Antenna coupling with antique Transmitter

[Buzzkf9cm]

I have a problem with a replica Antique MOPA Transmitter. The transmitter is a MKIII "Tinker Box" use in Station X at Bletchley Park. I built the transmitter using information I got from corresponding with various people that had first and second hand knowledge of the radio. I gleaned the measurements using photographs from several publications. The photo attachment shows the original coils and the antenna coupling coil seems to be wound too far from the tank coil. It was my belief the antenna was a doublet fed with 600 ohm ladder lead. The output valve is an 807 with 675 volts to the plate and 45 MA under load. This works out to about 32 watts input. Using a small 7 1/2 watt light bulb across the antenna terminals, I only get about half brillance. This is probably about 3 watts output. I don't want to change the configuration to a pi network or change the coil lay-out. I want to maintain the original circuit. Any suggestions would be appreciated.


TNX Gary, KF9CM

[MichaelR]

Hi gary,

Are you sure these are the antenna coupling coils

Mike

[Skywave]

This TX would have almost certainly have been used to feed a balanced antenna via a 600-ohm balanced line. What you need, therefore, is an antenna coupler to transform the O/P Z of the coupling coil to the Z of the TX line. Have a look in any Amateur Radio publications of the period 1930 - 1950: for example, The Radio Amateurs' Handbook (ARRL) ed. 41 would be a good choice.

Oh - and don't overlook the importance of a LPF betwen the TX O/P and the antenna coupler: again, the publications I refer to will give you the necessary guidance.

Al / Skywave.

[Buzzkf9cm]

I'm attaching a schematic that I used to build the transmitter. There is no known schematic other than a hand drawn representation of the transmitter. I used several MOPA schematics from the 1936 to 1940 Armature Hand Books to assemble my schematic shown in the attachment. If you will notice the output configuration appears to be a dual type balanced and unbalanced. This was agreed upon by several sources. Perhaps there was a tuner involved in the original configuration at station X.

[dave walsh]

The Doublet Aerial seems likely Gary [looking at period military theory] but there is no chance I could solve the coupling conumdrum! However your "Station X" website is very interesting indeed. Dave W

[Skywave]

Now things are becoming clearer!

Re: the O/P coupling coil. As the cct. stands, there is no way that that coupling winding will match a 600-ohm line: the turns ratio, pri. to sec. is 1:1. Therefore, if it is to feed a balanced antenna with 600-ohm feeders, 300-ohm ribbon or even 75-ohm coax, an external matching unit will be required; balanced or unbalanced as the feeder to the antenna requires.

As it stands, the top pin of the ANT connector looks like it could be used to feed a random length of wire with the series 150pF cap. to tune this aerial to resonance.

Al / Skywave

[OErjan_S]

hmm, there seems to be 3 coils but only data for two?? odd, any indication on the small coil next toL2?
that smaller one is likely the antenna matching part of circuit. but odd configuration for 600ohm, as Skywave say, likely to work for inverted L or random wire.

hmm, if that was a winding between a split L1 to give a balanced output...

[MichaelR]

Hi gary,

How are you tuning up the transmitter.

The circuit shows a 12v bulb which I assume you use to tune the grid circuit. A class c amplifier needs sufficient drive to get the output power you are expecting.


Mike

[Buzzkf9cm]

On the schematic L2 are three separate coils; 1.7 MCS, 2.5-4 MCS and 4-8 MCS.
L1 is the oscillator coil. (MCS is MHz before they gave Hertz the credit)

In retrospect I have built 2 Paraset replica spy radios and the output/tank circuit is set up in that radio is very similar to the MKIII. The exception is in the Paraset they don't bring out the aerial cap to a terminal like the MKIII does.
TNX for the insight Skywave.

I think the capacitor output terminal is meant for additional capacitance to match a long wire antenna and not to feed a balanced antenna.

If you click on this LINK
You can see my projects and my ongoing Station X endeavor.

de Gary, KF9CM

[Sean Williams]

Hi Gary,

This sounds wrong to me - the idea of a balanced antenna system just does not ring true, given the age and simplicity of the transmitter - much more likely to be an end fed antenna.

Anyway, basics first, what sort of grid current do you have on the 807? - 1-2ma for full o/p

Have you tried the old ferrite/brass trick on the coils - (ferrite rod passed in the coil, then a brass rod) noting change of output power.

There isnt much to go wrong with this sort of thing, so you cant be far out!

Cheers
Sean

[OErjan_S]

where have you hidden L3 and could there be a mistake in the circuit?
i can see 5 RFC and 4 coils, one is for the bulb, then you have the two large coild for the tuning and a smaller coul for coupling to antenna.

[Station X]

This is essentially a spy set and it would be very unusual for such a set to use a balanced output. The information I have quotes the RF Output as about 25 Watts.

[MichaelR]

Graham,

Gary has built a replica set this does not mean that the circuit is as was used in the original.

Moreover the unit he is replicating is that which would have been used at bletchely park hence it would have used balanced line.

The only similarity to Beltchley park is that it is genericaly a MOPA design, (master oscillator, power amplifier)

Personaly the only way you can control grid drive from what I can see is by adjusting the grid tuning control. Any body who has setup a valve power amplifier knows that it is an iterative process constantly adjusting the loading , PA TAnk coil and grid tuning coil to "move" the PA valve into an operating point whereby you get the output with the valve in spec from an operation point of view.

I agree with graham with regard to the L2 coil, the output impedance of the PA valve with at approx 600V Anode voltage and screen and Anode current of 45Ma is about 13Kohms.

regards
Mike

[Station X]

I knew that Bletchley Park had receivers, but I did not know that transmissions were made from there using a MK.III transmitter.

As you say there is no grid current control, so the lamp would just indicate that the crystal was operating correctly and that the grid circuit was tunned for maximum drive.

I understand that the operating instructions for the set are stiil available, so I assume these would give details of the antenna arrangements? Unless of course the Bletchley transmitter was unique.

Gary has obviously done far more research than my cursory efforts.

[MichaelR]

good point

[Buzzkf9cm]

This attachment is a partial schematic that was gleaned from tracing the wiring on the original MKIII. This task was preformed by Mr. David White Curator at Bletchley Park. As you can see this is a very basic circuit and only uses one RFC on the PA valve The tank circuit coils are all hot. There is no grid control other than the 1.7 to 2,5 MHz switch position, which introduces the variable oscillator capacitor. In the power supply circuit the only difference is there is no bleeder resistor. There was a series 7.5 k ohm resistor between the HT section and the oscillator. The original uses 400 VDC for the HT. The output does say "Doublet Aerial" ,but it seems to be a bit unbalanced and more suited for an end feed array.

With a shortage of parts that Robert Hornby and Arthur Newton ( the designers of the MKIII) had to work with, It worked very well.
I designed my circuit using refining techniques from that era and it sounds pretty well on the air. During the war the US sent a lot of parts and machine tools over to you guys under lend lease and you guys, being very resourceful, made excellent use of these items. Some of the most brilliant inventions came out of Britain during WWII. I'm kind of bias anyway, with the last name Giles My ancestors didn't come from France

de gary, KF9CM

[MichaelR]

Hi Gary,

Looking at the cct from David White, the oscillator does not have any Anode Voltage feed when in the 1.7mhz to 2.5mhz position.

How did they adjust the grid tap position of the Anode Coil and the switch (2.5 to 7.5) tap position.

I applaud your motivation to reproduce these sets, after nov 11th remembrance it really does stress the need to never forget what our previous generations had to do.

Mike

[Buzzkf9cm]

I think there's a connection between the bottom of the coil (where it's labeled "bottom") and that right angle point. It is also missing an extra variable capacitor that should be across the junction of sw1 and sw2 to the anode side of the coil. I think I'm going to construct a unbalanced to balanced tuner with a RF current meter from that era.

[Buzzkf9cm]

I did some calculations on the output coil (20 uh) and it has an XL of 430 to 650 ohms throughout the 2.5 to 4 MHz range. I don't know how this stacks up in the final analysis of this discussion, but this might be a factor in the original design. If you calculate the variable capacitor (150 pf), it tracks that range as well. If you look at the schematic there are two configurations one could use, the antenna leads across the output coil and the antenna leads across the ground and far side of the coil. The latter being the most desirable (IMO) because of the return path to ground and the series tuned circuit. Does this hold any logic or am I way out in the Moors?


de Gary,KF9CM

[Skywave]

I did a few calculations too - and came up with the following:

A typical short 'longwire' at 3.5 MHz will present an impedance at the TX end in the region of 15 - j200 ohms. Since the O/P transformer has a 1:1 turns ratio (or very nearly, on some of the other bands) and the O/P Z of the PA is about 13k-ohms resistive, (see Post above), these are the two impedances we need to match.

A calculation of the necessary matching network * gives a loading coil of about 20 uH and a tuning capacity of 110pF. The swing on the cap. should be enough to accomodate the freq. range 2.5 to 4.0 MHz without changing the value of the loading coil.

These figures are commensurate with the values shown in the cct. diag. for this freq. band.

This convinces me that this TX was designed to be used with a random (short) wire - which also fits in with Graham's (Station X) thoughts that this was probably a spy-set design.
I now withdraw my first thoughts, published above, re: "a balanced line would have been used" I was mistaken.

* I haven't gone into the maths. here; various standard reference books cover this topic adequately. I used the RSGB Handbook, 4th. Ed., page 11.28, for example.

Al / Skywave.