fisher MPX decoder troubleshoot

vintelectra

Active Member
Hi Fans: Restoring this fisher 500-c MPX decoder if find that there is no separation.
New tubes, recapped mylars, changed diode bridges and nothing.
Need help ! Tanks in advance, martin
 
Vint -- If you're getting sound from both channels in the stereo position, but no stereo of any kind, the problem is somewhere in the 19kHz pilot/38 kHz sub-carrier circuits. While the 400 and 500C are very similar designs -- and use the same decoder -- I'm not directly familiar with the 500C as I am the 400. Does the 500C include a stereo indicator of some type when receiving a stereo station? (I'm sure it does) If so, is it working properly? If it is, it operates off of the 19 kHz pilot signal, meaning that portion of the circuit can be eliminated, and the problem must exist in the 38 kHz doubler section. If the stereo indicator is not working correctly, then the 19 kHz pilot circuits and/or the stereo indicator issues must be resolved before going any further.

If you changed the diodes in the 38 kHz balanced modulator circuits, that is likely going to be the first place to look. Obviously, polarity is important with all 8 diodes, as is their characteristics matching that of the original diodes. Did the decoder work before you installed the new components? Adding that detail to the list of clues can be a significant help as well.

Let us know!

Dave
 
fisher MPX troubleshoot

Hi Dave: Thank you so much for your interest and insight.
The stereo indicator works well( lights up when tuning.)
The unit came half restored, with different diodes: 2 ceramic red blocks
with 4 leads ea. The Decoder did not have any separation when I got it, and
therefore I suspected this to be the problem, so I replaced them with the set
of diodes I pulled out of a 400 (for parts). I followed the connections of the removed diodes ( problem here?).
Note worthy: the four 1Meg resistors in the diode bridges seen in the
circuit print are not there . But they are not present in a similar working
unit which separates optimally.
I eagerly welcome your suggestions
 
The problem with working on multiplex adapters, is that they can be "working" perfectly, as in there are no defective components and all voltages are correct, but if they become misaligned, they simply produce a mono signal out of both stereo outputs.

FM stereo works by using a L+R main channel, and L-R sub-channel, that is contained within the total modulation envelope. The L-R is then broken into two such signals that are 180 degrees out of phase with each other, producing L-R, and -L+R. With a proper mixing circuit, these three signals (L+R, L-R, and -L+R) can then be combined so that +R and -R cancel each other out leaving 2L in one channel, and +L and -L cancel each other out leaving 2R in the other channel. The relationship these signals have with the 19 kHz pilot "timing" signal then determines which channel is L and which is R.

I mention all of this hardly to impress, but to show how the whole process is ultimately based on cancellation and addition. When these processes don't happen, the whole house of cards falls down, and you get nothing but mono out of each side. This is why I have always said that for proper FM stereo operation, the alignment of the multiplex adapter is rather critical for the "math" to work properly. In particular, if the phase of the various signals is altered by much at all -- as determined (in part) by the adjustment of the various tuned circuits -- then proper cancellation cannot happen, even if the amplitudes of these signals are correct. In normal operation, a properly operating adapter will display at least 30 db of separation, while it takes little mis-adjustment to have this deteriorate down to 10 db or less.

The bottom line is that to check all of these things properly takes a very good scope (as in one with ideally zero phase shift difference between the X and Y channels, and a good multiplex generator, so that the amplitude and phase of the various signals employed can be precisely known. Without proper test equipment, it is almost impossible to adjust an adapter for proper operation by ear. You might take a look at a recent thread I posted about "A different kind of Fisher" where I recount my recent efforts in aligning the adapter in my Fisher 400. Apparently, the pics I provided for that thread were lost in some of AK's new server work, so I will try to repost them to give you an idea of the types of signals being discussed here.

Without proper test equipment, the one thing I would have you carefully recheck, is the connections of the diode modulator circuits. The connections for ground, and application of the composite signal are obvious, but the connections to the 38 Khz doubler transformer is critical. Normally, these are connect one way for one diode set, and the opposite way for the other. Therefore, if these become reversed in one diode set, you will effectively end up with the same single channel being produced at the output of both stereo outputs, which would certainly have no separation.

Ultimately, the diodes -- contrary to what some might like to think -- rarely go bad. It is unbelievable how many people "tinker" with the adapter to try and resolve some issue they perceive, and almost always end up making it much worse. If the adapter was not working before you started your work, and there are no errors in your work or defective components installed, it is almost a certainty that someone tinkered with the settings, which if this is in fact the case, then no amount of component replacement will ever bring the stereo back again. The unit will need to be evaluated with proper equipment to ultimately determine its status, and readjust it for correct operation again. The good news is that once set, the units rarely give any concern, except when someone -- with the best of intentions -- tries to "upgrade" all the components installed. The ONLY components that should be looked at in this light are the input coupling cap, and the two output coupling caps. The composite signal coupling cap (a 1 uF electrolytic) should also be replaced as a matter of good maintenance. However, all other components should be considered as off limits, and are not handling audio signals anyway.

This may not be what you wanted to hear, but it is honest information. Double check your work closely, but be prepared to send the unit off if no mistakes are found. The good news is, that a capable shop SHOULD be able to align the adapter by itself, making a little more work for you in its removal and installation, but making shipment much easier. Of course, it's always best to align the entire receiver in one complete setting, as a poorly aligned tuner section will also produce bad FM stereo, and the very best results will always be obtained when the tuner and multiplex units can be tested together. However, very good results can be obtained from aligning the adapter by itself as well.

I hope this helps!

Dave
 
Before you send it to a shop: has the little blue pot been moved in any way? This is the separation pot and sometimes, if set completely CW or CCW, there could be no separation at all!

If it is already near center (or approximately), it is probably not your problem.

Patrice
 
fisher mpx decoder trouble shoot

Thank you Dave for this wealth of info. I did align the whole unit, and aligned
the mpx with a fisher 300 mpx gen and dual channel tetronix scope, even comparative signal tracing side by side with a good 500-c. The signals differs
at the output of the diode bridges and the separation pot show no effect. If the diodes are good then whats left is an erroneous connection. My alternative is to rebuild the bridges on a small board as opposed to clumped together.
The remaining question is the absence of the 1 meg resistors. Should I
replace them on the new board? Thanks. Martin
 
Hi Martin -- The 1 meg resistors are equalizing resistors, that were (apparently) omitted in later units -- possibly either because they were generally found unnecessary, or more likely, better diode matching capabilities made them unnecessary. If you are using diodes from a donor unit, then I would include the resistors if the donor unit used them. If they were not present, then omit them.

The 300 is what I use as well, and with it, you should be able to troubleshoot the adapter rather quickly. When the 19 kHz pilot signal of the 300 is connected to the X input of your scope, and the Y input is connected to the output of C216 on the adapter, are you getting a stable 2:1 Lissajous pattern? If so , then all is good up to the output of the 38 kHz doubler transformer. If not, there are problems before the diode modulators. The Lissajous patten obtained should be clearly defined, and quite stable.

You mention that you were tracing the signal through the problematic unit with a good unit along side. What sort of signal were you using and observing with that process?

Dave

Dave
 
fisher mpx decoder trouble shoot

Thanks again Dave. I got a very stable 2:1 lisajous picture.
For comparative signal tracing I connected the RF 100 FM MPX output
of the fisher 300 to both radios(loose connection) and followed the signal
simultaneously in each channel of the scope. The signal changes at the output
of the diode complex, and the final signals at the mpx outputs does not show the typical mpx signal appearance the good one shows.
Now I understand the function of the resistors, thank you.
So it seems to boil-down to erroneous connection of the diodes!
I got new zener diodes NTE5021A( which seems to be compatible
with the old ones. Do you think it would be a good idea to make a small board
with them and the carefully follow the circuit diagram for connections?

martin
 
Martin -- the diodes employed in the diode modulators are not Zener diodes, but typically germanium high speed switching diodes, as used in detector applications. A Zener diode would be wholly inappropriate for this application, since a Zener diode ultimately conducts in both direction with regards to voltage polarity, whereas proper operation of the diode modulators depends on the diodes completely cutting off during one cycle, and having a very low forward drop in the other, which is characteristic of a germanium diode device. Something along the lines of a 1N294 would be much more appropriate for the application.

If I may, where did you get a reference that the modulator diodes are Zener devices?

Dave
 
fisher mpx decoder trouble shoot

I got the info from a forum, dont remember which , but I am not surprised is faulty. I will get some 1n294 and try to match them . Or should I stick with the diodes from the 400.? whats your opinion?
 
If the diodes pass a basic ohmmeter check (conducts one way, doesn't the other), I think I would stick with the 400's diode arrays. Otherwise, the germanium devices will be your best bet. Being that they won't be particularly matched however, you will want to include the 1 Meg equalizing resistors with any aftermarket diodes you install.

Sorry you're having all these problems, but I have no doubt you'll be able to resolve the issue.

Dave
 
fisher mpx decoder trouble shoot

I just went over the connections and apparently are good. I"ll keep searching.
Thank you again for the tremendous help and knowledge. Hope we meet again

Martin
 
Martin -- As a reference point for you, with a signal tuned in from your 300, and it supplying a 1.0 volt 1 kHz signal to the left or right channel, you should be able to observe about 15 volts P-P of 38 kHz voltage being supplied to each set of 33K isolation resistors in the balanced modulators, relative to ground.

The composite voltage will represent about 12 volts P-P

Also, your scope should show a classic 1 kHz sign wave -- filled in with 38 kHz energy -- directly at the output of the diode modulator for the channel driven. There will be about 7 volts P-P at this point under these conditions.

I just finished an alignment on a KM-60, and thought I'd take some readings to pass on to you.

If the 38 kHz energy is there, the diode modulators are good and installed properly, and the composite signal is being applied to the modulators, then all the necessary information is there for the matrix stage to properly balance out the L and R signals.

I hope this helps.

Dave
 
Thanks Dave for info.I think I checked and got a good sine 38K, but I will do
it again this time with volt measurements. At this time I am working on its front end due to some problems there, as soon finished I will give you a follow-up
on your suggestion. Thanks again Martin
 
fisher mpx decoder trouble shoot

Hi Dave: Almost there! While I had aligned peaked the rf and If strip, I did not check for signal quality, this time I traced the abnormal signal to the front end and found a wrong value cap. This corrected the signal quality and voltage.
I measured , as you suggested the clean sine 38k at the resistors and I got a good 5 volts. Then moved the scope to the mpx outputs and I get some separation finally, but only 50%. The pot did not affect either channel. Maybe
I should realign the mpx? This is an example of how complex is to work in
previously tinkered unit. Martin
 
Amen to that. My 400 had been previously tinkered with as well.

When you measured 5 volts at the 33K isolation resistors, remember that my voltage was as taken off of a calibrated scope screen, and posted to you in a peak to peak value per side. If you were reading from an RMS responding meter, then that does in fact correspond to nearly 15 volts peak to peak.

You must get the tuner section proper correct first. A poor tuner section will deliver poor stereo from even the most perfectly aligned multiplex section. In particular, the bandwidth of the IF strip, and adjustment of the discriminator transformer can have a marked effect on stereo performance.

What cap was found to have the wrong value?

Dave
 
The 5 V was measured with calibrated scope ,peak to peak.
The cap was C 26 in the RF section. Was 24, call for 1000p, I placed an 850 pf

I redid the alignment of IF strip and now I am aligning the rf section, and find that the audio output signal ( from the modulated Fm), has two consecutive
peaks , and the maximum tuning signal does not correlate with maxim mun
audio sine taken from Rec.Out jack. You can reproduce this double signals
by slowly turning the dial. As you suggest, I want to adjust the RF to
optimmun and only then check the separation. Any suggestions on this point?

Martin
 
Martin -- In the 500C (like the 400), the tuning indicator is driven off the signal from the limiter transformer. Therefore, when the tuning indicator does not accurately indicate (i.e. peak at) the center of the FM channel, it is typically due to misalignment of the limiter transformer. However, I see that the 500C uses a single slug transformer in that location, versus the double slug version used in the 400, so in your case, there may still be uneven peaking frequencies earlier in the strip.

In all the years I have been aligning FM tuners, I use a crystal controlled FM marker signal only to yank an IF strip back into the ballpark when it has been so horribly tinkered with or worked on, that you have no idea what the alignment status of the strip is, other than it's very bad.

However, once that is accomplished, I ultimately use a weak signal station to touch up the strip (one that won't let the limiter limit), and it's during this portion of alignment that such issues as you are experiencing usually work themselves out. I always align the discriminator with a DC voltmeter however.

It is usually quite easy to get the first, second, and third IF transformers to align to close agreement, but by the time you reach the limiter and discriminator stages, it gets pretty hard using the prescribed Fisher method. That's why once in the ball park, I rely most on the unit's own tuning meter to achieve optimum performance up to and including the limiter, and then the voltmeter for the discriminator.

Finally, also be on the lookout for oscillation within the tuner section -- usually indicated by the unit appearing to be picking up a signal at some point on the dial, but with no antenna or artificial signal of any kind connected to the unit. The Fisher receivers pack a lot of gain in a relatively small area, so it can happen. If it does, it will through off your best alignment attempts. That's why I always use a weak station for the final touch up, as the best defense against such oscillations is in fact a well aligned IF strip.

One last thing -- are you sure your 300 is well calibrated and operating optimally? The manual has the procedures to check it out, but be aware that I have found an error in the calibration procedures (a fairly significant one at that), that apparently was never corrected by Fisher -- at least to my knowledge anyway. Likely, I would bet that's because very few of these units were ever re-calibrated after they left the factory, so in effect (and in today's vernacular), who knew?

I hope this helps!

Dave
 
fisher mpx decoder trouble shoot

Hi Dave : Such wealth of Key info. Perhaps you should write a booklet for the
restorer. I will follow your suggestions and send you a follow-up.
My 300 was all recapped and calibrated according with the manual
except for one step(I dont remember which) but the separation
shown in the good f 500-c I use for comparison , falls within spec.
I am not suprise the limiter is causing the problems, since I had to
rebuild it by hand, but I will follow your steps and let you know.

Thanks again Martin
 
Stick with it Martin -- I'm sure you'll get to the bottom of it all.

For the record, if your 300 was "blindly" re-calibrated using the 300 manual I have (with Appendix and Addendum), it will cause symptoms such as the Stereo Beam indicator on the 400 series of receivers (for example) to slam closed (overlap, or certainly close more than a normal stereo station would produce), and produce erroneous voltage readings during the multiplex alignment process as well. If you have either of these symptoms let me know, and I'll show you where and how to correct that in the manual.

Dave
 
Back
Top Bottom