$20 LCR ESR Transistor checker project

[Rober]

Hi Feliciano
One of my testers is using k-firmware and the other m-firmware. Both testers work fine. Now I'm going to test the modified color firmware of Youriy_K - it is a modified k-firmware and it looks like an output of testers with mega644. It looks like very nice and I wanted test it. The only published files are -.hex, -.eep and makefile. To compile I'm need the changed and/or modified files.

[Rober]

Thank you so much, just I found your mail and will test now

[Rober]

All three versions are standing on head.

[Yuriy_K]

All three versions are standing on head.

Try the following option from personal mail

[Rober]

Hi Yuriy_K
Unfortunately all three versions are flipped  horizontally and  for 8 MHz not necessary, because no tester with 8MHz-quarz

[ricktendo]

All three versions are standing on head.

Try the following option from personal mail

Would you be so kind as to share your makefile for the mega328_color_kit with me too?

I would like to compare to mine and see if I can get as many features as possible and have it fit the flash.

[Yuriy_K]

Would you be so kind as to share your makefile for the mega328_color_kit with me too?

I know several different modifications of the ST7735 that require customization of my sources. To increase the number of menu items for the 328 processor, I removed or replaced some of the Karl-Heinz approaches. For this reason, the sources are not transferred. You need to create a working version on the latest version 1.13k yourself and send me the Makefile or try my firmware from page 277. Based on the results of these tests, i can understand what changes need to be made and the source texts for your display.

[ricktendo]

I do not have one for 1.13k but I do have one for 1.45m, these are the modifications I make.

Code: [Select]

cd TransistorTester-source/trunk
sed -i -r 's/\#CFLAGS \+\= \-DNO_FREQ_COUNTER/CFLAGS \+\= \-DNO_FREQ_COUNTER/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DPAGE_MODE/CFLAGS \+\= \-DPAGE_MODE/' ./mega328_color_kit/Makefile
sed -i -r 's/CFLAGS \+\= \-DFONT_8X12thin/\#CFLAGS \+\= \-DFONT_8X12thin/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DFONT_8X16thin/CFLAGS \+\= \-DFONT_8X16thin/' ./mega328_color_kit/Makefile
sed -i -r 's/CFLAGS \+\= \-DBIG_TP/\#CFLAGS \+\= \-DBIG_TP/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DINVERSE_TP/CFLAGS \+\= \-DINVERSE_TP/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DNO_TEST_T1_T7/CFLAGS \+\= \-DNO_TEST_T1_T7/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DWITH_XTAL/CFLAGS \+\= \-DWITH_XTAL/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DWITH_UJT/CFLAGS \+\= \-DWITH_UJT/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DWITH_PUT/CFLAGS \+\= \-DWITH_PUT/' ./mega328_color_kit/Makefile
sed -i -r 's/OP_MHZ \= 8/OP_MHZ \= 16/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DNO_LONG_PINLAYOUT/CFLAGS \+\= \-DNO_LONG_PINLAYOUT/' ./mega328_color_kit/Makefile
sed -i -r 's/CFLAGS \+\= \-DANZ_MESS\=25/CFLAGS \+\= \-DANZ_MESS\=30/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DEXT_NUMERATOR\=10/CFLAGS \+\= \-DEXT_NUMERATOR\=10/' ./mega328_color_kit/Makefile
sed -i -r 's/\#CFLAGS \+\= \-DEXT_DENOMINATOR\=1/CFLAGS \+\= \-DEXT_DENOMINATOR\=1/' ./mega328_color_kit/Makefile
Main thing I want to do is increase the menu count but from downloading your makefile from 277 I think I know why mine is shorter

[Yuriy_K]

I do not have one for 1.13k but I do have one for 1.45m, these are the modifications I make.

Here is the firmware that Rober tested today.
According to him, everything works fine.
Only I do not know what changes he made to the circuit in his TTtester, except for those that are published above.

[Rober]

@ Yuriy_K.
5 years plus (your last question) . A little mistake I have seen  in both versions. When you test a FET, at the symbol are gate, drain and source labeled - but at the end of text, only gate is labeled. For drain and source stand a "?". It doesn't have to corrected - only for your infirmation.

[madires]

The '?' or 'x' for a JFET's drain/source pins indicates that drain and source are symmetrical. It's not a bug, it's a feature.

[tsmith35]

Oh, that's really interesting. Myself, I never thought about such a thing since I'm so used to silicon components having a polarity of sorts. But I see where one textbook says, "Although the JFET is a symmetrical device whose source and drain can be interchanged, it is useful in circuit design to designate one of these two terminals as source and the other as drain. The circuit symbol achieves this designation by placing the gate closer to the source than to the drain."

From a practical point of view, I can see where designating source and drain may be necessary for circuit design since component orientation could be an issue with three connection points.

[madires]

There is a way to tell source and drain for a symmetrical JFET apart, but the Transistortester can't do that. The gate-source capacitance is much lower than the gate-drain capacitance. And not every JFET is symmetrical.

[magic]

Speaking of capacitance, how accurate is that pF measurement and junction capacitance reporting? Does the measurement technique provide stable/reproducible readings near zero or does it have visible zero drift like the LC100-A? Does it show any reading for less than 1pF?

Is it true that this function still only exists in K firmware?

[madires]

Speaking of capacitance, how accurate is that pF measurement and junction capacitance reporting? Does the measurement technique provide stable/reproducible readings near zero or does it have visible zero drift like the LC100-A?

The lower limit for the normal capacitance measurement (also used for diodes and so on) is:
- k-firmware: 25 pF
- m-firmware: 5 pF

The meaurement can be fairly stable with a good tester (stable Vcc).

Does it show any reading for less than 1pF? Is it true that this function still only exists in K firmware?

The k-firmware offers an optional software feature called SamplingADC which allows a special measurement of low capacitance values down to a few fF. For the m-firmware there's an LC meter hardware option with a similar measurement range down to a few fF.

[Yuriy_K]

Does it show any reading for less than 1pF?
Is it true that this function still only exists in K firmware?

An example of measurement on 1.13k firmware and comparison of measurements.

[magic]

Thanks, this doesn't look too terrible. Am I to understand that 0.01pF stability is realistic with this meter?
Those cheap resonant LC meters easily drift more, even after warmup. Furthermore, mine clamps negative readings to zero, making it harder to correct for drift.

Yes, this is the mode I meant. Reading the 1.13k manual gave me impression that it's this mode which is used for junction capacitance (it even mentions testing at 0V and -5V bias), I didn't know that normal mode can be used too.

[dark_hawk]

Hi,

I have an LCR-T4 tester I've bought from China ages ago and just now found out it has an original better firmware and a 300 page thread on eevblog.
I swapped the 8M crystal to a 16M and upgraded the firmware to 1.13k found on that yandex disk. And things seem better in all aspects.

I'm thinking about getting Hiland M644 or the T7 with that fancy color screen.
I have a few questions that I tried to search answers on the thread but couldn't find any.
1- What would be the extra benefits I get If I switch to one of those newer versions?
2- Which one should I get? The Hiland M644 or the T7 or another model?
3- RDSon readings of my T4 seems off. I think maybe because it's being tested with 5V instead of the 10/12 volts I usually use to get RDS.
This is made worse with mosfets that have lower gate capacitances like the IRF740B with reads 4 ohms RDS while the value using a 12v/1.6A load is around 0.5-0.6 ohm.
Are the newer versions using higher voltages when testing for RDS since it can test zeners up to 35? Or is it the still the same 5 volts.

[Feliciano]

These gadgets works with 5V for the microntroller, so your Rds measurements are with Vgs 5V. (And the ctesters able to test zeners use a divider to measure/calculate Vz).

64KB versus 32KB allows more functions to be enabled on your firmware at once, and some of those requires extra hardware (like the hardware extensions on the Hiland M644, or you can DIY).

[dark_hawk]

Could you please elaborate as to the extra functionality I will get with a 64KB chip/Extra hardware?

[Feliciano]

I mean, more things enabled on your own particular firmware build.

There are plenty of features you can choose from. You can look at the config files of the m-firmware or the makefile of the k-firmware for grasping how many (similar or different) features each one has to offer.

Every function uses memory, some more, some less. A larger font user more flash, and a color screen requires more code than a monochrome screen (and deppends on the particular controller), etc.

Let's take one example: to show the E12 or E24 or E96 standard values next to a resistor measurements, so you can say whether the particular piece you're measuring is within range or not. Of course E96 with colour codes will use more memory than E12 only text. You can also choose to show standard values for capacitors.

Another example: there is an option to monitor DUT placed between Terminals 1 and 3. You can monitor L, or C, or R, or R and L, or R and C, or R, L, and C. The memory requirements are different.

So maybe somebody wants to enable every single feature the m-firmware or k-firmware has. I don't whether 64KB of flash will suffice for that, neither I know whether the RAM would it be enough too. Also keep in mind some features requires extra hardware.

Bottomline: me and others look at the source code, check which things like better than others for our job/hobby, make the file, check whether fits our unit or not, until we find the best tailored binaries for our likings.

[faustino]

Hi everybody
Right now where can i find the Hiland M644 online? In Banggood it says its in backlog.
Thank you

[Yuriy_K]

Measurement example of a MOSFET transistor rarely encountered depletion type

[Feliciano]

Does anyone knows the maker of the attached Zener? (The 6P2's logo resembles General Semicondutor, but I don't think it is).
I would ask the same for the V05, but I don't notice a logo 

[indman]

SML4735 - ZENER 6.2V