistors, and noticed a nuance. It immediately shows, for example, 4470k, and then at first faster, and then it grows more slowly ... after a minute, about 4495k ... if you wait even longer from about 2 minutes, it can grow to 4497k ... and stops.
Interested in the question: is this normal for measurements? if so, is it really necessary to measure large resistances for 2 minutes? Or is it some kind of accumulation in the controller (or a failure ... although everything else measures correctly ...)? When can a measurement be considered more or less correct? At the very end or when it starts to "slow down"?
I don't attach a photo, but I can, if necessary)))
I measured on different megaohm resistors ... 0.125W, 0.5W, 1W ..
The denominations are also from 2M to 4.7M ... everything is about the same in time ..
I apologize if this has been explained before (tell me the page?) .. just for the first time I measured megaohms, so it was embarrassing ...
Tester G328A(BGR), firmware 1.13 color(from Yuriy_K).

[gipetto]

my jt lcr t7 jumps around resistance measurements also 3.3kohm resistor
3278
3286
3290
3290
3289
3286

I have calibrated it with the sorted 3 pin wire but i can't find the procedure i recall reading about how to calibrate it with a capacitor. i've searched through the ttester.pdf

[madires]

There was a need to measure megaohm resistors, and noticed a nuance. It immediately shows, for example, 4470k, and then at first faster, and then it grows more slowly ... after a minute, about 4495k ... if you wait even longer from about 2 minutes, it can grow to 4497k ... and stops.
Interested in the question: is this normal for measurements? if so, is it really necessary to measure large resistances for 2 minutes? Or is it some kind of accumulation in the controller (or a failure ... although everything else measures correctly ...)? When can a measurement be considered more or less correct? At the very end or when it starts to "slow down"?
[...]
Tester G328A(BGR), firmware 1.13 color(from Yuriy_K).

From Karl-Heinz' documentation for the k-firmware:

If the ATmega has more than 8 KByte flash memory, the voltage measurement at the resistors will be delayed until no more changes are detected or the time limit is reached. With this method big capacitors are no more detected as resistors by mistake and the DC resistance of big inductors will be measured correctly.

I'd guess, an unstable Vcc (5V) could trigger that feature too.

[Obelix2007]

... calibrated it with the shorted 3 pin wire but i can't find the procedure i recall reading about how to calibrate it with a capacitor. i've searched through the ttester.pdf

... you can find it at page 40, capter 3.3 Selftest and Calibration.

Regards

Horst

[gipetto]

I've read that already. I think my firmware does not have the WITH Samplin-
gADC = 1 enabled in the makefile. thanks anyway

[richtoy]

Hi,

I have one of these cheap testers and decided to spend a rainy afternoon attempting a firmware update 

The version of the device I have is LCR-TC1 with original 2.12k firmware; a ATMEGA324PA version with 16MHz crystal with exactly this schematic  https://www.eevblog.com/forum/oshw/chinese-tc1-transistor-tester/?action=dlattach;attach=1355798;image

So from reading I see that I first need to update the firmware in U4, the STC15L104W.  The first thing I noticed is that the TX & RX are reversed in the P1 & P2 connections.  In the original schematic U4 pin 5 (tx) goes to P2 and pin 6 (rx) goes to P1 whereas is the schematic above they are reversed!

I downloaded the latest stc-isp (V6.91K) and eventually managed to update the firmware using the method described with the addition of the holding the button on the device down whist power is applied to force the 15L104 into boot mode.

Checking target MCU ...
  MCU type: STC15L104W
  F/W version: 7.2.5Q

Current H/W Option:
  . Current system clock source is internal IRC oscillator
  . Current frequency: 11.051MHz
  . Wakeup Timer frequency: 27.089KHz
  . Do not detect the level of P3.2 and P3.3 next download
  . Power-on reset, use the extra power-on delay
  . RESET pin behaves as IO pin
  . Reset while detect a Low-Voltage
  . Thresh voltage level of the built-in LVD : 2.42 V
  . Inhibit EEPROM operation under Low-Voltage
  . Hardware do not enable Watch-Dog-Timer
  . Watch-Dog-Timer pre-scalar : 256
  . Watch-Dog-Timer stop count in idle mode
  . Program can modify the Watch-Dog-Timer scalar
  . Do not erase user EEPROM area at next download
  . Do not control 485 at next download
  . Do not check user password next download
  . TXD is independent IO
  . TXD pin as quasi-bidirectional mode after reset
  . P3.3 output HIGH level after reset
  . Reference voltage: 1231 mV (Range: 1150~1320mV)
  . Testing time: 2017-9-28

  MCU type: STC15L104W
  F/W version: 7.2.5Q

Adjusting frequency ...         [1.031"]
Adjusted frequency: 12.011MHz (0.090%)

Re-handshaking ... Successful         [0.141"]
Current Baudrate: 38400
Erasing MCU flash ...  OK !      [0.359"]
Programming user code ... OK !      [1.375"]
Programming OPTIONS ... OK !      [0.063"]

H/W Option upgrade to:
  . Current system clock source is internal IRC oscillator
  . Current frequency: 12.011MHz
  . Do not detect the level of P3.2 and P3.3 next download
  . Power-on reset, use the extra power-on delay
  . RESET pin behaves as IO pin
  . Reset while detect a Low-Voltage
  . Thresh voltage level of the built-in LVD : 2.42 V
  . Inhibit EEPROM operation under Low-Voltage
  . Hardware do not enable Watch-Dog-Timer
  . Watch-Dog-Timer pre-scalar : 256
  . Watch-Dog-Timer stop count in idle mode
  . Program can modify the Watch-Dog-Timer scalar
  . Do not erase user EEPROM area at next download
  . Do not control 485 at next download
  . Do not check user password next download
  . TXD is independent IO
  . TXD pin as quasi-bidirectional mode after reset
  . P3.3 output HIGH level after reset
  . Reference voltage: 1231 mV (Range: 1150~1320mV)
  . Testing time: 2017-9-28
MCU ID : F2D400EC18041F

  MCU type: STC15L104W
  F/W version: 7.2.5Q

  . Set frequency: 12.000MHz
  . Adjusted frequency: 12.011MHz
  . Trim error: 0.090%


  Complete !(2023-04-12 17:49:36)

However, after successful flashing of the STC15L104W the tester no longer works.  It turns on as far as the title bar "M-Tester" but does nothing else and will not turn off!  Is this normal?

My intention is to build the full firmware on a RPi using the options that between lines 534-667 of the Clones.txt file in the firmware folder.  Has anyone done this?  Any pitfalls?

What is the procedure for flashing the main firmware?  Is the connection made via the U4 bit with EN shorted as power is applied to get the ATMEGA324PA into the correct mode?

[madires]

That's to be expected without the OSHW firmware. The 'main' firmware is programmed via the ATmega's ISP interface using an ISP programmer, e.g. USBasp or Arduino with the ISP sketch.

[elecdonia]

]In this Chinese firmware, it looks like self-calibration has been cut off. I have tried many times, but only this result appears

I have read your post before. Those resistors in my kit are good: 680, 680, 680.01.

I looked inside my Transistor Tester clone which contains same LGT8F328P MCU as yours.

Edit: Sorry, I was looking at wrong photo. Your LGT8F328P unit does have same PC board as my unit

 However my unit has a totally different PC board (photo attached).

All of my Transistor Testers with Chinese firmware go straight into "self test" immediately after powering up with 1-2-3 shorted. This differs from OSHW firmware where "Start" button must be pressed a second time to enter self-test.

I intend to rework my LGT8F328P unit for ATmega328P with the instructions you provided. Thanks! 

[oitar]

Any suggestions regarding resetting the fuses would be welcome

Another method for recovering a bricked AVR is using an external clock/oscillator. I did a quick search and found these:

https://www.eevblog.com/forum/beginners/bricked-atmega-1284-with-avrdude!-can-it-be-saved/
https://www.avrfreaks.net/s/topic/a5C3l000000UMBiEAO/t099084
https://www.avrfreaks.net/s/topic/a5C3l000000UboeEAC/t159156
https://classbproject.com/fixes-and-hacks/recovering-a-bricked-avr/

Connected it to an external 1 Mhz source(arduino), the mega324PA even has 125.0 kHz coming out of its PB1 pin. Running avrdudess with -b 1200 -B 56 (16 kHz SCK). Still unable to read signature, or change fuses.

Yes, I got it... Unbricking the avr(smd, 44 pin mega324PA) in-circuit is possible. Luckily most of the needed pins are present on the J7 header(H7 plus 2.1 board). Had to unsolder the onboard 10 kOhm reset resistor(to avoid frying the cpu with the 12 V). Also unsoldered the 6 pin display board(quite unnecessary I think), used arduino as prommer, and plenty jump-wires on abreadboard.
Don't want to clog the forum, but if someone needs it will provide step-by-step. Currently playing with the 1.48m.

[dirtmover]

I have the same item like this. Mod to Atmega328 with 1.48m firmware. Here is the result:

Got the exact same device delivered last week, T7-PLUS-V2 with V3.1E FW . Only found this discussion a couple of days ago, wish I'd have found it before ordering.

So far noticed the capacitor esr, inductance and zener breakdown voltage is quite a bit off

Can you point me to some details of this mod.

[elecdonia]

Got same device delivered last week, T7-PLUS-V2 with V3.1E FW . So far noticed the capacitor esr, inductance and zener breakdown voltage is quite a bit off

Does the PC board in your unit look exactly like either of my 2 attachments? Tan board or blue board?

Note that both units in my photos have 32-pin MCU (8 pins per side). Also there is a 16MHz crystal connected to the MCU. In both of my photos the MCU is “Logic Green” LGT8F328P.

I am asking because more companies are manufacturing Transistor Testers all the time. I’m trying to stay current with all the different variations which people are receiving.

[dirtmover]

Does the PC board in your unit look exactly like either of my 2 attachments? Tan board or blue board?

This is mine, looks the same as your tan one

[hoangtran]

I have the same item like this. Mod to Atmega328 with 1.48m firmware. Here is the result:

Got the exact same device delivered last week, T7-PLUS-V2 with V3.1E FW . Only found this discussion a couple of days ago, wish I'd have found it before ordering.

So far noticed the capacitor esr, inductance and zener breakdown voltage is quite a bit off

Can you point me to some details of this mod.

I shared here in this post: https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg4804496/#msg4804496