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

email us
Forums | Crystal Radio Think Tank ** Advanced Forum | New crystal set, a radio that has it all? Post Reply Send Topic To a Friend
Author Topic
homebrewPosted - 13 November 2012 8:46  Show Profile  Email Poster  Edit Message  

Edited by - homebrew on 11/13/2012 8:53:44 AM

neazoiPosted - 13 November 2012 14:55  Show Profile  Email Poster  Edit Message  
do you think that a ceramic switch close to the coil, with short lengths of taps cables connected directly to the coil (no twisting) will add significantly to the losses?

what do you think of my overal setup?


_J_Posted - 15 November 2012 1:17  Show Profile  Email Poster  Edit Message
By multi-band I wondered if you meant break the BCB into multiple sub-bands, or BCB and some other band(s)?

Switching coils add significant losses. The switch structure insulation losses are only one of the many losses of switching a coil. The mere routing of coil tap connector wires around the coil seems to cause enormous losses. Coil taps themselves introduce losses. These would not be so noticeable if the coil design already has enormous losses though.

About the overall setup, I am still unclear about the objective of this, maybe you expect to switch between architectures and compare performance without the retuning. It is an interesting experiment but I could anticipate several problems with this expectation. One is underestimating the losses of switching and the unused but coupled in components. Would you be OK with comparing different radios instead of switching one radio between different architectures? That would give you a more realistic comparison. What about building each of these architectures separately and figure out how they behave before combining them into one experiment. Then you will be able to access the performance cost you suffer by combining and switching them all in one.

Don’t count on the * note under 3. Connecting the antenna to a tuned circuit changes its tuning substantially. An antenna that is not cut exactly to a quarter wave is reactive. Short antennas are like a capacitor. When you connect a capacitor to a tuned circuit, it tunes it to something else. Detector circuits also have reactance too. Wires connecting stuff together has a little reactance. That being said, the Q of these coils is not likely to be high enough for you to notice much of that detuning. It is pretty hard to get a crystal set to tune sharp.

There were several things you seem to expect that I would think are unlikely. These are:
From 8. I got the impression you expect tuning of the multiple tanks to track.
You seen to expect the vastly different architectures to perform more or less optimally with the various parts switched in or out.

John Davidson

Garry NicholsPosted - 15 November 2012 8:12  Show Profile  Email Poster  Edit Message

You may find the many technical articles at this site useful. There is a LOT of information here:

I'm not sure that you are familiar with the basic principles and constraints of crystal set construction.

If the coils are at least medium quality (such as well made basket weave coils OR solenoidal coils which are wound with the wire turns spaced about one wire diameter and on a low loss form) then the next things to improve are a good quality detector diode, an audio matching transformer and special sensitive headphones, such as sound powered (balanced armature) phones.

Your circuit does not seem to indicate these refinements.

Perhaps your intention is to cover a lot of frequency spectrum with circuits that are only average in performance?

Changing the circuits around to the many configurations that have already been tried will not improve performance as much as proper matching to a quality diode, audio transformer and phones.

Unused components close to active components on a test setup will interfere with the active components. It would be better to remove them from the test setup and put them aside.

I agree with John, that it is very difficult to get two separately tuned stages to tune and track properly with a single ganged variable capacitor. (Trimmers will not suffice. They are OK for amplified sets because the tuning need not be exact, only close enough!) I think that the few crystal sets that do this make performance sacrifices in favor of the convenience of single knob tuning.

All high performance two stage (double tuned) sets that I can remember use separate variable capacitors which are tuned separately.

Two stages with exactly matching coils and variable capacitors will tune at different rates and will tune somewhat different parts of the band because they have different loads placed upon them. Their tuning depends on what is hooked up to them. In the first stage it will be antenna and ground. In the second stage, detector and audio load. Each stage's load is very different, as John has stated.

A famous high performance set:

A famous two stage set with dual cap (single knob) tuning:

I think the single knob set is intended for use in urban strong signal locations. Anyone please correct me if I am wrong. Thank you!


homebrewPosted - 15 November 2012 9:34  Show Profile  Email Poster  Edit Message  
A double tune set with one knob works OK anywhere. You will still get good DX.
The two knob version gives a distinct advantage adjusting between selectivity and sound levels in different parts of the band.
My normal set up is two variables with dial string pulleys rubberbanded together which makes individual tuning easy enough when needed.
There are exceptions. My Tuggle tuner on a ferrite core is one.
Garry NicholsPosted - 15 November 2012 14:7  Show Profile  Email Poster  Edit Message

How are selectivity and sound level in a double-tuned set varied by having a separate tuning capacitor for each stage?

Wouldn't the amount of coupling between the two stages be what varies selectivity and sound level?

I can see how in the single stage of the Tuggle tuner two independent caps could vary selectivity and sound level.


_J_Posted - 15 November 2012 15:4  Show Profile  Email Poster  Edit Message

Re: For a double tune set with one knob works OK anywhere. You will still get good DX

Do you mean the one knob controls the tuning of both tanks (and they track -well enough to get good DX), or one knob controls the antenna tank AND the antenna decoupling cap?

John Davidson

krystalloPosted - 19 November 2012 15:21  Show Profile  Email Poster  Edit Message
Hey Neazoi:

I agree with many of the points presented here by other folks.

The ones I'd emphasize are :

Your radio seems to have too many traps,only trap what you NEED to.

Switch points can cause a loss of Q.

Unused taps are a bad idea. Bring copper plated flat toothed alligator clips ( order from Radio Shack) TO the coil. This works well on basket weave coils.

IMHO, most coils are OVER tapped. I generally tap to the var cap stator end of the coil, MAYBE at the 75% point, definitely at the 50% and definitely at the 10 % point up from the var cap rotor end of the coil.Further resolution of taps didn't seem to do much .

OVERALL,I think a lot of this depends on what you are trying to do. If this is simply an "experimenter's set up", fine.

If you are trying for SERIOUS results and good DX , you can eliminate 50 % + of all of this.

GOOD crystal sets ARE NOT complex. But they HAVE to be built VERY WELL.

Unfortunately, for best results this involves seeking high Q components and applying them properly. There are NO free gains in physics , placing a bunch of "options" in a circuit probably is NOT going to take you to optimum performance.

You will experience a lot of "different" results, with some even being bizarre.

But the more you put IN a crystal set, the more potential for loss.Use ONLY what you HAVE to get the job done.LOSS is the enemy to be defeated in a crystal set. Cut loss wherever you can .

My suggest for a good working crystal set ( if that's what you looking for) is as follows:

The "Tuggle" type circuit. It requires three var caps ( four if you need a trap). The var caps really should be ceramic insulated,IMHO .

Each var cap used is FOUR gang 500 pf (NOT 365) ceramic insulated var cap wired "wiperless". I REALLY like how these work.But they can be hard to find and are quite pricey

If cost is an issue , single gang 500 pf ceramic var caps can be used.

If you really need to save money or simply can't find ceramics, single gang bakelite 365 pf var caps are "OK" ( but NOT optimum).

But be advised for every cost you cut, you usually cut the radio's radio's Q . Once Q is lost, it cannot be "made up for" by trying to do it "that much better" some where else in the radio.

A crystal radio is a "chain". Watch out for "weak links" ANYWHERE in circuit from the antenna all the way to the ground.

Explore low loss coil designs (ie basket weave ).

Use the correct diode.Here in the U.S. this means having a choice between using a 1n34 or FO- 215 .

Match phones to diode, especially if you are using "magnetic" type headphones.The Bogen T-725 is an inexpensive "beginner's grade" AUTO transformer.

You can use "Hi impedance"/"piezo"/"ceramic" type EARphones. They are inexpensive and CAN work FANTASTIC- IF you get factory good copies. But they can be quite problematic at times as well( once in a while they arrive dead). I usually buy 4 or 6 at a time just to be sure. You might "get away" without a transformer when using these EARphones or piezo HEADphones.
Audio "bleed" (fixed or variable) resistors are required with any piezo type reproducer.

ABSOLUTELY spare NO effort to fully optimize your antenna and ground system for specifically for CRYSTAL SET operation. This means a SINGLE wire feed from an END fed (inverted "L") antenna. Keep feed line suspended IN AIR and off surfaces .No shorted coax feeds (etc). Every nanowatt counts.

Many folks neglect their ground .You almost CAN'T "OVER DO" a crystal set ground.

FWIW tuning a double tuned crystal set takes practice. Move one var cap a bit and sweep back and forth with the others.Ball drives or VERY big knobs really help.All knobs will/should interact.

So to sum a "good" crystal set can be built for "reasonable " money, but the higher end designs will take some investment.

Since you have already built some crystal sets , you know what a THRILL it is to actually listen to one .Hopefully you have already copied some Dx. ( please tell me what you've logged so far).

And since you are an amateur radio op you should have a good back round on many of the issues covered here.

de N1NQC

Garry NicholsPosted - 19 November 2012 16:59  Show Profile  Email Poster  Edit Message

Do you think that he needs to go to ceramic caps if he is not using litz wire coils? I think he mentioned using solid wire and basket weave.


homebrewPosted - 20 November 2012 5:17  Show Profile  Email Poster  Edit Message  
On a solid wire oatbox set I do not see a improvement using silver plated ceramic variables in the parallel tank but can tell a difference when used in series with the coil as a antenna tuner.
Subbing the same variable in a double tuned set anywhere changes the optimum spacing point and I can usually tell a improvement.
I do not consider a ceramic variable a necessity. Other things will give you a bigger bang for your buck but ceramics are a good investment if the price is right and they will eventually be a plus when you get to the Litz coils.
I do recommend upgrading the insulators on any cap that allows it to be done easily.
As to your querry about tuning a double tuned set two variables rubber banded together works best for me. I set the coupling for a compromise between max selectivity on the local boomer in the middle of the band and max volume best tuning on weak stations at each end and normally do not change it. Minor tuning adjustments in selectivity and volume can be made by changing the two variables separately.
Bolting the frames together also changes the tuning.
I was referring to using one knob to control the tuning of both tanks. Keep in mind my set up of two separate variables with dial string pulleys rubberbanded toghther.
Two variables with a common frame will usually work good and pull in the same DX.
Like I said, there are exceptions.

Edited by - homebrew on 11/20/2012 5:34:21 AM

Edited by - homebrew on 11/20/2012 5:37:20 AM

Edited by - homebrew on 11/20/2012 6:12:12 AM

krystalloPosted - 20 November 2012 8:12  Show Profile  Email Poster  Edit Message
Hey Guys,

IMHO, I PERSONALLY DO think that IF you can get them and IF you can afford them, ceramic var caps a generally worth it, maybe not so much in a low Q oat box type set, but more so in a higher Q basket weave type coil(or certainly w/ litz).

I feel that if you enclose FOUR high Q var caps w/ ball drives in separate cabinets w/ ceramic stand off terminals on the outside, you are set for a lifetime of serious radio experimentation.True, a tall and expensive order- but you'll never worry about var caps ever again.

"Stand alone" var caps were common in the 1920's - 50's and some "precision" lab types were VERY expensive.

I feel there was an audible difference in using a ceramic vs a bakelite in a basket weave,at least in MY radios.Maybe YOUR mileage varies.

To ME EVERY bit counts in a crystal set ("yards by inches".

CAN you use bakelite? Sure.Will it be as good ? To ME , no. But we've been around on this before and I have heard the formulas on how some feel you can use bakelite. No problem.

All I know is that I get a very respectable solid wire dx operation on ceramic and not as well on bakelite.Period.I live in a BRUTAL big city A.M. market ( Boston).

Maybe in a rural setting on strong widely spaced dx bakelite is more "tolerable" . Dunno.

FWIW, the "gurus" (Schamarder,Tongue, et al) "back in in the day" maintained that "Nothing ruins tank Q quicker than crumby var caps". There was NO reference to what type of coil wire was used. Hmmmm.

Even if you feel you CAN'T hear any difference (due to low coil Q), getting these caps on hand now is a good move if you are EVER considering upgrading later.

These caps like SP phones(ie Big Cans), they are getting harder and harder to find and they prices aren't going to go DOWN.

My big litz "competition" grade sets kind of "hit a wall" and worked almost TOO well (if such a thing is possible). They grabbed 95% or better of what the ICOMS heard !

So I "moved on" intentionally into a "handicap" and tried to optimize a solid wire architecture. I feel I did quite well in this respect, but it seemed to "require" ceramic on the basket weaves.

I just go for the best when I build. I understand that ceramic is a hard to find and sometimes expensive item.

I suppose you could try and do as well in ALL other aspects of the set and simply "eat" "whatever" loss in encountered (audible or not, per one's "viewpoint").

So I'd say only go as nuts as you want to or can afford. Better to build a "good" set and have fun than to hold out for a GREAT set that may never get built.


golfguruPosted - 20 November 2012 16:30  Show Profile  Email Poster  Edit Message  
Where are ya, neazoi?


_J_Posted - 20 November 2012 16:40  Show Profile  Email Poster  Edit Message
Hi HB,
Re: one control for 2 tank variables:

Since one variable has a complex reactive antenna whose Z varies with frequency coupled to it, and the other varable has a detector that has capacitance with attached body reactance coupled to it I did not expect the two tanks to track very well. Mine never do. That is why I suggested to Neazoi not to count on that. Then, of course, if you can rig the pulleys and strings to pre-emphasize that out, or if tuning is not sharp, maybe it still kinda works?

John Davidson

golfguruPosted - 20 November 2012 16:47  Show Profile  Email Poster  Edit Message  
I have found that the A/G tank tunes so broadly that it probably is "not that hard" to keep the tuning tracking reasonably?

It is not as if you are trying to track two "razor sharp" resonances together?


Garry NicholsPosted - 20 November 2012 20:0  Show Profile  Email Poster  Edit Message
golf' ;

How tight was the coupling for the "approximate tracking" situation that you mention?


Do I recall a Ben Tongue article indicating that phenolic caps did fairly well below the high end of the MW band?

In personal correspondence, Ben mentioned to me that a good quality variable with ceramic insulation was a lot of the way towards the silver plated Holy Grail caps. I had asked because I could not find any silver plated caps. But I do have a few with ceramic insulation.

My medium grade double tuned set did quite well with a somewhat large and somewhat wide space phenolic in the antenna stage and a ceramic in the detector stage. (I figured the better cap belonged in the detector stage because its Q could be maintained higher than the antenna stage which was swamped by the antenna and ground loading.)

That was with #21 solid copper Rook coils, a Schottky or FO-215 diode and transformer matching to so-so sound powered phones.

The set was also fun to use and got lots of stations with a 1N34A diode (tapped about 1/3 up the detector coil) and my old Brandes 2k DC antique phones.

I mention this to put some perspective on the conversation. We don't want to discourage less experienced new comers from building sets because they are not using the very best components.


Topic is 4 Pages Long:    1 2 3 4

Click Here To Close Thread, Administrators & Moderators Only.

Show All Forums | Post Reply