Radio8z part 15 tx design

this is the one i was referring to http://www.part15.us/node/2791

Robert,

I recall that you have asked for this and apologize for not responding. I had thoughts about writing a short article on the construction and operation but this has been on the back burner for a long time and didn't get done. I do have a schematic with parts values which I will post here later today. This should be sufficient for an experienced experimenter in case they want to replicate the transmitter.

Hope this will be of interest even if you don't build it.

Neil

yes it will. thanks.

Here is a link to the transmitter schematic:
Note: Aug. 20, 2012 The schematic was edited to correct the omission of C12.

Schematic Link

Some notes:

This transmitter is intended to be used with an external base coil loaded 3 meter antenna. Due to the filtering action of the output network which attenuates harmonics it can be used with an unloaded 3 meter wire antenna but it may be difficult to get a DC input up to 100 mW.

Transistors Q1and Q2 are hard to get and for Q1 a 2N3904 or 2N2222 can be used but the efficiency will drop a few percent.

Q2 can be almost any NPN non Darlington power transistor (such as TIP41A) and it should be heatsinked.

The values of C9, C10, and C11 were chosen for the particular crystal used and may need to be changed according the the crystal used.

The RF TRANSF. on the output is optional and is used primarily to isolate the signal/power ground from the antenna system ground. The gain in efficiency due to a better match to a 30 ohm load is offset by the loss in the transformer. The core material is powdered iron and no part spec. is given since the core used was from my junk box. The transmitter will work without it.

The DC input power is set using R1 and is measured by measuring the DC voltage across R12, dividing this by 22 giving IC. This is multiplied by the DC voltage across Q1 collector to emitter. Typical values with a base coil loaded 3 meter antenna are IC = 41.5 mA, VCE = 2.37 Volts giving a RF power delivered to the antenna system of 85 mW for an efficiency of 86%.

Neil

.

Neil, thank you for your great generosity in sharing your intellectual property.

I have printed the information for study purposes, and very much admire the skill built into the design.

There are times when its pleasant to sit and stare at good circuits.

Carl, thanks for your kind words. You will probably notice that the audio section is very much the Ramsey AM-25 design with a few changes. Why re-invent the wheel?

The "magic" in the RF section is in the component values which were selected to approach class E operation.

The final transistor is a bipolar and most designs favor an enhancement mode FET. I found, through experimentation and simulation, that the bipolar gives higher efficiency. Also, the comparatively low collector to emitter capacitance makes the operation more predictable.

Neil

FWIW, the 2N4047 listed for Q1 & Q2 cross-references as an NTE128. Mouser, among others, seems to have plenty of these in a variety of forms including TO-220 case. Assuming the cross reference is accurate, that may solve the hard-to-get nature of these transistors.

When I saw the 1k input impedance I immediately thought of the Ramsey AM25, and I have wondered in the past about that choice for an input impedance for a consumer product.

Although professional audio outputs tend to be on the low side, often designed to drive a higher impedance, there are many consumer products with output impedances higher than 1k, and wouldn't that result in signal reduction and possible distortion?

To see it from another point of view, more typical audio inputs on consumer gear tend to be 10K, 20k, even 50k.

A discussion on this point will be informative.

Carl wrote "When I saw the 1k input impedance I immediately thought of the Ramsey AM25, and I have wondered in the past about that choice for an input impedance for a consumer product."

First let me state that this doesn't cause any problems when connected to the line output of my equipment so I didn't change this. You are correct that a 1k input Z would reduce the signal from a high Z source but this has not been experienced with my equipment.

Secondly, the type of operational amplifier circuit used (U1B) is a non-inverting configuration which has an infinite Z at pin 5. But this amplifier is also used to amplify a DC component from R1 which sets the DC output to provide power to the final stage. The resistance seen from pin 5 looking through R5 is the value of R5 added to the resistance from the wiper of R1 to ground which can vary from zero to a bit below 1k ohms. Since the pot R1 is variable the Z at pin 5 seen by the audio signal will vary between 10k and approx. 11k ohms. To audio, this variable resistance is seen at the wiper of R2.

U2 and following stages serve as an audio amplifier and a DC amplifier so the AC and DC gains need to be set properly but are not independent. R5 sets the DC gain. Thus it is not simple to change the values of R1, R4, and R5 to increase the resistance to gnd. at pin 5. If R2 is made larger then the audio gain will not be a linear function of the wiper position of R2 and would be difficult to set. R2, at 1k, appears to be a compromise between input Z and good adjustment of audio gain. The circuit was likely designed this way to reduce parts count.

A simple way out of this is to add another non-inverting op amp stage at the input but since the present circuit shows no noticable effect on the line level signals from my equipment I didn't pursue this.

Neil

Appreciate your audio comments, Neil.

During the brief time I fiddled with the AM25 I don't recall having any problems with that input.

But here's a related question: would it be true to say that an unbalanced audio input is, by definition, a consumer or at best a semi-pro feature.

That is to say, is not "pro" audio equipment defined by balanced audio in and out?

Or, maybe those distinctions are a thing of the past.... ?

First of all, there's the "miracle"
that you - Neil - were able to
design and build the circuit.

For me, a non-engineer, that's amazing
by itself. Being a little dramatic -
having that kind of knowledge is sort
of a miracle.

Anyway - now that I got that out - thank you
for letting us see the design.

It was really nice of you to do that.

Bit by bit, I learn a little here and there.

Thanks again,
Bruce, DRS2

Hi Neil,

What crystal is used, and for what transmitting frequency?

I am glad there is so much interest in this and even if you never build it this is an opportunity to learn some of the inner workings of a transmitter.

Carl,

For short runs of let's say 10 feet or less the single ended audio connections work fine with the exception of the lousy RCA plugs and jacks. If longer runs are needed then two factors come into play. One is the cable capacitance which tends to attenuate high frequencies and the other is the probability of having hum due to differences in equipment ground potentials (ground loops). With differential connections the common mode ground difference signals are gone and the lower capacitance per foot of twisted two conductor wire is less than for coax.

Ken,

The crystal I used is from a junked digital board and it is not a very good one. It takes quite a bit of capacitance to pull it on frequency and it is very temperature sensitive drifting 3 Hz per degree F. I had to add an oven to keep it on frequency. It is cut for 1680 KHz. To do it right I would get a crystal from ICM cut for a fundamental mode of the frequency of choice with a load capacitance of about 20 to 30 pf. I think Carl has bought crystals and maybe would comment about this.

Neil

Crystal is not a girl's name, although that would be nice.

The crystal experience I have is as follows:

I got crystals for 13,560, kHz the Big Talker shortwave project, from ICM (International Crystal Manufacturing Co.) cut to the fundamental.

Previously I had ordered such a crystal from PCS Electronics but so much time went by with zero delivery that I switched to ICM.

By the time the PCS crystal, marked as "13.560 kHz", arrived, I tried it and it was actually 13.555 kHz, exactly 5 kHz too low.

I never notified PCS because it was $9 and I was busy.

As for the AM crystals, I also used ICM because of their previous experience with LPB specialized crystals that are cut (I think, but double check this) 6-times higher than fundamental and divided down.

I have never ordered an AM band fundamental crystal, but there are members who have.