Fisher 505-T MPX Stereo Seperation Issues

[jd2005]

This is the problem Description

The FISHER 505T receiver was thoroughly renovated. The FM tuner and Multiflex tuner were tuned to what looked like proper tuning ...

The diagnosis was that the FM separation between the two channels was very low. Compared to other receivers, it turned out that this was indeed the case. There is no real seperation between L and R (theer is some slight).

This is circuit of the FM-IF and MPX stages. (where do i see the American vs European FM (de-emphasis) selection ...)

Hfe_fisher_450-t_service_en-012 (1)

• jd2005
• Jul 22, 2017

There is a kind of a filter at the MPX output. I guess it is for attenuation as well as frequency compensation.

Hfe_fisher_450-t_service_en-012 (2)

• jd2005
• Jul 22, 2017

I have bypassed the filters and conncted directly to the L and R outputs. well ... the seperation is much more obvious. however bypassing the filter casues high frequency and also hi level distortion.

I do not have an MPX FM generator, so the tuning was done via a broadcasting station to the FM-IF level and a scope was used to tune the multiplex decoder frequencies. everything seems to work fine ... there is no malfunction and is this the way it functions ? strange isn't it ...

The possible solutions I thought of:
1. Inserting a small circle of PLL Decoder and bypassing the current decoder (not nice...., it's not a restoration but it will work)
2. Purchase / loan FM STEREO Signal generator to make precise tuning. A story of 300-400 dollars and more.
3. Like 2. But troubleshooting by a cheap FM STEREO FM transmitter (purchased). and a two-channel audio signal generator (purchased) - to inject differenet L and R channles in order to be able to distinguish between them.

I would be happy to receive further suggestions for identifying and solving the problem.

Thanks

 

[dcgillespie]

If you've got a good 38 kHz presence at the collector of Q403, and the balanced detectors Q405 and Q406 are good, then all you're really left which is alignment issues. Have you connected your scope to the 19 kHz and 38 kHz sources within the decoder to confirm what kind of Lissajous pattern you're getting? If you can post those, then it will be easier to see if alignment is really the issue.

Dave

 

[jd2005]

Oh yes. 19Khz Checked. 38Khz Checked. eventually i have borrowed a FM signal generator from a friend and was able to further adjust the MPX.

IMG-20170728-WA0001

• jd2005
• Aug 4, 2017

well... the improvement was very slim but still noticeable. Should live with that.

IMG-20170728-WA0003

• jd2005
• Aug 4, 2017

That is how they have designed the circuit in those old days and that's the best that it can deliver (L-R separation i mean).

IMG-20170728-WA0013

• jd2005
• Aug 4, 2017

 

[dcgillespie]

JD -- Not at all true. I just finished up a thread in this forum where by I convert one of Fisher's Crosby based vacuum tube stereo tuner/preamp units (one of their "oops" units) to operate correctly with one of their "WX" MPX sub-chassis decoders installed. This is one of their classic 3 tube GE/Zenith based MPX decoders that Fisher used in all of their "Wide-Band" tuners throughout the 60s. A very similar version of it -- the MPX-65 -- was used in all of their MPX receivers of that day as well. Like the SS circuit you are dealing with, these are a classic time-division designs that were in large part responsible for the enviable position that Fisher achieved and held onto when it came to FM Stereo MPX performance. For the record, the stereo separation performance produced from that project can be seen in this L only scope shot here:


A R only scope shot looks identical to this one. These are the result of over-the-air (broadcast) MPX test signals transmitted to the 202-T's tuner section. The measured separation from both L to R and R to L is 38 db at 1 kHz.

The scope shot you provide shows virtually no separation at all, so something is very wrong. A couple of questions:

1. How are you injecting the MPX signal? Directly into the MPX section, or is it being received over-the-air through the tuner section from your generator? It is always best to do separation testing using over-the-air signals, as then the characteristics of the final FM detector are taken into account. If you are directly driving the MPX section, make sure you are following any Fisher instructions when doing so as they commonly required an R/C network to be placed in the signal lead when driving the MPX section directly to account for the characteristics of their Wide-Band ratio detector stage in the tuner section.

2. In using over-the-air signals, the proper alignment of the tuner section -- and particularly that of the ratio detector transformer -- must be verified, as a poorly aligned tuner section will not let the MPX section perform properly.

3. If your pilot and sub-carrier signals are good, and both of the L-R demodulators are good, then you are invariably coming down to a phase problem -- either the 38 kHz oscillator is not being properly triggered by the pilot signal, or (more likely) the phase of the injected composite signal into the 38 kHz demodulators has been disrupted.

4. Regarding the phase of the composite signal, it is injected into the L-R demodulators (via the Sec CT of the 38 kHz osc transformer) by the composite coupling cap shown at the bottom center of the full schematic. I can't make out what the value of the cap is, but it is invariably an electrolytic cap, and when it develops a problem, separation problems are the inevitable result. You cannot rely on using a simple cap tester to verify the integrity of this cap. If it is the original cap, then I strongly suggest testing it by lifting one lead of the old one and substituting in a known good one to check the performance of the original cap. The vacuum tube decoders use a 1 uF electrolytic cap in a similar position of their design, and it was a known failure point over time in those units. When it goes, the L+R audio would commonly make it through the cap just fine, but the L-R sidebands would not. Because the audio would get through, the cap's failure would not make itself immediately apparent. But of course, the poor coupling of the sideband information produced all manner of separation problems due to the cap's deterioration. Check the integrity of this cap carefully to determine if it is either the problem, or can be eliminated as a potential one. If this cap has failed and you have tried to compensate by aligning the MPX section while it was still in the circuit, then replacing it will likely necessitate your having to realign the section again for it to operate properly.

I hope this helps!

Dave

 

[jd2005]

Hello Dave,

Thank you for your post.

I wish i could reach the same separation that you have received... 38db... wow !
The service manual in my case says that separation should be around 30db...

And yes. the tests were done by injecting "over the air" thru the antenna input.

I have "recaped" all of the electrolytic caps on the tuner module.

Do you refer to the coupling capacitor C412 or to C413 ?

Here's a better image link of the tuner schematic.

https://drive.google.com/file/d/0Bz32O2xTTVQtbjMzbmRPcHBVdVk/view?usp=sharing_eil&ts=59846f8c

Anyway i get what you mean. I will have to again track the signal this time through all of the test points that are shown in the schematic.

I will get onto this again when i have time and the signal generator once again.

Thanks!

JD

.

 

[dcgillespie]

I'm sure you'll get to the bottom of it. Since you are sending your test signals through the tuner (which is best), do make sure that the tuner portion is operating properly as a poorly aligned tuner section will not let the MPX section perform properly as mentioned. Also, when you say that the tuner module has been recapped, I don't know if that includes the MPX section or not, as I've never had a 505-T on my bench (just prefer the vacuum tube gear), but if it doesn't, then be sure to check the condition of the composite coupling cap as mentioned as well.

A separation specification of 30 db would be typical of the design used in this unit, as it is for virtually all of the dual demodulator diode type designs. I would consider the 30 db specification to be the absolute optimum separation such designs are capable of, with most working examples only "approaching" 30 db of separation in my experience. The WX and MPX-65 vacuum tube designs I mentioned use dual fully balanced ring diode demodulators for detection of the L-R and -L+R sub-channel signals, which inherently produces superior separation performance. The separation performance of these decoders is specified by Fisher as being 35 db, but when the stars align, I've actually seen even closer to 40 db of separation -- but that takes perfectly matched diodes, balancing resistors, etc. Separation levels of 33-35 db are not at all uncommon for those decoders with nothing more than routine maintenance (recapping) work.

Good luck with it!

Dave

 

[larryderouin]

Dave; Fisher used an "ALL IN ONE BOARD" for AM-FM IF, Muting and FM MPX. It's an extremely packed board. And it's a ROYAL, GOLD PLATED, PAIN IN THE A** to work on. They used this board and variations of it on most of the Solid State stuff from 1968 ON. The only one's that don't have this board that I have are the 500-T,550-T, 700-T, and 160-T. I generally leave them alone unless something is FUBAR in the tuner.

JD; You have my sympathies!

Larry