Thursday, December 30, 2021

DBM transformers

The progression of DBM transformers:

The top one is wound on the type core The Chief designer and I used in the original 40/80 receiver. He wound it himself!

The middle one was the next one in the progression. Designer In Training wound it for him. He bragged she had excellent vision.  Good hand eye coordination too!

The bottom one I wound today. I used an eye piece and stumbled through it. For reference the wire is #34.

Two a day is about all I want to wind. Anywho time to build some DBM's. Our previous test told us the smaller ones performed better. The new one should be spectacular????

Sunday, December 19, 2021

Reflex - RF and AF share an amp


You can look the screen shots over and see both RF and AF is passing threw the stages. 

This was my original circuit. A quick look at the scans will show I changed the value of C3, bypassed the output with C4 and added a choke. This change removed a lost of noise in the output.

Tuesday, December 7, 2021

In search of the elusive Q!

 We hear about the Q sometimes as though it is impossible to comprehend. I will attempt to help clear some of the allusion. The Q is simple a ratio of resistance and reactance. First a sim , then a test.


I made these coils and they were just laying on the bench so I will use one of them for the test. 

The jig is quite simple except the circuit load the generator and the test is invalid. I tried using a 1K resistor to eliminate the loading. It sort of worked. It takes 10 volts in to produce 1mv across the test circuit.

without the resistor I get about the same input with 1 volt.

Using a transformer the results are much better.

I learned a couple of thing from this test.

1. The test signal should be less than 1 MHz.  What range is your meter good for? My meters are rated 100KHz so I would look at using maybe 50KHz to 100KHz. 

2. I show 2 caps in the circuit. One is adjustable and one is fixed. You can use the fixed to get close and then tune the circuit with the adjustable. 

3. the RF transformer needs to be in range too. Watch the core rating.


What about that Q. Examine the test circuit and you will find the input is in parallel with a coil and cap in series. The first meter monitors the resistor voltage. The coil and cap will have Er times Q volts across them. The voltage on the coil and the cap are 180 degrees out of phase. In a resonant circuit the coil charges the cap and the cap charge the coil. It is like a juggler passing his balls from one hand to the other. Look at the charge and discharge path for the series resonant circuit and you see the input signal is in series with it. So we have Es in series with and equal to (El - Ec). And this is Q in a nut shell. El can be equal to Es or it could be 100 times Es. What determines this value? 

Circuit resistance does. If the coil reactance and resistance have a ratio of 100 the Q is 100 and the voltage gain is 100. How can this be? Remember the circuit has +jx and -jx values which are equal and therefore cancel. The current limiting factor in the circuit is resistance. Remember E=I*R. In a series circuit I is the same throughout. 


E reactive is amplified by the coil based on the current limit the resistance establishes. 

That's it.

The coil field builds and store energy limited by the resistance. (In theory it will take all the power source can supply.)

The coil field collapses and attempts to maintain the current limited by that same resistance. (Of course in practice we see losses but this where the elusive Q comes in.)

A purely reactive circuit would have a Q of infinity. 

A circuit with the resistance equal to the reactance would have a Q of 1.

And so the journey to finding Q begins.

Thursday, November 11, 2021

Low level crystal detector Vs high level detector And a dab of Z matching.

Before we can examine a detector we must define the type - high level or low level? The diode has 3 sections in its characteristic curve. It can be cut-off, conducting or in the twilight zone called the square law region.


The vacuum tube has the same three regions which may be used for high or low level detection. In this case we see a low level detector operating in the square law region.


In a high level detector the diode is switched on and off. Then the RF is filtered to develop the audio. We would see this type detector in a tube or transistor radio with the signal amplified before being detected.

The following screen shots show the small and large signal detectors.

Now the Z match in one page. 

Notice when the Zin matches Zout half the power is lost in the generator. In a crystal set working in the micro volt range this would be a  problem. See the alternate definition of Z matching on the right hand column.

Look at the load on the printouts. This chart relates a 100 Ohm source to loads from 1 Ohm to 10000 Ohms. The 10k load receives a higher %age of the power. This would be good for the small signal detector. The higher the better? 

Look back at the large signal detector. It could drive a high Z head phone? 

So the first step is define the project and stick with the design. Don't talk apples and oranges. 

A crystal set is limited by the signal and power levels. A TRF or superhet would be a different animal.

While you are thinking about these things consider the antenna system and feeder system. A transmitter hookup would require a large signal approach.  A receiver could be more small signal approach.

Sunday, November 7, 2021

A coax maker explains the Z here

 Just putting this here before I misplace it.



Friday, November 5, 2021

What determines coax Z. What determines twin lead Z? Is all coax 50 Ohm? Is all twin lead 300 Ohm?

 These scans from the calculator should tell the story.

It appears to be a matter of construction. So why 50 OHMs?

Z matching? Some points to remember.

 The old saw is to obtain max power transfer you must match the source and load Z.

The part that is left out is WHEN Zin = Zout half the power is lost in the generator!

Consider the 1000 Mega watt power generator feeding the grid. If it was Z matched HOW WOULD THE GENERATOR SURVIVE 500 Megawatt. The answer is it would not. The wires would melt and the flash over would be exiting. I saw this once. Thankfully the lead is a fusible link. It produced the flash in a confined space.

Anywho, Just look the sims over and see if it make since.

the delay produced by the Tline with matched Z.note the 50% signal loss.

lower the gen Z and the output increases. matching.uh?
Raising the load Z is not so good.
A lower gen Z reduces losses. Look at the rise in output.

If matching ALL of the circuit only losses half the signal what is the Tline Z?

In other words the Tline is 50 Ohm? So why is there no loss produced when I insert the Tline?

Hint: The Tline is not resistive.

Hint: The Tline is not a flat 50 Ohm. Actually it may vary much more than one might think.

So when they say the Tline is 50 Ohm what does it mean?

Maybe more on this later?

Sunday, October 31, 2021

Single supply op amp in sim


Single battry supply for an op amp

Using an op amp  for a simple radio. Single 9 volt supply. An example of how to "split" the supply ans create a reference for the inputs.


Friday, October 29, 2021

Active antenna for shortwave

This little circuit is a start point for an active antenna.
I labeled some tie pints to help with the hookup.

The gain is good.
Look at the current drain. (as designed)
Subbing a J112 increases the current drawn but the gain is still about the same.
Checkout the frequency response. My target is 3 - 10 meg. Ok for that range? Using a J112 here.

Using the 2SK2539 reduces the current draw. You could use any JFET but the performance will vary.

What would happen if we vary R13, R14 and C9?

What would happen if we vary R16, R17 and C12?

Does it need a RF gain pot?


Friday, October 8, 2021

Update on 1.5 volt receiver

A question about this circuit started me looking at it. It uses a 9 volt supply but why? The audio amp and Vcap tuning require 9 volts. The detector uses 1.4 volt.

 Using this portion of the circuit with my 1.5 volt hearing aid amp give good results. 

I would substitute a 200 ohm resistor for R1 and use a lower value pot for the regen. My test determined R2 could be around 10K. With the 100k pot it operates close to the pots low limit.

I did assemble the DBM and Lambda Diode set. That is still an option.

I am considering a discrete component Gilbert Cell. While this is a good option it would require a higher supply. 

Someday maybe the Gilbert Cell, Lambda Diode, and earbud amp will find their way onto a board. The Lambda Diode does not need a tapped coil so it would be possible to make a socket to use plug in coils. The Gilbert Cell provide conversion gain. This could be the beginning of an all band receiver. It could use one Vcap and plugin coils.

Some food for thought.

Tuesday, September 21, 2021

Adding a Lamda Diode Oscillator and V-Cap


The four transistors produce a Vcap to tune the oscillator. The sim is stepping the 10k pot from 1k to 10k in 1k steps. The 100p cap (C2) set the range while the pot produces band spread. If C2 is a variable air cap you could tune with it and use the pot as a band spread. You could use a switch to select C2 as a band selector.

* I need to do some adjusting on the circuit but this is a good start. Putting it here so I don't loose it.

Friday, September 3, 2021

3 volt earbud amp

 Back to the project. I will be going with the 3 volt supply. My earbud is rated at 1ma. This amp look good.

The input is 500uv for 1 ma output. 

The battery drain is a little less than 1 ma. R6 could be adjusted to vary the Q current. If R6 was 1 meg it would probably be good but for I'll go with 100k and expect to need a volume control. 

Time to clean off the bench and gather parts for a build.