Saba FM multiplex adapter 12 coil issue

      Saba FM multiplex adapter 12 coil issue

      Hi,
      I repair lots of Saba automatic radios in Canada.

      Today, I am struggling to revive a Saba FM Multiplex Adapter 12, and have an issue with the coil L5 (doubler 19->38 kHz).

      The doubler works, but after some minutes the 38 kHz voltage drops significantly. If I cool the coil (with cool spray) the voltage is back up again…(supply voltage is correct and stable, the 19 kHz voltage at the oscillator coil is perfectly stable).
      Therefore I conclude that the coil L5 is faulty (insulation failure when it heats?).

      Does anyone have an idea of the specifications of this coil (listed in Saba documentation as part 3872 507 111)?
      If I could get hold of its specs, I could perhaps have a replacement made (by Reinhöfer, for instance)…

      Of course, a spare, or an adapter out of service (but with this coil OK) would also solve the problem ;) ...

      Thank you in advance for yout insights!

      Hi Chris,

      These are your options:
      A)
      1. Remove L5 from the circuit. Then measure the ohmic resistance between the center tap and each of the two terminal taps. You may assume that the center tap is actually "in the middle", thus ohmic resistance to both side should be similar. If the wire is broken inside the coil (your observation suggests that) you will find at which side its broken.
      2. Measure the inductivity in the frequency range > 5 kHz, < 50 kHz with a quality RLC-bridge from center tap to both terminals each. Inductivity measured over the terminal taps shall be 4x the inductivity measured between center and terminal tap.
      3. With these measurements you may figure out in which half of the coil the wire is broken or at which of the three taps you have a bad solder joint. If it's just a bad solder joint, you have a good chance to fix that by removing one turn and resolder.
      4. Above measurements give you the required data for reconstructing the coil (determine the wire gauge additionally).

      Nevertheless, its tedious and if you can't make the coil by yourself also costly.

      B) Obtain a spare SABA stereo multiplex adapter 12 and replace it as a whole. Probably difficult to find one.

      C) Or replace the SABA multiplex adapter 12 by a transistor equivalent. Projects were described in the web. Requires good electronics skills.
      Attention: SABA multiplex adapter works with 230 V supply (!!!). Replacement transistor / IC decoders for 7-pin and 9-pin tube socket SABA stereo decoders usually work with 6.3 V heater voltage and have different socket wiring. When such replacement is used the connecting socket wiring (and socket) in the radio must be changed to make it compatible with the transistor replacement unit.
      i.e. ebay Canada item number: 405377045202 (9-pin socket version for 6.3 V).


      Good Luck, Best Regards

      Reinhard

      Dieser Beitrag wurde bereits 4 mal editiert, zuletzt von „oldiefan“ ()

      Hi Reinhard,

      Thank you for your suggested options.

      A) I do intend to take L5 out and make some measurements. However, I don't think the wire is broken, as it does not completely stop working. My best assumption at this stage is that the insulation between windings is faulty, and this places a resistance in // with the inductance when hot, thus dampening the resonance.
      If I have the specs of the coil (via measurements or documentation) I will try to wind a new one or have it wound by Reinhöfer.

      B) I'd love to get a spare adapter 12... actually I often look for some, since they were provided as options and we find many radios in which they were not installed... Do you know of a source? I don't see them often on ebay...

      C) I have done that in the past (based on the description in jogis-roehrenbude.de, based on the Loewe Opta decoder). It worked very well. But here I am working for a client, I cannot invest the same amount of time ;)

      Thank you, Best Regards,

      Chris

      Hi,

      Here are some measurements of L5. Inductances are measured at 10 kHz. Resistances are measures with a multimeter, accuracy not better than 10%.
      Looking at the schematic, the three connections are Schwarz (+230V), Rot (output, connected to C17), and Grün (plate of ECF80 penthode).

      - Between Rot and Grün: 5,5 mH, 5.4 ohm;
      - Between Schwarz and Rot: 1.38 mH, 2,5 ohm;
      - Between Schwarz and Grün: 1.41 mH, 3.0 ohm.

      How I analyse that:
      - the overall inductance (5,5 mH) matches the value obtained via calculation (resonance at 38 kHz with cap 3,3 nF => L = 5,3 mH);
      - the fact that the sum of the two halves do not equal the overall can be explained by the fact that there are not two independent inductances in series, but a unique coil;
      - the tap is intended centered, the difference between 1.38 and 1.41 mH is probably not intentional, it is influenced by the ferrite core used for tuning;
      - the slight difference in resistances is probably due to the fact that one half is wound inside the other, therefore the wire length differs.

      These measurements have been made with a coil previously tuned to resonance (oscillator at 18,965 kHz, as per specs). Tomorrow I'll try to measure the overall inductance with tuning all the way to one side and all the way to the other.

      Unfortunately I am not equipped to measure the wire gauge.

      I don't know if the following information would be sufficient to have a new coil made:
      - overall inductance between 5,.. and 5,.. mH
      - center tap
      - should stand DC current up to 10mA
      - should stand AC voltage up to 100Vss

      Thank you for your insights
      Bilder

      • 20241222_220223 small.jpg

        167 kB, 1.000×1.776, 14 mal angesehen

      Indeed, the measured inuctances 5.5 mH vs. 1.4 mH fit exactly to half of the number of turns on an identical core. Inductance is proportional to the square of n.

      Best regards
      Christian
      **************************************************
      2 + 2 = 5 (für extrem große Werte von 2)

      Dieser Beitrag wurde bereits 1 mal editiert, zuletzt von „chriss_69“ ()

      @dl2jas - additional measured values: adjusting tuning, the inductance varies from 4.8 mH to 5.55 mH

      @oldiefan - good point! I'll mention it to the coil maker

      @kugel-balu - yes, I had tried with two new ECF80 ;)

      @chriss_69 - you're right! Although I suppose the theoretical value would be 1.3 mH, but the measurement is not accurate enough.

      I have just sent a request to Reinhöfer. Do you know of other coil makers I could inquire from?

      Thank you, best regards,

      Chris

      Dieser Beitrag wurde bereits 1 mal editiert, zuletzt von „Chris Don“ ()

      Actually, in doubt I had replaced C16 and R16...

      One thing I didn't mention, though, is that at the beginning I was getting no 38 kHz at all, due to a faulty C15. It is only after replacing C15 that I got the 38 kHz back... but slowly decreasing when L5 was heating.
      I wonder if the fact that the coil had been feeding the ECF80 plate (for years, perhaps) without oscillation isn't the cause of its damage.

      Chris