EMRFD Message Archive 15490

Message Date From Subject
15490 2019-10-02 02:31:00 Ashhar Farhan teflon toroids
I was trying to build a 60 MHz band pass filter for an up conversion receiver. I first tried the Mircometal toroids. The losses were quite bad, after struggling for a few hours, I decided to measure the toroid's Q on 60 MHz using the series trap method.

T50-6, L=1uh (although I had just wound 14 turns to hit 800nH), Q= 98
T30-6, L=1ufh(14 turnes), Q=70

At this point, I decided to try something else. I changed from swg28 gauge to 20 swg enamelled wire. I wound an inductor using 1/4-20 bolt as a former with 20 turns. This gave an inductance of 550 nh. The Q was surprisingly good at 288. I had to recheck this with the single resonator filter to be sure. These inductors would be too big for my purpose. I needed toroidal inductors so that they are self-shielded.
I took a 1/2 inch teflon rod that is used locally in some electrical work. I drilled a central hole with a 1/4 inch drill and cut them into slices of 1/4 inch. With this I wound a toroid with 14 turns. This resonated at 65 Mhz with a 10 of capacitor (as a trap). The inductance was measured as 700nH and it has reported a Q of 350!
At this point, I switched back to the filter, and three of these toroids were used to make a triple tuned circuit. This is a modified 3 section low pass filter with an inductor in series with 10 pf caps. There is probably some interaction between the toroids, I had to twist the middel toroid to eliminate a 'double' peak. The loss of the filter is a suprisingly low 2.5 db with a 6db bandwidth of 3.9 MHz. 
This filter was built using squares cut out on a copper clad board using the back plane copper as the ground plane. The filter has bad response above 500 MHz probably due to construction practice. A quick 'tack-and-check' test shows that this can be improved with introducing shields between the inductors.

To sum it up : 
A. Teflon is a great medium to work with, it is easy enough to make them yourself.
B. Shielding is still useful, even for toroids.

- farhan


15491 2019-10-02 06:06:28 Clutter Re: teflon toroids
I don't know if this would help you, Farhan- but the powdered iron series toroids
have included phenolic core versions for many years. Whether this would help,
as compared with Teflon, which is a superior dielectric, remains to be seen- but
they have been available all along. FWIW I looked on the kitsandparts.com
website and I see that he offers only T37-0, T106-0, and T130-0 sizes. I have
a few of those tan-colored toroid cores in my own collection, mostly in smaller
sizes. Anyway Tnnn-0 generically calls out a phenolic toroid core, no powdered
iron present in the core.

73, David K3KY


15492 2019-10-02 07:33:58 Ashhar Farhan Re: teflon toroids
Thanks David,
I will get some of this in a month. Someone is travelling on the 1st November. Ill evaluate them too.

- farhan

15493 2019-10-02 09:19:02 Rob Re: teflon toroids

pictures?


15494 2019-10-02 09:22:34 Jim Strohm Re: teflon toroids
Tap washers for the 21st century?

73
Jim N6OTQ

15495 2019-10-02 13:10:24 jgnoci@gmail.com Re: teflon toroids
 I needed toroidal inductors so that they are self-shielded  -

I presume you realize that toroidal coils are not self shielding...Wound on a high permeability toroidal (ferrite) ring, they have low leakage fields and so may behave as shielded, but, for example,  powder iron cores with permeability in the twenties and below exhibit enough leakage flux to make many pole high factor filters possible only with inter-inductor shielding. Toroidal coils on cores with permeability = 1, or teflon cores, offer no shielding at all...
Joe
15496 2019-10-02 17:43:24 Ashhar Farhan Re: teflon toroids
Joe,
I didnt understand the lore well enough. Thanks for pointing that out. I have been looking out for Doug Demaw's book on inductors on ebay. That's the recommended text, they say.
73, f

15497 2019-10-04 12:57:07 VU3SXT Sandeep Lo... Re: teflon toroids
Plzz document, measuring  toroid's Q 
with series trap method...
with pictures of your homebrewed torroids...

Our MIL grade BEL used threaded teflon rods even for 80 Meter LPF

15498 2019-10-04 13:01:07 VU3SXT Sandeep Lo... Re: teflon toroids
>
>>
>


[Non-text portions of this message have been removed]
15499 2019-10-04 13:25:49 VU3SXT Sandeep Lo... Re: teflon toroids


15500 2019-10-04 21:40:51 Bill Carver Re: teflon toroids
My humble opinion Fahran: "lore", even from one as revered as DeMaw, may not always be exact. Your 6m BP filter is one case where, as Joe pointed out, low permeability toroids have significant external fields, something lore doesn't stress.

Another "lore" point: "inductance of a toroid is only a function of the number of turns". Absolutely true for very high permeability cores, but only approximately true for u=10, and not true for u=1 Teflon-cored coils. Without changing the number of turns, even with red u=10 cores you can make small changes in inductance and distributed capacitance/self resonant frequency by spacing the turns together, or further apart (assuming your core is not full, preventing wiggle room).

I think DeMaw had a big box of red and yellow cores and particularly when minimum loss/highest Q wasn't a big deal he sometimes used cores that were not really optimum. If you're fighting for lowest insertion loss BP filters, for example, you might find that a lower permeability core gives higher Q. For example, on 20m its common to see yellow (u=8) cores, but black (u=6)  often produces significantly higher Q. And depending on how much room you have, going from a -37 core to -50 can produce a significant Q increase.

For even seemingly simple HF filters, if you want highest Q you should take a look at the Amidon data book: Bill Amidon put many, many hours into winding coils and measuring Q to produce the Q v.s. frequency curves in the "Orion-Powder and Ferrite Coil Forms" booklet. It might be instructive to take a DeMaw designed coil and see what Amidon's Q prediction for the specified core, the look at alternate core material and sizes to see what their Q might be. Even then you aren't likely to match inductances, and Amidon's were usually FULL 360 degree windings yielding maximum inductance for that core and wire size.

I guess I'm saying that the "art" of inductances is sometimes different from "lore". Both, of course, are only approximations to serious theoretical analysis, which in turn is only an approximation to reality. But when you're struggling for every bit of performance you can get it can pay to dig deeper and perhaps develop your own art/lore!

73 - Bill


15501 2019-10-05 07:34:54 Ashhar Farhan Re: teflon toroids
Bill,
Thanks for the note. I am mystified that the Q of an inductor is specified without specifying the frequency. Now, I am building amps to up each tone of my two tone tester to measure the IMD.
Won't more turns lead the greater flux? Not sure if this is even a we'll formulated question...
73, f

15502 2019-10-05 21:47:36 Bill Carver Re: teflon toroids
I had some 5.2 MHz telephone company filters that had matching toroidal inductances, -2 material, stuffed inside the little drawn cup that surrounded the toroid. Obviously so signals couldn't couple from input to output as easily.

An external shield is not a shorted turn since it doesn't go through the center of the toroid. It's not a "turn". But anything in the mangetic field of the toroid can change its inductance and losses.....essentially the shield becomes a shorted turn. If it was perfectly conductive it would change the inductance but not the losses.

Tinned steel shield might not be the best choice! But copperclad PCB, spaced away from the coil, probably won't do too much damage to Q.

Bill W7AAZ


15503 2019-10-06 03:34:40 Roelof Bakker Re: teflon toroids
Hello Bill,

I have measured the Q of T-50-2 and T-50-6 toroids mounted flush on a copper clad PCB earth plane and
also mounted vertical at a distance. No difference in Q was found, so ever since I have been confidend
to mount them flush on a ground plane if this was convenient.

73,
Roelof Bakker, pa0rdt
15504 2019-10-06 04:50:13 Sudipta Ghose Re: teflon toroids
Hi Group!
I was thinking on this line about 5-6 months back when I was building my latest Linear PSU using a toroidal step-down xmfr. The mounting for such xmfrs (as they are heavy vis-a-vis the small ferrite cores) always consists of steel hardware. Being the owner of a homebrewed toroidal xmfr I elected to use para-magnetic materials like stainless steel hardware to mount the xmfr like a small section of pipe, washers alongwith nuts and bolts. I thought of copper hardware but then the strength and stability was the point that steered me away. There is absolutely no 'wham' when I switch on this 40A PSU housed in a recycled steel cabinet.
 
Decades ago, Ed Weatherhold, W3NQN, kindly gifted me some 88mH surplus toroids from the Chesapeake and Potomac Telephone Company in W-Land. I brought them out from my junk box to find that they are completely sealed in tin containers. However, I have used some of them to build a CW filter after removing the container which was evidently the shield.

Thanks to all for whetting a new thought process.

Kind regards and 73,
 .  
Sudipta Ghose VU2UT
--
One of those ... ...
15505 2019-10-06 07:08:48 Ashhar Farhan Re: teflon toroids
Roelof,
We are discussing intercoupling as well as the Q. Stray intercoupling can dramatically lower the stopband attenuation as well cause ripple in the passband. There are easily observed in a triple tuned circuit with linearly wound, old fashion, inductors with air core.
The Q I reported was quite surprising for me. I don't remember this good a Q in any toroid. I began with using water faucet washers (we called them water gap washers) almost 12 years ago.  At that time, the passband appeared good enough on my oscilloscope. I didn't have a sensitive RF power meter then. However, the w7pua/w7zoi power meter changed it all. It is the observation that is at the root of our practice. Measurement has no substitute. More on this in a day.

73, f

15506 2019-10-06 08:25:25 Sudipta Ghose Re: teflon toroids
Hi Group!
I was thinking on this line about 5-6 months back when I was building my latest Linear PSU using a toroidal step-down xmfr. The mounting for such xmfrs (as they are heavy vis-a-vis the small ferrite cores) always consists of steel hardware. Being the owner of a homebrewed toroidal xmfr I elected to use para-magnetic materials like stainless steel hardware to mount the xmfr like a small section of pipe, washers alongwith nuts and bolts. I thought of copper hardware but then the strength and stability was the point that steered me away. There is absolutely no 'wham' when I switch on this 40A PSU housed in a recycled steel cabinet.
 
Decades ago, Ed Weatherhold, W3NQN, kindly gifted me some 88mH surplus toroids from the Chesapeake and Potomac Telephone Company in W-Land. I brought them out from my junk box to find that they are completely sealed in tin containers. However, I have used some of them to build a CW filter after removing the container which was evidently the shield.

Thanks to all for whetting a new thought process.

Kind regards and 73,
 .  
Sudipta Ghose VU2UT

--
One of those ... ...
15507 2019-10-06 08:35:09 jgnoci@gmail.com Re: teflon toroids
Hi Sudipta,
There would be no 'wham' as the magnetic field is well contained within the core - note what was said in the posts - it depends on the leakage field and with high permeability cores that field is low. I don't think 50Hz power transformer cores can be mentioned in the same technical breath as RF Powdered Iron cores in terms of leakage fields...

Joe
15508 2019-10-06 08:37:59 Nick Kennedy Re: teflon toroids
I recently plotted the response of a 5 element 40 meter LPF I'd built using a couple of T68-6 toroids. It looked pretty good but I was surprised to see the two zeroes for the 2nd and 3rd harmonics completely missing. (Each inductor does have the appropriate parallel capacitor.) I'm blaming it on coupling between the toroids although I haven't proven that yet. They are in-line and almost touching. 

I need to re-do the build with more spacing (it's Manhattan style) and even a vertical shield between the toroids if necessary and see if that does the trick.

73-

Nick, WA5BDU
 

15509 2019-10-06 10:19:58 Ashhar Farhan Re: teflon toroids
Vertical shielding is quite good enough to show you that it is needed. You could also try twisting the middle toroid to a right angle from the end ones.

15510 2019-10-06 10:42:37 swift_glen Re: teflon toroids
Ashhar,
Was very impressed with Harold Johnson's build of VHF helical resonator,
(included in EMRFD's CD article). He used polystyrene as a coil former.
He cautions about using "pure" polystyrene rather than derivatives like
polystyrene-acrylonitrile. 
Dielectric constant of Teflon (2.1) may be lower than Polystyrene (2.5),
which should result in slightly higher self-resonance for an inductor.
It seems that his resonator Q was very high, if he could draw arcs
off the coil of a FET oscillator!
His helical resonator was built inside copper pipe (with end-caps) so
all fields were contained.
15511 2019-10-06 12:48:51 Bill Carver Re: teflon toroids
I just took a prewound coil from my junkbox coil, T68-2 core with Qabout 28 turns of 0.70mm diameter wire. It had a of 285 at 6.9 MHz (the frequency the Q meter was already at when I turned it on).

I moved a piece of single-sided G10, 1.6mm copperclad near the coil its inductance changed and had to change the tuning capacitance. Both change in inductance and capacitance of coil changed. I didn't try to figure out what was dominant, just thinking of Q. The Q dropped to 240 when the G10 side (not copper side) was against it.

That's not the end of the world.......certainly not like having a shorted turn on the toroid. But that is a 15% drop in Q. It would be interesting to use different u toroids and see how the coupling to adjacent copperclad board with u=8 and u=6 material. I have seen some commercial PCB where they removed the copper circle underneath toroids. When I mount a  toroid against a PCB I use a nylon washer just as a not-intellectual "habit", hi. But this test indicates that if the copper side is up, there would be a similar coil-to-copper spacing, and a similar Q reduction.

Bill - W7AAZ

.

15512 2019-10-06 13:25:33 VU3SXT Sandeep Lo... Re: teflon toroids
another hambrew project for Farhan, Hi hee...
mix copper-powder with conductive carbon paint & fine coat on PTFE sheet, & use as shield... 
Kidding :D
here orientation does not matter much !


replacing tap-washers with Teflon rods was long back suggested to me by...
now silent-key VU2DEV

73, signing off...







15513 2019-10-06 13:37:16 VU3SXT Sandeep Lo... Re: teflon toroids


On Sat 5 Oct, 2019, 1:55 AM VU3SXT Sandeep Lohia, <sandeeplohia12@gmail.com> wrote:


15514 2019-10-07 13:16:20 STEVE Re: 40 M 5 element LPF swept

Hi Nick,

Did you rebuild your filter and sweep it again?

I made a quick Tonne Elsie software 5 element 40M LPF and built it just now, using two T50-2 toroids placed horizontally and about 1/4" apart (no shields or other attempts at wide spacing). Filter is about 1.5" long x 1/2" wide. Swept with the VNWA it clearly shows the two nulls. I didn't try with the toroids vertical, maybe that's how your first version was made. Since Yahoo won't allow attachments, if you email me I can send you the plot of mine if you'd like.  ratzlaffsteve@gmail.com

73,

Steve AA7U

15515 2019-10-09 13:21:12 Bill Carver Re: teflon toroids
Oh "for sure" the Q varies with frequency. The shape of Q v.s. frequency changes with distributed capacitance and winding technique and wire size.  I don't remember what loss dominates on the low frequency side, and what loss dominates on the high frequency side. Wes probably knows without having to look it up!

But every coil has maximum Q in the middle of some frequency range. As I commented before, often the peak Q is obtained with a lower mu core than you normally see used....black u=6 cores on 20m for example, instead of the commonly used yellow u=8 cores.

Does it matter? For many things, not very much. Although if you want your filter to really work to its catalog shape, you may need to factor in component Qs. Especially, as you know, when the filter itself has a high Q (IE, quotient of center frequency divided by 3 dB bandwidth).  A textbook "Handbook of Filter Synthesis" by Anatol Zverev, often referred to as just "Zverev", has tabular information on filters compensated for various degrees of inductor Q, and showing the insertion loss.

If you want a Chebychev filter with, let us say 7148 Khz center and 3 dB down at 6950 and 7350 KHz, the "filter Q" is 17.9.  It will be possible to obtain the theoretical Chebychev passband shape with component Q's of 20. It is also possible to get the same shape with component Q's of 100, or 200. The only difference between those filters will be the insertion loss. As your inductor Q's approach the "filter Q" the insertion loss increases "toward infinite loss". Don't ask me what shape a filter has at infinite insertion loss, hi.

I used a Q meter in a back room at college, and it was so interesting that I tracked down a Boonton Q meter and have found toroid construction very interesting both on low insertion loss RX filters, and in making lowpass filters for QRO amplifiers and an antenna tuner where coil heating was an issue.

Actually, that college exposure turned into a crazy situation: I now have two Boonton 260A's and an HP4342A. All working fine. The Boontons use vacuum tubes, but are very well engineered and work just as well as the HP.  It wouldn't ship well, but can you carry a fairly large boat anchor home from your next USA visit?

Curves in the Amidon book are interesting to see.

73 - Bill



15516 2019-10-09 20:43:31 Ashhar Farhan Re: teflon toroids
bill,
i might do that. i really like test instruments that i can fiddle around with if they break down. i travel with a tiny backup and the rest is just the 'junk' in two suitcases.
to the reading list, i would really add the IRFD. It is now free with the CD of EMRFD as a PDF. There is an excellent discussion of Q in the chapter 3 (Filter Basics). I was reading this just morning.  It explains why the rule-of-them about unloaded Q should be ten times the loaded Q. Another small gem : The number of ripple bumps inside a BFP's passband tells you the number of poles in the chebyshev!!
thanks everyone for the enlightening discussion.
- f