Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Apr Tue 23, 2019 11:33 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 29 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: My results-"Improved" Cockaday 4-Circuit Tuner
PostPosted: Sep Fri 08, 2006 8:20 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11100
Location: Vieques, PR, USA
I got my 4-circuit "Cockaday" tuner playing this afternoon. If first impressions mean anything I think we have a winner! Thats somewhat in contrast with some of the observations in the previous thread. I hope that doesn't mean my comparison is based on my other projects being duds. 8)

This is the improved version and has some extra features...
1. Adjustable grid coupling cap - and yes it makes a noticeable difference.
2. Adjustable plate bypass cap - a VERY worthwhile addition.
3. Pot for B+ adjustment - still learning to work it with the 00A.
4. Supposed to have an adjustable grid leak but I don't have a spare one so I'm making up a set of different values to tinker with.

My initial test was using an 00A with about 24 volts B+ available, headphones only with no amp (thats under construction too). I was picking up stations that I have never heard on a one-tuber and find it to be very selective. I used a really good tuning cap and the coil is wound with #18 so that probably contributes a lot to that. I could not do the bank-winding on the antenna coil with enamelled wire and had to resort to #24 to make it fit. I don't see where this should make a whole lot of difference since its just a loading coil. I didn't even need to use my external inline trap to get rid of the local station with this circuit.

One of the features in the 4-circuit article was the so-called "automatic" set-and-forget regeneration control so that you don't have to walk the controls up the band until you find the sweet spot. They also mentioned it might take a while to find that point between juggling the front panel control and the plate bypass trimmer. So far I agree with that !

This set is going to be fun to tweak out and learn how to operate. It already impresses me. I'll be posting more info as I get a handle on it.

-Bill

Image

Image

Image


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 9:38 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1086
Location: Zip : 80751
Great as usual Exray!!! It looks fantastic :P


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 9:40 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 34326
Location: Sandpoint, IDAHO 83864
One question for you Bill. Being that you do such fantastic work building sets like this, I was wondering just how many hours you put into something like this?
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 9:42 pm 
Member

Joined: Jun Fri 02, 2006 9:58 pm
Posts: 97
Location: Elsberry, Missouri, USA
MOST BEAUTIFUL JOB EXRAY,

I just can't see the logic behind the design of this receiver. Is it documented as to why?

What is the adjustable grid cap adjusted for?

I presume the antenna loading inductance is adjusted for maximum signal---I can understand that.

What is capacitor H adjsted for?

I presume capacitor E is adjusted to resonate with coil C, OK.

My biggest problem is understanding the parallel RC network from plate to grid. What is capacitor F adjusted for?

_________________
John Amery
1251 Deer Run Road
Elsberry, MO 63343-4011
KF0HX


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 11:36 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 740
Location: U.S.
Needless to say, beautiful set. Excellent craftsmanship, as attested by the excellent performance of the receiver.

As stated in the other thread, I was rather skeptical of certain claims surrounding the Cockaday Four-Circuit Tuner. The two performance reports, one from Billy's great example and my quick test effort, were disappointing. Considerable emphasis was placed in the original article on stringent construction practices, especially in regard to the the coils. The coils indeed seem to be exceedingly critical. I would speculate that the average 1920s amateur would make rather substandard coils and be therefore met with disappointing results. That's probably why this circuit was driven into near extinction.

Bill, I feel I must thank you for demonstrating what this circuit is truly capable of. I must revist this circuit and be similarly impressed.

Recently there has been somewhat of a debate surrounding the 01A and 00A tubes. Any chance you could drop in an 01A for a comparison? I'm sure several (including me) would be interested in the results.

73's


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 11:43 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11100
Location: Vieques, PR, USA
First... Curt...its impossible to say how much time. The reading and parts scrounging and thinking thru the layout is definitely the most time-consuming. Once thats done it might take a day to make up the panel and board and mount parts. The wiring took about 5-6 hours yesterday.

johnamery wrote:
MOST BEAUTIFUL JOB EXRAY,

I just can't see the logic behind the design of this receiver. Is it documented as to why?


Thanks John but you're putting me in a spot to explain too much in the way of technical details. My understanding is somewhat homegrown and superficial - but I don't mind saying something wrong and have a brighter bulb straighten me out! In other words, don't take anything I say as an absolute answer.

That disclaimer out of the way...we touched on this in the other thread. Consider a circuit like this one as being: 1) a grid leak detector only, 2) an oscillator, 3) a regenerating detector. The difference between the three lies in a combination of biasing and amount of feedback.

Quote:
What is the adjustable grid cap adjusted for?


I'll have to get back to ya on that :) I notice initially that I can indeed find a peak for weaker signals but I need to work with the thing a bit more to realize the implications. I know what you're thinking (from the other thread).

Quote:
What is capacitor H adjsted for?


Look at it somewhat like a throttle cap in a conventional modern circuit. Its bleeding off some of the rf when in regeneration to sort of 'set' the range of the absorption ckt which is the front panel control.

Quote:
I presume capacitor E is adjusted to resonate with coil C, OK.


On this aspect I'm totally lost because the values don't indicate resonance at say the low end of the band. In theory its supposed to 'absorb' some of the regenerated field in a controllable manner but I still don't understand why it wouldn't be tuned to about the same resonance as the F-B tank. I have some ideas but I need to play with them. This is on my list of things to figure out.

Quote:
My biggest problem is understanding the parallel RC network from plate to grid. What is capacitor F adjusted for?


I guess you meant L-C and not R-C. In this case F-B is the tank for the main tuning. G is the coupling cap - normally 250-ish mmf but here its a trimmer ~50-500. I is the grid leak return. They are just dumping in the output from the plate back into the bottom end of the main tank. If you look at the ckt from an RF loop standpoint the 'cold' end is not really cold. Thats ONE of the feedback loops. But don't forget cap H. Thats another RF loop.

Now when compared with a typical tickler winding and throttle cap you have the same basic thing. Here the rf loop is to the colder end of the coil instead of coupling via tickler. And the control is obtained via that E-C absorption winding (primarily) rather than via cap H. Same dog with different spots. H *could* control it just as well.

If I'm guessing right, the idea was to let it regenerate as 'normal' but let the absorption ckt act as a high-q circuit to squelch the feedback rather than controlling the feedback by means of swamping with the throttle cap, etc.

Bill


Top
 Profile  
 
 Post subject:
PostPosted: Sep Fri 08, 2006 11:56 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11100
Location: Vieques, PR, USA
DXer wrote:
As stated in the other thread, I was rather skeptical of certain claims surrounding the Cockaday Four-Circuit Tuner. The two performance reports, one from Billy's great example and my quick test effort, were disappointing. Considerable emphasis was placed in the original article on stringent construction practices, especially in regard to the the coils. The coils indeed seem to be exceedingly critical. I would speculate that the average 1920s amateur would make rather substandard coils and be therefore met with disappointing results. That's probably why this circuit was driven into near extinction.

Bill, I feel I must thank you for demonstrating what this circuit is truly capable of. I must revist this circuit and be similarly impressed.

Recently there has been somewhat of a debate surrounding the 01A and 00A tubes. Any chance you could drop in an 01A for a comparison? I'm sure several (including me) would be interested in the results.

73's


Hi DXer...
Well, my first go at it was only a "lets see if it works" effort. I think there was a sincere intent by Cockaday to emphasize decent coils...18 wire instead of weenie 26 for example...and you could apply this logic to many ckts and achieve improved results. The Q with 18 is probably double what would be achieved with 26. 18 cotton -covered wire and that diameter/length ratio is about the max that you can get (within reason). I'm not sure they knew that for sure back then - its still argued among the xtal radio crowd but is in the right ballpark for optimal :)

I did pop in an 01A just for grins but made no effort to reverse the filament polarity set up. It was clearly not as good as the 00A. This is my first 00A set so I'm having to learn its behaviour as well.

-Bill


Top
 Profile  
 
 Post subject:
PostPosted: Sep Sat 09, 2006 8:17 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11100
Location: Vieques, PR, USA
Some details - John Amory I know you're curious about the grid coupling cap.
I have a station on 1370 about 1/4 mile away from the shack. One of my 'proofs' of a radio is seeing how close I can sneak up to 1370 without being overwhelmed. There's a semi-local station on 1480 about 10-15 miles away and its always a good test. Some 1970s era transistor radios can't pass and many of these old 20's circuits cannot cut it. Little TRF sets can't separate them. I almost always have to use an inline trap to tone down 1370. Not so with this guy.

Since I used a trimmer in this ckt it gave me an opportunity to do some twiddling and I'm quite surprised. By tweaking this coupling value I can go from 1480 not heard at all to 1480 being heard perfectly with no interference and can even eek out a distant 1430 station albeit with QRM. I'm sure this has as much to do with the L-C tank being efficient as it does the actual value of this particular cap.

In checking the low end of the band I get my best 'peak' on that cap in the same general ballpark setting. Selectivity isn't an issue lower in the band in my location.

After several T&E settings and rechecking the value of the trimmer the readings consistently fell in the 50-100 pf range, mostly to the lower side 50-60-70 pf. I would have no reservations at all about subbing the trimmer with a fixed 50 pf cap but I'd still point out to the next guy that the value is going to be dictated by other factors like the tank itself.

Now the grid leak resistor. I've been working with a 2 Meg just because it was handy. I can completely remove it and the difference isn't all that drastic....I lose some of the effect of the filament rheostat...so there's obviously some other leakage going on that somewhat serves the purpose. It has been said that the 00A likes a lower value like 500k and in this case that try failed me. It didn't allow me to hear 1480 cleanly. Maybe it makes for a more sensitive detector but in this case sensitive might equal overload. I'll stick with 1-2 megs.

hehe...I'm running with three dead 9-volt batteries...I should have checked them with something other than my tongue. I'm at 16.7 volts B+ !!! And 0.33 ma current draw.


Top
 Profile  
 
 Post subject:
PostPosted: Sep Sat 09, 2006 8:53 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 34326
Location: Sandpoint, IDAHO 83864
I think that being that the 00A is a soft tube, there will always be some grid conduction. That could explain why it works without a grid leak resistor. I notice several old circuits from 1910 to 1920 that used audions and they did not use a grid resistor at all. One book I have around here someplace said the tube has an imperfect vacuum, thus the grid leaks off its charge thru the tube.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Sep Sat 09, 2006 11:29 pm 
Member

Joined: Jun Fri 02, 2006 9:58 pm
Posts: 97
Location: Elsberry, Missouri, USA
You gotta be right curt, if that 00A is soft (does that mean gassy), that's the only way yo can get by without a grid resistor.

Question about grid resistor loading RF circuit?

What size inductance or capacitor is being used in
the tuning before the grid leak resistance?
What is the reactance of the L or C?
How does that compare with the load R?
That turns out to be a pretty high Q doesn't it?
Is the inductor better Q than that with infinite loading?

_________________
John Amery
1251 Deer Run Road
Elsberry, MO 63343-4011
KF0HX


Top
 Profile  
 
 Post subject:
PostPosted: Sep Mon 11, 2006 12:23 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 248
Location: Liberty, Texas USA
A nice looking set Bill, and I'm glad you are pleased with its performance. I wish I could say the same thing about mine.
exray wrote:
hehe...I'm running with three dead 9-volt batteries...I should have checked them with something other than my tongue. I'm at 16.7 volts B+ !!! And 0.33 ma current draw.

I think mine works just as well with less than that, but I don't remember exactly how much.

_________________
What we do today is the continuing history of radio - br


Top
 Profile  
 
 Post subject:
PostPosted: Sep Mon 11, 2006 4:34 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 11100
Location: Vieques, PR, USA
johnamery wrote:
You gotta be right curt, if that 00A is soft (does that mean gassy), that's the only way yo can get by without a grid resistor.

Question about grid resistor loading RF circuit?

What size inductance or capacitor is being used in
the tuning before the grid leak resistance?
What is the reactance of the L or C?
How does that compare with the load R?
That turns out to be a pretty high Q doesn't it?
Is the inductor better Q than that with infinite loading?


Hmmm.... I don't really have all the answers to those questions.
The tank circuit is typical for the era. I'm a little on the 'fat' side for various reasons but the coil is about 300 uH and the tuning cap 530pf at max. If I get in a surgical mood I had considered removing some turns. I followed the plans and didn't bother to recalculate. For modern BCB use that combo would do much better if the coil were 200 uH. Amazingly I was able to hear a station on 1620 so I'm not really losing any part of the band - its just that I've got a lot of dead rotation at the low end. Our local nav beacon is on 390 and I actually tried for it because with the values given it should just about make it! Given the choice I'd rather hear 160 meters than the beacon. But, read the question below...

The reactance varies with freq, of course :)

I think the tank is a good hi-q one. I had used this tuning cap previously in a crystal set and it was pretty good. You'll note that I had replaced the original stator insulators with plexiglas, I'm sure the coil could be better with 18 CC wire rather than enamel but I suspect that the difference might not be audible.

Maybe the poor selectivity when trying the 500k gridleak vouches for the fact that the tank is pretty sharp?

***********

Now, I've got a question.

I'm still trying to figure out the 'tuning range' of the absorption tank. Just for reference that winding only measures 100 uH and the plans asked for a 350, 460 or 500 pf cap. Clearly this alone doesn't reach the lower end of the band. Is "mutual inductance" whats at work here and giving enough effect to make that circuit effectively tune so much lower? That might explain why the inductance of the rf tank is on the large side and suggests that maybe I oughta not mess with it?

And if that's true...I'm really guessing now...back to the "automatic tuning"....I'm starting to suspect that the ideal fixed setting of that absorption ckt is one that is intended to track with rf tank frequency changes by virtue of the variations in mutual coupling/inductance.

Does that make sense?

-Bill


Top
 Profile  
 
 Post subject:
PostPosted: Dec Fri 29, 2006 3:23 am 
Banned

Joined: Dec Wed 13, 2006 4:03 am
Posts: 1209
Location: 18.1000N 65.4763W
Bumping this back up.

I finally got a grid dip meter and checked the "stabilizer" circuit resonance against the grid circuit. Sure as shootin' it tunes at exactly TWICE the desired frequency. And it actually tracks quite well for the "automatic" tuning through a good portion of the band.

Playing with the knobs and the GDO and all my other gadgets I have ruled out mutual inductance as being a player in the scheme...well, other than the frequency pulling aspects :( but thats also one of the advantages of the automatic tuning.

So, I'm still at Square One trying to figure this one out. Apparently it was 'known' that this LC tuned to the SECOND harmonic is able to goof the circuit enough to serve as an effective regeneration control. It seems that the circuit may remain in regeneration and the stabilizer, when tuned accordingly, squelches the circuit somehow without behaving like a trap at the desired freq. I've poured over the books and can't find any discussion on what the effects of a 'second harmonic' type of wavetrap might be, nor have I ever noticed anything in other circuits by tuning a trap like that.

Any fresh ideas?

-Bill


Top
 Profile  
 
 Post subject:
PostPosted: Apr Tue 24, 2007 3:43 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1488
Location: Jackson, TN
Hi all,

I've been silently lurking for a while, but finally have something new to talk about!

I am fascinated by the Cockaday and finally got around to breadboarding it. For this version I took some liberties, but believe that functionally it is equivalent. I'm using a type 30 tube. At first, it was very difficult to get anything out of, it wants to spread my local blowtorch over the whole band. But with a lot of tweaking I'm starting to see some success. Here are some observations...

1) I can never completely stop it from oscillating. I get a whistle at every 10 kHz, but it's hard as heck to tune it so I can actually hear the program. I've tried the throttle that Bill mentioned - no help. Tried a variometer in series with the feedback line from plate - made oscillation worse. I found that using 3 turns instead of one on the antenna coil seemed to boost dx stations enough so that many of them can be copied, but the oscillation / distortion is still present. Sure enough, the stabilizer cap does help to minimize it though.

2) I can not get any real tuning effect from changing taps on the loading coil. I can only use a very low tap or the blowtorch will dominate.

3) Grid leak value (tried 1 meg & 3 meg) does not seem to make any significant difference.

4) Stabilizer. There's been much discussion about how it works. First, it will tune in on fundamental or fundamental X 2. Example: my blowtorch at 1390 will scream when the stab cap is set at the 20% area. Given that the calculated stab tuning range is about 1000 - 3000 kHz, this is right where you would expect to find the fundamental. On weaker stations, however, it seems that the stab cap can be set at roughly F x 2 and that it does respond and reduce the regen effect. Also, as noted earlier, the stab setting may be left in a certain position for a range of frequencies, say over a 100 kHz range.

5) I'm running at 45V B+, this may be high, I'll try lower values.

6) Tuning is as sharp as a razor, I need to add a vernier.

Anyhow, I'm definitely interested in continuing to work with this set. I have seen glimmers of picking up stations on every 10kHz step, but this is fleeting and hard to repeat. The real challenge right now is how to limit oscillation without compromising the design!

General note on the type of regen feedback used on this set. I don't know what it's called but I've seen it on other old sets. ie take plate voltage righ back to grid (through a cap). As I think about this scheme, I have a couple questions:

1) How the heck to control it since the ac component on plate gets impressed right back to grid?

2) I believe that this scheme can feed back audio as well as rf. I guess the 250 pF grid cap would block most audio. It raises the question about using a higher cap value, say .01, that would pass audio to grid also. This thing would act like a reflex?

Enough rambling. I'd love to keep this discussion going, a bunch of folks built these in the 20's, they must have been seeing something.

Tim


Top
 Profile  
 
 Post subject:
PostPosted: Apr Tue 24, 2007 2:53 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 1020
Location: St. Louis, MO, USA
Awhile back, I found an old home brew cockaday radio. It need a lot of work but I was able to finally get it working. It uses a '99 tube. I found out that the B+ plate voltage was very critical. With 22 volts, all I heard was one or two faint stations, but when I reduced the voltage to 18 volts, I was able to get several stations.

- Dennis


Top
 Profile  
 
 Post subject:
PostPosted: Apr Tue 24, 2007 4:58 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 34326
Location: Sandpoint, IDAHO 83864
I have seen battery advertisements for B batteries back in the early 1920's where nearly every cell is tapped and brought out to a binding post on the top of the battery. Since some of the early circuits and tubes were very sensitive to B+ voltages, it stands to reason why the batteries were constructed this way.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: Apr Tue 24, 2007 7:57 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 7673
Location: 13 Critchley Avenue, PO Box 36, Monteith Ont, P0K 1P0
Tkilboy wrote:
Hi all,

.....

1) I can never completely stop it from oscillating.

.....

I'm running at 45V B+, this may be high, I'll try lower values.

.....

Tim


Exray had the same troubles on this set and ended up cutting the B+ way back IIRC.

I think he also ended up at about 18vdc.

cheers


Top
 Profile  
 
 Post subject:
PostPosted: Apr Tue 24, 2007 8:13 pm 
Silent Key

Joined: Jan Thu 01, 1970 1:00 am
Posts: 34326
Location: Sandpoint, IDAHO 83864
It should be at the top of the ten commandments for regenerative set building, but you don't see it mentioned much. But anybody with experience with them knows exactly what I mean when I say to always go with the lowest voltage on the detector. I have had dismal results when going over 30 volts and always have had best results with many tubes at around 18 to 20 volts. Eighteen seems to be some magic number when it comes to plate voltage on regenerative detectors.

Note- this is when triodes are used. If you are using screen grid tubes, the plate voltage can go higher, but keep the screen voltage down to the roughly18 volts.
Curt

_________________
Curt, N7AH
(Connoisseur of the cold 807) CW forever!


Top
 Profile  
 
 Post subject:
PostPosted: May Thu 03, 2007 12:16 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 572
Location: AUSTRALIA
If anyone is interested, I discovered I have what I believe to be a homebrew "Cockaday" 5 tuber (minus tubes) in my collection. I bought it as a junker because even though the panel was broken, it had a heap of audio trannys. It may have been a kit once as the coils seem to be professionally wound, especially the pile wound loading coil, as it seems to require precise crimping of the wire where the turns overlap (or crossover). [Does anyone know if there was a special tool to accomplish this type of winding?]
There also appears to be a professional, engraved, bakelite battery terminal strip at the rear. If any one wants the pics. and circuit they can email me as I don't know how to post them. The radio appears to have been modified/upgraded (inexpertly) at some time but I thought some of the circuitry may still be original enough for you guys to answer some of your questions about biasing and voltages. It has an 18v supply to the first tube (in all, 5 separate supplies 6,9,18,45,90). It is also equipped with a pushpull output but no wiring to the grid of the second tube. Maybe a single-plate-to-pushpull-grids tranny was changed to a standard one after burnout? Somebody mentioned "razor sharp tuning" which is catered for with trimmer type tuning condensers (duel knobs). I'll leave to you lads to come up with some theories. Cheers, Phil.


Top
 Profile  
 
 Post subject: Re: My results-"Improved" Cockaday 4-Circuit Tuner
PostPosted: Jul Fri 24, 2015 5:59 am 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 156
Location: Lancaster, Ca 93536
Hi Bill,
I really would like to build up this "Cockaday". I've been reading everything and wanted to start. I was wondering if you had the pictures you posted originally. Seems everypicture I want to see is gone. Really sucks
Thanks
Jack McMullen KG6INX


Top
 Profile  
 
Post New Topic Post Reply  [ 29 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 4 guests



Search for:
Jump to:  




























Privacy Policy :: Powered by phpBB