Author Topic: temperature stable inductor design  (Read 6563 times)

0 Members and 1 Guest are viewing this topic.

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
temperature stable inductor design
« on: February 06, 2018, 06:00:36 pm »
I have posted this in the general section but i feel its appropriate for this forum.

I did some research and i found a general rule that a inductor will have a temperature coefficient of inductance about 70 percent of its mu in ppm. I tested two inductors i had and found that they had approximately 500ppm drift.

I did find inductor materials that have a mu as low as 14. If the book i found is correct then it might be possible to make something like 10ppm inductors by using these cores.

However forum members mentioned that mechanical winding stress, drift of leakage inductance due to wire position and air gap distance from core expansion might have strong effects. I believe that these parameters lead to high nonlinearity and overall contribution to the drift.

Furthernore there is the drift of interwinding capacitance due to the tempco of the dielectric properties if the glue used to hold down the windings. I also thought that the enamel used over the wire and the plastic or epoxy coating of the inductor core could lead to drift that lasts in the period of days... Plastics are not very stable.

So, my plan was to obtain some low mu core materials and run some tests.

First off, do you think i will get better performance with torroidal or cylindrical cores? Does the airgap stabilize things?

Do you think that trying to silver plate windings on s core is a worthwhile persuit? I recall a professor telling me that in the v2 rocket there were inductors that were basically silver plated windings on ceramic cores for high stability. I do not know if this was a design decision to increase viibrational resistance or if it was better in other regards for wire i.e. it is the only way to achieve a high vibration resistance air coil with minimum dielectric between the windings.

This lead me to think that it might be worthwhile to plate a winding on a conventional core. I am not sure about how to go about doing it but i thought it might be more stable then enamel wire in the ppm region. This way there are no mushy materials between the conductor and the core, save for the paint coating of the inductor. Any ideas how to silver plate epoxy or paint with a durable coating suitable for precise electonics?

Im not too sure of where i am going with this, i thought that my low drift inductor standard should be an air core inductor, so i could see how things behave without extra mu but i am not sure how to fix it i.e. some kind of bobbin? I imagine if its soldered to a pcb there would be more stress then if it was hanging off some aligator clips in my thermal test chamber.

I have a feeling that with different ferrite materials there may be strong variance in the drift contribution from different factors and there might not be an all fits one solution.

Any insight? And i also heard some core materials i.e. iron powder will age, which is no good.
 

Online Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: temperature stable inductor design
« Reply #1 on: February 06, 2018, 06:25:02 pm »
Might do a search on the General Radio standard inductors. Could be wrong, but I think they were toroids wound on inert ceramic forms. IET still sell's 'em. I also remember something about NIST doing inductors with the wire laid into grooves ground into cylindrical inert ceramic tubes.
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: temperature stable inductor design
« Reply #2 on: February 06, 2018, 07:02:20 pm »
I should hook some inductors to 4263B, put them into TEC-box and run some tempco measurements. Any ideas for measurement settings? :)
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: temperature stable inductor design
« Reply #3 on: February 06, 2018, 08:44:28 pm »
I just do a sweep of 20c and measure the average, my box is not advanced enough at the moment to do logging or setpoints, simple power adjust.

Problem with ceramic forms is I want stable inductors in the uH region so I can make higher order stable passive filters in the 100KHz-1MHz range, I need to determine if there are better solutions then the gyrator.
 

Online Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: temperature stable inductor design
« Reply #4 on: February 06, 2018, 10:01:32 pm »
Small coils of heavy gage Invar wire?
 

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: temperature stable inductor design
« Reply #5 on: February 06, 2018, 10:03:00 pm »
I just do a sweep of 20c and measure the average, my box is not advanced enough at the moment to do logging or setpoints, simple power adjust.

Problem with ceramic forms is I want stable inductors in the uH region so I can make higher order stable passive filters in the 100KHz-1MHz range, I need to determine if there are better solutions then the gyrator.

What are the requirements? Inductance range?  Current? At what frequency? What level of a DC current should be tolerated? Size? Temperature range? Q?

Cheers

Alex
 

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: temperature stable inductor design
« Reply #6 on: February 06, 2018, 11:49:28 pm »
I'm not sure. I wanted general materials properties/design information for my experiments.

I wanted more of a lay of the land of inductor stability rather then a specific design. Kind of like the information found in keithleys low measurement handbook. There does not appear to be much information about inductors... for instance this forum is crazed about stable resistors and there is quite a bit of information available about capacitors already..

there just seems to be some hatred of the inductor passive in the precision community, everyone would rather avoid its use with creative ways then explore it because it seems to be the most nonlinear, unstable and generally unwieldy device in electronics.. just look at how much effort was put into gyrators and switched capacitor filters. It seems to be relegated to either coarse measurements which leave people satisfied with some small amount of dB drift (which is large in ppm) for radio purposes, where signal strength does not matter so much, so long the information is conveyed, or for power electronics purposes, or for HF rejection/EMI applications where its basically used to give someone confidence in passing a standard or to keep something kinda working in case there is a problem to prevent catastrophic operation modes.

I would like to make a playbook that gives you options in using the inductor in passive filters for measurements that voltnuts would find interesting. If you read the literature about resistor stability, you get a good idea of what the 'theoretical limits' for stability are for resistive dividers, etc. But nothing of the sort really exists for inductors. I thought that there might be some areas where they are the superior choice, i.e. in circuit LCR measurement at high frequencies with other high level signals present for active impedance analysis, so that a device does not get swamped by dynamic range violations in an ADC or clipping a fast active filter. Not to mention they provide very robust protection for band limited AC measurements.

The idea for using low TCR wire to wind the inductors is a good idea, they would limit the dimensional changes and stresses imposed by the wire, however it is lossy right?

I am interested in mu core materials because of the large size of air coiled inductors, I would like to focus on at least somewhat methods practical for regular circuit use. It would be hard to justify using something the size of a shoe box to make a front end for a LCR meter or something, in most cases... >:D

I guess its mostly because in the region where inductors are wanted, dB measurement starts being used, which is typically far coarser. No one measures 0.001dB.

Also, the use of passive filters would remove intermodulation distortion products in the semiconductor ADC, preamp, active filters, though I am not sure how measurable they are at lower frequencies.
« Last Edit: February 07, 2018, 12:15:15 am by CopperCone »
 

Offline ap

  • Frequent Contributor
  • **
  • Posts: 282
  • Country: de
    • ab-precision
Re: temperature stable inductor design
« Reply #7 on: February 07, 2018, 02:44:40 am »
A good source for Standards are always documents from National Institutes such as NIST, PTB or the like.
Lay of land, sort of, although pretty old : http://nvlpubs.nist.gov/nistpubs/jres/7/jresv7n2p289_A2b.pdf
Basic, and in some instances certainly outdated, but good overview.
Many similar docs out there.
Metrology and test gear and other stuff: www.ab-precision.com
 

Online Vgkid

  • Super Contributor
  • ***
  • Posts: 2710
  • Country: us
Re: temperature stable inductor design
« Reply #8 on: February 07, 2018, 04:06:53 am »
Look here, info on the design on the GR-1482.
https://www.ietlabs.com/pdf/GR_Experimenters/1952/GenRad_Experimenter_Nov_1952.pdf
Gotta love the genrad experimenter articles.
If you own any North Hills Electronics gear, message me. L&N Fan
 
The following users thanked this post: zhtoor

Online Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: temperature stable inductor design
« Reply #9 on: February 07, 2018, 04:51:32 am »
Inductors are pretty simple- keep the size from changing and you'll have stability. Thus my suggestion of Invar wire. It's lossy and you want mechanical stability, both of which indicate using a heavy gage. It's hard to plate, maybe it could be silver plated for RF work.

Or, design in terms of a system. Copper coils are known to have a certain temperature coefficient and this can be cancelled out by specific capacitor types. I think polystyrene was a good match, but can't remember. That approach is sometimes better than trying to have everything zero.
 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: temperature stable inductor design
« Reply #10 on: February 07, 2018, 07:44:32 am »
Fired ferrite-ceramic cores should be quite dimensionally stable (we used them as high temperature spacers at work a few times). Not so sure about their magnetic properties though, I'd expect those to change quite a bit.

Might be worth investigating if there are any low value inductors with a thin noble metal layer sputtered onto a cylinder which was afterwards cut into a spiral. That should give good dimensional stability over a 0-50°C temperature range without breaking the bank too much (as far as manufacturing cost goes).
 

Offline Henrik_V

  • Regular Contributor
  • *
  • Posts: 155
  • Country: de
  • “ground” is a convenient fantasy
Re: temperature stable inductor design
« Reply #11 on: February 07, 2018, 09:11:01 am »
Have a detailed look at the tricks old RF designers used back in the old days where stable LC oscillators where build. Same apply for stable filters.
The main trick is to choose the rigth combination of filter/oscillator elements, such that the TC mainly compensate. 
I remember that caps come in much more different flavours (aka TCs) than you can easely find them nowadays.
Need a lot of skilled selection... nothing for mass production.
Today one would shift the influence of filters in the range where it doesn't matter. 
A fellow just shaked GR inductors to test it's stability due to transport. I think a puplication is already in the pipeline... 
I love my GR standards..great gear,  especially the ones with the smaller SN  (and the enamel? sign)  :D
Greetings from Germany
Henrik

The number you have dialed is imaginary, please turn your phone 90° and dial again!
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: temperature stable inductor design
« Reply #12 on: February 07, 2018, 11:32:48 am »
I would think it reasonable to wind a standard sort of inductor, like the GR part above, then wind a specially constructed one (whether due to materials or to geometry; e.g., a gapped torus with a bimetallic strip to reverse the tempco) to null the tempco.  The first challenge is making each component have a reliable, consistent, repeatable tempco.  The second is making them track, and getting the proportions right.

Like I said before, there's absolutely nothing preventing this from being done, but it will take a lot of research.  You aren't going to find the plans just laying around for how to make a standard part.  (Even the SI derived unit definition is terribly impractical, amperes being defined by geometry.)

It seems it would be much better to spend that research time and money tweaking the application.  Don't calibrate to a standard inductance, calibrate to a standard resistance times a 90 degree phase shift at a given frequency.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: temperature stable inductor design
« Reply #13 on: February 07, 2018, 04:16:30 pm »
I am interested in mu core materials because of the large size of air coiled inductors, I would like to focus on at least somewhat methods practical for regular circuit use. It would be hard to justify using something the size of a shoe box to make a front end for a LCR meter or something, in most cases... >:D

Have a look at pot type cores from MnZn ferrites, which are made specifically for tuned circuits at 0.1-2 MHz range, these may have temperature coefficients below 5ppm/C (for example EPCOS M33 and Ferroxcube 3D3 soft ferrites have permeability tempco below 2.5 ppm/C between 25C and 55C and initial permeability around 700) .

Cheers

Alex
« Last Edit: February 07, 2018, 04:45:17 pm by Alex Nikitin »
 

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: temperature stable inductor design
« Reply #14 on: February 08, 2018, 12:56:18 am »
why a pot type core?

Also, according to what I read, the tempco of the inductance is 70% of the mu, so that would be around 300ppm. This does not beat the gyrator circuits
« Last Edit: February 08, 2018, 02:31:59 am by CopperCone »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: temperature stable inductor design
« Reply #15 on: February 08, 2018, 12:34:25 pm »
why a pot type core?

Also, according to what I read, the tempco of the inductance is 70% of the mu, so that would be around 300ppm. This does not beat the gyrator circuits

You keep repeating that, but you don't give a reference.  And despite having been told otherwise on numerous occasions, and a number shown above completely blowing up such a claim.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: temperature stable inductor design
« Reply #16 on: February 08, 2018, 05:11:08 pm »
What number? The reference is bellow

I found some cores with known materials, I will wrap them with some anneled copper wire and take measurements.


I think you confused TCL and TCmu but im not sure what number blew up the claim. Im ultimately interested in TCL.
« Last Edit: February 08, 2018, 05:20:42 pm by CopperCone »
 

Online Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: temperature stable inductor design
« Reply #17 on: February 08, 2018, 06:35:29 pm »
That's interesting. Now, of course, I want to know why!
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: temperature stable inductor design
« Reply #18 on: February 08, 2018, 10:06:37 pm »
That's not a reference, it's a screenshot of a scan.

For all I know, it could've been published by an internet hack.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
« Last Edit: February 09, 2018, 01:21:31 am by CopperCone »
 

Offline Edwin G. Pettis

  • Frequent Contributor
  • **
  • Posts: 451
  • Country: us
  • The plural of anecdote is not data.
Re: temperature stable inductor design
« Reply #20 on: February 09, 2018, 01:39:43 am »
This data sheet on the GR 1482 series of inductors gives some clues as to construction, containment, stability and TCL which is approximately 30 PPM/°C.

https://www.ietlabs.com/pdf/Datasheets/1482.pdf
 

Online Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: temperature stable inductor design
« Reply #21 on: February 09, 2018, 04:53:53 am »
Also, go to https://www.ietlabs.com/genrad/experimenters/1950s and look at the November 1952 Experimenter for some internal construction details. Naturally these are much larger than you want, but the article might trigger some thoughts or clues.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14072
  • Country: de
Re: temperature stable inductor design
« Reply #22 on: February 09, 2018, 09:44:16 am »
An important clue from that scan is that it claims that the TC gets larger with a large effective Mu. This points towards an inductor with gap and the rather obvious point that AL depends more and more on the gap and less on the core once the gap is large enough.
 

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: temperature stable inductor design
« Reply #23 on: February 09, 2018, 09:52:33 am »
I've just measured the inductance change on a 125uH coil with a MnZn ferrite core. For 20C (from 23C to 43C) change the inductance has changed from 125.00uH to 125.08uH (measured by HP4192A at 100kHz), so the measured tempco was around 640ppm/20C = 32ppm/C .

Cheers

Alex
 
The following users thanked this post: chuckb

Offline CopperConeTopic starter

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: temperature stable inductor design
« Reply #24 on: February 09, 2018, 05:05:40 pm »
Thats fascinating. Did you see hysterisis when it cooled down? And how big is your ferrite?

Do you know the rated permeability of it?

Interesting that a common core has the same stability as a standard inductor

I wonder how much stability could be gained by going through the whole song and dance that the standard inductors use. I mean their pretty much air coiled, since their on ceramic with no magnetic properties.. So it looks like the core material adds very little. Guess we need to know its design permeability

All the manganese zinc ferrites i can find have a mu between 400 to 4000.. I wonder if that book is filled with a bunch of baloney

are you sure you dont have a nickle zinc core? They have mu that would agree with the book
« Last Edit: February 09, 2018, 05:47:24 pm by CopperCone »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf