Author Topic: $20 LCR ESR Transistor checker project (Read 3452894 times)

tom666 · « **Reply #975 on:** May 08, 2015, 01:51:26 pm »

Extended fuse F9 = 01

because some numerical values containing undefined bits.

caius · « **Reply #976 on:** May 08, 2015, 02:25:10 pm »

Quote from: tom666 on May 08, 2015, 02:20:09 pm

Quote from: caius
I can't see how to program the EEPROM of the MCU with my TL866CS.
Code memory = TransistorTester.hex
Data memory = TransistorTester.eep
Config = Fuses

OK, thanks.
A question : if ATMEGA168 EEPROM size is 512 bytes, how could I program 2KB of data of the .EPP file?
Anyway my programmer doesn't recognize MCU, I double-checked ICSP diagram and it's right.There are two GND points on tester ICSP header but I connected only one to the programmer, I don't think it's due this.

tom666 · « **Reply #977 on:** May 08, 2015, 02:37:08 pm »

File contents is in hex format.
Just plug only one GND. Check the presence Vcc (5V)while programming.

caius · « **Reply #978 on:** May 08, 2015, 03:00:38 pm »

Quote from: tom666 on May 08, 2015, 02:37:08 pm

File contents is in hex format.
Just plug only one GND. Check the presence Vcc (5V)while programming.

I have not yet tried to program it since programmer doesn't recognize ATmega168 even reading it (ID error).There is something wrong for sure, I suspect the programmer itself, I'm powering the MCU through it (it can deliver up to 120mA on the ICSP VCC pin).Part name on chip is scratched off but it shoud be a ATmega168.

caius · « **Reply #979 on:** May 08, 2015, 07:04:08 pm »

OK, I succesfully programmed the ATmega168 on my transistor tester with Arduino as ISP and AVRdude, all went fine.
I noticed that the 1.09k firmware adds some new features like reverse leakage current (ir) measurements and Vloss measurement, are these reliable?
Besides,it seems that with this firmware valid test points for capacitors are only TP1 and TP3.If you use TP1 and TP2 you get "no unknown or damaged part" message, if you use TP2 and TP3 you get wrong measurements.Do you confirm this?
Thanks again for you help.

P.S.
With this firmware ESR measurement is more accurate but there is always a difference of .7 Ohms compared to my Bob Parker ESR meter.
Is there anything I can do on the hardware to improve it?

tom666 · « **Reply #980 on:** May 09, 2015, 07:19:05 am »

I think it will be a problem with the connection TP2 on the MCU port.
Check that the value of R2 to R7 are correctly (680, 470k) and their connection to the TP1 to TP3 (mainly TP2).

caius · « **Reply #981 on:** May 09, 2015, 08:50:46 am »

Quote from: tom666 on May 09, 2015, 07:19:05 am

I think it will be a problem with the connection TP2 on the MCU port.
Check that the value of R2 to R7 are correctly (680, 470k) and their connection to the TP1 to TP3 (mainly TP2).

I check connection and resistors values and they are OK.I think it's something releted to this 1.09K firmware or I really need to modify hardware.I look at this manual:

http://radioskot.ru/_fr/6/ttester_eng105k.pdf

Can the hardware changes be made also on my PCB revision or are they already inplemented?Layout is quite different as well as silkscreening of components.

tom666 · « **Reply #982 on:** May 09, 2015, 11:48:32 am »

Correction of the Chinese version testers (Booster edition 2.x) you will find since version 1.06k of user guide (attached pdf file contained a specific section of this manual).

The only one difference is that your version does not contain 8MHz crystal. Then it should be added blocking capacitor 100nF directly to the MCU pins. Other error is (this does not affect the functionality) the use of zener diode D1 instead of a source of precision 2.5V reference voltage. If you would like to use the new features it is appropriate to replace the MCU for ATmega328.

Despite these minor deficiencies, I think the tester should work after upgrade firmware without hardware modifications.

caius · « **Reply #983 on:** May 09, 2015, 03:02:46 pm »

Quote from: tom666 on May 09, 2015, 11:48:32 am

The only one difference is that your version does not contain 8MHz crystal. Then it should be added blocking capacitor 100nF directly to the MCU pins. Other error is (this does not affect the functionality) the use of zener diode D1 instead of a source of precision 2.5V reference voltage. If you would like to use the new features it is appropriate to replace the MCU for ATmega328.

Ok, I'll try an ATmega328.I presume that with this more capable MCU I can try latest firmware, which one in particular?.Are the fuses values the same you suggested me for 1.09k upgrade??
Anyway I could measure (out-of-circuit) on my revision Booster Edition 2.3 PCB two 160nF capacitors instead of 100nF ones silkscreened on Booster 2.2.I will replace these with correct values as well as put a 100nF by-pass capacitor across VCC and GND of MCU and replace diode D1 with a precise shunt voltage reference.

P.S.
A quick question:
Are the GM328 and T3/t4 component testers based on same schematics of mine?
Thank you again for you help.

tom666 · « **Reply #984 on:** May 09, 2015, 05:03:59 pm »

For MCU M328 you can use the latest version of the software. For example, the final v1.11k or development version 1.12k (author Karl-Heinz Kübbeler). You can also use the alternative version of the software by Markus Reschke (current v1.17m).

Everything is available in the SVN repository:
https://www.mikrocontroller.net/svnbrowser/transistortester/

Setting fuses for M328 (applies to all versions of software):
RC operation, CLK 8MHz
lfuse:0xe2 hfuse:0xd9 efuse:0xfc (0x04)

8MHz crystal - full swing crystal
lfuse:0xf7 hfuse:0xd9 efuse:0xfc (0x04)

8MHz crystal - low power
lfuse:0xff hfuse:0xd9 efuse:0xfc (0x04)

"GM328" is identical with standard schematics diagram, but the "T4" have different mapping of MCU ports with the G-LCD - connection as "strip grid board" (using of MCU ports PD1 to PD4).

Clone of GM328 (by Dietrich L.):
https://www.mikrocontroller.net/topic/345717#3856301

Schematics diagram of LCR-T3 & LCR-T4 (by Roland Elmiger):
https://www.mikrocontroller.net/topic/248078?page=6#3784106

kubi48 · « **Reply #985 on:** May 09, 2015, 05:23:36 pm »

The version 1.11k and also the developer version 1.12k can also be used with a mega168. Of course not all features like support of graphical display or the menu functions are available for the mega168. For 1.11k and 1.12k version shorter current pulses are used for the ESR measurement. So the ESR can be measured for lower capacity values like 100nF. The ESR result of the TransistorTester match to a 10kHz measurment, not to a 100kHz result, which is often used. If you try to compile the source for a mega168 target, you should use a new avr-gcc revision like 4.8.2 , because the code optimizing of newer versions is better. Otherwise the code may not match to the 16k flash memory.
If you wish to use the full selftest with a mega168, you must deselect one of two hFE measurement approches first (NO_COMMON_COLLECTOR_HFE or NO_COMMON_EMITTER_HFE).

caius · « **Reply #986 on:** May 09, 2015, 07:31:21 pm »

Quote from: kubi48 on May 09, 2015, 05:23:36 pm

The version 1.11k and also the developer version 1.12k can also be used with a mega168. Of course not all features like support of graphical display or the menu functions are available for the mega168. For 1.11k and 1.12k version shorter current pulses are used for the ESR measurement. So the ESR can be measured for lower capacity values like 100nF. The ESR result of the TransistorTester match to a 10kHz measurment, not to a 100kHz result, which is often used. If you try to compile the source for a mega168 target, you should use a new avr-gcc revision like 4.8.2 , because the code optimizing of newer versions is better. Otherwise the code may not match to the 16k flash memory

OK, thanks for explanation

Quote

If you wish to use the full selftest with a mega168, you must deselect one of two hFE measurement approches first (NO_COMMON_COLLECTOR_HFE or NO_COMMON_EMITTER_HFE).

I just noticed that self-test in not implemented in V1.09K like in the old V1.05K

caius · « **Reply #987 on:** May 09, 2015, 09:35:13 pm »

Quote from: tom666 on May 09, 2015, 05:03:59 pm

For MCU M328 you can use the latest version of the software. For example, the final v1.11k or development version 1.12k (author Karl-Heinz Kübbeler). You can also use the alternative version of the software by Markus Reschke (current v1.17m).

Everything is available in the SVN repository:
https://www.mikrocontroller.net/svnbrowser/transistortester/

Setting fuses for M328 (applies to all versions of software):
RC operation, CLK 8MHz
lfuse:0xe2 hfuse:0xd9 efuse:0xfc (0x04)

8MHz crystal - full swing crystal
lfuse:0xf7 hfuse:0xd9 efuse:0xfc (0x04)

8MHz crystal - low power
lfuse:0xff hfuse:0xd9 efuse:0xfc (0x04)

"GM328" is identical with standard schematics diagram, but the "T4" have different mapping of MCU ports with the G-LCD - connection as "strip grid board" (using of MCU ports PD1 to PD4).

Clone of GM328 (by Dietrich L.):
https://www.mikrocontroller.net/topic/345717#3856301

Schematics diagram of LCR-T3 & LCR-T4 (by Roland Elmiger):
https://www.mikrocontroller.net/topic/248078?page=6#3784106

Thank you for your exhustive explanation, I can say you are a real expert of these component testers

According you, which could be the next step of hardware/software development?

tom666 · « **Reply #988 on:** May 09, 2015, 10:59:11 pm »

Mainly many thanks to deserve Mr. Markus Frejek, Mr. Karl-Heinz Kübbeler (kubi48) and Mr. Markus Reschke (madires).
I only collect the information available on the Internet.

zomi · « **Reply #989 on:** May 10, 2015, 05:24:36 pm »

Tom, have one of the fish8840 and trying to change the lcd..
Broke the original one ..

Added the trimmer..
and changed the resistor..
But I just get a display full of blocks..

Do I need to reprogram it to work properly ?

try the mod you showed here.

https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg644091/#msg644091

tom666 · « **Reply #990 on:** May 11, 2015, 07:28:38 am »

Quote from: zomi

... have one of the fish8840 and trying to change the lcd..
Added the trimmer..
and changed the resistor..
But I just get a display full of blocks..
Do I need to reprogram it to work properly?

Yes, you need to use firmware designed for character 2x16 LCD:
https://www.mikrocontroller.net/svnbrowser/transistortester/Software/trunk/mega328/
Fuses:
lfuse:0xf7 hfuse:0xd9 efuse:0xfc (0x04)

Note:
For battery voltage measurement (divider R8, R15) you can modify in Makefile this settings to:
CFLAGS += -DBAT_NUMERATOR=66
CFLAGS += -DBAT_DENOMINATOR=33
In principle, however, you can use the firmware from that folder without editing Makefile - value of Vbat will not be OK.

Anks · « **Reply #991 on:** May 14, 2015, 07:17:45 am »

Could anyone clarify if this tester can do ESR in circuit and if there are any modifications or specific firmware that has to be made/used.

The specific tester in question is this http://www.ebay.co.uk/itm/Mega328-Transistor-Tester-Diode-Triode-Capacitance-ESR-Meter-MOS-PNP-NPN-LCR-T3-/121595483585?pt=LH_DefaultDomain_3&hash=item1c4fa7d5c1

Any help much appreciated.

tom666 · « **Reply #992 on:** May 14, 2015, 08:55:14 am »

It is necessary to make a firmware upgrade:
https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg659867/#msg659867
or
https://www.mikrocontroller.net/topic/248078?goto=4120737#4105488

Use the actual version from the SVN archive designed for this type of tester (LCR-T3):
https://www.mikrocontroller.net/svnbrowser/transistortester/Software/trunk/mega328_T3_T4_st7565/
Fuses:
lfuse:0xf7 hfuse:0xd9 efuse:0xfc (0x04)

I recommend to read the manual:
https://www.mikrocontroller.net/svnbrowser/transistortester/Doku/trunk/pdftex/english/

nassosxlvbros · « **Reply #993 on:** May 15, 2015, 07:42:59 am »

Hi to all from Athens!
After reading a lot for this tester I decide to get one!

So I bought this tester,4 days ago,plug in a 9V battery and worked fine....Suddently after a dozen of measures refuses to power on (only the green LCD light goes on when I press test button)...I left it,check battery,connections,bad solderings etc,but all seems OK.,then I put it away....Next day worked well for some time, but after 4-5 measurements stops again....
What is going on

nassosxlvbros · « **Reply #994 on:** May 15, 2015, 07:52:01 am »

coil measurement....and after that,tester is dead....

firepower · « **Reply #995 on:** May 15, 2015, 08:10:10 am »

Quote from: firepower on April 27, 2015, 05:06:19 am

I found this one GM328A firmware V1.12 with input port protection and 0.1% resistors. Menu has Transistor, Frequency, f-Generator,10-bit PWM, C=ESR@TP1:3, rotary encoder, Self Test, Contrast, Show Data, Switch Off

U3 (V05) & D1 (KK 1450) may be similar to the protect circuit attached. Voltage regulators, 5V (AMS1117-5.0) & 3V3 (AMS1117-3.3) for LCD, U4 voltage reference TL431

got my tester today and have updated my original post with more details above. Its Hardware PCB ver 1.11

One thing i have noticed is the red LED seems a bit point less, it always on and only goes off when button is pressed. Is there any other purpose for it?
Think I may just remove it if it has no other use, Back light is bright enough to let me know its on.

Gyro · « **Reply #996 on:** May 15, 2015, 09:16:43 am »

Unfortunately the Red LED is part of the power switching circuit - it's in the collector circuit of T1 which holds the power on once the button is released. Check the schematic in the official AVR ttester manual.

If you simply remove the LED then the tester will turn off as soon as you release the button. Shorting it should work, but the current path is still there.

The T3 and T4 testers have been a bit more economical by connecting the LCD backlight LED in this circuit position (though I think they've lowered R7 a bit to give it more current) instead of simply powering it off the 5V rail.

firepower · « **Reply #997 on:** May 15, 2015, 09:41:31 am »

Thanks for the info, I checked the circuit, My R7 is 2k2 and circuit recommends 3k3 so that should help reduce current and dim the LED.

paulie · « **Reply #998 on:** May 15, 2015, 09:54:16 am »

Quote from: nassosxlvbros on May 15, 2015, 07:42:59 am

Next day worked well for some time, but after 4-5 measurements stops again.... What is going on

Did you discharge that big cap before testing again? Although it hasn't happened with me this is said to damage the unit. Another reason to avoid the slick but more expensive units with soldered in MCU. Socket versions are getting hard to find now but ordered with a few spare Mega328 chips can be a life saver.

nassosxlvbros · « **Reply #999 on:** May 15, 2015, 10:38:39 am »

Quote from: paulie on May 15, 2015, 09:54:16 am

Quote from: nassosxlvbros on May 15, 2015, 07:42:59 am
Next day worked well for some time, but after 4-5 measurements stops again.... What is going on

Did you discharge that big cap before testing again? Although it hasn't happened with me this is said to damage the unit. Another reason to avoid the slick but more expensive units with soldered in MCU. Socket versions are getting hard to find now but ordered with a few spare Mega328 chips can be a life saver.

Of course I discharge every cap before measure.Tester does this even when I test resistors.....any ideas?


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: $20 LCR ESR Transistor checker project (Read 3452894 times)

Share me