Sweetheart Radio

[peter.sables]

As you may know I've a collection of WWII Spy radio's. Well this one is a new addition. At its time it WAS the smallest radio in the world. 4 1/2" X 4 1/2". That was in 1943... It had 3 1T4 Valves.

It is a receiver only. 50,000 were made, 8000 were dropped in Norway, the rest, well all over the place. At the time radio's were banned. No one had or were not meant to have one. They were dropped one month before "D" Day. So that people knew what was happening, if not the day it was happening.

It was an important radio. Its common name was a Sweetheart radio, but also called the Propaganda radio. To us it was simply called a 31/1.

This one is still having work done on it, It was a total mess. But it will work again. It being a copy.

Pete G4MRU

[dave walsh]

There's a page about it in the 1944 SOE catalogue reproduced in "The Secret Agents Handbook" I see. 30v HT.
Dave

[peter.sables]

Hi Dave,

Yes lots of info about this radio. It will run down to about 20v. So not bad. 1T4 valves! there must have been better ones about, even then.

Pete

[neon indicator]

RCA 1T4 appears to be same as later Philips DF91
The "better" (1/2 filament current anyway) was DF96 in 1954.
I think given the constraints of 1942 and project, the 1T4 was maybe best choice. What other valve would they have used.
A condition of the design was they had to use Domestic non-Military components.
I have the schematic redrawn in Eagle if anyone wants the CAD file I'll email it and and my new version.

Amazingly the 4.5V flat torch batteries are still made but few retailers stock them. I don't know exactly what the 30V Hearing aid battery was. I doubt it's in production.
I'm making a fairly close clone using Russian 1j24b valves (1991 production! Maybe 1960 design based on 1955 1j17b), less than 1/4 original filament current, 11mA vs 50mA.
http://www.radiomuseum.org/tubes/tube_1j24b.html
I'm using modern parts, but it's basically same design. Also I can fit the batteries in the case...

Some links also in my article to "Sweetheart".
http://www.techtir.ie/construction/sweetheart-clone

1966 Portable Russian radio using many 1j24b http://www.radiomuseum.org/r/aleksan...26r12p_12.html

[val33vo]

Interesting schematic I cant quite understand how v1 functions but I suspect it will be some sort of regenerative design whereby it will lessen the damping on the tuned circuit however as its anode is only connected to the HT ( via a de-coupling network ) and its cathode is grounded I cant quite see how it works maybe somebody could enlighten me here

regards

val33vo

[neon indicator]

It's badly drawn, The V1 anode also feeds L3, C5 and C8.

There is a properly drawn one in Eagle on the link I posted above. I can email anyone the file.

I've attached it here, but the Forum SW may resize it too small.

It's regenerative/positive feedback to increase Q. If carefully adjusted it may act as detector too. At lower levels of feedback with a stronger station the V2 may demodulate via slope detection.

The difference with my updated design:
3 x AA Alkaline instead of slightly larger Zinc Carbon Torch battery (but more capacity).
12 x CR2032 Coin cells as 36V HT instead of small 30V Zinc Carbon Hearing aid battery. There is more capacity and less voltage droop on the CR2032. About 8c each in quantity.

Regenerative feedback may be on G2 rather than Anode. I'll try both. Certainly the the 1j18b Hergé radio works best using screen grid feed back. Picks up UK MW in Mid West Ireland during daytime.

Modern ceramic earpiece out of Maplin €3 or £2.

Miniature "polycon" variable capacitors stripped from 2 x €2 MW/FM "pound shop" radios as the two variable caps. The original style are just about obtainable but expensive and large.

Reverse engineering of the original coil design (without one in my hand) reveals it is not aircore. In fact you can see a core adjuster protruding rear of case. A feasible smallish number of turns added to the MW ferrite rod /coil of the cheap radios mentioned creates L2 and L3 (7uH and 14uH). L1 is 85uH. The MW ferrite coil is somewhat more than this, but has a 1/3rd tap. So you can use the tap to earth and switch aerial to either end of L1 (the original MW coil) depending on aerial wire length.

The stripboard/veroboard has a cut-out so the 3 x 1j24b Russian Rod Pentodes take up 3mm less depth (PCB + soldered component leads). This allows the 3 xAA cells in a standard holder to fit on top in the case.

The 1j17b, 1j18b, 1j24b are not that dissimilar to IT4 /DF91 or DF96, except only 11mA 1.2V filament instead of 50mA 1.5V filament. And much smaller. The 1j18b works at over 100MHz, and the 1j24b is likely better, but has lower Anode current limit.

The "Sweetheart" doesn't appear to have anything unusual or innovative for the era, a very standard kind of design. The differences are Packaging and the low HT. Usually 45V, 67V or 90V HT battery pack used for those valves.

The end point of 90V Zinc Carbon is about 56V and 45V pack is about 28V

Since the HT is maybe less than 1mA, perhaps the 30V HT battery replaced 1/2 way between its fresh voltage and end point?

[julie_m]

I think the horizontal connection from L3-C5 joins the vertical connection from R2-V1, just because there isn't an exaggerated crossover symbol there as there is everywhere else wires cross.

Modern practice is to emphasise junctions with a dot and only ever to let three "wires" meet at a junction, so there can never be any confusion -- if you need four wires joined, you can use two 3-way junctions.

[neon indicator]

Agreed "ajs_derby".

Also logic suggests that V1 heater is drawn connected wrong to be consistent with the other heaters. I've put that the more likely way round on the new Eagle version of the schematic. As well as fixing the junction.

Also the 30V HT is in series with the 4.5V LT, so V1 is about 34.5V HT, The other valves 33V HT and 31.5V HT

[G4XWDJim]

I 'restored' an original one of these some time ago for a friend and did a write up on the American forum http://antiqueradios.com/forums/view...758680#p758680 It worked very well and was a compact litle set.

Jim

[neon indicator]

It's a lovely job and a good write-up, Jim. One of the Polish radios often confused with "Sweetheart" (they are only slightly similar on outside, but inside are superhet and dual band) recently sold for nearly £1000. I imagine the actual "Sweetheart" are not cheap in good condition. Military stuff with any covert connection seems to command higher prices than domestic of same era.

[peter.sables]

Hi Jim.

Im following all this and now mine is not far off. A problem!! I see on the photo of yours that C7 is a 1.o uf. were it should be a .1 uf. That could be why its so big.

Nothing really, just thought it worth a mention.

Pete

[G4XWDJim]

In the original C7 was 1.0 uf so I followed that. C13 was 1.0uf electrolytic originally but I upped it to 2.2 simply because I had a suitable component. The owner wasn't interested in complete originality so I felt at liberty to make minor changes.
It would have been worth reconfiguring the output stage to better drive the DLR 5 phones I gave him rather than the crystal earpiece of the original design.

Jim

[neon indicator]

I assumed that either they couldn't fit a 1uF or the schematic was wrong regarding C7. I increased it to a 1uF 50V electrolytic in parallel with ceramic on the Tatiana. See http://www.techtir.ie/construction/tatiana-design

I increased some 0.1uF to 220nF as the 220nF 50V ceramic parts are small. Some I have 0.1uF 100V X7R ceramic

[peter.sables]

Hi Jim

mm. Not sure about that but if it works, fine. Yes the 2.2 mf cap is fine. Linking out out C12 and R10 will do the job but note sure if the valves will list long.

Pete G4MRU

[G4XWDJim]

Originally Posted by neon indicator

http://www.techtir.ie/construction/tatiana-design

NI,
I think your Tatiana looks good. Not too sure about the polyvaricon capacitors but lets see how they go. Nice and neat and small with useful anticipated coverage.
I await the results with interest.

Thanks for the compliments about my efforts. It was enjoyable to do and I was paid well for my efforts which is always a bonus. The owner's a vintage car enthusiast and he raced me around the Prescott Hill Climb circuit in an open vintage car with only rear brakes. phew!

I think if I built a 'replica' I'd fit an RF stage on it with only one audio amplifier stage. Although this one was good I think it could have been better but I didn't have full authority to modify the original Norwegian design too much. It was supposed to be a sort of 'restoration'.

Best

Jim

[peter.sables]

Hi

Well project complete. No real problems to speak of. Apart from small points

The heater chain should be laid out as shown, The, shall we say swopped pins on V1 are correct as it is on the first in the heater chain to balance the full chain.

C7 Should be .1 uf and not 1.0 uf

C13 Should be a 1 uf as shown. 2.2 uf is OK but not right

Thank's to all who gave me input and help. It works just fine

Pete

[neon indicator]

Originally Posted by neon indicator

Redrawn Schematic

So this redrawn schematic is correct?

https://www.vintage-radio.net/forum/...9&d=1315310053

[peter.sables]

Hi

No its wrong. V1 should have its pins switched over. It will still work as is but not has well.

ie:= V3 in at pin 7 out pin 1 V2 in at pin 7 out pin 1 V1 in at pin 1 out pin 7 (earth)

Pete

[neon indicator]

Why would f+ and f- be swapped compared to the other valves when "earth" is LT+?

Surely f+ goes to "earth" as the rest of the heater chain is more negative? Hence R1 goes to f- which is the "f+" of v2

Ignore the orientation of valve graphic or pin numbers. They are the wrong pin numbers as I didn't have a 1T4 / DF91 in the library then. Go only by pin names.

[peter.sables]

Hi

Yes indeend. you have answerd it yourself.

I swapped over the pins myself and did not realy notice any diferanc myself so put it back.

Pete