Stella ST425a LW & MW Transistor radio

At 5 1/2 x 3 3/4 inches and 1 1/2 inch deep, quite a small design for the time, 1964. It appears to have been made by Philips as the circuit diagram for their L1G30T model is the same, just the outside finish and leather case that differs. Both employ a compartment with a switch behind the battery door to hide the permanently wired ear-piece. It took me quite a while to sort out what the switch was for as my find had the ear-piece missing, I thought at first it was a tone switch. Looking to be quite rare as even the radio museum web site only has drawing of it. The Stella has a diamond (Rhinestone) installed to the front above the volume control.

The repair
I nearly gave up on repairing this one, it was a tough old boot. The 3 OC81's gave plenty of audio when a signal was applied to the volume control, but no RF anywhere to be found in the RF stages.

I could get a small signal on the base of TR1, but little elsewhere. The RF detector diode X1 was open circuit, I swapped all three AF115's  RF transistors, I think as I applied heat from my iron to the print, components were starting to breaking down. The last IF transformer had an open circuit secondary winding, and when removed fell to pieces, it's tiny about 5 mm square. At this point I decided to abandon the RF section in favour of fitting a TA7642 integrated circuit with just 3 pins and a few bias components fitted to a small printed circuit board that I etched myself overnight. They can be brought on-line from a UK seller for about £1.30, including postage, or if you have more time 20 can be purchased from China for about £1.99.

The circuit to replace TR1,2 & 3.

All I needed to do was clear the redundant components from the printed panel to give me room to fit the new one over. I took the switched MW/LW signal that once went to TR1 base and fed it to the input pin of the new I/C,  (Pin 2), pin 1 goes to earth and 3 is the output signal to be fed to the volume control.

The broken IF transformer (I could repair the broken wire, but the base print fell away)

The panel after RF stages removed. (leaving just enough room for new overlay print)

I know this all sounds like cheating, but what you must remember is this radio is probably rare because all the other repairmen have given-up and thrown their repair away. This repair restores it to working again with 2 bands and all the original controls still work, so on the outside she looks the same.
The original IF coils would have been tuned to 470 Khz and this makes the stations take up less room on the tuning dial, these days there are only 2 or 3 stations on these bands anyway and well spaced out, so no real loss.
Before you do any work the tuning wheel and cable along with tension spring will need to be removed to gain access to the print, take lots of photo's to see how to re-fit cable, can be quite a tricky job, much patience is needed.

The printed board I etched myself and fixed in place by double sided sticky tape

The TA7642 only requires 1.5 volt to power it, and this will provide nearly 1 volt RMS audio on it's output, more than enough to drive the OC81 transistors. I took the +ve from that and fed it to bottom 1.5V battery in the chain of 4, via a tin plate and a hole on the battery box. After fitting all this I still had no RF from the aerial coils and decided to remove and check the tuning capacitor, having already checked the ferrite rod coils. It turns out the earth pin had parted from the print unseen below the component, I gave out such a scream, lol. Having put that right with an extra piece of wire, the set sprang back to life again after it's long sabbatical break.

The original inside showing the phone switch on right (I removed that too)

The case & cabinet
The leather case is one of the best I've ever come across from this period, I think it survived the years because the opening was at the bottom and not the top like most portables, the thin leather frays and breaks. Because this radio was designed to gain quick access to the battery cover, so's to bring out the ear phone, it made it stronger. I used Bakelite polish on the plastic and a toothbrush to get to awkward places, there was a small chip centre top on front that I filled with white acrylic modelling paste. Otherwise there was some yellowing to the plastic where the Sun came around leather top, a little Vim powder (quite rare now) rubbed on that.

Leather in nice condition - thankfully

here

The alternative version in 1964 with Philips logo on it (Carry case was alt. too)

L1G30T Philips Radio ( Paul Rumpf in Australia  writes:- 

I have had one of these since 1969. I started working at Philips Hendon works in Adelaide, South Australia, in 1964 and was a development engineer in the Components Laboratory. This department produced HF coils, Deflection system components, wound coils and transformers, etc., and other electronic items which were made for the television and radio assembly lines.

The L1G30T came into my hands when I left the company in 1969. The radio was, I believe, produced in Europe, and was intended to be an imported radio. I believe that there was a resistance from the Australian management to imports which would compete with the factory production schedules

my MW only version

.Recently the radio stopped working: the sound became distorted and operated on and off. I searched for information for this radio and found it was based on the L1G30T. Australia never used LW broadcasting and the L1G30T was a cut down version for those countries that did not need the LW Band. Examination of the plastic moulding shows the cut-outs for the band switching buttons, but the ferrite aerial is directly soldered to the PCB.

Noldy's comment about his difficulty in repairing his L1G30T are interesting and need comment. He said he could not get any sense or signals in the RF section. I suggest that this because the radio is a superheterodyne receiver and if the local oscillator fails, the there is no IF (455Khz) signal to be found. It is a pity he wrecked the guts and replaced it with a High Gain amplifier and effectively converted it to a TRF receiver.

The original coils were a 1960 development from Dutch Philips. These coils are VERY small AND delicate. They can be removed satisfactorily BUT one MUST remove EVERY trace of solder from the pins before trying to move the coil. In Australia, there were two modules designed with this coil design. The modules are UF 40X and UF30X. the UF3 is a 3 transistor superhet design with self oscillating mixer. The UF4 is a 4 transistor design with bandpass IF's and separate oscillator. The wire used to wind these coils was typically a 3 x 0.03 mm litz wire.

My L1G30T has 3 x AF117 transistors in the RF section, but the audio amp is different. The preamp is a AC125 and the output transistors are a pair of AC128's. Otherwise the design is identical to that of the published circuit diagram.

The loudspeaker in my radio has no magnet shield as in the published pictures. It has a small Alnico magnet and the impedance is 27 ohm. My belief is that the original speaker is a 2 ½ inch diameter. In my radio, a 2 ¼ inch speaker is fitted. It is glued to an adaptor ring to allow it to be fitted into the moulded case.

Looking closely at the speaker I found that the flexible tinsel wires from the tag-plate to the cone showed signs of green discoloration. Gently probing one wire caused it to disintegrate. My experience of loudspeakers is that braided conductors are used, but in this radio, tinsel conductors were used. I punched a small hole in the cone, mounted a small nickel plated eyelet, and soldered it in place with a circle of copper wire on the 'other' side to retain it, and reconnected the eyelet to the shorter tinsel on the back side and used a fine copper wire to connect the eyelet to the voice coil conductor. Then swamped the whole arrangement with contact adhesive. The loudspeaker is OK.

As a passing comment, it was known that manual assembly operations using females and fine copper wires, did suffer from the menstrual state of the ladies. There is apparently a change in the acidity of sweat during the cycle. The radio still did not work however.

I have an old Tektronix 564 oscilloscope. I checked the local oscillator output at the junction of R27 and C6. The receiver was set to around 700 KHz. At the R27/C6 junction there was a signal of around 0.5 volt peak to peak. This signal varied in FREQUENCY as the tuning dial of the radio was changed. The frequency was measured and found to be about 1.1 to 1.2 MHz. This is the correct LO frequency for the 455 KHz IF. The LO was therefore functioning. The amplitude of the LO signal at T1 collector was about 2 volt P-P.

At the collector of T2, the signal amplitude was around 0.15 volt peak to peak when the radio was tuned to a broadcast station at 693 KHz. This frequency was about 455 KHz, and when the tuning dial was adjusted, this frequency did not change but the amplitude reduced as the tuning was de-tuned. This showed the IF signal was present at T2 collector. The picture of the PCB track for T3 collector does show (not very clearly) that the track width becomes very narrow just where the track deviates around the pin number 1 of the 2nd IFT ( coil L10a). This track was broken; transistor T3 was loosely mounted and over the years, vibration and movement broke the track. Repairing this track fixed the problem.

The voltage at T3 collector was measured and found to be around 2.5 volt peak to peak. The frequency is the IF (455 KHz). I tried to measure the RF input signal amplitude at T1 base, but the received signal level was too small. At T1 collector, the LO signal swamped the IF signal. At T2 collector of course, the LO signal has been filtered out t\by the action of the IFT1 (coils 6,7, and 8).

The coil design has 3 pins on one side and two on the other. Philips Australia drawings show the pin numbering as: 1, 5, 2, 3, 4,; going from the bottom left and then clockwise; just like the numbering of valve base pin numbering. Pin 5 is between 1 and 2.

Not much more to say about the radio, except that it is probably very rare, being a prototype marketing model that was not sold in Australia. Attached are a couple of pictures.