$20 LCR ESR Transistor checker project

[elecdonia]

OK, I managed to brick my T4. I flashed the firmware using:

EDIT2: Oh, I erased the entire flash memory and now I need to burn the bootloader again, right?

A bootloader is not needed when programming the AVR MCU with the ISP header.
(On the other hand a bootloader is required to program Arduino boards via a serial or USB port. It’s easy to get confused about this.)

Upgrading transistor tester firmware always requires programming the flash and the eeprom.

Generally the fuse bytes require no changes (except possibly when changing the frequency of the crystal from 8MHz to 16 or 20MHz - could the experts please comment on this?)

Programming of the fuse bytes is required when a virgin (never used before) AVR MCU is programmed for the very first time. The factory default is 1MHz internal clock. Therefore the fuse bytes must be changed to “external crystal oscillator” or else the transistor tester will operate at only 1/8 of its normal speed.

-Elecdonia

[vklimk]

To prevent the failure of U7 I intend to experiment with placing a resistor of perhaps 4K7 or 10K between the high end of the 40 ohm sensing resistor and the current sensing terminal of U7. Hopefully this will prevent failures of U7. The worst case test will be to have the tester powered up and then apply a short circuit between K and A.

It will cause quite a big additional power consumption by device (additional 40mA or something like that) (my misunderstanding, I didn't catch the place of that resistor you proposed).
You can put protective diode or zener diode in parallel to that 40 ohms resistor. Diode will not allow over voltage on U7 sensor pin, it will cut voltage to 0,7v.

But we have to think about a case user shorted A and K terminals - in this case U7 will produce voltage near (Ubat - Vschottky) = 4v and it will be applied directly to our diode. The diode may be burned if we do not limit current. We can put 100 ohms resistor between output capacitor of U7 and K-terminal - it will limit current through diode (Ishorted = (4v-0,7v)/100ohms = 33ma). Voltage drop on this resistor in normal situation (measuring zener diode) will be equal to (100 ohms * 5mA) = 0,5v - it is very low and it should not affect zener detection functionality of TT.

[MrAl]

This is a reply to the very first post...


Hello there,

If you get a scope and frequency generator you can test for all those things and get the error down to a very low value.  This is especially useful for cap ESR which is an important specification.

Also, these little meters will never be able to read inductors except for the very small ones maybe, and maybe only air core inductors.  That's because inductors with metal cores require some excitation current to be able to get a good idea what the inductance is, and the inductance changes with DC bias so you need a way to change that also.  Sometimes the only way to measure inductors like this is to measure their characteristics while they are actually in the circuit they are going to be used in, such as a buck converter.
These harder to measure inductors include 50/60Hz line frequency inductors and switching regulator inductors as they usually have some type of specialized metal core which makes them very DC bias sensitive as well as frequency sensitive.  You can usually measure ESR with an Ohm meter for these types though.

To measure cap ESR, you can put some resistance in series with the cap and use a square wave to energize the circuit.  Looking across the cap with the scope will show a waveshape that can be analyzed that will reveal the capacitance and the ESR.  The ESR shows up as a very vertical part of the wave and measuring the current thought the resistor on channel 2 of the scope and cap voltage on channel 1 will show you the ESR by simply dividing the voltage delta by the current delta for that vertical part.  A bad cap ESR shows up very vividly this way.

[indman]

MrAl,why did you provide us with a full copy of the message February 23, 2013 from the 1st page of the topic? You could just make a short link to this post in the topic.

[madires]

Generally the fuse bytes require no changes (except possibly when changing the frequency of the crystal from 8MHz to 16 or 20MHz - could the experts please comment on this?)

If you just change the frequency the fuse bytes don't need any changes..

[madires]

Also, these little meters will never be able to read inductors except for the very small ones maybe, and maybe only air core inductors. 

The Transistortester isn't an LCR meter, but in can measure L, C and R to some extend. Get one, compare the results with a proper LCR meter, and you'll see that the Transistortester isn't that bad.

[rsjsouza]

Hello there,

If you get a scope and frequency generator you can test for all those things and get the error down to a very low value.  This is especially useful for cap ESR which is an important specification.
(...)

I agree with your remarks about the potential pitfalls when measuring inductors and capacitors, but the keyword is convenience: for a quick check these testers are exceptionally useful. 

[elecdonia]

Also, these little meters will never be able to read inductors except for the very small ones maybe, and maybe only air core inductors. 

The Transistortester isn't an LCR meter, but it can measure L, C and R to some extent. Get one, compare the results with a proper LCR meter, and you'll see that the Transistortester isn't that bad.

For years I’ve used the LC100-A to measure inductors & capacitors. The LC100-A excels at measuring small capacitors & inductors - uH & pF. That said, the LC100-A is also capable of measuring large values: 100,000uF, 50H. I recommend that everyone who enjoys using their transistor tester should consider adding the LC100-A to their test gear collection.

The LC100-A is a modern digital version of the legendary “ grid dip meter ” used by ham radio enthusiasts for nearly 100 years. It measures the unknown L or C by connecting it in parallel with an internal calibrated C or L. This forms a tuned LC circuit. An IC comparator with positive feedback drives the LC circuit into steady oscillation. Typical oscillation frequency is 1-500kHz. The internal MCU-based frequency counter measures the oscillation frequency and uses clever math to calculate and display the uH or pF of the unknown L or C.

Many Asian electronics vendors sell the LC100-A. Although quality/accuracy varies by vendor, the low cost ( < $25 USD ) allows exploring this device on a budget. The original LC100-A manufacturer is a company named MingHe. I have a genuine branded MingHe LC100-A which is quite accurate, often well within 2% when compared to high-end LCR meters. Try to find a vendor who shows a photo of the LC100-A with a MingHe label on the PC board. This is the one to buy.

Comprehensive discussions of the LC100-A may be found here:
https://www.antiqueradios.com/forums/viewtopic.php?f=8&t=365161
https://www.eevblog.com/forum/testgear/lc100-a-a-precise-lc-meter-for-3-75$/msg4596301/#msg4596301

-Elecdonia

[elecdonia]

Generally the fuse bytes require no changes (except possibly when changing the frequency of the crystal from 8MHz to 16 or 20MHz - could the experts please comment on this?)

If you just change the frequency the fuse bytes don't need any changes..

Thanks!
I’ll soon be updating the firmware in one of my transistor testers. I also intend to experiment with changing the crystal from 8MHz to 16MHz.

-E

[indman]

For several years I’ve been using a device known as the LC100-A to measure inductors and capacitors.

This information is certainly interesting to many on the forum, as well as information about measuring ESR using a generator and an oscilloscope. But I want to ask without offense - what does this information have to do with the project being discussed here? There are relevant topics for discussing these devices on the forum.

[elecdonia]

For several years I’ve been using a device known as the LC100-A to measure inductors and capacitors.

This information is certainly interesting to many on the forum, as well as information about measuring ESR using a generator and an oscilloscope. But I want to ask without offense - what does this information have to do with the project being discussed here? There are relevant topics for discussing these devices on the forum.

I agree: in this topic we should focus on the transistor tester.

My LC100-A post was intended mainly to call attention to a low-cost alternative device for measuring C and L. I was responding to several questions posted here about measuring L & C with the transistor tester.

I use my transistor testers several times per day. But I go to my LC100-A when L or C must be measured precisely, especially with RF circuitry.

To me the transistor tester represents an inexpensive device which every electronics enthusiast should have on their workbench. I am very appreciative of the efforts of the folks here on this topic who continue refining & maintaining the transistor tester project.

-E

[Yuriy_K]

Also, these little meters will never be able to read inductors except for the very small ones maybe, and maybe only air core inductors. 

The Transistortester isn't an LCR meter, but in can measure L, C and R to some extend. Get one, compare the results with a proper LCR meter, and you'll see that the Transistortester isn't that bad.

For several years I’ve been using a device known as the LC100-A to measure inductors and capacitors.

Both authors inattentively read messages with examples of measurements. I gave measurements of small capacitances < 1 pF.
When it comes to measurements of small and large inductances, it is necessary to take into account the different measurement methods and frequencies at which measurements are made.
The top row shows the readings of the TTester when the turns of inductance are compressed and the readings of the measurement of large inductance. Below are the readings of this inductance on different RLC meters at different measurement frequencies.

[elecdonia]

MrAl,why did you provide us with a full copy of the message February 23, 2013 from the 1st page of the topic? You could just make a short link to this post in the topic.

What is the method to insert links to other posts which are inside the EEVblog forums? The only instructions I can find in the forum help documents are for linking to URLs. I can do that OK. But I cannot yet link to an individual forum post. 
Thanks in advance!

[indman]

elecdonia, this is easy to do if you copy the link "Re: $20 LCR ESR Transistor checker project" in the browser and paste it through the globe icon "InsertHiperlink" 
For example, here is a link to the 1st post in the topic
https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg195093/#msg195093

[elecdonia]

To protect U7 I will try adding a series resistor (4K7 or 10K) between the high end of 40 ohm current sensing resistor and U7 current sensing pin.

It will cause quite a big additional power consumption by device (additional 40mA or something like that).

You can put protective diode or zener diode in parallel to that 40 ohms resistor. Diode will not allow over voltage on U7 sensor pin, it will cut voltage to 0,6v.

I will test your suggestion to connect diode in parallel with 40 ohm resistor.
Advantage: Addition of parallel diode doesn’t require cutting a PC board trace.

However, addition of 10K series resistor should not cause power consumption to increase. The input resistance of U7 current sensing pin is much larger than 10K. Therefore the series resistor is not expected to alter the operation of U7.

[vklimk]

However, addition of 10K series resistor should not cause power consumption to increase. The input resistance of U7 current sensing pin is much larger than 10K. Therefore the series resistor is not expected to alter the operation of U7.

Agree. I have updated that message recently, please have a look. I think it is necessary to add one more component for a special case I didn't think about. See a drawing there as well.

By the way, if the input resistance of U7 current sensing pin is much larger than 10K then almost all voltage from capacitor will be applied to U7 current sensing pin (it will be similar to a simple resistor divider), 10K resistor will not protect U7 from overvoltage.

[elecdonia]

However, addition of 10K series resistor should not cause power consumption to increase. The input resistance of U7 current sensing pin is much larger than 10K. Therefore the series resistor is not expected to alter the operation of U7.

Agree. I have updated that message recently, please have a look. I think it is necessary to add one more component for a special case I didn't think about. See a drawing there as well.

By the way, if the input resistance of U7 current sensing pin is much larger than 10K then almost all voltage from capacitor will be applied to U7 current sensing pin (it will be similar to a simple resistor divider), 10K resistor will not protect U7 from overvoltage.

Yes, adding a small series resistance (100R) between the filter capacitor and the ZIF socket will reduce the instantaneous peak of the transient current to 300mA when a direct short is placed across K and A test pins.

Regarding current sensing pin of U7, it is possible this device has ESD protection, but dumping a capacitor charged to 30V directly into it would likely cause permanent damage. I accidentally destroyed the SRV05-4 protection device in a much older transistor tester (GM328 AY-AT type) with a capacitor which I (incorrectly) thought was discharged. The ATmega328 MCU was undamaged, so the protection circuit did protect the MCU. I temporarily removed the faulty SRV05-4 but will replace it as soon as the replacements arrive. I also ordered replacement U7 devices at the same time. I will be testing the revised circuit you recommend as soon as these arrive.

[vklimk]

Yes, adding a small series resistance (100R) between the filter capacitor and the ZIF socket will reduce the instantaneous peak of the transient current to 300mA when a direct short is placed across K and A test pins.

300mA will be applied for a very short time and will not damage the diode.
And we can increase this resistor to limit the current more. But we have to know triggering voltage for zenner diode detection to find the best value for the resistor.
Can you find this voltage?
Just remove burned u7 and connect voltage supply to K and A terminals to emulate voltage drop of zenner diode. Set 30v and slowly decrease it, TT will start detecting zenner diode at specific level. It will be near 25v most probably. Having triggering voltage we'll be able to find better value for that resistor.

[indman]

I also ordered replacement U7 devices at the same time. I will be testing the revised circuit you recommend as soon as these arrive.

For me, your efforts to restore this clone from FNIRSI look ridiculous - it's like trying to make a chocolate candy out of shit!? FNIRSI has released a crude, low-quality product to the market, which does not correspond to the level of this project in any way. It is desirable to return such products to the seller, to demand a refund of the money spent for wasting time.

[elecdonia]

Hardware upgrade and update mod m-firmware 1.43m,k-firmware 1.13k for clone LCR-TC1!

I wish I had purchased LCR-TC1 last year, back when the Asian manufacturers made them with ATmega324 MCU. Sadly it appears all LCR units available today contain an unknown MCU (frequently mislabeled ATmega328). I know of 2 ways to ID the fake MCU, but it requires seeing a photo of PC board (never shown by the Asian vendors):
   1) Pinout is totally different than Atmel MCU
   2) There is no crystal

I would have loved to do the ATmega644 mod described in detail in the following post. It has all the advanced features in a self-contained unit powered by a rechargeable battery!
https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg3578117/#msg3578117

FYI the Asian vendors are now offering bare ATmega324 and ATmega644 TQFP-44 MCU chips for $20-30 USD each, plus shipping. I suspect they learned the bare IC chip all by itself can now be sold for more $ than a fully assembled transistor tester. As a long-term Atmel AVR fan I’m glad I bought several tubes of AVR MCU (DIP-40) with large flash size several years ago before they became expensive, and then unavailable from USA distributors like Digikey & Mouser.

New plan: I’ll build a “full featured” transistor tester from scratch. As a consolation prize I may use a much larger TFT display panel. It will be the “bench version” rather than “handheld.”

[elecdonia]

I also ordered replacement U7 devices at the same time. I will be testing the revised circuit you recommend as soon as these arrive.

…FNIRSI has released a crude, low-quality product to the market, which does not correspond to the level of this project in any way.

Maybe FNIRSI had no official relationship at all with these recent LCR-TC1 clones?
Could it be that the LCR clone makers simply thought it would look better to display FNIRSI on the self-calibration screens?

When I ordered the LCR-TC1 (where U7 failed after I used it a couple of times) I was hoping to receive the ATmega324 unit.

From what I’ve read about FNIRSI, the stuff they make is very cheap. Some of it is said to be “functional” to a certain extent. But mostly it is just “cheap.” I won’t buy any more FNIRSI gear until their reputation improves. I own way too much test gear already. Most of it is in the “restored vintage” category.

[MrAl]

MrAl,why did you provide us with a full copy of the message February 23, 2013 from the 1st page of the topic? You could just make a short link to this post in the topic.

Oh it must have come out that way when quoted the very first post.  Not sure why the images show up again like that.  Makes it longer.
I can try to get rid of that part i guess.

[MrAl]

Also, these little meters will never be able to read inductors except for the very small ones maybe, and maybe only air core inductors. 

The Transistortester isn't an LCR meter, but in can measure L, C and R to some extend. Get one, compare the results with a proper LCR meter, and you'll see that the Transistortester isn't that bad.

For several years I’ve been using a device known as the LC100-A to measure inductors and capacitors. I highly recommend the LC100-A for measuring small capacitance and inductance - uH and pF. That said, the LC100-A is also capable of measuring 10,000uF and 50H.

The LC100-A is a modern digital version of the legendary “grid dip meter” used by ham radio enthusiasts for nearly 100 years. It tests the unknown L or C by connecting it in parallel with an internal calibrated C or L to form a tuned LC circuit. It then uses an IC comparator with positive feedback to drive the LC circuit into steady oscillation. Typical oscillation frequency is from 5kHz to 500kHz. Finally an internal MCU-based frequency counter measures the frequency of oscillation and uses clever math to calculate and display the uH or pF of the unknown L or C.

As a result of my personal experience with the LC100-A I recommend that everyone who enjoys using their transistor tester should also try out the LC100-A.

There are numerous Asian vendors for the LC100-A. Although the quality and accuracy of the LC100-A implementation varies by vendor, the cost ( < $25 USD) allows one to explore this useful device on a budget. The original manufacturer of the LC100-A is a company named Ming-He. I have a genuine branded Ming-He LC100-A which is amazingly accurate, often within 2% when compared to high-end LCR meters. If you can find a vendor who shows a photo of the LC100-A with a Ming-He label under the PC board, then this is the one to buy.

Here is a link to a more comprehensive discussion of the LC100-A from the Antique Radio forum:

https://www.antiqueradios.com/forums/viewtopic.php?f=8&t=365161


-Elecdonia

What is the excitation current adjustment range?

I cant see how any one meter will be able to measure all inductors, although it may still be useful.
For example, some buck converters will output 100ma while others 100 amps.  That means the filter inductor may be seeing either 100ma DC or 100 amps DC.  I've dealt with that entire range in my career so i can see how things can vary and how the inductance can vary.
Im not saying other meters cant help, but they will never be able to test the entire range of inductors available.  That's all.

[MrAl]

Hello there,

If you get a scope and frequency generator you can test for all those things and get the error down to a very low value.  This is especially useful for cap ESR which is an important specification.
(...)

I agree with your remarks about the potential pitfalls when measuring inductors and capacitors, but the keyword is convenience: for a quick check these testers are exceptionally useful.

Yes of course i understand that, but i wanted to make it clear that no one tester can test every inductor out there and some require a large DC bias to test properly.  For a quick example, saturation current.

I just wanted to make it clear that there will be inductors that are harder to test so they would have to be tested in the actual circuit they will be used in or a prototype of the same circuit.  That way they are subject to the same bias and frequencies and all that.

[indman]

I just wanted to make it clear that there will be inductors that are harder to test so they would have to be tested in the actual circuit they will be used in or a prototype of the same circuit.  That way they are subject to the same bias and frequencies and all that.

None of the people who understand the meaning of electronics does not argue with this opinion, it's true, but as correctly observed by dear rsjsouza - the key purpose for this project is a QUICK TEST of a part, removed from the board or a new (unknown) part. For more detailed testing and analysis you will certainly need more serious equipment.