TeensyLCR – A Teensy based DIY LCR Meter

[HaiLun]

Very nice project. 

I'd like to add a good transistor hFE tester to this.  Something that could test with base current > 1.0 uA, VCE > 1 volt and read betas higher than ~600.  These are the parameters of most of the current transistor testers.

It would be great if it had a thermistor attached to the transistor and it measured hFE at different temperatures as the transistor heated up.
 

VCE > 1   ………… Are you sure 1V is enough?

[drksy]

What's the total cost to build one from scratch, only counting the electronics, no enclosure, no kelvin alligators or other mechanical/misc stuff?

[WeTec]

I'd like to add a good transistor hFE tester to this.  Something that could test with base current > 1.0 uA, VCE > 1 volt and read betas higher than ~600.  These are the parameters of most of the current transistor testers.

I had the idea to add a DC output like Agilent E4980A option 001. This can also be used to measure RDSon of MOSFETs. A DAC module could easely be implemented. I had this in mind: DFRobot GP8211 1-Channel 15-bit I2C to 0-5V/10V DAC Module (https://www.dfrobot.com/product-2753.html).

[WeTec]

What's the total cost to build one from scratch, only counting the electronics, no enclosure, no kelvin alligators or other mechanical/misc stuff?

• PCB with most of the passives preassembled: 80$ (incl. shipping+taxes) for 5 pcs.
• Display: 10$
• Teensy 4.1: 32$ (was a christmas pressent, so 0$ for me)
• Most of the ICs where salvaged from other PCBs or samples. So 0$ for me. About 87$ at Digikey for 1 unit.
• BNC connectors: 16$ (Salvaged from other PCBs, so 0$ for me.)
• Cheapo 4x4 keypad: 1$. I used a selfmade keyboard PCB: 7$ for 5 pcs.
• rotary encoder: 2$

I paid about 100$. In addition to the fun of building it, it was absolutely worth it. If you are going to build it from scratch it will cost about 235$ + additional cost for kelvin alligators clips and other stuff. Depends what is laying around and in the junk box. 

[WeTec]

Some improvements have been made:

• Frequency resolution is now 10 Hz (was 100 Hz).
• Minimum frequency is now 10 Hz (was 100 Hz). Now one can measure components at mains frequencies like 50/60/100/120 Hz.
• Added list sweep function. Perform automatic sweep measurement by sweeping the frequency linearly or logarithmically. Users can view data by a table and obtain up to 100 points in one sweep. If a USB flash drive is connected, sweep data is stored as text file automatically.
• OPEN and SHORT correction function performs correction over 38 pre-set frequencies. The correction for all other frequencies are interpolated from those pre-set frequencies. The correction data is stored permanently on the EEPROM.

[elwar85]

Very cool project. I'm interested in building one of these but would like to make some adjustments to the PCB form factor (to fit into a specific handheld enclosure). I gather it was designed in EasyEDA. Would you consider adding the schematic/PCB files to the repo?

[WeTec]

Would you consider adding the schematic/PCB files to the repo?

Schematic and Gerber files are on Github already. Recently I made some adjustments to the built-in calibration routine and the hardware (tuning some resistor values). The repo is not up to date yet. I'll keep you updated.

[elwar85]

I've seen there's a PDF of the schematic and the Gerber files, but this doesn't let me make changes to the layout. I'd have to rebuild both in EasyEDA to be able to change the PCB dimensions and move things around a bit. Was just hoping you might be able to save me that job!

Will await your update before I attempt a build anyway, if there's been hardware changes.

[coromonadalix]

@elwar85   you're still lucky to have source code and schematics ....


What would be the Teensy minimal board revision(s)  the latest 4.1  that i would need to buy     or some older 3.5 3.6 ? i have one of theses laying somewhere


thks and continue the work   

[WeTec]

Recently I've updated the calibration routine. Everything you need is a pretty good AC voltmeter with 1VAC range (I recommend to use 34461A ) and a couple of 0.1% resistors. You first adjust the output with the AC voltmeter. Voltage and current ranges are adjusted by using 0.1% precission resistors and the output. I achieved good accuracy after the adjustment. The reading error is less than 0.5% of all 0.1% resistors I have. Detailed results will follow.

In order to achieve higher output voltage levels, changes are made to the output and input filters. Check Github for the latest schematic. Output level can now be set between 0.03 Vrms and 2.5 Vrms with 1 mV resolution.

[WeTec]

What would be the Teensy minimal board revision(s)  the latest 4.1  that i would need to buy     or some older 3.5 3.6 ?

I recommend to use Teensy 4.1 because of way more CPU speed. I do a lot of calculations every 200ms. The floating point unit helps alot, but I'm not sure if the older models are fast enough.

[WeTec]

You don't actually need a voltmeter like the 34461A for adjustment. Your regular handheld multimeter from the drawer is sufficient to adjust the output voltage of the LCR meter. The input voltage and current ranges are adjusted by applying a voltage to the 0.1% resistors. So the error of the voltmeter is equal (and with same sign) on the voltage and current measurement. The error is compensated by applying Ohm's law.
I used my Extech EX330 to adjust the output. The EX330 has an accuracy of 1.5 %+-3 digit at the 4VAC range.
I did some measurements with 0.1% resistors from 100 Ohm to 1 MOhm @100 Hz, 1 kHz and 10 kHz test frequency. For this purpose I built four resistance decade boxes with 0.1% resistors (one decade per box, 100 Ohm, 1k, 10k and 100k).
100 Hz: Rs error of all results (100Ohm - 1MOhm) is <= 0.1%.
1 kHz: Rs error of measurements up to 600 kOhm is <= 0.1%. From 700 kOhm, the error increases to -0.25% at 1 MOhm.
10 kHz: Rs error of measurements up to 90 kOhm is <= 0.1%. From 100 kOhm, the error increases to -8% at 1 MOhm.

[WeTec]

I still have some issues to get precise results of the phase between voltage and current. Especially at test frequencies >10kHz and very low or very high impedances. The main reason is the phase error of the gain stages. A phase measurement with an oscilloscope (probe before and after the gain stage) confirmed the simulation.

[WeTec]

In fact, the parasitic capacities are way more complex than shown in the image above. For instance, there are some additional capacities at the analog switch. Hard to simulate. However, I have determained some phase correction values by measuring the phase of resistors (where I assumed no reactance). I use this values to compensate the phase deviation over frequency and gain setting. The overall phase error is not perfect but good enough to make some considerations about accuracy. I have created an accuracy chart which shows the impedance accuracy over frequency. This is based on resistor measurements only, but I assume that capacity and inductance results ar not far from that. Furthermore, I decided to specify only up to 80 kHz. 90 kHz is possible but, well, not very precice.

[viveksingh]

Can you share some video or working of this project (TeensyLCR meter)?

[WeTec]

Sorry, I don't have sufficient video equipment for recording. Working is very similar to use a regular DMM. The device is ready for measurement a second after power up. Default settings after power up are

• Function: Cs/Rs
• Level: 1 Vrms
• Frequency: 1 kHz

Just connect your DUT and results are displayed. Changing settings is easy via touch menu. Values can be entered via keypad.