$20 LCR ESR Transistor checker project

[indman]

No, it is not DIP MCU.  I don't understand why it should it be DIP MCU?

No, you can't use this tester. A poor quality display is used here.It will not run at 16MHz and installing an adapter for the ATmega644 would not make sense.

[Pygmalion]

No, it is not DIP MCU.  I don't understand why it should it be DIP MCU?

No, you can't use this tester. A poor quality display is used here.It will not run at 16MHz and installing an adapter for the ATmega644 would not make sense.

I simply cannot understand you. You mentioned GM328 clones, which all have 8MHz (according to your own document). Now 8MHz is suddenly no longer good.

I don't care about the ATmega644. I just want a USB-powered device that has higher precision than directly powered DevKit 644 without a linear regulator.  And whose firmware can be upgraded.  And I don't care about display as long as values are correct.

[indman]

I don't care about the ATmega644. I just want a USB-powered device that has higher precision than directly powered DevKit 644 without a linear regulator.  And whose firmware can be upgraded.  And I don't care about display as long as values are correct.

You can buy any option you want, no one's stopping you. I just offered you mine and wrote the nuances that you may need in the future.If you are not interested, you can ignore my messages.

[madires]

No, it is not DIP MCU.  I don't understand why it should it be DIP MCU?

Many of the current models come with 'alternative' MCUs, such as APT32F172K8T6 or LGT8F328, markings removed or falsely marked as ATmega. The firmware is poorly ported. Those testers are junk! The best bet to get a tester with a genuine ATmega is to look for the DIP version, because the alternative MCUs aren't available as DIP.

[Pygmalion]

Many of the current models come with 'alternative' MCUs, such as APT32F172K8T6 or LGT8F328, markings removed or falsely marked as ATmega. The firmware is poorly ported. Those testers are junk! The best bet to get a tester with a genuine ATmega is to look for the DIP version, because the alternative MCUs aren't available as DIP.

Oh, thank you, now I understand! The picture shows an ATmega328, but it could be something else. Is it possible to check if it is a real ATmega328 after receiving the product? Because I can demand a refund if the product doesn't match the description.

And after the refund, I could solder the real ATmega328, which I also have on hand.

With an MCP1702-500 there's no need for an external 2.5V reference. The reference makes only sense if it's an order of magnitude better than the voltage regulator. The switching noise of a wall wart can cause measurement issues (-> add an additional LC filter).

I understand the part about the LC filter, and that's great advice. I have done some research. With a ripple of 50mV for a typical switching regulator, 47uH and 10uF LC filter (I already have these two on hand), the cutoff frequency is 7.3kHz and the attenuation is -45dB or 180x, which reduces the ripple to 0.3mV, on pair with linear regulators.

I wish I understood other nuisances of power supply as well as this one.

[madires]

Oh, thank you, now I understand! The picture shows an ATmega328, but it could be something else. Is it possible to check if it is a real ATmega328 after receiving the product? Because I can demand a refund if the product doesn't match the description.

From the Clones file:

Code: [Select]

- MCU in DIP is usually an genuine ATmega.
- APT32F172K8T6 has different power pins:
  - 32-LQFP/QFN: 18=Vss, 19=Vdd (ATmega328 32-TQFP: 5=Gnd, 4=Vcc)
- APT32F172K8T6 uses SWD port for ISP:
  - 5 pins: Vdd, Vcc, F_SDAT, F_SCLK, F_RST
- Testers with APT32F172K8T6 often lack a quartz crystal.
- LGT8F328 has slightly different pins:
  - QFP32L: 21=PE2/SWD      (ATmega328 32-TQFP: 21=Gnd)
            18=PE0/SWC/APN4 (ATmega328 32-TQFP: 18=AVcc)
- LGT8F328 uses SWD port for ISP:
  - 5 pins: Gnd, Vcc, SWD, SWC, Reset


[Yuriy_K]

No, it is not DIP MCU.  I don't understand why it should it be DIP MCU?

Yes, this is a normal T-tester, look at the presence of six sockets for connecting the programmer - below the processor on the back side.

[indman]

Yes, this is a normal T-tester, look at the presence of six sockets for connecting the programmer - below the processor on the back side.

Yes, this is an ordinary tester on Atmega, but the ugly display used in it with an additional broadcast of the data makes it almost the same garbage as clones with fake MCU

[Pygmalion]

Yes, this is a normal T-tester, look at the presence of six sockets for connecting the programmer - below the processor on the back side.

Yes, this is an ordinary tester on Atmega, but the ugly display used in it with an additional broadcast of the teams makes it almost the same garbage as clones with fake MCU

I wish you to be more specific about what is wrong with the display. What does "additional broadcast of the teams" mean?

Does this essentially mean that in your opinion there isn't a good China clone that runs on 5V, even though some of them have both a linear regulator and a proper ATmega MCU?

[indman]

I wish you to be more specific about what is wrong with the display. What does "additional broadcast of the teams" mean?

Look carefully at the scheme of a similar clone that you showed before. Green is a microcircuit, which is an additional translator of the display data. The original GM328A with power from the 9V source uses another display that does not need this complication of the scheme. As the display  with an additional translator has a slightly different color palette, they are less contrasting and cannot maintain work at a higher clock frequency, which is often very important for color display, since with increasing frequency the image is updated much faster.

[RoGeorge]

What does "additional broadcast of the teams" mean?

Does this essentially mean that in your opinion there isn't a good China clone that runs on 5V, even though some of them have both a linear regulator and a proper ATmega MCU?

The "additional broadcast of the teams" is probably just a mistranslated expression.

The microcontroller is powered with 5V, while some displays are made for 3.3V.  Other displays are made for 5V, depending of the display part number/model.  However, some schematics will "abuse" a 3.3V display by sending to it 5V signals from the microcontroller.

Controlling a 3.3V display with 5V signals might work in practice, but might also produce unwanted side effects.  Proper design will be to use a level-shifter circuit, an IC that shifts the 5V signal from the microcontroller into 3.3V signals for the 3.3V LCD.

In conclusion, when some circuits work with 5V signals (the microcontroller) while other parts of the schematic work with 3.3V signals (the LCD), the proper design will be to use a level-shifter IC.

Most of the tester clones are designed for the lowest possible price, so they will not have a level shifter.  Instead, they will "force" 5V MCU signals into a 3.3V LCD (with some resistors).  This won't fry the LCD, but it might affect the accuracy and precision of the measurements.

[indman]

The "additional broadcast of the teams" is probably just a mistranslated expression.

No, I wrote everything correctly. The translate is correct. Look once again at the scheme of the clone that I showed.
Coordination of levels is carried out by other microcircuits.

[RoGeorge]

I'm afraid "additional broadcast of the teams" doesn't make sense in English.  I'm not trying to be rude, or to win an argument, just that, in English it doesn't make sense.  I don't know what to understand.

"Teams" usually are made of people, like a "football team".
"Broadcast" usually means to transmit to everybody, by radio/TV, or by voice.
"Additional" means some extra quantity.

My best guess is by "teams", you refer to the "data bus".  And another guess, by "additional" you mean the difference between 5V and 3.3V?  I don't know.

Again, I'm not attacking you.  I'm very grateful for all the advice you provide, and for all the software.  Thank you! 

[indman]

My best guess is by "teams", you refer to the "data bus".

Yes, you are right. I changed the word "teams" to "data" in my previous posts.

And another guess, by "additional" you mean the difference between 5V and 3.3V?  I don't know.

"Additional" - because when using the classic ST7735 display, this control chip is not needed!
So understandable?

[indman]

Does this essentially mean that in your opinion there isn't a good China clone that runs on 5V, even though some of them have both a linear regulator and a proper ATmega MCU?

I don't follow all the new clones released by my Chinese friends. But if I vitally needed USB power i would still choose the classic 9V powered GM328A(DIP formfactor) clone variant. For USB power supply you can easily and quickly use a special adapter like the one shown in the photo below. Or something similar and simple in circuitry.

[Pygmalion]

I know that some of you may find me annoying, but let me explain my motivation.

I hate chargers and I hate batteries. In my ideal world, everything should be powered or charged by a standard USB-A or USB-C connectors. (Thanks, EU!) I never buy original RPi Zeros and RPi Picos, but (more expensive) Chinese clones with USB-C connectors, my bike always has a dynamo...

But now I have an idea. I have a few of those fancy PD / QC 3.0 chargers. I have a challenge for you all:

Does anyone know how to design a circuit that negotiates a higher voltage with the fancy USB charger and sends that voltage to the linear regulator, but in the case of a regular 5V USB charger, to the LC filter?  I could combine that with Markus' design.

EDIT: The selection of 9V seems quite simple, set D+ to 0.6V and D− to 3.3V. The problematic part is measuring the voltage and passing it to the linear regulator if 9V and to the LC filter if 5V.

[Pygmalion]

What does "additional broadcast of the teams" mean?

Does this essentially mean that in your opinion there isn't a good China clone that runs on 5V, even though some of them have both a linear regulator and a proper ATmega MCU?

The "additional broadcast of the teams" is probably just a mistranslated expression.

The microcontroller is powered with 5V, while some displays are made for 3.3V.  Other displays are made for 5V, depending of the display part number/model.  However, some schematics will "abuse" a 3.3V display by sending to it 5V signals from the microcontroller.

Controlling a 3.3V display with 5V signals might work in practice, but might also produce unwanted side effects.  Proper design will be to use a level-shifter circuit, an IC that shifts the 5V signal from the microcontroller into 3.3V signals for the 3.3V LCD.

In conclusion, when some circuits work with 5V signals (the microcontroller) while other parts of the schematic work with 3.3V signals (the LCD), the proper design will be to use a level-shifter IC.

Most of the tester clones are designed for the lowest possible price, so they will not have a level shifter.  Instead, they will "force" 5V MCU signals into a 3.3V LCD (with some resistors).  This won't fry the LCD, but it might affect the accuracy and precision of the measurements.

Thanks you very much for the explanation, now it makes much more sense. Actually, I know quite a bit about these displays; I even have a github project and a regular website dedicated to high speed (parallel) communication of the Raspberry Pi with various displays and also how to customise these displays for 3.3V communication.

Of course, it should be much easier to switch from 5V to 3.3V; a simple voltage divider will suffice. Without a level shifter or voltage divider, this will definitely affect the display. But I don't understand how this can affect the measurement?

[RoGeorge]

Measurement in this device are often made by connecting the unknown component to +5V (through the microcontroller pins at logic 1).  The current is then measured, or maybe the raise or the decay time in an RC circuit is measured.

The 5V has to be perfectly flat and constant during a measurement.  If, for no matter what reason, the +5V voltage has spikes, or small voltage variations, or noise, these might lead to innacurate results, because all the calculations are made assuming the voltage is constant.

Possible reasons for voltage variations while a measurement is in progress:
- voltage ripple (at mains frequency when using a transformer + regulator, or at high frequency when using a switching power supply)
- any variation in current consumption will also affect the +5V regulated voltage, not much, but voltage regulators are not perfect.  When the load current varies, voltage varies, too, thought +5V will vary only with a little bit.
- most input lines in almost any IC, will have internal protection diodes (clamping diodes), like this:  https://electronics.stackexchange.com/questions/324285/internal-clamping-circuit-at-the-input-pins-of-an-mcu.  So, if you feed a 3.3V input with 5V, that voltage will pass through the internal protection diode of the input, and it will arrive at the +3.3V supply, creating perturbances in current and in voltage.
- even the microcontroller itself creates current spikes, when its internal gates switches between 0->1 or 1->0.  Current spikes induces noises in any high impedance traces, and induces voltage spikes in voltage regulators.
- different MCU instructions during the executed code also have different current consumption, the power save states of the MCU also means very different current, and so on.

There are many reasons why the voltage will not be a perfectly constant 5.00000000V at all times during a measurement.

These variations are all in theory, and each type is mitigated in one way or another.  Which one of these possible perturbations will produce visible measurement deviations (and which one will pass unobserved) is hard to tell, other than by experimenting in practice.

Anyway, this tester is not an ultra high accuracy scientific instrument.

The merit of this tester is that it is very clever, clever in hardware and clever in software.  It is very easy to use, and a very practical tester to have in the lab.  Its precision is good enough for day to day practical electronics.

So, don't overthink the precision, accuracy and resolution.  Your initial plan (to use the modules you already have) seems like a good plan to me.  In case it doesn't work from the first try, I would start another topic, to ask for advice about the particular design you may have, because this topic is not about designing new clones of the tester.

Build your own if you have the time, and if you like tinkering with electronics.  If you just need a tester, and not interested much in experimenting (and maybe sometimes failing and fixing) while making your own tester, then buy one.

If I were to buy today, I would probably ask indman if he has any for sale, or which one he recommends (because he's a developer of this tester).

[indman]

If I were to buy today, I would probably ask indman if he has any for sale, or which one he recommends (because he's a developer of this tester).

Thank you for your feedback, but I am not the developer of this project. I just to the best of my knowledge and ability a little help to the respected authors that you know.
Madires  is sometimes difficult to answer numerous questions about the project alone.

[madires]

Yep, support takes a lot of effort. Besides this thread there's another on mikrocontroller.net and the issues section on github. I'm thankful for any help, and indman is a great contributor.

[Pygmalion]

So, don't overthink the precision, accuracy and resolution.  Your initial plan (to use the modules you already have) seems like a good plan to me.  In case it doesn't work from the first try, I would start another topic, to ask for advice about the particular design you may have, because this topic is not about designing new clones of the tester.

Build your own if you have the time, and if you like tinkering with electronics.  If you just need a tester, and not interested much in experimenting (and maybe sometimes failing and fixing) while making your own tester, then buy one.

I actually like tinkering. And since the first transistor testers actually used Arduinos and Arduinos can be powered via USB, one could argue that it should work reasonably well with a normal USB power supply too. Nevertheless, I took the time to develop my own power circuit. If anyone is interested, I've posted it in a separate question here: https://www.eevblog.com/forum/projects/voltage-based-automatic-switching/

[ltz1000a]

Measurement in this device are often made by connecting the unknown component to +5V (through the microcontroller pins at logic 1).  The current is then measured, or maybe the raise or the decay time in an RC circuit is measured.

The 5V has to be perfectly flat and constant during a measurement.  If, for no matter what reason, the +5V voltage has spikes, or small voltage variations, or noise, these might lead to innacurate results, because all the calculations are made assuming the voltage is constant.

Possible reasons for voltage variations while a measurement is in progress:
- voltage ripple (at mains frequency when using a transformer + regulator, or at high frequency when using a switching power supply)
- any variation in current consumption will also affect the +5V regulated voltage, not much, but voltage regulators are not perfect.  When the load current varies, voltage varies, too, thought +5V will vary only with a little bit.
......
There are many reasons why the voltage will not be a perfectly constant 5.00000000V at all times during a measurement.
......

If you require high-precision voltage, ADR1000 and ADR1399 are excellent options.

[indman]

If you require high-precision voltage, ADR1000 and ADR1399 are excellent options.

That's a good one! These sources are good but for this project they are clearly redundant.
It is quite sufficient to use the original MCP1702-5002 or LDK320AU50R

[Pygmalion]

Well, I had an interesting discussion with an analogue signal expert on StackExchange with a very different opinion, but I don't want bother you with that.

Because of that I have decided to construct the device in such a way that I can experimentally determine who is right: I will build my own simplified DevKit-644 Transistor Tester that has both a 5V input through a simple LC filter and a 9V input through a modern linear regulator. I will then take measurements with both inputs and with expensive laboratory equipment and compare them. (Such a test would also be useful for Chinese devices...)

Can you suggest some benchmark tests for the precision of the device? Apart from the obvious tests like measuring a capacitor, resistor or coil?

Also, can someone explain to me why the DevKit-644 uses two MCP1702-5002 linear regulators? I realise that one is powering a relay and a Zener test supply, but by my calculations each of them (only) draws about 30μA of current and are not working at the same time. If 500mA supply is tight, I would opt for a more powerful LD1117-V50 linear regulator.

Finally, I would like to know if anyone is interested in my project and if so, where I could share my thoughts and results?

[madires]

Also, can someone explain to me why the DevKit-644 uses two MCP1702-5002 linear regulators? I realise that one is powering a relay and a Zener test supply, but by my calculations each of them (only) draws about 30μA of current and are not working at the same time. If 500mA supply is tight, I would opt for a more powerful LD1117-V50 linear regulator.

One is powering the boost converter for the Zener test. And the other one powers the whole tester circuit. That decouples the boost converter power-wise to reduce it's impact on the main circuit. It also provides flexibilty and sufficient current for hardware options. The DevKit is meant to play with.

Finally, I would like to know if anyone is interested in my project and if so, where I could share my thoughts and results?

A new thread for discussing your ideas would be nice.