Resistance and Inductance of Motors and Controller Parts.

liveforphysics · Nov 1, 2010

I bought a very very cool LCR/ESR/ESL meter from the member "Texaspyro" on this forum.

After a little baseline calibration work, I got a pretty decent baseline zero.

Only a machine that uses 4-wire kelvin type clips coupled to a pretty clever instrument can take readings like these.
Here it is checking the resistance of a very short path through a thick chunk of freshly cleaned copper. (trying to get as close to zero as possible)

You notice it reads 85uOhms.
That's 0.000085Ohms, or 0.085mOhms. That's plenty good enough for our purposes with EV parts.

This is how it zeros for inductance
That's 67pico Henrys. That's 0.000000000067Henrys, or just 0.000067uH. That's a small enough amount, it might actually be the correct amount of inductance for a 0.5mm distance apart on a solid copper block. Either way, it's about x1000 more precision than we need for EV applications.

Next Post will contain motor data.

Post after that will be cap data.

liveforphysics · Nov 1, 2010

Motor winding data:

Big HXT 130KV motor from hobbycity 12tooth 14pole:

"8600w" hobbycity motor

Biggest hobbycity inrunner

Various smaller RC outrunners:

liveforphysics · Nov 1, 2010

Cap testing data:

First off, the caps that blew me away are these little "HUG 100V 1000uF" caps. Physically identical looking (other than the brand), and labeled with identical specs to the Cheng-Cheng caps, yet you need over 10x of the Cheng-Cheng caps to equal the ESR of one of these el-cheapo HUG caps, which have some shockingly good test results. I tested about 8 caps, and they all vary about +-15% in ESR and capacity, but even the ones that tested the worst were still outstanding performers.

Here is another great performer for it's size, the ZL series from Rubycon. For such a tiny physical size, these things have awesome specs. You could fit 10-20 of these things in a normal infinion case, and end up with a fantastically low-ripple voltage stable and rugged controller.

If you're looking for the ultimate in voltage stability and protecting the FETs from voltage spikes:

Honda Insight IMA controller cap bank:

Another great surge cap:

Yet another great surge cap:

Here is an example of how all the combined caps on a methods controller read:
You will notice if for example he had used the HUG caps rather than the Jameco caps, his combined resistance would be ~8mOhm rather than 13mOhm, and the controller's ripple voltage would be reduced by 39%. This is where knowing the actual values for caps becomes so handy.

Jeremy Harris · Nov 1, 2010

Outstanding work, Luke. This will really be a tremendous help when it comes to trying to solve some of the thornier problems we have with driving low resistance, low inductance, motors.

BTW, what's the black rectangular 20uF, 3.4mohm cap? Looks like a pretty good thing to have right on the power bus to me, but that probably means it costs and arm and a leg..............

It looks like the best budget option may be an array of those cheap caps with such a good spec. I reckon distributing the caps all along the power bus has to be a good way to go to try and keep ripple right down.

Jeremy

Tiberius · Nov 1, 2010

Yes, well done Luke.

The motor L and R results are particularly useful. Do you know if those models are Y or Delta?

Nick

liveforphysics · Nov 1, 2010

Jeremy Harris said:
Outstanding work, Luke. This will really be a tremendous help when it comes to trying to solve some of the thornier problems we have with driving low resistance, low inductance, motors.

BTW, what's the black rectangular 20uF, 3.4mohm cap? Looks like a pretty good thing to have right on the power bus to me, but that probably means it costs and arm and a leg..............

It looks like the best budget option may be an array of those cheap caps with such a good spec. I reckon distributing the caps all along the power bus has to be a good way to go to try and keep ripple right down.

Jeremy

They are some snubbers made for protecting big IGBT's from freewheeling voltage spikes.
Extremely fast, extremely low resistance, and extremely high ripple current.
BigMoose sent me a box of them.

Very good stuff.

They aren't cheap, or small, but they can offer a level of protection that other caps can't, so when it's needed (like trying to operate with kila-amp phase currents on a BLDC controller), it's the way to go.

liveforphysics · Nov 1, 2010

Tiberius said:
Yes, well done Luke.

The motor L and R results are particularly useful. Do you know if those models are Y or Delta?

Nick

Thank you Nick.

Not a single motor I tested had any visible star connection points that I could see. So, I would ass-u-me they are all all delta.

Since I can measure the resistance and inductance of a single tooth winding set now with some degree of accuracy, It should be fairly simple to do a few sample max copper fill test-wound teeth, maybe a 4-turn, a 8-turn, and a 12-turn, and a 24-turn, plot the results, and we can experimentally confirm our guesses on the non-linear relationship between the inductance and turn-count (should be X^2 I think?). Then, once we've got that slope figured out, it should be fairly easy to tune a motor's wind to achieve the desired performance for the intended operating voltage without a bunch of trial and error runs to get the configuration right.

Tiberius · Nov 1, 2010

liveforphysics said:
... the non-linear relationship between the inductance and turn-count (should be X^2 I think?).

In theory it should be proportional to the number of turns squared. At these frequencies and geometries I'd expect that to be almost exact.

The saturation current will vary as the number of turns, but the measurement will be at low level so it won't see that effect.

What will affect it though is the magnetic circuit. So the inductance of a stator winding will be very different when the rotor is in place. What would be interesting is to see if the meter can detect a slight variation in inductance according to rotor position. I don't know what level you would expect to see. Turning by hand will only generate a few 100 mV so it shouldn't damage the meter but it could upset the reading until it settles

Nick

liveforphysics · Nov 1, 2010

Excellent idea Nick, I will see if I can get a minima and maxima inductance value dependent on rotor position.

Fligh High · Nov 1, 2010

Fantastic measuring device. Lfp I hope you will keep showing more examples of whatever that can be measured with this thing ! because it is really interesting.

Do you think it can show differences in induction in layered board / layered busbar design compared to non-layered or doesn't it work like that ?

For some reason I expect to see space-invaders or pacman appearing on that screen !!! It looks so 80's computer like

It's already worth having in a glass display for the looks alone .....

liveforphysics · Nov 1, 2010

Fligh High said:
Fantastic measuring device. Lfp I hope you will keep showing more examples of whatever that can be measured with this thing ! because it is really interesting.

Do you think it can show differences in induction in layered board / layered busbar design compared to non-layered or doesn't it work like that ?

For some reason I expect to see space-invaders or pacman appearing on that screen !!! It looks so 80's computer like It's already worth having in a glass display for the looks alone .....

Thanks

I love that everytime it turns-on, it thinks it's 1989.

This thing was $$$$ back in the day! Like $25,000usd of 1980's money! According to the asset tag sticker, this one used to be owned by the OWAN generator company.

I would love to check induction equal cross section and linear distance between a laminated bus bar and solid bus bar. If anyone has some litz wire they want measured and quantified for resistance and freq vs linear inductance etc, feel free to send me some, or really any parts you've got that you've not been able to get specs on. Now that I've got it all setup and dialed in, it really only takes a minute to get a reading on most anything.

I actually designed my first ever workbench/desk, which was inspired by getting this meter. Now I will have a home for my scopes and power supplies and meters and things.

7circle · Nov 1, 2010

It would be interesting to see if there are small changes in inductance of motor coils when the magnets are lagging or not sitting in their low cogged energy potential state.

Great you could get to calibrate effectively.

It looks like you can change the test frequency from 1000Hz.

Also wondering if you use a calibrated polycarbonate capacitor in series with a battery you could measure the battery inductance at 1000Hz.

I used to have a job calibrating and repairing a similar capacitor tester called DIGIBRIDGE.
Then the company had got a project to build a production tester. That measured the dissapation factor and capacitance at 1000Hz while super-imposing 50/60Hz 400Vac.

Take care with it don't shove a charged capacitor in it.

liveforphysics · Nov 1, 2010

7circle, you're brilliant!
Know cap value in series to test cell(or pack) ESR! Drool!

liveforphysics · Nov 1, 2010

So, after working all night and getting off work at 10am, I couldn't sleep, so I built the bench that I had sketched up at work.

CAD Sketch:

Starting to make the legs.

Coming along well.

Becoming a home for some tools

I'm happy with the addition. Now I just need to get a chair. (I've never owned a chair, and this is the first desk/workbench I've ever had in my life. lol I've worked just kneeling on the floor for years).

bigmoose · Nov 1, 2010

Note enough bench space.... you will fill it by the end of the year!! :mrgreen:

You must always have 5 sq feet of unused bench space... for safety sake... and so you can buy another instrument!! :mrgreen:

Great job Luke! I got my HP up and running today. I do not have a set of good kelvin leads like yours. I ordered them, but they have yet to arrive. I made my own "almost kelvin" leads up. The reference point is a couple of inches upstream of the clips.

That said, here are the readings on similar capacitors that I promised you. I REALLY love those Sprague 735P low inductance, low ESR caps. They rock on my instrument. I will need to get more familiar with it. I am printing the manual tomorrow. It came with a cal certificate, but I would like to verify that against a couple of known good standards. I think your instrument is newer, and has more precision (reads more decimal places) than mine on the secondary readings. I was playing with mine today on battery ESR also. I got that feature to work. Fooling around with the batteries today, I decided to rebuild 3 of my old drill packs.

Almost forgot, a tip about workbench surfaces. I have found that a 1/4 in sheet of masonite, that brown pressed board. Shiny and hard on one side, and with a back that is textured makes a great bench top. It seems to be reasonable with respect to ESD properties also.

View attachment 5
View attachment 4
View attachment 3
View attachment 2
View attachment 1

johnrobholmes · Nov 1, 2010

You guys really give me tool envy. What an awesome asset.

amberwolf · Nov 1, 2010

liveforphysics said:
(I've never owned a chair, and this is the first desk/workbench I've ever had in my life. lol I've worked just kneeling on the floor for years).

That's usually how I work on stuff, because then I can spread the project around me and work on it, instead of being limited to just a bench or other work surface. Plus, I *have* to pick up my stuff when I get done or else the dogs will play with the bits and pieces...if I used workbenches I would leave stuff cluttered all over them instead. :roll:

Actually, I *do* both of those: use a workbench/table/etc., *and* leave stuff all over it when I'm done, often losing tools and stuff for months or years because I suddenly need the work surface for something else, and have to rake all the previous stuff into a box.... :oops:

liveforphysics said:
If anyone has some litz wire they want measured and quantified for resistance and freq vs linear inductance etc, feel free to send me some, or really any parts you've got that you've not been able to get specs on.

I'm tempted; I have a few of the soda-can-sized caps from various linear power supplies that I am curious how good they still are (some are 30-40 years old, maybe 50, but probably don't have more than a few hundred hours use on them, if that). Also some Elna caps out of Sony 1" VTRs that I'm unable to find specs on, that used to be filters for the brushless tape reel drive motors and power supplies. (there are also lots of solid-state-relays in there that I wish I had a pinout and specs for, but can't find anywhere, and IR doesn't respond to inquiries about them).

Jeremy Harris · Nov 2, 2010

bigmoose said:
Note enough bench space.... you will fill it by the end of the year!! You must always have 5 sq feet of unused bench space... for safety sake... and so you can buy another instrument!!

You beat me too it, Bigmoose! I was going to suggest doubling the size Luke thought he needed.

Jeremy

PS: I'm on my 5th generation workbench come storage rack design; each time I make them bigger than I think I'll need, each time I run out of space within a few months......................

liveforphysics · Nov 2, 2010

LOL!

I was thinking doing a 6ftx2ft bench, with a 6ft x 1.5ft mid-shelf, and a 6ft x 8in top shelf would be over-kill. Soooo much bigger than my lap, which served me well for that last couple decades. lol

Speaking of more equipment, as soon as the wallet recovers, I'm thinking about getting a good bench DMM to go with my kick-ass LCR meter (but I don't want to spend more than a couple hundred

).

Thinking maybe something like this:
http://cgi.ebay.com/Guildline-9578-Solartron-7081-8-Digit-Multimeter-LEADS-/130450253449?pt=LH_DefaultDomain_0&hash=item1e5f70e689#ht_500wt_1156

A computer option like this also seems attractive:
http://cgi.ebay.com/Signametrics-SMX2040-AC-DC-Digital-Multimeter-PXI-Card-/220679156440?pt=LH_DefaultDomain_0&hash=item336180b2d8#ht_1623wt_1153

But then I was thinking maybe just a lot of el-cheap 4.5digit multi-meters like this might be better for me, because it would give me 4 separate channels to observe voltage or current with reasonable 4.5digit resolution:

http://cgi.ebay.com/Lot-4-True-RMS-Digital-Multimeter-DM-441B-/200271212258?pt=LH_DefaultDomain_0&hash=item2ea1182ee2#ht_769wt_1139

Or maybe a kick-ass 4-channel USB logging scope/multi-meter?
Or a scope with ground isolation between channels would be kick-ass too. It's super lame being tied to the same ground plane when you want to over lay scope traces.

liveforphysics · Nov 2, 2010

Here is another of that 8+ digit multi-meter, but with some error message (wait @ F), does that message mean anything to you guys?

http://compare.ebay.com/like/400164472019?ltyp=AllFixedPriceItemTypes&_dmd=Gallery&var=sbar&rvr_id=161825727313&crlp=1_263602_304692&UA=%3F*S%3F&GUID=7c0922221250a0aad0b7ed87fd724b00&itemid=400164472019&ff4=263602_304692

bigmoose · Nov 2, 2010

Luke, I am biased AGAINST computer connected instrumentation. Sooner or later you goof up or the circuit goofs up. At the voltages we are working with, and with potential ground loop problems, the probability of FUBARing the laptop is just too high in my opinion. The dollar value may be the same, but I have invested way, way too much time in the software, files and system build on my laptop.

That said, I got my HP 3456A voltmeter (6 1/2 digits) for a couple of hundred bucks, and I didn't wait and shop for the best price. All these old HP items have the GPIB bus so you can automate your readings if you want with Labview software, which is what I use.

Also I have secured auction material from these two sites. Trouble is they do not ship. You or someone must go and retrieve it. Most of my power supplies are from these auctions as was my tab spot welder. Statistically 80% work out of the box, 15% need power supply repair, and 5% is totally trashed and fit for the dumpster. You know I'll go and retrieve for you within my driving area.

http://gsaauctions.gov/gsaauctions/gsaauctions/
http://www.govliquidation.com/index.html

liveforphysics · Nov 2, 2010

bigmoose said:
Luke, I am biased AGAINST computer connected instrumentation. Sooner or later you goof up or the circuit goofs up. At the voltages we are working with, and with potential ground loop problems, the probability of FUBARing the laptop is just too high in my opinion. The dollar value may be the same, but I have invested way, way too much time in the software, files and system build on my laptop.

Ehh, I've got 3 outdated netbooks just laying around. I run only open-source or cracked bootleg software, and the data I want on them is also copied other places, so it's not really like I loose anything other than a paper-weight if it fails and wrecks a netbook. But I do see your point there my friend.

bigmoose said:
That said, I got my HP 3456A voltmeter (6 1/2 digits) for a couple of hundred bucks, and I didn't wait and shop for the best price. All these old HP items have the GPIB bus so you can automate your readings if you want with Labview software, which is what I use.

Also I have secured auction material from these two sites. Trouble is they do not ship. You or someone must go and retrieve it. Most of my power supplies are from these auctions as was my tab spot welder. Statistically 80% work out of the box, 15% need power supply repair, and 5% is totally trashed and fit for the dumpster. You know I'll go and retrieve for you within my driving area.

http://gsaauctions.gov/gsaauctions/gsaauctions/
http://www.govliquidation.com/index.html

What a cool site! I will keep an eye out on the lots for awesome things.

Thinking about it more, it does seem like a 6.5digit meter would be more than I would really ever need, and those HP's do seem to have great specs. I might just pick one up. (after next payday, lol, right now I'm pretty short again. lol)

Ricky_nz · Nov 2, 2010

One of my workmates had a dell laptop and somehow and he somehow got 24V between 2 of its ports (not sure which).

He had the accidental damage insurance option that was part of the deal when he brought it.

The surprising bit was that they covered the replacement mother board

.
I thought that insurance was for things like dropping it not using it as a resistor!

Edit: I prefer to keep things separate or only have one connection to the laptop.
I'm going to buy a USB isolator to use when testing my motor controller.

grindz145 · Nov 2, 2010

I need one of these LCR meters. It's officially on the EBay shopping list. :twisted:

Jeremy Harris · Nov 2, 2010

I've just done some quick calcs on the current ripple versus PWM frequency for that HXT motor, adding 10mohms to allow for lead resistance, FET resistance and cell Ri (probably an underestimate). Anyway, for 40mohms and 26uH I get the following plot of current ripple vs frequency:

I'm guessing that we want the current ripple to be below about 10%, which is easily met by the 15kHz PWM frequency of the XieChang controllers. This is reassuring for those looking to use large format FETs, because generally the big packages don't switch very fast.

Thanks for the raw data, Luke. As soon as I have a bit more spare time I'll look into this in more detail. Knowing how the inductance varies under different conditions would be very useful, but probably hard to measure. I have a suspicion that the inductance will massively drop as the core gets close to saturation, which may be the reason we see those very large current spikes that trip the XieChang over-current protection.

Jeremy

Resistance and Inductance of Motors and Controller Parts.

100 TW

100 TW

100 TW

100 MW

10 kW

100 TW

100 TW

10 kW

100 TW

100 mW

100 TW

100 W

100 TW

100 TW

1 MW

10 MW

Administrator

100 MW

100 TW

100 TW

1 MW

100 TW

10 kW

1 MW

100 MW

Similar threads