$20 LCR ESR Transistor checker project

[MrJohnnny]

No worries! The official Transistor Tester firmware (k and m version) is open source!!! Unfortunately some clone vendors modify the source and claim ownership while violating the copyright of the original authors. It doesn't make much sense to try stopping them, because that would require a lot of money for lawyers. Therefore my advice is to buy only hardware compatible clones, since those can be upgraded to the current firmware version. If you buy a clone with a modified hardware/firmware, you might never get any update from the vendor. And a self-made tester is best one anyway

PS: The next m-firmware version will have a squarewave signal generator (up to 2MHz for 8MHz MCU clock) operated by the rotary encoder supporting dynamic turning velocity.

OK, thanks for response. In order to thank authors I decided to build it myself, even when it will came out not as good as bought one, I feel proud using it  (and I did, I use it now very often), without giving money to Chinese for a thing, that they didn't invented    I'll like to try m-version, however, with such stuff I'm dealing with having an inductance meter is very helpful. Seems like I need to think about making another one for m-version exlusively, hmmm...

[timb]

But why, if the tester calibrates against a precision voltage reference?

It doesn't calibrate against a precision reference as I understand it. It simply uses the ADC's reference input which is either VCC or 1.5V.

The 2.5V precision reference is optional. From my understanding of the PDF linked on page 1, it's only used for measuring the battery voltage accurately.

Either way, the more accurate the resistor, the more accurate the calibration.


(Incidentally, the Chinese MK-168 I got uses 0.1% resistors, which shocked the shit out of me. Well, unless they're just painting a purple band on some cheapies!)


Sent from my iPhone

[timb]

Another thing I noticed was the lack of a Zener diode, which I thought was required for a voltage reference? The other thing of note is SMD capacitors C1 (missing), C6, C7 and C8 on the back. There’s plenty of board space on the front, so I thought that was a weird decision. The last oddity is what appears to be a bodge resistor connecting pin 2 of the ATmega to VCC.

I assume the "bodge resistor" you refer to is the little surface mount package.  I believe that is on pin 27, not pin 2, if I remember how to count pins on a DIP.

Spot on. I was looking at the board with the chip out of the socket when I wrote that.

So this is where the precision reference should go. Lucky I've got about 6 REF5025's in a sealed bag around here somewhere. Time to customize!


Sent from my iPhone

[markce]

I've got a MK168 as well. It was amazingly accurate in both R end C measuments, but no zener in place.
Looked it up, and it's only used as added function in calibration mode. The voltage regulator used on this
board nice and much better than an ordinary 78l05. Given the construction and the accuracy, I'll not mod it
anywhere soon. Pleased with it as it is.

[timb]

Yeah, I was just going through the README in Markus’ firmware and it says this:

The external 2.5V voltage reference should be only enabled if it's at least
10 times more precise than the voltage regulator. Otherwise it would make
the results worse. If you're using a MCP1702 with a typical tolerance of
0.4% as voltage regulator you really don't need a 2.5V voltage reference.

The regulator used in the MK-168 is the AMS1117-5V, which isn’t quite as good as the MCP1702 and others, but like you said, more than acceptable I think. I might still install the reference for the hell of it.

[madires]

It doesn't calibrate against a precision reference as I understand it. It simply uses the ADC's reference input which is either VCC or 1.5V.

The 2.5V precision reference is optional. From my understanding of the PDF linked on page 1, it's only used for measuring the battery voltage accurately.

The optional external 2.5V reference is used to determine the offset of Vcc (5V). That offset has an impact on the self-adjustment of the internal bandgap reference and all measurements.

[timb]

Right. I was just stating that the optional 2.5V Ref wasn't used as the ADC's reference input or anything.

(And of course by battery accuracy I meant VCC.)


Sent from my iPhone

[Scaramanga]

Hello everyone!

I just bought one of these , version 2.4 2013/07/01

i just wanted to ask:

I saw people in the first pages installing decoupling caps near by the chip, while others don't. What are the benefits of it?
Sorry if i sound silly but i'm not familiar with µC world
Thanks in advance

[madires]

I saw people in the first pages installing decoupling caps near by the chip, while others don't. What are the benefits of it?
Sorry if i sound silly but i'm not familiar with µC world

The decoupling caps keep the voltage of the power supply stable if the load, i.e. the current draw of the MCU, changes. You can read more about that at https://en.wikipedia.org/wiki/Decoupling_capacitor for example.

[Scaramanga]

ok thanks for your response 
I'm new at electronics, after thinking about it i don't think i will put more money on this gadget. I'd better spend my time & money into learning more about electronics  rather than trying to modify this low-cost gadget
i just removed the zener as suggested & i'll leave it as it is. I'll calibrate it if ever i find a low-tolerance film capacitor.

[moemoe]

If anyone here should be interested, I'm running a centralized order on my variant of the board. Price per board is 3.50€, shipping (worldwide) 3,10€ or 5,05€ for >2 boards.

It's designed to fit into the http://www.reichelt.de/SP-6000-SW/3/index.html?&ACTION=3&LA=446&ARTICLE=33838&artnr=SP+6000+SW Strapubox 6000, equipped with a http://www.mouser.de/ProductDetail/ELECTRONIC-ASSEMBLY/EA-DOGM162W-A/?qs=sGAEpiMZZMv8dIMJ2cKI/LMZh%252b6hoZs2 EA DOMG Display (mostly because it's much smaller than a normal 2x16 Display).

All Data is available on https://github.com/maugsburger/avr-component-tester and some images on the first posts here: http://www.mikrocontroller.net/topic/343407 and http://www.mikrocontroller.net/topic/314097

In opposite to most china versions it can test zener diodes up to 30V, too.

[madires]

i just removed the zener as suggested & i'll leave it as it is. I'll calibrate it if ever i find a low-tolerance film capacitor.

It doesn't have to be a low-tolerance film capacitor, any film cap >100nF is fine. I use an 1µF MKT for example.

[timb]

So I built my reference last night! The REF5025 requires a low ESR cap between 1 and 10uF on the output for stability. To keep things nice and compact I dug through my box of donor boards and found a nice 10uF Tantalum SMD on an old motherboard. With a bit of work I was able to solder it to the output lead and ground ring of the SOIC breakout board I was using. I also added a 10uF ceramic across the input and output leads for good measure.

Here it is with a 555 for comparison:


Plus the obligatory close up shot:


I'll mount it to the top of the MK-168 board with some double sided tape and tap GND and VCC directly from two components in front of it, then run a small lead directly to the reference input.

I know this is a bit overkill, but I've got a ton of these REF5025's leftover from a client project awhile back. (These are the high grade versions, too!)


Sent from my Smartphone

[Scaramanga]

It doesn't have to be a low-tolerance film capacitor, any film cap >100nF is fine. I use an 1µF MKT for example.

ah i see

so the quality factor is about the reactance/esr at given frequency

is a mkp capacitor from fluo lamp ok ?

[timb]

Is there an easy way to disable continuous mode with Markus’ firmware? I find it’s a bit wasteful for me, since I’m not testing a string of components at a time. Also, I’ve got to say that this firmware feels a lot slower than the version in the trunk directory. Maybe it’s something in my settings?

Edit: I set #define CYCLE_MAX to 1 in config.h, but it still seems to go more than once sometimes (I know if I hit the button after a measurement it’ll go again, but I don’t touch anything and it’ll go again once and awhile it seems; mainly after I’ve just measured a component).

On another note, I got my reference installed. Works great!





That big capacitor that’s folded over next to the reference board is a 10uF I added to help filter the output of the voltage regulator. Turns out they have *no* capacitor directly on the output! (I’ve got it soldered to the pin of a resistor which is the first thing the regulator’s output connects to [first solder connection coming from the regulator’s tab in the second picture].) This should help keep the regulator’s output a lot more stable.

[texaspyro]

That big capacitor that’s folded over next to the reference board is a 10uF I added to help filter the output of the voltage regulator. Turns out they have *no* capacitor directly on the output!

If you are using an LDO  or low quiescent current regulator,  they pretty much require caps on the input and output or else they turn into dandy oscillators.  Also the output cap cannot have too high OR LOW of an ESR value.   And there may be limits on the capacitance that they require (again can't be too high or low).  Always check the data sheet when using an LDO or low quiescent current type of regulator.

[timb]

That big capacitor that’s folded over next to the reference board is a 10uF I added to help filter the output of the voltage regulator. Turns out they have *no* capacitor directly on the output!

If you are using an LDO  or low quiescent current regulator,  they pretty much require caps on the input and output or else they turn into dandy oscillators.  Also the output cap cannot have too high OR LOW of an ESR value.   And there may be limits on the capacitance that they require (again can't be too high or low).  Always check the data sheet when using an LDO or low quiescent current type of regulator.

Yup, exactly. That’s why I was surprised there wasn’t one. They were filtering the input from the DC Jack, the input to the LDO and the pins between VCC and GND, but that’s about it. I checked the datasheet for the AMS1117 and it requires one, hence adding the 1uF electrolyte.

TI actually makes some “cap free” LDOs, and a bunch that are stable with ceramic outputs, which can save on the expense of buying tantalums.

[timb]

Hmmm, my tester seems to be off by an order of a magnitude on capacitors with the 1.14m software. 100uF caps are reading 10uF! Any idea what could be causing this?

It helps if you set the fuse bits for an 8MHz Crystal instead of the internal 1MHz resonator!

So, my actual and real problem is that I can’t get it to measure inductors at all.

[madires]

is a mkp capacitor from fluo lamp ok ?

Haven't tried one yet, but it should be fine.

[madires]

Is there an easy way to disable continuous mode with Markus’ firmware? I find it’s a bit wasteful for me, since I’m not testing a string of components at a time.

Please press the push button when powering on a little bit longer (>0.3s). For powering off it's the same procedure.

[madires]

It helps if you set the fuse bits for an 8MHz Crystal instead of the internal 1MHz resonator!

So, my actual and real problem is that I can’t get it to measure inductors at all.

The internal RC oscillator is 8MHz too, but the clock prescaler is set to 8 by default. Are you using an ATmega 168 or 328?

[timb]

It helps if you set the fuse bits for an 8MHz Crystal instead of the internal 1MHz resonator!

So, my actual and real problem is that I can’t get it to measure inductors at all.

The internal RC oscillator is 8MHz too, but the clock prescaler is set to 8 by default. Are you using an ATmega 168 or 328?

Right, but I had the fuse bits on the Mega328 chip set to the defaults, which is internal @ 1MHz. So I had uploaded the hex and eep with make upload, but forgot to do a make fuses afterwards.

That's why the thing was running so dog slow and not measuring things correctly.

So I did get it to measure a large (500uH) inductor last night. But it's seeing my smaller (4.7uH) ones as capacitors. Is there some sort of limit on the size of inductors it can measure?


Sent from my Smartphone

[madires]

So I did get it to measure a large (500uH) inductor last night. But it's seeing my smaller (4.7uH) ones as capacitors. Is there some sort of limit on the size of inductors it can measure?

Yes, the lower limit is about 100µH.

[timb]

So I did get it to measure a large (500uH) inductor last night. But it's seeing my smaller (4.7uH) ones as capacitors. Is there some sort of limit on the size of inductors it can measure?

Yes, the lower limit is about 100µH.

Thanks! Can you go into more detail on why it's limited to that? I'd be interested in helping to add support for lower and more precise inductor measurements.


Sent from my Smartphone

[texaspyro]

My Chinese version with the graphics display does a fairly decent job on inductors starting out around 20 uH