saba: 14; Multiplex-Decoder

Hello,

I recently purchased a SABA continental 3050US with Stereo Decoder (Breisgau Radio Chassis with Dual 1008 turntable and remote control).  The unit was not working.  Checked and replaced several valves, replaced the capacitors specific to North America (60Hz) and cleaned the autotuning mechanism that was jammed.  The unit is working very well with the exception of the Stereo Decoder (I think).  The sound is phenomenal.

The Stereo works fine for about 30 minutes, before the stereo indicator (EM87) slowly opens completely and then there is no sound.  The quality of the stereo signal degrades slowly after about 20 minutes before being completely unacceptable by the time it stops working altogether (after 30 minutes).  If I shut the system down and let it "cool" and start it up, the stereo is fine again.  I have tested this several times and it is very consistent. 

If I switch to MONO, it works perfectly, even after the stereo stops working.

I have checked pin4 on the decoder and get about 28 VDC (a little less then the 30VDC specified).  I have researched a little and found that the transistors in these early units degrade over time and develope whiskers (AF117 & AF118).  I have ordered some AF117 and AF118 transistors (and AF127 which is a substitute for the AF117), but suspect they might have the same problem.  I have removed the shield from T3 to see if this would improve things.  Same result.

Does this sound like a decoder problem?  Are the transistors the most likely problem?  Is there a better way to fix the problem then just replacing with NOS transistors?  Are there modern transistors that can be substituted with some circuit modifications?

Thank you for your help.

Peter

To thank the Author because you find the post helpful or well done.

Regards

I have a similar problem, wasn’t Saba and wasn’t decoder part.

But the symptoms were similar.

Trouble with heating, in cold condition working well, after some time of heating period stops work.

Probably some of the discrete element goes leaky or shorted after heating period (capacitor, electrolyte, transistor etc.)

I remember that was a nightmare to fix this trouble.

I’ve use a cold spray, a hair dryer applied to some parts of the unit, trying to heat them before failure occurs.

But you can at first try to check all critical voltages on unit, compare them with the diagram values.

Then compare all this values after failure occurs.

If this happen in audio part of the unit, in the part where there is no DC voltage that could be measured, it can be a nightmare to solve.

Good luck.

To thank the Author because you find the post helpful or well done.

 

Check C14 o R17. If C14 is leaky and go short, then T2 will be cut-off, and then.. no sound at all.

If R17 is very out of value, then the same symptom

It's a very good idea check with an oscilloscope the 38KHZ signal (stereo pilot tone) 18VSS. 

Good Luck!

To thank the Author because you find the post helpful or well done.

I finally solved the issue.  It took quite a bit of time (several months, working off and on) and I learned several things.

The intermittent problem finally went to complete failure of the stereo signal.  Originally after ~30 minutes, the stereo signal would fade out and go to Mono.  After several weeks of use, the stereo signal was gone completely even on start up.  So I hoped that this would make solving the issue easier...It did not...

1.  I checked all of the transistors.  I built a simple circuit that tested the existing transistors as a switch (using an led to indicate it was on).  All of the original ones passed.  I tested for any shorting between the pins (tin whiskers).  No evidence of any shorting at all.  There is much information on the internet about tin whiskers developing in AF117 and AF118 transistors.  Can I trust my transistor check circuit, perhaps at different frequencies, this test is invalid?

2.  I started checking all of the resistors on the circuit to see if any were out of tolerance.  All were fine.

3.  I checked all of the capacitors on the circuit, using a newer multi-meter as well as a vintage Solar capacitor analyzer CBB (test under voltage for leakage etc).  All caps passed.  Can I trust my vintage capacitor checker?

4.  I checked all of the diodes with my newer multi-meter in diode mode all passed, with forward voltage drops of ~0.35 to 0.385 volts.  there is no information that I can find that shows these AA119 diodes to be prone to failure, only a single reference to the forward voltage drop increasing over time, but no values.

5.  I took a chance and ordered some 2N2207 (AF118) as well as some OC45 (AF117 with no shield lead) transistors to replace the original transistors.  Expensive.

6.  Replaced the transistors.  Some function returned to the stereo decoder.  The signal level fluctuated wildly on the EM87 at about 50%.  This seems like I am moving in the right direction.  Perhaps the equivalents aren't really as good as the original types?

7.  Thinking that I had some success, I ordered some true AF117 transistors as well as some true AF118 transistors (NOS and VERY expensive).  I replaced the transistors with the NOS transistors.  No stereo signal at all now.  This was very discouraging.

8.  Thinking that there might be something wrong with the NOS transistors, I returned the 2N2207 transistors and the OC45's (hoping to get back to a situation where something was working).  No sucess.  I started to think that perhaps I had destroyed one of the transistors with too much heat as I was soldering and un-soldering (de-solder station).  So I removed the OC45 transistors and returned the NOS AF117's.  Success.  Stereo is solid no fluctuation.  I am happy.  I used the auto searching to scan down the dial to the next station....Suddenly, no more stereo signal and I can not get it back now.  Now I am unhappy.

9.  I go back to basics.  I was never happy that I didn't find a single component that looked like it was bad.  I only had suspicions about the transistors having tin whiskers, and only suspected that perhaps my Solar capacitor checker was not great.  I had tested everything so I decided to check everything again.  As luck would have it the first thing that I test is the diodes.  I accidentally set my multi-meter to resistance instead of diode mode.  As I was checking diode GR11, the forward resistance measured 50K.  Expecting to see a forward voltage drop, the number jumped out at me.  I decided to check the forward resistance of all the other AA119 diodes.  They were all about 7-10K.  I put my meter back into diode mode and check the forward voltage drop it reads fine (I think...at 0.35V, similar to all the others).  As I return to resistance mode and check GR11 again, I start to see forward resistances of either 10K or 50K, depending upon how I place the probes on the leads.  If I "pull" gently on the leads, the resistance is 50K, if I push gently it is 10K!!!!  Here is my intermittent problem....MAYBE.

10.  I don't have any germanium replacement diodes.  I check my stock and find some schottky diodes with similar forward voltages at small currents.  I replace GR11 with the schottky and check the decoder in the system.  It works! Solid stereo signal, no fluctuations, change station everything is fine.  Now I am happy, we have found the intermittent component.  So I decide to replace all of the AA119 diodes with the same schottky's.  Signal strength improves and stays solid.  Now I am very happy.  Replace the cover and listen to the wonderful FM stereo sound of the SABA with an almost fully closed EM87!

Note:  In the end I left the 2N2207 (AF118) replacements in the system.  I left the NOS AF117's in as well.  I didn't have the heart to go back to the original components again and risk damaging something.  Leave well enough alone.  All of the polarized caps were replaced with new caps, as well as most of the others.  No re-alignment was necessary.  I suspect that all of the old components that I removed were fine. with the exception of the AA119 in GR11

Lessons learned:

2N2207 good replacement for AF118 in this decoder.

OC45 good replacement for AF117 in this decoder.

schottky diode (I need to find the part number and post) good replacement for AA119 in this decoder.

You can pretty much replace all the capacitors (at least I was able to) without re-aligning the decoder (I left the caps on the inductors original).  The SABA 14 decoder appears to be extremely tolerant of changing components.

Solar Capacitor analyzer works excellent.  I can trust it.

Every AF117, AF118, 2N2207 I had shows no signs of tin whiskers (not shorts), Is this phenomenon as common as people think?  It does not appear so, I can find no evidence of tin whisker issues in any of the ones I own or bought.

Always check the forward resistance of an AA119.  It appears as though the junction or attachments of the leads can degrade over time.  Is this some sort of different expansion rates as the diode gets heated, that over time leads to failure of the lead attachment???

 

Thank you for all of your suggestions.  It took quite a bit of time to find the problem (and a little luck).  Perhaps someone will now have an idea of how to double check an AA119 diode.  Or know where to quickly start looking if you have an intermittent or non-existent stereo signal coming out of your decoder.

Peter

To thank the Author because you find the post helpful or well done.

1N60 is a general replacement for Germanium diodes, especially AM detectors. In many cases the 0.6V drop isn't significant and 1N4148 will also work (FM discriminator).

If there is an emitter resistor and it's not an output stage often a Silicon PNP will replace an AF117.

The cheap Russian P416A seems to replace AF117 fine, but on Bush TR130 the BC557 or 2N3906 worked fine too as AF117 replacement without changing bias. The AF127 is a replacement too and allegedly not to prone to Whiskers proble,.

I had 1963 Ever Ready and the AF117 are fine. But later Hacker RP18 and Bush TR130 all the AF117 had failed. (The AF114 in the Hacker fine though).

To thank the Author because you find the post helpful or well done.