$20 LCR ESR Transistor checker project

[indman]

Ok,flodo, the picture in a JPG format will suit you?

[flodo]

For sure! Thank you very much

[seanspotatobusiness]

PD6 remains at 5 volts when I select "switch off". The LED at the top of the PCB turns off whilst the rotary encoder is held down but the voltage on PD6 stays on and when the rotary encoder is released, the screen refreshes.

Remove the ATmega, press and hold the rotary encoder and measure the voltage at PD6.

With the ATmega removed and the rotary encoder held down, the voltage is ~4.5 V at PD6. When the rotary encoder is released, this decreases to 4.2 V. I hope that says something useful!

[madires]

It does. We can exclude the ATmega, and the next step is to to check T1 which is switched by PD6.

[seanspotatobusiness]

It does. We can exclude the ATmega, and the next step is to to check T1 which is switched by PD6.

I wonder if I might have damaged something. It turned out that pins of the transistor highlighted below were shorted by a bridge of solder. Now that I've removed the bridge, the tester only comes on if I press and hold the rotary encoder down. As soon as I release the rotary encoder, the tester turns off. Would this short-circuit between the base and collector of that transistor cause damage to another component (maybe that transistor itself)?

[EEV2K]

Hi eevblog,

I bought a TransistorTester similar to T4 (but not identical, LCR-T4) awhile ago but the LCD is broken and I tried to replace it with Nokia 5110 (PCD8544).  Markus's version supports Nokia 5110 so I flash the board with ComponentTester-1.33m but it's not working.  By the way, I am new to avr so there will be mistakes.

Hardware:

Code: [Select]

//---------------
//TransistorTester/ComponentTester T4 Clone with PCD8544
//Atmega328 8MHz
//PCD8544 (bit-bang SPI)
//---------------
//Nokia 5110:
//1: VCC - 3.3V
//2: GND - GND
//3: SCE - GND
//4: RST - 10K - PD4 (2)
//5: D/C - 10K - PD3 (1) - Data/Command
//6: DN(MOSI) - 10K - PD5 (9)
//7: SCLK - 10K - PD2 (32)
//8: LED - 100R - 5V
// Note: it does not work when RST=PD0, and DIN=PD1, PD0 and PD1 are serial RX and TX, maybe that is the reason

display     config-<MCU>.h   default   remark
                               328
  /RES        LCD_RES          PD4       optional
  /SCE        LCD_SCE          GND       optional
  D/C         LCD_DC           PD3
  SCLK        LCD_SCLK         PD2
  SDIN        LCD_SDIN         PD5
Later on I also added a serial port interface to try to figure out what is wrong.

Files changed:

Code: [Select]

Makefile - No change
config.h:
  Line 082: uncomment #define HW_REF25 to enable hardware reference
  Line 307: comment out #define SW_PROBE_COLORS to diable color coding
  Line 396: uncomment #define UI_SERIAL_COPY
  Line 406: uncomment #define UI_SERIAL_COMMANDS
  Line 689: uncomment #define SERIAL_HARDWARE
config_328.h:
  Line 099: uncomment #if 0 .. #endif ;to disable LCD_ST7565R
  Line 203: change to #if 1 ;to enable LCD_PCD8544 or comment it out
  Line 209: change to PD4
  Line 210: comment out #define LCD_SCE ;since this is connected to GND
  Line 213: change to PD5
Clones - No change
variables.h - No change so far

At present, after writing the firmware with avrdude to ATMega328P, I push the blue test button.  A short push, the back light on LCD turns on, so there is power, but nothing on the screen.  Voltage measured 5.05V at the ISP programming port, ~3V at LCD.   Serial port shows "Component Tester v1.33m".  Once the button is up, it's off.   I am not sure if this is normal or maybe POWER_CTRL(PD6) is not working properly.

A long push, back light turns on, nothing on the LCD, serial port shows "Component Tester v1.33m" and after a short while "No component found!".  I think this means EEPROM is working and the logic of the firmware is being followed.

To help figure out things, I modified main.c so I can test individual components of the firmware and the hardware:

Code: [Select]

#if defined(__AVR_ATmega328__)

  #if (CPU_FREQ != 8000000)
    #error <<< KK: Error CPU FREQ is not 8MHz >>>
  #endif
 
  #ifndef LCD_PCD8544
    #error <<< KK: Error LCD is not PCD8544 >>>
  #endif
 
  #ifdef SPI_HARDWARE
  #error <<< KK: Error using SPI hardware instead of bitbang >>>
  #endif
 
  #ifndef SPI_BITBANG
  #error <<< KK: Error not using SPI bitbang >>>
  #endif
 
  #if (SPI_SCK != PD2)
  #error <<< KK: Error SPI_SCK != PD2 >>>
  #endif
 
  #if (SPI_MOSI != PD5)
  #error <<< KK: Error SPI_MOSI != PD1 >>>
  #endif
 
  #if (POWER_CTRL != PD6)
  #error <<< KK: Error POWER_CTRL != PD6 >>>
  #endif

  #if (TEST_BUTTON != PD7)
  #error <<< KK: Error TEST_BUTTON != PD7 >>>
  #endif

  #ifndef HW_SERIAL
  #error <<< KK: Error serial port (HW) is off >>>
  #endif

  #ifdef SERIAL_RW
  #error <<< KK: Error serial port (RW) is on >>>
  #endif

  #ifdef HW_TOUCH
  #error <<< KK: Error touch support is on >>>
  #endif
 
  #ifndef LCD_SPI
  #error <<< KK: SPI for LCD is off >>>
  #endif

  #define KK_DEBUG
  #define KK_DEBUG_OVERRIDE
 
#else
  #error <<< KK: Configuration Error >>>
#endif

void KK_Debug1(void)
{
  //KK: 100Hz toggle PD1
  DDRD = (1 << PD1); //Make PORTD1 an output pin by setting it to 1
  while(1) //infinite loop
  {
    PORTD = (1 << PD1); //Turns ON LED at PD1 by setting it 5V
    _delay_ms(5); //5ms delay
    PORTD= 0x00; //Turns OFF All LEDs
    _delay_ms(5); //5ms delay
  }
}

void KK_Debug2(void)
{
  //KK: serial: output A to Z
  uint8_t cc = 65;
  Serial_Setup();
  while(1) //infinite loop
  {
    Serial_WriteByte(cc);
    _delay_ms(50); //50ms delay
    cc++;
   if (cc>90) cc=65;
  }
}

void KK_Debug3(void)
{
  //KK: LCD: output A to Z
  uint8_t cc;
  LCD_BusSetup();                       /* set up LCD bus */
  LCD_Init();                           /* initialize LCD and clears the LCD*/
  UI.LineMode = LINE_STD;               /* reset next-line mode */
  while(1) //infinite loop
  {
    LCD_Clear();                        /* display was initialized before */
    cc = 65;
    while(cc<91) {
    LCD_Char('X');
      _delay_ms(50); //50ms delay
      cc++;
    }
    _delay_ms(2000); //2s delay
  }
}

void KK_Serial_WriteBits(uint8_t cc)
{
uint8_t i=0;
while(i<8)
{
if (cc & (1<<i))
Serial_WriteByte('1');
else
Serial_WriteByte('0');
  i++;
}
Serial_WriteByte('b');
Serial_WriteByte('\r');
Serial_WriteByte('\n');
}

void KK_Debug4(void)
{
  //KK: LCD: output A to Z
  uint8_t cc, bb;
  Serial_Setup();
  LCD_BusSetup();                       /* set up LCD bus */
  LCD_Init();                           /* initialize LCD and clears the LCD*/
  UI.LineMode = LINE_STD;               /* reset next-line mode */
  while(1) //infinite loop
  {
    LCD_Clear();                        /* display was initialized before */
    cc = 65;
    bb = DDRD;
  KK_Serial_WriteBits(bb);  //1 is output, so DDRD should be 00111100b, hardware serial is used so bit 0 and 1 is probably ignored
    while(cc<91) {
    LCD_Char(cc);
      _delay_ms(50); //50ms delay
      cc++;
    }
    _delay_ms(2000); //2s delay
  }
}

void KK_Debug_SW(void)
{
  KK_Debug4();
}


/* ************************************************************************
 *   the one and only main()
 * ************************************************************************ */

/*
 *  main function
 */

int main(void)
{
  uint8_t           Test;          /* test value */
  uint8_t           Key;           /* user feedback */
  #if defined (HW_REF25) || ! defined (BAT_NONE)
  uint16_t          U_Bat;         /* voltage of power supply */
  uint32_t          Temp;          /* some value */
  #endif

  #ifdef KK_DEBUG_OVERRIDE
  KK_Debug_SW();
  #endif


  /*
   *  init hardware
   */

The code above replaces the beginning of the main function.  So the #error's are tested after config.h and config_328.h are loaded and there is no unexpected configuration errors as far as I can tell since make compiles the code just fine.  I used KK_Debug1(), to test all the pins that I modified to make sure the connections are good.  Frequency measurement of the output measured 100Hz, so 328P is running at 8MHz. KK_Debug2() showed serial port is working but KK_Debug3() and KK_Debug4() is not working - nothing is displayed on the LCD.  Serial output of KK_Debug4() showed DDRD is 00111100b.  This, I think, showed the port direction are probably configured.  To make sure LCD is connected correctly, I tested it with examples that come with Adafruit_PCD8544 library in ArduinoIDE.  It displays animated smiles correctly.

So now I am stuck and I need help with what to do next.

Thanks.


edited: typo last paragraph disabled -> displayed

[madires]

I wonder if I might have damaged something. It turned out that pins of the transistor highlighted below were shorted by a bridge of solder. Now that I've removed the bridge, the tester only comes on if I press and hold the rotary encoder down. As soon as I release the rotary encoder, the tester turns off. Would this short-circuit between the base and collector of that transistor cause damage to another component (maybe that transistor itself)?

Yes, T1 could be damaged. The ATmega should be protected by the base resistor R8 (limiting current) and its internal clamping diodes.

[madires]

At present, after writing the firmware with avrdude to ATMega328P, I push the blue test button.  A short push, the back light on LCD turns on, so there is power, but nothing on the screen.  Voltage measured 5.05V at the ISP programming port, ~3V at LCD.   Serial port shows "Component Tester v1.33m".  Once the button is up, it's off.   I am not sure if this is normal or maybe POWER_CTRL(PD6) is not working properly.

The first thing to investigate is the power control circuitry. Find the transistor which switches power and look for the ATmega pin switching that transistor. We'll take care about the display later.

[seanspotatobusiness]

I wonder if I might have damaged something. It turned out that pins of the transistor highlighted below were shorted by a bridge of solder. Now that I've removed the bridge, the tester only comes on if I press and hold the rotary encoder down. As soon as I release the rotary encoder, the tester turns off. Would this short-circuit between the base and collector of that transistor cause damage to another component (maybe that transistor itself)?

Yes, T1 could be damaged. The ATmega should be protected by the base resistor R8 (limiting current) and its internal clamping diodes.

Do you happen to know whether it can be replaced with any of the folloing three NPN transistors:

BO337-25
2N222 A 331
A42 B331

or will I need to get another 9014? Thanks.

[madires]

Any jelly bean NPN should work.

[flodo]

Here is the sourcecode for U4's alternative firmware: https://github.com/atar-axis/tc1-u4

[seanspotatobusiness]

Any jelly bean NPN should work.

I removed the 9014 (the one closest to the LED at the top left) and it seems okay. Using a DMM diode test I get 0.7 V with positive electrode at base and negative electrode at either the collector or emitter. The DMM reads O.L. with the probes reversed.

When I connect power to the tester, it behaves the same way it did before I removed the transistor. If I press and hold the rotary encoder it comes on and switches off as soon as I release the encoder.

Does it seem less likely now that putting in a new transistor will fix the tester?

Edit: I inserted a BC237B and now it comes on when I press and release the rotary encoder but it ignores whatever components I connect and goes straight to a screen where it flashes "Vext=0mV". It then stays there without switching off and doesn't not respond to the rotary encoder. Should I try another transistor?

[EEV2K]

The first thing to investigate is the power control circuitry. Find the transistor which switches power and look for the ATmega pin switching that transistor. We'll take care about the display later.

Power control is now working.  I remove the original backlight but didn't put anything back in its place.  So once I put an LED there, it works fine.  So right now, if I push the test button, it will stay on for 20 or 30 seconds, longer if I leave a component in the test socket.  What to do next?  Any suggestions on how to figure out why the LCD is not working?

[Gtx21]

I'm replacing my TL-431 with the LM4040.

Edit: leaving this for history, but the diagram and comparison below is wrong as the data sheet view is bottom up. So don’t use this post.

Pinouts of LM4040 & TL-431:


There is a post referenced as post #4052 in several thread replies which is supposed to be the correct layout on the AYAT PCB of the LM4040, but post #4052 doesn't have anything to do with this swap.  Post #4048 https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg1306265/#msg1306279 however has a picture, which if I take to be the "old post #4052" and assume to be correct,  contradicts the pinouts above.  (Note that the - and + on the LM4040 are oddly labeled and are not the traditional - and + of the Anode and Cathode of a Diode (which they conveniently drew for us)).

To me, it looks like it should go as:

TL431	LM4040
Pin 1 (Ref)	Pin 3 (NC)
Pin 2 (Anode)	Pin 1 ("-", Anode of Diode)
Pin 3 (Cathode)	Pin 2 ("+",Cathode of Diode)

Here is the picture @Pepe10000 posted:



Am I reading things wrong and if so, how?

[pepe10000]

Hi.

Watch this:

[madires]

Edit: I inserted a BC237B and now it comes on when I press and release the rotary encoder but it ignores whatever components I connect and goes straight to a screen where it flashes "Vext=0mV". It then stays there without switching off and doesn't not respond to the rotary encoder. Should I try another transistor?

Maybe a problem with the rotary encoder?

[madires]

Power control is now working.  I remove the original backlight but didn't put anything back in its place.  So once I put an LED there, it works fine.  So right now, if I push the test button, it will stay on for 20 or 30 seconds, longer if I leave a component in the test socket.  What to do next?  Any suggestions on how to figure out why the LCD is not working?

Interesting! That clone uses the backlight LED instead of the dedicated indicator LED. The most common problem with displays is the pin assignment. I'd guess that you have double-checked all signal lines already. If you have a scope/DSO please check all signals at the display. Do SCLK and SDIN look correct? The idea is to verify that the level shifting is working fine. In case the signals aren't clean try to lower the value of the resistors. IIRC, Karl-Heinz suggests 2k7 in his documentation.

[Gtx21]

Hi.

Watch this:

 
Actually, that’s opposite of the data sheet I have.

http://www.ti.com/lit/ds/symlink/lm4040-n.pdf

Edit: Holy crap!  I’m glad I asked.  I see it now. Bottom up! Thanks @pepe10000!

[EEV2K]

The most common problem with displays is the pin assignment. I'd guess that you have double-checked all signal lines already. If you have a scope/DSO please check all signals at the display. Do SCLK and SDIN look correct? The idea is to verify that the level shifting is working fine. In case the signals aren't clean try to lower the value of the resistors. IIRC, Karl-Heinz suggests 2k7 in his documentation.

Thank you.  I got the LCD working.  It turns out that SCLK pin got loose with all the plugging and unplugging during testing.  I re-solder it and now there is something on the LCD screen.  Contrast 66 was too high for my setup, 33 is too low, 55 seems to be about right.  I didn't test the contrast setting further.  The "m" in v1.33m is missing some dots on the right side but the "m" in "Component Tester" is displayed correctly.  I am not sure if suggest there are still problems with display or not.

With a MAX232CP based serial port, a LM317 based 5V to 3V DC-DC converter, and a red led where the backlight used to be, it uses 41mA without back light and about 51mA with backlight when it is on but not actually testing anything.  Is this about right?

I am going to add some relay to discharge capacitors before testing similar to figure 2.3 of "TransistorTester with AVR microcontroller and a little more Version 1.13k".  Based on the posts here, it seems to be a good feature to have.  My question are, does it have to be a mechanical relay or can it be a CMOS analog switch like DG202/DG212.  These RDS(on) of 115ohm.  Would they make it impossible to measure low resistance resistors or ESR on inductors and capacitors?  Another thing is would CMOS analog switch itself be susceptible to residual voltage in capacitor?  How high is the residual voltage on a capacitor?  Is the voltage depending on the voltage that was initially applied to the capacitor to charge it or is it something intrinsic to the capacitor?  DG202/212 can operate between 4.5V to 18V and absolute max is 44V.  Why does section 2.2.1 page 11 says "A complete protection is not possible," is it referring to Figure 2.2a, Figure 2.3, or both?

[Azure]

I am going to add some relay to discharge capacitors before testing similar to figure 2.3 of "TransistorTester with AVR microcontroller and a little more Version 1.13k".  Based on the posts here, it seems to be a good feature to have.  My question are, does it have to be a mechanical relay or can it be a CMOS analog switch like DG202/DG212.  These RDS(on) of 115ohm.  Would they make it impossible to measure low resistance resistors or ESR on inductors and capacitors?  Another thing is would CMOS analog switch itself be susceptible to residual voltage in capacitor?  How high is the residual voltage on a capacitor?  Is the voltage depending on the voltage that was initially applied to the capacitor to charge it or is it something intrinsic to the capacitor?  DG202/212 can operate between 4.5V to 18V and absolute max is 44V.  Why does section 2.2.1 page 11 says "A complete protection is not possible," is it referring to Figure 2.2a, Figure 2.3, or both?

The residual voltage is the charge (voltage) stored across the capacitor after the source charging it is shutdown or removed (when not being powered or disconnected from circuit).  So this can be up to the rated voltage of the capacitor, which for most capacitors over 10V up to typically 1,000V or more are a problem for the tester electronics and any minor protection circuits.

A cmos switch will no work (RDSon and peak voltage) as you suggested might be a problem,

It is best to learn the habit of always discharging a capacitor before testing it with the transistor tester.

[madires]

The "m" in v1.33m is missing some dots on the right side but the "m" in "Component Tester" is displayed correctly.  I am not sure if suggest there are still problems with display or not.

That's most likely a timing or signal issue.

With a MAX232CP based serial port, a LM317 based 5V to 3V DC-DC converter, and a red led where the backlight used to be, it uses 41mA without back light and about 51mA with backlight when it is on but not actually testing anything.  Is this about right?

Seems to be in the ballpark.

I am going to add some relay to discharge capacitors before testing similar to figure 2.3 of "TransistorTester with AVR microcontroller and a little more Version 1.13k".  Based on the posts here, it seems to be a good feature to have.  My question are, does it have to be a mechanical relay or can it be a CMOS analog switch like DG202/DG212.  These RDS(on) of 115ohm.  Would they make it impossible to measure low resistance resistors or ESR on inductors and capacitors?  Another thing is would CMOS analog switch itself be susceptible to residual voltage in capacitor?  How high is the residual voltage on a capacitor?  Is the voltage depending on the voltage that was initially applied to the capacitor to charge it or is it something intrinsic to the capacitor?  DG202/212 can operate between 4.5V to 18V and absolute max is 44V.  Why does section 2.2.1 page 11 says "A complete protection is not possible," is it referring to Figure 2.2a, Figure 2.3, or both?

Since a relay handles more current than a CMOS switch it's better to use the relay. The residual voltage you've mentioned is caused by an effect called "dielectric absorption" (https://en.wikipedia.org/wiki/Dielectric_absorption ), and for an electrolytic cap it can be 10-15% of the original voltage. Input protection was an afterthought for caps and resulted in the two simple circuit options which have some drawbacks. A proper input protection would require a complete redesign which would make the tester more complex and expensive. I'm a friend of discharging caps via a resistor before checking them with the tester.

[+Mikie]

My AY-AT clone doesn't turn itself off if left alone for several minutes and also doesn't turn off if I try to switch it off manually. If I use the rotary encoder to self "switch off", the screen refreshes and the tester doesn't switch off. Does this suggest a fault? I hope to replace the microcontroller with one flashed with 1.33 (m-firmware) whilst it the one it came with is running 1.12.

Yes, there's something wrong with your tester. Please check the power control signal (ATmega328 PD6). It should change to low when powering off.

I've finally completed and boxed mine, and it turns off on long(ish) encoder press while displaying result. Two clicks on the encoder to the left enters the menu. 1.33m flashed.

I have some questions on this:

- First, is this the correct behaviour?
- I've read the ttester pdf and the menu options seem different. Kinda makes sense because different firmware, but what are the equivalents for "m" firmware?
- Shorting the probes for test does something but there is no indication to connect a capacitor. Is this correct?
- I didn't see the ESR result when measuring capacitors, so I tried the "ESR" entry in the menu. It does measure ESR, but the tester never turns off, neither by itself, nor by long press or anything else. I can't get out of it at all. Had to short the battery to turn it off (didn't want to unscrew the box open).
- What are the options "Adjustment", "Save" and "Load"?

[madires]

Yep, it's explained in the README file. And it will give you also the answers to your other questions.

[+Mikie]

Oh, found it. Didn't notice it before because it had no extension...

Thank you for this little toy, it is really useful. And thank you for forcing me to learn some AVR programming, although 10 years late 

For what it's worth, attached is my compiled firmware 1.33m for AY-AT clone with 20MHz crystal and LM4040. The resistors I used are only 1% though. Use it to check if yours works ok, but if you want better accuracy - you'll have to compile your own. The box is 3D printed from a Scrachi design on Thingiverse, only a little adjusted to expose SMD test pads.

[001]

amasing!
It looks like funny oldtimer or cute LEGO bricks 

Sorry for my lasy ass (i didnt read all tread), but tel me please  whats the difference between this project vs original old german project vs china clones?

Thanx