Testing with an LCR meter

[Stephen1966]

A short while ago I purchased a cheap handheld LCR meter, the DER EE DE-5000, and I'm now start to take some useful measurements with it which I would like to share.

A 9V powered LCR meter isn't going to offer the same range of functions or capabilities as a spec'ed-out bench-top meter costing a hundred times as much. This much is guaranteed! I bought it originally, to perform some rudimentary analysis of a new output transformer but I was also looking for a cap tester with a little more functionality than my DMM and in that area I am very happy. And for a tester costing as much as a good DMM for anyone (i.e. someone like me) who isn't in a mass production environment or as rich as Croesus, I can say it provides a nice step up from accepting on faith the data (if there is any at all) in datasheets and a ballpark device that is going to allow us to make some more informed choices when dealing with components and troubleshooting.

https://www.deree.com.tw/de-5000-lcr-meter.html

I'm going to use this post to lay out some of the testing I've been doing with it, to describe how I set the device up to take measurements and to record some of those measurements and (wish me luck!) the maths that go with it to arrive at usable data when building amps.

Please feel free to contribute.

[neskor]

I have one, best purchase in a while

[sluckey]

I bought a Proster BM4070 from Amazon last year for only $36. Played with it a couple days. I like the flip-up display. Been in my closet ever since then.

[Stephen1966]

I bought a Proster BM4070 from Amazon last year for only $36. Played with it a couple days. I like the flip-up display. Been in my closet ever since then.

What I like about the DE-5000 are the kelvin clips. When you get it fresh out the box it has these alligator clips on cables 2 or 3 cm long. One of the first things I did was grab a pair of kelvin clips and do the mod.



Much more practical...

Time to upgrade Steve?

[sluckey]

Time to upgrade Steve?

Oh no. I just bought it to play with because I was bored.

[pompeiisneaks]

I bought that one too, and have planned on the mod to the clips at some point too. Really dig mine.

~Phil

[cys]

This thread is timely for me as I picked up a DE-5000 a few weeks back, ostensibly for more capacitor testing capabilities. Have yet to use it.

[Stephen1966]

Testing caps is straightforward.

First calibrate the meter.

We want to check the dissipation factor (tangent delta) to see if the cap is good or not.

I'm testing a CDE 47uF/63V electrolytic - 476TTA063M

We need the datasheet:

TTA-2944741(47uF-63V).pdf

In the datasheet, we look for the Test frequency (120Hz) and the voltage rating for our cap WVDC 63 (V) [Edit: it's highlighted in the dataheet]

We take the meter out of Auto mode, into capacitance mode and set the frequency to 120Hz.

We are looking for a dissipation factor (tangent delta) of .1 or better.

1673531233344.jpg

This cap is good.

ESR and phase angle can be found there as well.

[Stephen1966]

At the moment, I am testing my power transformer and I want to share with you what I think the LCR meter is not good for.

Testing winding ratio

This is normally done by applying a voltage on the primary and measuring across the secondaries. The advantages (or disadvantages) of this is that

• The resulting voltage gives the actual voltage of the transformation

• You risk overloading your DMM if the secondary voltage is high

• The reading will include core, resistive, and capacitive losses

• The calculation of ratio, Vs/Vp = Ns/Np, will be greater than the actual turns ratio

It's therefore suggested that the turns ratio can be calculated from the primary and secondary inductance.

https://www.hioki.com/global/learning/c ... -test.html
https://www.hioki.com/global/learning/u ... ers_8.html
https://youtu.be/dabYCMeZBQs

Primary inductance (Lp) is measured across the primaries with the secondaries open; and secondary inductance (Ls) is measured across the secondaries with the primaries open.

Testing mine at 100Hz and with the primaries being the 230V tap to common. I arrived at

• Lp = 1.344H

• Ls = 8.049H

The formula I then used, was: Ns/Np =√(Ls/Lp)

Substituting this, I get:

Ns/Np =√(8.049/1.344) =√5.989 = 2.447

However, I know the transformer is rated for 700V (350/0/350) on the secondaries and I should be arriving at a factor of ~3 to give me the required transformation from 230V:700V. Sure enough...

230 x 2.447 = 563V

I can fudge the numbers if I reduce Ls to around 63% of its value (0.85H) and this gets me to 708V but this is not good enough.

The transformer, the Tube Town, tt-dmb100-pt, (https://www.tube-town.net/ttstore/tt-du ... ormer.html) also has taps for 120V and 240V so out of curiosity, I ran the tests again but with the 120V Lp being tested at 120Hz to account for the N. American supply. The numbers were even worse.

120V @ 120Hz, Lp = 0.45H

• Ns/Np = √(8.049/0.45) = √17.88 = 4.23

4.23 x 120 = 507.6 V

240V @ 100Hz, Lp = 1.485H

• Ns/Np = √(8.049/1.485) = √5.42 = 2.33

2.33 x 240 = 559.2V

I've seen another way of calculating the turns ratio measuring the impedance value Z on the primary side after connecting a resistance to the secondary side, (N = √[R/Z]) - see links above - but because of the magnetic leakage, I believe this would at best, be an approximation too.

Merlin recommends calculating the turns ratio from voltage as well and if you have a variac, you can keep the voltage low enough on the secondary side to avoid blowing up your DMM.

This was an experiment just out of curiosity and the figure I am really interested in is actually leakage inductance (LL) which is going to contribute to my experimental design of an RC snubber for the rectifier in my 183. Still it was a good opportunity to find my way around the power transformer and I want to supplement this with a test for the mutual inductance, which I am hoping will give me some indication of the efficiency of the transformer.

The formula, Ns/Np =√(Ls/Lp) is actually an inversion of the formula Np/Ns =√(Lp/Ls) I found for step down transformers - I couldn't actually find the formula for step up transformers - so it's entirely possible it is my shaky maths that are the real problem. If you know any better, I am more than happy to stand corrected. And who knows, with a better formula, I might arrive at better results.

[cys]

The inductance in the formula Ns/Np =√(Ls/Lp) needs to be the magnetizing inductance. The inductance measured in the primary with the secondary open is the total inductance, which includes both the magnetizing inductance and the stray inductance. If one measures the stray inductance of the primary with secondary shorted and subtracts that from the total inductance measured with the secondary open, one gets the magnetizing inductance of the primary (Lmag-p). Doing the same measurements on the secondary will give Lmag-s. Hopefully this will give the desired result...

[Stephen1966]

The inductance in the formula Ns/Np =√(Ls/Lp) needs to be the magnetizing inductance. The inductance measured in the primary with the secondary open is the total inductance, which includes both the magnetizing inductance and the stray inductance. If one measures the stray inductance of the primary with secondary shorted and subtracts that from the total inductance measured with the secondary open, one gets the magnetizing inductance of the primary (Lmag-p). Doing the same measurements on the secondary will give Lmag-s. Hopefully this will give the desired result...

Thanks. Let's see if I have this straight then...

This is a schematic of the transformer:

tt-dmb100-pt.jpg

Shorting one side and measuring the other, and vice versa is the same as measuring for leakage inductance. So if we do that

Np = Lp - LLp = Lmag-p
Ns = Ls - LLs = Lmag-s

When Np = primary windings; Lp = Inductance measured on the primaries with the secondaries open; LLp = Leakage inductance measured on the primaries with the secondaries shorted. And likewise for the Secondary side...

And then,

Ns/Np =√(Lmag-s/Lmag-p)

Further thoughts... would you short all the auxillary taps: the filament, bias and 12V windings, too?

[cys]

Yes, I would short everything. Also, it would be best to solder the shorts rather than use clips as otherwise (more) error is likely to be introduced, see: https://www.voltech.com/support/technic ... nductance/

[Stephen1966]

Yes, I would short everything. Also, it would be best to solder the shorts rather than use clips as otherwise (more) error is likely to be introduced, see: https://www.voltech.com/support/technic ... nductance/

That's great thanks. The same page is also offered as a pdf

measuring-leakage-inductance.pdf

I was thinking about the soldered short as well. Great minds think alike and all that... For meters in a production environment the idea of an offset setting in the meter is good. Not sure we need it at our scale, or if this meter will do something like that, but there are some settings I haven't explored yet. The 'Rel %' button might do something like that.

It has to be tomorrow now but today I took delivery of a AC/DC 9V power supply. The official power supply is close to a hundred bucks if you can find one, and this one cost less than a cinema ticket. https://www.datart.cz/napajeci-adapter- ... n3110.html The advantage of the AC/DC supply is that it disables 'APO' Automatic Power Off. Meaning that after you calibrate it, you don't have to keep fiddling with it every few minutes between measurements to keep it switched on AND CALIBRATED. It's probably going to require another mod though. Normal 3.5mm barrel jacks seem to come only in 10mm and 9mm length flavours. The official wall wart has a barrel length of 12.5mm and 10mm isn't quite long enough to sit in the socket securely, though it has been sitting here quite happily for the last fifteen minutes or so with no issues. Depending on the kind of jack plug you find, it won't take much to mod either the plug or case around the socket. Highly recommended you get one. The jack is centre positive: 3.5mm outside diameter, 1.35mm inside diameter.

[Edit: with a scalpel, I carefully shaved off some of the sleeve around the DC jack. It now fits securely and I can move the meter around, it's no longer an intermittent contact and in fact, it takes a little effort to remove the jack from the socket. Hint - first, I practiced on a spare to ensure it wasn't going to destroy the jack.]

[Stephen1966]

Taking account of the magnetizing inductance (Lmag) then I can't say the results are any better... This part of the experiment though, gave me a chance to observe the leakage inductance (LL) which is going to become useful later on.

Today's primary and secondary inductances (Lp and Ls) are somewhat higher than they were yesterday. I don't know if this is due to the arrangement of cables coming off the transformer or because the meter is now operating at a full 9V from the power supply or because the core has become somewhat magnetized from running the tests. Calibration was done and this time, I let the meter run for a minute or so to let the readings settle down and for the leakage inductance tests, shorts were made with solder joints (not just twisted together). It's hard to say, I just ran the test for LLs under battery power, and got the same result as from the wall supply.

L - LL = Lmag

Ls (@ 100Hz and 120Hz) = 9.2H

Lp (120V @ 120Hz) = 0.45H
Lp (230V @ 100Hz) = 1.387H
Lp (240V @ 100Hz) = 1.482H

LLs (@ 100Hz and 120Hz) = 0.054H

LLp (120V @ 120Hz) = 0.002H
LLp (230V @ 100Hz) = 0.005H
LLp (240V @ 100Hz) = 0.005H

Lmag-s = Ls - LLs = 9.2 - 0.054 = 9.146H

Lmag-p (120V @ 120Hz) = Lp - LLp = 0.45 - 0.002 = 0.448H
Lmag-p (230V @ 100Hz) = Lp - LLp = 1.387 - 0.005 = 1.382H
Lmag-p (240V @ 100Hz) = Lp - LLp = 1.482 - 0.005 = 1.477H

Feeding the magnetizing inductances into the equation to find the turns ratio (Ns/Np)

Ns/Np =√(Lmag-s/Lmag-p)

120V
√(9.146/0.448) = √20.42 = 4.52
120 x 4.52 = 542.4V

230V
√(9.146/1.382) = √6.62 = 2.57
230 x 2.57 = 591.1V

240V
√(9.146/1.477) = √6.2 = 2.49
240 x 2.49 = 597.6V

All of the taps were open when I measured Ls and Lp. All the secondaries were shorted when I measured the primaries (the unused primaries were left open).

1673708873118.jpg

1673708891877.jpg

It isn't a very impressive set of results but I am hoping the leakage inductance on the secondaries is somewhere close to actual and will allow me to take the rectifier design a little closer to spec.

We might be able to find more useful results if we test a transformer with less complicated windings. As far as determining coil ratio though, it looks like voltage-out / voltage-in (Vs/Vp) is a more reliable method.