$20 LCR ESR Transistor checker project

[indman]

And now, Chinese friends have launched an updated version of DSO-TC3 on the market, in which they added a signal generator and also taught how to measure the ESR of capacitors.

This is how the new DSO-TC3 measures the ESR of capacitors. While the Chinese friends have a lot of work to achieve similar results on the classic AVR. In the meantime, the ESR of this device shows the weather on Mars and not the actual equivalent resistance values.
However, like other similar testers that are not built on the original ATMEL.
In the photo, for comparison, the measurement of the same capacitor with the mod m-firmware 1.48m.

[Yuriy_K]

This is how the new DSO-TC3 measures the ESR of capacitors.

The photo does not show which capacitor is being measured. Such readings are typical for ceramic capacitors. Here are the measurements on the firmwares of both authors.

[indman]

The photo does not show which capacitor is being measured.

The photo shows the measurement of an electrolytic capacitor 4.7uF(not ceramic). Is it really not clear that such a electrolytic capacitor of any company cannot have ESR = 0.11Ohm ?
Here are some more comparison photos for example. Transistor KП103 DSO-TC3 defines as a resistor.

[Fuzzy Star]

The photo shows the measurement of an electrolytic capacitor 4.7uF(not ceramic). Is it really not clear that such a capacitor of any company cannot have ESR = 0.11Ohm ?

I have some 10uF/16V Sanyo Os-con polymer electrolytics that measure ESR=0.11 Ohm
https://en.wikipedia.org/wiki/Polymer_capacitor

[indman]

Fuzzy Star,and the photographs that I provided do show polymer capacitors? Or does DSO-TC3 define them as polymer?

[Fuzzy Star]

Fuzzy Star,and the photographs that I provided do show polymer capacitors? Or does DSO-TC3 define them as polymer?

"electrolytic capacitor 4.7uF(not ceramic)... such a capacitor of any company"
Polymer capacitors are electrolytics, too and they don't look different. We can't be sure from your photo what type of electrolytic is it.

does DSO-TC3 define them as polymer?

Do these testers can detect and display capacitor type?

[indman]

Polymer capacitors are electrolytics, too and they don't look different. We can't be sure from your photo what type of electrolytic is it.

I assure you that the photo measured is an ordinary electrolytic capacitor, not a polymer one. If you can't tell the difference between an ordinary electrolyte and a polymer one, I'm sorry.

[oitar]

Ok, I bricked it, took a week
Maybe this will help someone avoid doing the same:
The Windows installation of WinAVR for some reason installs an empty folder for Programmers Notepad. So, used WinAVR Portable. Didn't [apparently] need to install the extra toolchain, but had to change the msys-1.0 file inside C:\WinAVR_Portable_V14_1\WinAVR-20100110\utils\bin to get everything to compile without errors.
The 1.44m firmware compiled fine with minimal set of features, the 1.48m was giving an error, so when I compared the 1.44 and 1.48 I saw both #define SPI_BITBANG, and #define SPI_HARDWARE were commented off. Enabling #define SPI_BITBANG made it compile without error..

That's when I decided to try SPI_HARDWARE, and also manually changed the low fuse from 0xff to 0xf7 using Avrdudess(I saw inside the Makefile: LFUSE_CRYSTAL = -U lfuse:w:0xf7:m, since I have a mega324pa, running 16 MHz external crystal)... //famous last words...

Whichever of the above did the trick, I cannot program anything on the cpu anymore, everytime avrdude "Detect" shows different device signature, and any attempt to change fuses or firmware gives an error. And of course the screen stays blank.
Any suggestions regarding resetting the fuses would be welcome , but it looks like- apart from cpu swap(which would be practically impossible, since i have no hot air station/experience), it's only good for the dust bin.

[madires]

You can debrick an ATmega324 by high-voltage (parallel) programming or maybe JTAG. Only a few ATmega programmers support the high-voltage mode, but some (many?) universal programmers do.

[ekoloski]

Fuse bits are dangerous...

It looks like that fuse config set the AVR to expect a low power crystal oscillator. If there's a crystal on there that just isn't being driven right, maybe you can rescue it by setting a slower bitrate and trying to program the fuse? With avrdude it's the -B option (substitute in your parameters for the programmer you're using, and experiment with the -B flag which is the bitclock period in uS): avrdude -p m324 -B20 ... -U lfuse:w:0xff:m

Otherwise as the previous poster mentioned, parallel programming is the way to recover it. The catch is that you'd need to remove the chip from the board first and you've stated that's a no go. Maybe there's someone local who can help out with the desoldering if necessary?

A number of programmers out there support hvpp. I personally use an older AVR Dragon, but some people even make their own programmers out of another AVR. If you are able to get it removed it would be something similar to:

avrdude -p m324 -P usb -c dragon_pp -U lfuse:w:0xff:m

[oitar]

Thanks for the replies! Why is it necessary to remove the chip for parallel programming? If I ensure all peripherals are disconnected, can't I do it in-circuit? I've already written it off, so wouldn't mind trying anything, if it will help others.
Sorry for wasting the time of one of the project creators !

[ekoloski]

I'd remove the chip out of caution, since I don't know what the schematic for this board looks like. HVPP uses 12 volts on the Reset line, if anything else on the board is connected to it and is not tolerant of 12V it could be damaged.

Beyond that, anything else connected to the pins could load it down and interfere with the programming. And lets not forget that there are a LOT of pins to connect (21 of them!) for parallel programming.

If there really isn't anything else on those pins that could get damaged or interfere, then you're right. The chip could be programmed in-place.

For reference, below is a snapshot from page 368 of the datasheet (this is a Mega324, right?) which shows the pins for parallel programming. Note that in addition to each named line all of PB is connected for data.

[Fuzzy Star]

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/

[Vitaly_Ne]

Hello everybody.
It was necessary to measure the capacitance of the coaxial wire (about 1m), I plan to make a probe for the oscilloscope, but I noticed something odd.
Compared with the device RM219.
Those. ordinary capacitances measure +/- the same way (taking into account the error), in the photo just for example the capacitance is 1 microfarad (yellow) and 10nF (green) ...
But that's it when measuring a piece of cable - the readings differ twice!??? I measured other pieces 50cm ... and 1.5m ... also a huge discrepancy ...
Moreover, measurements of other components are normal ... if there are discrepancies, then they fit perfectly into the errors (and my needs).
Usually I measure something with either a device or a tester and there have never been any problems ... but here an incomprehensible situation arose with the measurement of the cable .. Maybe someone came across? Is it in different ways or features of measuring instruments?
I want to know what to believe anyway...
Or maybe someone will also conduct a similar test (wire range from 0.75 - 1.2m coaxial), will you have discrepancies?

It’s just that later I’ll calculate the compensation capacity and I don’t know what evidence to build on ...
The cable is marked RG58A / U (but I doubt that this is the original ...)

Tester G328A(BGR), firmware costs 1.13k (from Yuriy_K from page 277)
In principle, I can try to flash 1.44m (Thanks to indman,
  for help in compiling in due time ...), only the USBASP-programmer threw it somewhere, but I can search(But I don't think it's a firmware issue...) ...

p.s. According to the videos on the network, who made such probes, the capacitance is within 30-45nF for people ... i.e. the readings of the tester in this case are greatly overestimated (I don’t understand the reason), but the readings of the device 219 are, as it were, normal ...
This is probably some kind of feature of measurements by a tester of coaxial cables?

[madires]

The Transistor Tester doesn't use an AC signal to measure capacitance. Instead it charges the DUT to a specific voltage, counts the time, and calculates the capacitance from that. It measures the DC capacitance. For a transmission line, e.g. a coax cable, a proper LCR meter with an AC signal is recommended.

[Yuriy_K]

This is probably some kind of feature of measurements by a tester of coaxial cables?

The indications should not depend on the methods of measurement. Here are examples of measurements on different instruments...

Attention! Added,  for the TTester on both firmware, there is a dependence of the readings on the connection of the leads to the measuring terminals.

[Feliciano]

I measured other pieces 50cm ... and 1.5m ... also a huge discrepancy ...

The capacitance of a coax cable is function of its length (think about the capacitance of two conductors in parallel of variable lenght separated a constant distance by a dielectric, the more the length, the more charge can be stored). In fact, if you look at the manufacturer datasheet it will tell you the typical pF/m or nF/m.

[Vitaly_Ne]

The Transistor Tester doesn't use an AC signal to measure capacitance. Instead it charges the DUT to a specific voltage, counts the time, and calculates the capacitance from that. It measures the DC capacitance. For a transmission line, e.g. a coax cable, a proper LCR meter with an AC signal is recommended.

Yes, that's understandable. Unfortunately, I do not have such devices yet. I thought to measure, not perfectly accurately, but at least to decide on further calculations ...
But such a nuance turned out ...

[Vitaly_Ne]

The capacitance of a coax cable is function of its length....

What depends on the length, I know that.
According to the datasheet, the manufacturer claims 85pf per meter.
When I measure a long piece and a short ratio with a tester, this is how it turns out.
Also, if I measure with a multimeter, the length ratio is the same (+/- error) ...
But Yuri wrote that the readings should be the same, but for some reason I have different ones .. moreover, they simply measure the capacitors EQUALLY ... and resistors, etc.
But it was with the cable that such a nuance arose ...
I myself wonder what my problems are ... is it a multimeter? those. According to Yuri's screenshots, his tester readings approximately coincide with mine (70-80pF)

[Vitaly_Ne]

The indications should not depend on the methods of measurement. Here are examples of measurements on different instruments...

Here you have "ideal" ... The tester calibrated ... Everything shows correctly ...
It happens that you need to believe his testimony (they are closer to the datasheet) .. but the multimeter turns out to be a failure for me ...
But why exactly when measuring the cable This is incomprehensible to me ... after all, all other measurements for them practically coincide ...

[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.

[hojnikb]

• We already had a short discussion about DurandA approach. You can go back to the posts of July 23, 2022 of this thread
• Unless the vendor shows an internal photo (and even in that case we cannot be 100% certain), you cannot know for sure what you're going to get untill you have it in your hands. For instance, last July I bought one T1 and one T7 from Amazon, for $20+ each, from different sellers. One was the real ATmega, the other was a 32pin clone

Guess i'll have to roll the dice and see what i get. It's cheap though, so it won't matter that much.

Just got mine after 3 weeks. It has an unmarked microcontroller. By the looks of it, it seems like a 32pin deal



Is there a way to know which of the clones this is ?

[Obelix2007]

Yes, that's understandable. Unfortunately, I do not have such devices yet. I thought to measure, not perfectly accurately, but at least to decide on further calculations ...

... the ComponentTester is better than you think, correct calibration is important ...

... I carried out the measurement with different devices and documented the results in photos.

Greeting Horst

[Fuzzy Star]

This is probably some kind of feature of measurements by a tester of coaxial cables?

Your photo says: "Measurements of just capacitances are more or less the same ... but when measuring a cable such a spread"
The capacitances you use to compare are 1uF and 10nF. But the capacitance of the cable you're trying to measure is less than 100pF. It's such a low value to say 'spread'. Results are 40pF and 70pF, it's only 30pF difference. With such a low value, your probe capacitances and the connection methods affect the results.
How many pF the Tester shows when there is nothing connected to the probes? (keep the probes connected to the ZIF socket)

Use a 50-100pF capacitor to compare your DMM and Tester. Probably you will see a similar 'spread'.

[Vitaly_Ne]

... the ComponentTester is better than you think, correct calibration is important ...
... I carried out the measurement with different devices and documented the results in photos.

Of course, the tester is an excellent device. Calibrated, I think, is also correct (I have been using it for a relatively long time and figured it out a little) ...
Here you are in the photo, I also look at the instrument readings more or less the same ...
So I have problems with the multimeter ... because. my readings are 70nF closer to those measured by Yuriy_K and you...
___________________
p.s.  I just measured a piece of double-sided textolite and ....
in general I have a COMPLETE misunderstanding ((((
The readings are almost the same (I leaned the contacts several times) +/- a few units of difference ..