Accurate inductance measurements on the cheap

[SingedFingers]

I am currently doing this, which is horrible:



I tune to the approx frequency of the circuit in question, record F, solder the inductor on the test jig (deadbug FR4), then adjust the air variable capacitors until the scope shows lowest voltage trough. There are two air variable caps in parallel, one 380pF and one 20pF which are on a calibrated scale made using a capacitance meter in my UT61E. Then I plug this into my calculator and solve for L. I can get a reasonably accurate figure down to about a 50nH resolution which isn't bad for a pile of junk.

However this is incredibly time consuming and requires a large rig set up and I have a project which involves winding about 20 of them soon and I just don't fancy the job.

I came across a pile of cash and decided, hey I'll get one of those nice AADE meters that everyone seems to have. Nope, turns out the originator is unfortunately deceased and they are out of production.

Ergo, is there another method of doing this on the cheap, be it some Chinese product, a quick hack job or something that means I don't have to set this up every time? I've seen a few projects on the web which involve a PIC16F628 and some discretes which are possible but none of them seem to have been reproduced or verified in any way.

Ultimately I need to measure HF band toroids, preferably around the operating frequency to a reasonable level. I can trim out the rest worst case when peaking filters etc.

[salbayeng]

I tried to do something similar to try and find the inductances to use in a 433MHz Tx, of the order 20-70nH, I used a spectrum analyser and a little jig, but couldn't really get any obvious sharp resonances.

Your method

the scope shows lowest voltage trough

is much more accurate if you use a two channel scope and look for 90 degree phase shift (or use lissajous and look for the circle between a left leaning ellipse and a right leaning ellipse). The problem with the method is the very low Q due to 50 impedances. I've used the 90deg method to investigate the inductance vs current of some power line coupling transformers I was building, and it's better than 1% relative accuracy.

The old school method uses a GDO , you just parallel your unknown coil with a known 1% silver mica, put it on a piece of wood and look for a dip on the GDO , you can couple a few loops around the GDO coil to a freq meter too.
https://en.wikipedia.org/wiki/Grid_dip_oscillator

[DimitriP]

...pretend you designed a board for this and announce you will start taking orders "soon"
...wait for a chinese reproduction...
...order it from ebay for 12.95

(or just build it and don't anyone about it )
http://www.qsl.net/va3iul/L_meter/L_meter.htm
 

[HeyTom]

If you want really cheap, I made this from some pic based projects I found and combined.   The results are consistent, but I'm not sure about the absolute accuracy.

Really, the pic is used for frequency counting and display, so any micro controller could be used.  The main part is the LM311 and C1 (which has to be known)  See attached.  I have Kicad files, and MPLab project  I can post them here if there is any interest.

It's pretty small and uses a rechargable LiPo battery.

[cdev]

My first suggestion would be to check out the web pages for the AVR Transistor Tester project. Not long ago I read that the developers were working on some improvements to the firmware which improved the measurement performance on low value industors quite a bit. Maybe that is working now. If so, thats likely to be your best bet.


There is a fairly inexpensive inductance meter that is recognized to be very good for the money which can be bought or around $125-150 USD from a Japanese firm. (DE-5000)

For somebody who is serious about inductance measurement at some point it makes sense to plunk down the money and buy one, as it seems to be an exceptionally good value..

[cdev]

You can use a DDS function generator and a scope or thrown together diode RF detector probe and a multimeter to measure the response of filters. You could also do something similar with a Raspberry Pi's GPIOs.

[vk6zgo]

I am currently doing this, which is horrible:



I tune to the approx frequency of the circuit in question, record F, solder the inductor on the test jig (deadbug FR4), then adjust the air variable capacitors until the scope shows lowest voltage trough. There are two air variable caps in parallel, one 380pF and one 20pF which are on a calibrated scale made using a capacitance meter in my UT61E. Then I plug this into my calculator and solve for L. I can get a reasonably accurate figure down to about a 50nH resolution which isn't bad for a pile of junk.

However this is incredibly time consuming and requires a large rig set up and I have a project which involves winding about 20 of them soon and I just don't fancy the job.

I came across a pile of cash and decided, hey I'll get one of those nice AADE meters that everyone seems to have. Nope, turns out the originator is unfortunately deceased and they are out of production.

Ergo, is there another method of doing this on the cheap, be it some Chinese product, a quick hack job or something that means I don't have to set this up every time? I've seen a few projects on the web which involve a PIC16F628 and some discretes which are possible but none of them seem to have been reproduced or verified in any way.

Ultimately I need to measure HF band toroids, preferably around the operating frequency to a reasonable level. I can trim out the rest worst case when peaking filters etc.

You have the wrong variable---used a fixed capacitor & vary the sig gen frequency.
Ditch the 50 Ohm termination,too!

Don't be too fussed about accuracy--commercially made inductors often vary 5-10% or more,between individual examples.

Another way is to look at a LR time constant curve,driving the LR circuit with a square wave.

[SingedFingers]

Thanks all. Will read this thread over the day.

I used the 50 ohm termination as it gave a sharper dip at resonance. This might be an error. Good idea with the fixed capacitor. I will try that. I can switch a couple of ranges in and out.

I have two generators; a cheap manual sweep generator and a 2019A. The latter is useless for sweeps as it has no GBIP and the former has low resolution despite having a 3:1 reduction on it. Wondering if I should lose both of these and just get a reasonable DDS.

[kerrsmith]

I built an inductance meter using an idea based on the following W2AEW video (#199: Measuring coil inductance and IF transformer resonant frequency):

https://youtu.be/Ff5xOENID7w

I modified the circuit so it worked with an Arduino by adding a buffer circuit then a frequency divider using an 74HCT393 (to bring the frequencies down to something the Arduino could measure), it also had a LCD display so you could read the inductance directly. I tested it with lots of different value inductors and it does seem to be quite accurate and gives a measurement quite close to the value they should be (and as mentioned above commercially made inductors often vary 5-10% or more). I was quite pleased with it considering it was built with ordinary (cheap) components.

I also bought the following inductance meter kit (I like building kits):

http://www.banggood.com/DIY-M8-LC-Digital-Inductance-Capacitance-Meter-Kit-p-976437.html?p=2D100314686672015046

This kit also measures capacitors which is quite handy and again it seemed to measure values pretty close to the indicated ones on the components tested. For my circuits as long as the values measured are about right I am happy as I always need to tweak things anyway.

[TheDane]

...wait for a chinese reproduction...
...order it from ebay for 12.95

Atmel based component testers can be found online, already built or as kits.
There are closed source firmware, and open sourced versions - depending on the version/features the unit offers. (Frequency counter, 'high-voltage' measurement, etc.)
- but if you can program the Atmega µP, you can load anything into it 
It uses the bare GPIO of the µP and math magic to measure, so do not load it with a charged capacitor - it will blow the controller!

It's possible to for the ATmega168/328 version to measure inductances of 0.01 mH to 20 H (can be detected and measured) - a 5 times improvement to your current setup , and a LOT easier 
(if you mean resolution=range - I can't see specs for inductance resolution, but resistance resolution is 0.01 if the ESR is below 10ohm)

http://www.mikrocontroller.net/articles/AVR_Transistortester
Google "2016-DIY-KITS-ATMEAG328P-M328-Transistor-Tester-LCR-Diode-Capacitance-ESR-meter-PWM-Square-wave-Signal" for kits and inspiration

[G0HZU]

Ideally you want to measure the inductance at the operating frequency. One alternative/cheap option would be to make a bridge or a directional coupler and measure the phase angle of the reflection coefficient (on a dual channel scope?) at the wanted frequency. Then with a simple excel sheet you could work out the inductance from the phase angle. This way you don't need to resonate it with a capacitor but you would need to make a bridge or a coupler and have two bits of coax the same length to feed to the scope. Or replace the scope with an old vector voltmeter but these are big and can be expensive.

I would use a modern VNA for stuff like this but you can also buy the cheapo USB based VNAs for a few hundred pounds or make one of the N2PK VNA kits. The N2PK version only works to about 60MHz but it is a decent instrument from what I've seen. A good VNA will also be able to measure the Q of a typical toroid with acceptable measurement uncertainty. Not sure if the N2PK VNA could manage this as I've never used one.

[chris_leyson]

I am currently doing this, which is horrible:

As you a driving a 50 load the series resonant would be better than parallel. I found this technique works well for larger inductors http://www.dos4ever.com/inductor/inductor.html but I haven't tried it on anthing less than a few uH as the fuction generator I use will only manage 4MHz.