Multiplex adaptor EU de-emphasis

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
An external network with a zero at -13.333... krad/s and a pole at -20 krad/s is what you need. You'll need two resistors and a capacitor per channel to make that:

resistor R1 from stereo decoder output to load
capacitor C1 shunting R1
resistor R2 from input of load to ground

If the output impedance of the stereo decoder were zero and the impedance of the load were infinite, the correct values would be R2 = 2*R1 and C1 = 75 us/R1.

The output impedance of your MPX100 stereo decoder is not zero, but 3 kohm according to http://fisherconsoles.com/non console manuals/fisher mpx100 Operating Instructions.pdf Do you know the input impedance of the circuit you are driving with it? If so, I can calculate the required resistance and capacitance values.
 

PRR

Member
Joined 2003
Paid Member
Don't add an external network. As long as it is IN Europe, modify the internal network. (If you carry it back and forth, add a switch.)

The 75uS is the stuff circled on the schematic. Analysis shows the equivalent RC to be 76uS which is more than close enough.

Leave the resistors alone, they do other stuff. The deemphasis is the 0.001u+0.003u circled.

At a glance, the obvious intent of two caps is that you snip-out the 0.001u when you get to Europe. This actually comes to 55uS, which is far closer than it needs to be. If you are very fussy, snip one leg of both, get a 0.0027u (2.7nFd) cap that reads a bit low and put it in.
 

Attachments

  • FisherMPX100-circuit.gif
    FisherMPX100-circuit.gif
    39.3 KB · Views: 103
  • FisherMPX100-analysis.gif
    FisherMPX100-analysis.gif
    7 KB · Views: 100
Don't add an external network. As long as it is IN Europe, modify the internal network. (If you carry it back and forth, add a switch.)

The 75uS is the stuff circled on the schematic. Analysis shows the equivalent RC to be 76uS which is more than close enough.

Leave the resistors alone, they do other stuff. The deemphasis is the 0.001u+0.003u circled.

At a glance, the obvious intent of two caps is that you snip-out the 0.001u when you get to Europe. This actually comes to 55uS, which is far closer than it needs to be. If you are very fussy, snip one leg of both, get a 0.0027u (2.7nFd) cap that reads a bit low and put it in.
Thanks, but please have a look at the actual schematic (I suppose what you have is an older one) I was thinking that element marked PC1 & PC2 was the network ???
 

Attachments

  • hfe_fisher_mpx-100_sch.pdf
    733.6 KB · Views: 37

PRR

Member
Joined 2003
Paid Member
PC_ is 19KHz filter. (Check: 82K and 100pFd makes 20KHz, not 2.12KHz.)

R36 C25 alone makes 56uS which is in the right area of the spectrum. We must add tube plate resistance, shunt R28 and some of R42... if that came out to be 62.5K it would be 75uS, so this is the US/Can deemphasis.

For EUR: Change C25 C27 to 2/3 the value, 0.000,8uFd (800pFd; 750pFd is very-very close).
 
PC_ is 19KHz filter. (Check: 82K and 100pFd makes 20KHz, not 2.12KHz.)

R36 C25 alone makes 56uS which is in the right area of the spectrum. We must add tube plate resistance, shunt R28 and some of R42... if that came out to be 62.5K it would be 75uS, so this is the US/Can deemphasis.

For EUR: Change C25 C27 to 2/3 the value, 0.000,8uFd (800pFd; 750pFd is very-very close).
But what about the influence of the R38 (250K) ?
 

PRR

Member
Joined 2003
Paid Member
But what about the influence of the R38 (250K)?

Apologies for not showing my matchbook.

I assumed that the fine Fisher engineers got it "right". Fisher still had smart people in the 1960s. Multiplex decoders were a hot topic at the time and there was a lot of industry discussion. Note the significant implementation change between the two models cited in this thread. Also Fisher still had very smart dealers and customers who would point-out any mistake.

But the US market box was cut for a 75uS deemphasis, and you only have 50uS signals in your area.

But what about the influence of the R38 (250K)?

AND the 330K beyond it. ALL the resistors here, and the tube, enter into it. I stayed up late doing a rough check.

But the core concept can be: the resistors set the bass-mid gain, and an impedance. For simplicity and fidelity, a single cap sets the treble deemphasis. Whatever that cap is for 75uS, make it 2/3rd the value and it becomes 50uS. Hence 1200p becomes 800p.

Today's full analysis is long and has surprises. I had assumed 12AT7 for low plate resistance. But there's 7,670 Ohms unbypassed in the cathode leg. With rp this makes ~~400K at the plate. Shunted by R27 47K so an exact estimate of "~~400K" is not needed (and tube aging hardly matters). This 42K has another 47K in series. That node is shunted by 250K Level Control, and (at full-up) by about 350K from R42 and 12AX7 virtual earth point.

I worked all that out and it was a bit shy of 75uS. Still well within +/-2dB, but I'd expect Fisher to be real close. The total R should be larger. Maybe the Level pot is never normally full-up, is that reasonable? Well there's a lot of gain on tap for a box which should be unity gain. V4 works at gain of 4. V5 has some gain. Buffer V1 has some gain. The diode bridge will have significant loss, but it does look like Level pot must normally be not-full-up.

With these assumed values I get very near 75uS. Changing 120pFd to 800pFd gets very near 50uS.
 

Attachments

  • FisherMPX100b.gif
    FisherMPX100b.gif
    21.1 KB · Views: 85
Thank you PRR.
It was logic, nice and clear. Final and simplest movement shal be changing 1200pF with 800pF, perhaps one better silver-mica one . And all other problematic components like elcaps and mylar if necesary.
Then I shall pass to some decent restoration of The Leak Throughline 3 which I intend to mate with the fisher.
Victor
 
I assumed that the fine Fisher engineers got it "right". Fisher still had smart people in the 1960s. Multiplex decoders were a hot topic at the time and there was a lot of industry discussion.

You can say that again! I have 5 of the 6 volumes of "Audio Anthology" -- which are reproductions of articles which appeared in Audio Magazine from the 1950's to early 1960's -- in Volume Six there are seven articles on the new stereo FM technology and how to get your arms around it.

Poor Avery Fisher, they took his name off Philharmonic Hall.
 
Hi, I have just started the restauration on the decoder. What do you sugest to start with as it looks untuched inside but too much time passed over.
Shall I cange part of old components? The el-caps I allready ordered some new multisection from F&T....but what about the mylar ones in the signal path?
 

Attachments

  • mpx 100 schematic.jpg
    mpx 100 schematic.jpg
    208.7 KB · Views: 60
  • parts list.jpg
    parts list.jpg
    155.3 KB · Views: 52
  • mpx 100 3.jpg
    mpx 100 3.jpg
    194.7 KB · Views: 41
  • mpx 100 1.jpg
    mpx 100 1.jpg
    159.8 KB · Views: 45
  • Screenshot-2018-3-14 Capacitor - CE Mfg , 350V, 40 40 40 40uF.png
    Screenshot-2018-3-14 Capacitor - CE Mfg , 350V, 40 40 40 40uF.png
    291 KB · Views: 37
Administrator
Joined 2004
Paid Member
I would start with the electrolytics and from there would only replace parts as required to get the unit working properly.

I would go at a very careful pace and do only one thing at time, otherwise it's quite easy to end up with a non functional decoder if you don't.

If you need to replace demodulator diodes, note their orientation very carefully before replacement - they're different for right and left channels.
 
I would start with the electrolytics and from there would only replace parts as required to get the unit working properly.

I would go at a very careful pace and do only one thing at time, otherwise it's quite easy to end up with a non functional decoder if you don't.

If you need to replace demodulator diodes, note their orientation very carefully before replacement - they're different for right and left channels.
You are right, once at a time.
So I shall start with the power supply. In that case, more than electrolitycs shall I replace the old selenium bridge? I heard this is changing parameters in time.
 
Administrator
Joined 2004
Paid Member
Definitely put replacing the selenium rectifier at the top of the list, when they fail the magic smoke that comes out is a little bit toxic. You may need to slightly increase the value of R50 if the voltages are high. (Since a VTVM of unspecified accuracy was used if you are within a few % at the nominal line voltage you're good to go.)
 
Definitely put replacing the selenium rectifier at the top of the list, when they fail the magic smoke that comes out is a little bit toxic. You may need to slightly increase the value of R50 if the voltages are high. (Since a VTVM of unspecified accuracy was used if you are within a few % at the nominal line voltage you're good to go.)

Right, what bridge you suggest?
 
Administrator
Joined 2004
Paid Member
I have no specific recommendation, anything with a 1A 1KPIV rating is sufficient, and you could build one from 4 x UF4007 as well.

Hopefully no caps are required across the diodes (just 0.01uF 1kV ceramics across the diodes). The selenium rectifiers had pretty soft turn off characteristics and may be less noisy than modern silicon diodes - R50 helps a lot in this regard.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.