$20 LCR ESR Transistor checker project

[madires]

A reminder on the call for localized texts based on ISO 8859-2. If you like to support the project by translating the firmware texts to a language covered by ISO 8859-2 please see https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg2913368/#msg2913368 for details.

[powersola]

Hi all,

maybe I can ask for your help.

I tried to compile the 1.38m firmware for my AY-AT clone (marked Geekteches GM328A).
The file I get compiling under linux with avr-gcc 9.2.0 is "Program:   37020 bytes (113.0% Full)"
I see it's a toolchain problem, and tried even with a windows VM via WinAVR and the toolchains suggested somewhere in the thread extracted in the WinAVR folder (3.4.5.30, 3.4.4.24, for example) but the less I can get is somewhere 105.x%.

May I ask if somebody of you that have 5 minutes of spare time can try, under a known previously working environment, to compile my files that I attach? Or I messed something in config, or my environment is not set up in the right way.

Thank you,
Powersola

[madires]

With avr-gcc 5.4.0 and your config.h I get 105% too. Try to disable some feature you don't need, e.g. SW_DS18B20, SW_DHTXX and SW_UJT.

[powersola]

Thank you for the feedback!

I'm going to try again disabling something and I'll try again.
Good work, thanks so much.

Edit: Tried removing DS probes, onewire and UJT and reached 98.something %. Success!

I get blank screen on the tester, but it's the first compilation of this firmware from scratch so I think I picked some wrong choice in config files... I put on Atmega the precompiled 1.37m i found here some pages back that works, and I'll retry again someday with compilation.

EditEdit: Got 1.38 running, had just to fix some display parameters that were wrong. Great!

Powersola

[tooki]

Hey everyone!

To those with an AY-AT or clone (i.e. color TFT version), could somebody measure how much current it draws in operation?

Thank you!


(Background: At my work (where I am interning before starting an apprenticeship as an electronics technician in the fall), one project we are working on (to be mechanical construction practice for the electronics apprentices, and design practice for the design engineering apprentices) is to use off-the-shelf AY-AT kits (but populated with the correct voltage reference and 0.1% resistors) and build a custom housing around it, with added banana jacks and SMT pad area. Since this model supports square wave output, and thus might be used for far longer periods of time than when testing components, I proposed that we power it from 6 AA batteries instead of a 9V block, since the AA's are far more cost-effective on a Wh basis. But since that also makes it bulkier, I figured I'd try and find out how much power it actually consumes.)

If anyone has suggestions or feedback on this plan, feel free to chime in! I'm still working my way through this whole thread… 

[madires]

I don't have any AY-AT clone but my guess would be between 20 and 30mA on average.

[indman]

Yes, here is a picture of the DTU-1701, you can see J6 goes to the MCU which is different to the Bside ESR02.
I think these are a cheap clone to be honest, it was only £16.
DTU-1701

I bought himself DTU-1701. I painted a complete schematic diagram
https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Schematic%20diagram
Red color means changes made to the scheme:
1. Replace the frail  7150 on the LP2950-5.0(better MCP1702-5.0)
2. Replace  TL431 on LM4040A (R2A)
3. Added ceramics capacitors 10mkF on LP2950 input  and Vcc processor(pins 4-6)
4. Control divider resistors R21-R22 for voltage replaced 47kOm, control point moved to the collector Q2 (optional).
5. Capacitor C10 deleted - normal work without it.
6. Replace quartz at 16MHz (optional).
Revision board 3.1 - changes in comparison with earlier BSideESR02 virtually none except for a correct wiring ICSP connections-now all your contacts conveniently separated by the connector. Photo board to upgrade and after it will be lower.

Firmware:
New English and Russian k-firmware revision 813 for this clone  https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Firmware/Karl-Heinz%20K%C3%BCbbeler%20

[tooki]

I don't have any AY-AT clone but my guess would be between 20 and 30mA on average.

Well that's the thing, we want to know actual measured current draw, not guesses. (I have an LCR-T4, but since the AY-AT uses a TFT, their current draw will not be comparable.)

[blurpy]

To those with an AY-AT or clone (i.e. color TFT version), could somebody measure how much current it draws in operation?

My AY-AT uses about 27mA during component testing. With short spikes up to 30mA, and about 23mA during frequency generation.

[madires]

Now I can claim that it was an educated guess.

[indman]

Hey everyone!
To those with an AY-AT or clone (i.e. color TFT version), could somebody measure how much current it draws in operation?

Those with a dim screen on the AY-AT
In the power supply circuit of the screen backlite there are as many as two current-heating resistors in series 220 ohms and the amount 440 ohms
One stands on the red board (I replaced it with a short jumper).
The second one is on a board with a display.
For normal visibility of the screen it was necessary to change it to 100 ohm resistor.
The consumption of current on the battery became 47mA screen is perfectly visible.

[GnobarEl]

Hello all,

I'm thinking about buying one of those transistors tests. Before I made the purchase I decided to make a quick search on google, just out of curiosity and I find out this huge community. It blows my mind all the discussion related with this simple device. I thought they were all the same and I now know there are a lot of form factors, firmware...

Well, the point is, I tried to read it all but in the end I couldn't get a proper reply. Can some help me to choose the most valuable device?
All the suggestions are welcome!

Please, don't reply to me, "that has been answer before." I read it all but I couldn't decide.

Thanks for your help.

[madires]

If you prefer a basic model choose an AY-AT clone. For Zener check and extended frequency counter get a Hiland M664. For a nice case, LiPo battery, IR receiver and Zener check a TC-1 (or another model with the same case). There's no perfect tester model unless you build one yourself.

[GnobarEl]

Hi, thanks for your advice! That really helped me a lot. Its easier for me to study the differences between them and decide for one. In the end, I think I will be happy with any choice.

One more question. Can I use the "oficial" firmware in any of those products? Or should I have something in attention when I buy it to be able to update the firmware in the future?

Thanks!

[madires]

Basically the OSHW firmwares (k & m) should run on all testers. One special case is the TC-1 (and family) with a management MCU. But we have an alternative firmware for that MCU to make the tester compatible, and also a small circuit to replace the MCU.

[GnobarEl]

That's great!

Thanks for your great support. I'm sure that your clear answer will be helpful for other as well.

Thanks!

Best Regards,

[madires]

We all need some good news. The new m-firmware version is here!

v1.39m:
- Added Polish texts based on ISO 8859-2 (Thanks to Jacon).
- Fixed issue with preprocessor macro in SPI.c and syntax error in ADS7843.c  (reported by Bohu).
- Added option to store firmware data to Flash instead of EEPROM (DATA_FLASH, suggested by Vitaliy).
- Renamed cyrillic fonts to "win1251" and replaced Czech fonts with ISO8859-2 fonts provided by Bohu.
- Display of font for test purposes (SW_FONT_TEST).
- OneWire scan tool for listing ROM codes of connected devices (SW_ONEWIRE_SCAN).
- Option for probe-pair specific resistance offset (R_MULTIOFFSET, suggested by Vitaliy).

Available at https://github.com/madires/Transistortester-Warehouse/tree/master/Firmware/m-firmware.

[tooki]

To those with an AY-AT or clone (i.e. color TFT version), could somebody measure how much current it draws in operation?

My AY-AT uses about 27mA during component testing. With short spikes up to 30mA, and about 23mA during frequency generation.

Awesome, thank you so much!

Hey everyone!
To those with an AY-AT or clone (i.e. color TFT version), could somebody measure how much current it draws in operation?

Those with a dim screen on the AY-AT
In the power supply circuit of the screen backlite there are as many as two current-heating resistors in series 220 ohms and the amount 440 ohms
One stands on the red board (I replaced it with a short jumper).
The second one is on a board with a display.
For normal visibility of the screen it was necessary to change it to 100 ohm resistor.
The consumption of current on the battery became 47mA screen is perfectly visible.

Great info, thank you as well!


Much appreciated, guys!

[Pukker]

Yes, here is a picture of the DTU-1701, you can see J6 goes to the MCU which is different to the Bside ESR02.
I think these are a cheap clone to be honest, it was only £16.
DTU-1701

I bought himself DTU-1701. I painted a complete schematic diagram
https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Schematic%20diagram
Red color means changes made to the scheme:
1. Replace the frail  7150 on the LP2950-5.0(better MCP1702-5.0)
2. Replace  TL431 on LM4040A (R2A)
3. Added ceramics capacitors 10mkF on LP2950 input  and Vcc processor(pins 4-6)
4. Control divider resistors R21-R22 for voltage replaced 47kOm, control point moved to the collector Q2. (optional).
5. Capacitors C10 deleted - normal work without it.
6. Replace quartz at 16MHz (optional).
Audit board 3.1 - changes in comparison with earlier BSideESR02 virtually none except for a correct wiring ICSP connections-now all your contacts conveniently separated by the connector. Photo board to rework and after it will be lower.

Firmware:
New English and Russian k-firmware revision 813 for this clone  https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Firmware/Karl-Heinz%20K%C3%BCbbeler%20

Thanks Indman for sharing the upgrading modifications.
I ordered 100 pcs 470k and 680 smd resistors, format 1206, sorted and paired them with my Fluke and replaced the original ones
which had a lot of deviation. After selfcheck / calibration the measurement was improved.
Mentioning that calibration done with battery power, not the same is when working with an adapter.
Calibrate with the powersource wich you want to use.
Thanks for compiling the software. In version 8.13 it shows DC_Power_txt or something, when powered with an adapter,
translation not correct. Battery Power is indicated OK.

[indman]

In version 8.13 it shows DC_Power_txt or something, when powered with an adapter,
translation not correct. Battery Power is indicated OK.

Yes, this is my mistake when compiling English firmware. I fixed, now the DC Power Mode is displayed correctly. 
https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Firmware/Karl-Heinz%20K%C3%BCbbeler%20/English

[Pukker]

In version 8.13 it shows DC_Power_txt or something, when powered with an adapter,
translation not correct. Battery Power is indicated OK.

Yes, this is my mistake when compiling English firmware. I fixed, now the DC Power Mode is displayed correctly. 
https://yadi.sk/d/yW8xa5NJgUo5z/BSideESR02(DTU-1701)/Firmware/Karl-Heinz%20K%C3%BCbbeler%20/English

Wow, that is superspeed service. Thanks.

Edit:
Tested, works fine.

[Pukker]

Meter "Not calibrated"

When I do selftest with my DTU-1701,
working OK. I calibrate with use of short testleads,
because in real life, mostly I use the testleads.
0.1ohm resistor reads 0.11.
Very good, resistor measurement in that range is not easy.
Even an good multimeter can't do that.

Strange, when I use the same resistor without testlead,
but directly on the clamps on the meter, it says 0.0ohm.
Ok, missing the testleads, that is OK.

But, the meter says directly "Not Calibrated".
Looks like the meter sees negative resistance and "resets calibration"
Tested this over again and calibration lost again.
Now I know, no problem, but want to tell this strange behaviour.

I hope this is not mentioned before, I follow this forum for long time,
but can't remember what is written in 247 pages.

[Gandalf_Sr]

So I just ordered some LCR Research Pro 1 Plus tweezers for $300 but I own several of these $20 testers.

If someone (maybe me) was to turn the $20 tester circuit into a small PCB with a small OLED display, couldn't the Madires code be modified to run some pretty good tweezers?

I searched this thread for "tweezers" but the only hits I got were from 2017.

[indman]

Meter "Not calibrated"
But, the meter says directly "Not Calibrated".
Looks like the meter sees negative resistance and "resets calibration"
Tested this over again and calibration lost again.
Now I know, no problem, but want to tell this strange behaviour.

That's right, Karl-Heinz mentions this in detail in his manual on page 116 in section 5.5 SelfTest Function:
"Before the test steps begin, the zero resistance of the connected probes
is determined for all three combinations (T1:T3, T2:T3 and T1:T2). This zero resistances will be
subtracted for the future ESR and resistance measurements below 10Ω. Only zero resistance values
below 0.90Ω are accepted, because this correction values are not uses for measurement of resistor
values above 10Ω. If you use cables for measurement, you should only take cables with low resistance
values. If the later measured resistance results fall below the particular zero resistance for more than
0.2Ω, the tester will be resetted to ”uncalibrated”.
Madires in the m-firmware provided this option and made it possible to save and load 2 different calibration profiles.
For internal and external test contacts, which is very convenient to use. There is no need to carry out a new calibration every time.
   For k-firmware there is no such possibility yet.

If someone (maybe me) was to turn the $20 tester circuit into a small PCB with a small OLED display, couldn't the Madires code be modified to run some pretty good tweezers?

Would you like to have such a tester device as in this video, I understand correctly?

[vivaT]

i made these tweezers for 2 pin devices, very light and nice for smd stuff.
the font is a little small for this type of application, ideally it would have two kinds of font to improve the readability but thats a bit beyond my coding level.
If theres interested ill post pcb, bom ect