$20 LCR ESR Transistor checker project

[hapless]

Assuming that the same firmware worked with the old MCU, maybe there is a problem with the soldering: The MCU leads that connect to the display could be either shorted or disconnected. There's also a chance that your new MCU has an issue with that particular port, in which case you may need to try another ATMega.

[bitseeker]

It should be an option, i.e. you can choose between On/test push button only, rotary encoder, On/Test plus up/down push buttons or On/Test plus touch screen. Or you could use multiple in parallel.

Ah, yes. I thought you had a particular implementation in mind. Of course, it's intended to be configurable.

What I've meant with "invisible up/down bars" is using the top two character lines as up and the bottom two lines as down button, which would be handy for menus. Or the left three and right three rows to decrease/increase a frequency for example. Could be even better than visible buttons at the bottom.

I got it now. Yes, that would certainly save visual space. I suppose it might catch people off-guard if they don't read about it before loading.

My goal is to create a simple and intuitive touch interface. So I'm trying to do some brainstorming with you all.

Yep, I like the concept. Perhaps start with a test version that works as you've described. Then, folks who have a touchscreen could give feedback after interacting with it for a bit.

[hapless]

What I've meant with "invisible up/down bars" is using the top two character lines as up and the bottom two lines as down button, which would be handy for menus. Or the left three and right three rows to decrease/increase a frequency for example. Could be even better than visible buttons at the bottom.

I got it now. Yes, that would certainly save visual space. I suppose it might catch people off-guard if they don't read about it before loading.

To help with that, a disappearing hint (in the form of inverted arrows) could be shown where the hotspots are. Of course, the ability to drag a menu up and down and tap the selected option would probably be more intuitive for me personally, and probably wouldn't require such hints.

[belzrebuth]

Hi all,
I've got a EZM 328 build that runs the (SVN684) - v2 firmware.(I don't know any more about the firmware I've just copied the name of a folder I've probably placed the .hex I used to update my firmware the last time.
I think that was roughly 9-10 months ago..
My device is the same as this one : http://www.ebay.com/itm/2016-EZM328-GM328R-transistor-tester-ESR-frequency-meter-square-wave-genera-/201663141755?hash=item2ef40f577b:g:plMAAOSwFe5Xzxsd

So ,what would be the latest firmware version for this device?Am I already running it?

I have a tl866 too so I can use that instead of in-circuit programming I've used the last time.

Can I just write the latest .hex to the atmel mcu or I need to set fuses etc?

[stj]

i built that firmware, it's specific for the *real* GM328 hardware.
i say real, because some sellers are putting that name on other unrelated stuff now!!
for anybody about to buy from ebay, the GM328 does NOT have a colour display for a start!

[belzrebuth]

So is this the latest version or there is a newer one for this one?
Mine does not have the ezm electronics logo but it's identical to the one I posted.

[stj]

it's well out of date, i built that in january i think.
i'v not made a newer one though because of change of address resulting in my main computer ending up in a box. 

[pepe10000]

I discovered that changing the HT7550 by an MCP1702T-5002E/TO in an AY-AT Diy Kit produces the same measurement failure in low-value resistors as occurred in LCR-T4-H although less frequent.

The solution has been the same, place three capacitors of 100nF, two in the 5v power lines of the processor and another in the reset.

Attached photos of the process.

[hapless]

Have you tried with a single capacitor on AV_CC only? I can understand the ones on V_CC and AV_CC, but are you sure that the one on the reset pin actually helps with this?

[pepe10000]

I certainly have not tried to filter only the main voltage (VCC and AVCC), but as I said before in post #3649, in the Russian forum they advise to put all three, and since it cost me no work, I also installed the reset .

I am almost sure that the capacitor of the reset is not necessary, but if it improves in some stability I prefer to leave it on.

A greeting.

[madires]

The inductance measurement uses high current (exceeding the maximum of an IO pin) for a very short period when measuring a low inductance. Therefore the LDO's output cap should be large enough. I use 2.2µF with the MCP1702-5005 usually. An additional bypass cap at the MCU is also fine. The circuit diagram for one of the kits shows only a 100nF cap right of the 7550, which isn't sufficient.

[upsss]

Hi, I have two of these testers, AY-AT and T3-T4 with the same firmware, ver. 736 running @16MHz.  When measuring small inductors <100uH in this case ~30uH, the T3-T4 version looks like it display the correct inductance.  On the AY-AT version with the same 30uH inductor, it displays "50nH if 21nF" (yes "50nH if 21nF" not a typo) what does it mean?  Also, both testers displays some sort of frequency in kHz what does the frequency mean?  Any way to have the AY-AT version display the correct inductance?

Maybe it is all related to the previous post, I just verified that both version have a 10uF capacitors on the input and output of the Vcc regulators.

[madires]

You've run into the SamplingADC. It's explained in Karl-Heinz' documentation. The "if 21nF" indicactes a resonant circuit and that the tester expects a 21nF cap in parallel. The frequency is the resonant frequency of the LC circuit.

[upsss]

@madires, how do you explain the discrepancy between the two testers?

BTW, I read a couple of times Karl-Heinz's "explanation" in paragraph "3.10 Measurement of coils".  unfortunately I have no clue what he was trying to say, maybe in German it made more sense or maybe because I am only an EE.

[perieanuo]

@madires, how do you explain the discrepancy between the two testers?

Maybe pcb routing ?just guessing...


Envoyé de mon iPad en utilisant Tapatalk

[pepe10000]

@madires, how do you explain the discrepancy between the two testers?

LCR-T4-H and AY-AT Diy Kits, with 738 revision k firmware and filter capacitors installed in the power supply lines.

I do not notice that difference.

However between version k and m if there are small differences in the measurements.

Attached some pictures of a quick test.

A greeting.

[madires]

@madires, how do you explain the discrepancy between the two testers?

I could only guess without having both testers. The first things coming into mind are PCB layout, bypass caps and current draw of the display backlight and hardware options like the Zener test. Those things are the typical causes for strange readings.

BTW, I read a couple of times Karl-Heinz's "explanation" in paragraph "3.10 Measurement of coils" and unfortunately I have no clue what he was trying to say, maybe in German it made more sense or maybe because I am only an EE.

Please see chapter 5.4.2 "Measurement of small inductances with the sampling method".

[madires]

However between version k and m if there are small differences in the measurements.

Of course there are. Most measurement methods got some minor differences and often different ways of compensation. A few methods differ a lot. And the m-firmware doesn't support the SamplingADC (yet?).

[hapless]

@upsss: I'm afraid mine will also be only a guess, but I think that maybe the clamping diodes on the input have something to do with this. If it's not too much trouble, I'd ask you to remove the six-pin package, re-run self-calibration, and then try to measure the same part(s) again.

[upsss]

@hapless, I gave up trying to fix it and I put the AY-AT back in its acrylic case which is a pain in the neck to take apart and put it back together.  I did try to improve the bypassing of the Vcc regulator and the grounding of the processor but nothing helped.  I don't believe it has anything to do with the clamping diodes on the inputs. 

What I think is happening there is that the Vcc regulator is going into current limit when trying to measure inductors with low inductance and low winding resistance.  The way the tester measures inductors is by measuring the current rise time constant through the inductor and calculating the inductance.  I believe the reason that my T3-T4 doesn't exhibit this problem is that it has been heavily modified, see picture.  It is powered by a li-Ion battery with a step up 5V PWM converter which supplies directly the Vcc to the tester.  This 5V output can supply much higher peak current than the original Vcc regulator on the AY-AT tester.  As a result, I believe ALL original testers will exhibit this problem when trying to measure power inductors, basically any inductor with low inductance and low winding resistance.

 

[hapless]

I have two different ones that don't seem to be affected, or maybe my inductors just don't meet the criteria. Mine use the regulator that they came with, no boost converter in sight. Something makes your tester think that there is a capacitor in parallel with the inductor. Something tells me your tester is not the only only doing that. I also think that it could be due to a slight error in calibration...

I'd like to hear Karl-Heinz's take on this, too.

[madires]

As a result, I believe ALL original testers will exhibit this problem when trying to measure power inductors, basically any inductor with low inductance and low winding resistance.

If Vcc drops too much the MCU's brown-out detection would be triggered. We had that for the Zener test (boost converter) when powered by the same LDO, for example. The low inductance measurement was checked with several different Testers and no issues were found. Of course there could be fringe cases. We can't check all clones and all revisions/versions of each clone. Please add an additional bypass cap and check if the problem is gone.

[upsss]

Any inductor <50uH reads "xxnH if xxnF" any >50uH read fine and it is very consistent.  Otherwise the tester works perfectly.

[hapless]

What happens if you also put a 21nF cap in parallel with those coils (the one you calibrated with)?

Edit: Is there also a lower limit? I tried with a 300nH coil and still can't reproduce.

[indman]

hapless,here example of measurement of the minimum inductivity - 65nH