The Xtal Set Society
The Xtal Set Society
profile | register | search | main site


email us
Forums | Rap N Tap Clubhouse - Basic Crystal Radio Fun! | Crystal set aerial Post Reply Send Topic To a Friend
Author Topic
_J_Posted - 27 January 2014 1:47  Show Profile  Email Poster  Edit Message
Hi HB,
The 1:1 length to dia never worked for me. Though many (including me) tried, Q meters prove that no one (including Terman) ever discovered any math that even comes close to approximating high Q or low coil losses from coil dimensions or even shows the right trends. They simply don’t track my measurements and I verified this with a number of others.

The best 200 uH solid wire coil I ever made was 62 turns of #18 wire spacewound on a 3.5” dia coil 4.5” long. The Q peaked at the top of the BCB at around 500. I should be able to get higher Q with bigger coils, bigger wire 4” or 12” or 30” etc, but I never beat that coil with anything I tried, and believe me I tried many formats and schemes. I have repeatedly asked if anyone can wind a better 200 uH coil I would like to copy it and measure it. No one ever answered that challenge. Can you?

This implies that there exists a solid wire coil that ‘best’ -contrary to theory, which says increasing every dimension should increase Q. I was always puzzled by that.

John Davidson

_J_Posted - 27 January 2014 1:58  Show Profile  Email Poster  Edit Message
Hi Garry,
I usually leave the nylon in the coil out of convenience, nylon is not as bad as PVC. But it does add some losses, better if removed. I never use PVC coil forms, I use lower loss ABS. To remove the spacer, I sometimes touch the coil across a car battery a second and it gets it hot enough to make slight track in the coil form. I also tighten the coil up while hot and secure the ends. When it cools it gets pretty tight and holds the spacing pretty well. Not easy...

It seems to take a big percentage increase in Q (50% ish) to notice much performance gain, so the trouble may not be worth it in pracitce. ??

John Davidson

homebrewPosted - 27 January 2014 8:35  Show Profile  Email Poster  Edit Message  
J:
Rules of thumb are just starting points.
The ideal squared 1:1 coil is for close wound magnet wire in a small range of wire guages and former sizes using fixed L and C values.
Your coil is spaced and uses too large a guage of wire to fall within the 1:1 rule.
One rule given to me by my Elmer was larger wire on smaller formers for higher frequencies. He was refering to my short wave experiments but your top of the band Q measurements and large guage wire probably fall roughly into the fringes of that rule of thumb.
As you move away from the design limits previously unimportant items start to creep up in value sometimes exponentially throwing a monkey wrench into the works.


Edited by - homebrew on 1/27/2014 9:18:47 AM

Edited by - homebrew on 1/27/2014 9:47:11 AM

Edited by - homebrew on 1/27/2014 10:51:12 AM

Garry NicholsPosted - 27 January 2014 11:38  Show Profile  Email Poster  Edit Message
Thanks, Jon. Did not know that nylon has better dielectric properties than PVC. Also, the nylon cord would only touch a turn at one tangental point, whereas the PVC coated wire has the PVC surrounding each turn and the PVC touches tangentally.

A car battery -- yikes! (Don't do this at home!?)

What was that 200 uH test coil wound on?

Garry

homebrewPosted - 27 January 2014 14:33  Show Profile  Email Poster  Edit Message  
J:
It seems to me that heating a copper wire enough to melt the former would increase the resistance and lower the Q? I'll have to try it tho. I liked the way the wires stayed put on my earliest coils made from parafin soaked cardboard tubes. Would using one piece of wire on all the coils then replacing it with new wire work?
*
A 50% increase in Q sounds too high to me. But it takes a lot of increase in signal strength for me to say "that is louder". I do notice sharper tuning before I hear a increase in volume when swapping coils.
Circuit Q would be a big factor in how much a higher Q coil effects the volume.
Small increases in coil Q can usually be seen by looking at sharpness of tuning and voltsge across the Benny. My opinion is that any improvement in Q that does not cost a fortune is worth it.
_J_Posted - 27 January 2014 22:14  Show Profile  Email Poster  Edit Message
Garry,
I can't find my notes right now but I am pretty sure it was ABS, with spacer removed.

John Davidson

_J_Posted - 27 January 2014 22:24  Show Profile  Email Poster  Edit Message
Hey HB,
Not sure I ever measured the difference, but I expect melting the former would increase losses a bit. I do not expect heating copper enough to soften the former would affect its resistance (after it cools). Parafin is also good, but it coats the whole wire. Yes heating one wire then replacing it with another works, I have done that. I painted a thin coat of lacquer on the final wire to keep it bright.

And I was talking about the difference in tuning sharpness, not loudness. The thing is that a 50% coil Q increase is not a 50% system loaded Q increase, more like a 5 or 15% system Q increase. That’s the rub.

John Davidson

homebrewPosted - 28 January 2014 9:34  Show Profile  Email Poster  Edit Message  
I soaked the former not the finished coil. The thin layer in top leaves a small indentation when the wire is wound on it. Holds it in place nicely.
A couple seconds under a hair dryer will bond the former and bottom of the wire a bit better.
The circuit Q's I've measured vary all over the place. So increasing coil Q should effect
some circuits more than others?

Edited by - homebrew on 1/28/2014 10:32:12 AM

Garry NicholsPosted - 28 January 2014 10:22  Show Profile  Email Poster  Edit Message
There must be different types of parafin. Is there a best type? Probably with no coloring agents (I'm thinking "candles".).

Or are you guys using "coil winding" parafin? Where do you get it?

It appears that parafin is fairly low loss? What if the whole coil is coated like they used to be?

I suppose for a spaced solenoidal coil some longitudinal stripes of parafin could be drizzled along the coil in maybe 4 places to hold turns in place.

Although homebrew's idea sounds best. Coat the form and let that under layer hold the turns.

Garry

homebrewPosted - 28 January 2014 11:27  Show Profile  Email Poster  Edit Message  
There are several grades of candle parafin. Mainly the candle part and the outside harder grade.
I used canning parafin. No reason except it was available in mom's kitchen and grocery stores.
_J_Posted - 28 January 2014 18:39  Show Profile  Email Poster  Edit Message
Please correct me if I am wrong, but I think I recall bee's wax is even better? (Its awful being old!)

John Davidson

_J_Posted - 28 January 2014 18:42  Show Profile  Email Poster  Edit Message
My apologies to Ray, I think we hijacked his thread. Next time I post to this thread I won't, I'll start a new one and give Ray his thread back.

John Davidson

homebrewPosted - 29 January 2014 7:48  Show Profile  Email Poster  Edit Message  
Bees wax and parifin mixed was used by the early Tesla coil designers.
Some use the round toilet seal wax now.
There's probably very little difference as far as xtal set coils go.
raypatPosted - 29 January 2014 23:14  Show Profile  Email Poster  Edit Message  
Guys don't worry about hijacking this thread.The discussion is interesting. For what it's worth I have been working some more on the mystery set. I have what I think is a tuggle arrangement tuning the antenna now, along with the sliding coil added previously and I used an old coil and another variable cap in the antenna as a wavetrap. The tuning is a bit complicated now and I am not sure if I have improved things much. The wavetrap is definitely doing something as I can knock down 2GB quite well using it.

Ray

Garry NicholsPosted - 30 January 2014 19:15  Show Profile  Email Poster  Edit Message
Hi raypat;

Let's see if I'm following this correctly. The Tuggle tuner is inductively coupled to a Mystery Set stage which is where the diode and phones are?

The Tuggle uses a dual variable in an otherwise conventional parallel tuned circuit. What makes it different is that there is not a direct connection to ground. Ground is through one set of stator plates on the dual variable.

What this buys is that the stage is "more coupled to ground" at the low end of the band and "less coupled to ground" at the high end. This is caused by the varying amount of C created by the section of the dual variable hooked to ground.

Or is your Mystery project a separate set?

Garry

Topic is 6 Pages Long:    1 2 3 4 5 6

Click Here To Close Thread, Administrators & Moderators Only.

Show All Forums | Post Reply