Fisher tube chassis balance controls

[audmod01]

I wanted to also thank all the members who have offered encouragement and helped with photographs of critical areas of circuitry along the way. All have been helpful in every way.

I plan to go through the schematic we developed for the improvements and increase the size of the text I used in it. It will be helpful when examining a printed copy of the schematic. It will also help when viewing the schematic on a computer screen. After I finish those changes I will post the Rev. B version on the AudioKarma.org archive area if all goes well.

Have a great day everyone!

Joe

 

[audmod01]

I pulled out the steel cabinet that came with the X-101C yesterday and did some cleaning on it. It was rather dull and dingy. The color looked close to the color of the area of the chassis front that is behind the front door on the front trim panel. I began cleaning with some formula 409 and loads of dark brown began to come off. I believe this unit was in the home of someone who smoked cigarettes at some time in its existence. Just as much of the dark brown coating came off of the inside of the cabinet as on the outside. I finished up by cleaning with an old toothbrush and some waterless hand cleaner. The result was that the real color is relatively close to that of a wooden walnut finish cabinet, but with a wrinkle paint finish that resembles leather grain. It has a few scuffed spots that went down to the steel underneath, but overall does not look too bad.

One aspect of a steel cabinet is that it helped to protect the chassis and front trim panel during shipment. I have seen a few examples of wooden cabinets that arrived in pieces from rough handling in shipment. Good packing is essential to equipment arriving in good condition.

I need to decide if I want to use the steel cabinet and add wood veneer cladding or just look for a new wooden cabinet made to fit this amplifier.

Joe

 

[larryderouin]

Joe; Congrats on finishing the X-101C. I'm sure that with the work you put into it, along with Dave's improvements, it will work for a very long time.

All of the Solid State FISHER Receivers and the TX-100 that I have, that were not a console unit, came with a METAL case, and those that have a wood cabinet also have the metal case. If it was a console piece, there was/is no cabinet, but the heatsink for the output's was redesigned. The I.D. Plate was generally riveted to the upper case in the corner.
The metal case was standard, wood cabinet was optional.


The tube units generally came naked, unless you bought the optional wood cabinet.

Solid State units look better (to me) with a full wood cabinet. Cladding it in a veneer directly on the metal doesn't enhance the look of the unit IMO. Removing the upper 1/2 of the metal cabinet when inserting the unit in a wood cabinet will enhance cooling, but doesn't detract from the look of the unit.


Something to think about

Larry

 

[audmod01]

Larry;

Thanks for the comments and suggestions. I do remember seeing the first Fisher solid state units show up at the place where I worked in Austin, TX. Most all the ones we sold were in wooden cabinets. The same was true of the tube units. I think the store owner had made the decision that it was the only way he would sell them. The same was true for H H Scott and McIntosh units we sold. We had a few customers mount them in custom consoles and remove the wooden cases. I did see a few H-K Award series tuners, receivers and amplifiers go without wooden cabinets. I have an H-K A-300 and F-50XK FM tuner kit that I bought while working there. Both came without any metal or wooden case/cabinet.

The metal cabinet for the X-101C has the rear cut-out area in the top to receive a metal grill for ventilation. I also see that the unit can be inserted from the front or the rear. It came with two wooden rails for either side that screw to the bottom of the cabinet (inside between the chassis bottom and the metal of the cabinet) and have slots cut in them along with matching slots in the bottom of the metal cabinet that long screws pass through to fasten feet to the bottom of the cabinet. Those same screws also hold the chassis in the cabinet.

I agree that the wooden cabinets just look better than metal ones. I found a company that makes wooden cabinets new to fit these and some other Fisher X-1xx series amplifiers plus other models, receivers and tuners. For a new wooden cabinet, they are no more than what some used and scratched OEM Fisher cabinets go for on the internet.

I just finished painting the chassis bottom cover with some Krylon Satin Nickel paint and it is in the barn drying.

Joe

 

[audmod01]

Some More Testing

Today I had some spare time and decided to do some signal injection and signal tracing. I pulled over the Tektronix 2335 scope and my signal generator. I set the signal generator to provide a 10kHz square wave at a level of about 1VPP. I found I lost one channel right at the tuner input. So I spent some time cleaning all the input jacks. That corrected that problem. Then I found I had signal all the way through the Op Amp, but the volume control worked backwards. Some quick rewiring of the volume control corrected that.

I found that I had signal all the way to the driver and phase splitters but no signals to the output grids. I found first one B+ wire missing and then another one to the other tube, so wires from the A B+ source to both tubes plate circuits had to be added. B+ was missing from the AC balance controls because of this. Once those wires were added the driver and splitter circuits worked correctly and signals showed up at the grids for each output tube. This missing set of wires was due to previous modifications in which the owner had wires cut completely away in this case. I just failed to spot this earlier in the restoration.

I feel confident now that when I install the new output tubes I will be able to hear audio! I will make final AC balance adjustments with the output tubes in circuit along with the grid bias adjustments.

Joe

 

[audmod01]

Loud Hum/Buzz both channels working

I finally installed the new set of output tubes and after setting bias to hold output tube conduction to about 25mA I discovered that as long as only one channel has its output tubes installed the sound from the equipped channel sounds good with no noticeable hum or buzz. However as soon as I equip the other channel's output tubes here comes a very loud and strident hum/buzz sound. It makes no difference which channel is equipped or unequipped.

I have gone over B+ connections religiously now and did find one other problem and corrected that. Now the voltages on plates look to be about right. The problem is somehow just related to the driver and phase inverters plus output tubes. If I disconnect the input and tone/volume circuits of the preamp stages by taking out the repurposed reverb jacks the problem is still there.

I am now beginning to wonder about the speaker wiring. Perhaps I did something wrong there. It is time to try something different.

Joe

 

[audmod01]

Loud Hum/Buzz related to Neg. Feedback Loop

I found that if I ground the junction of the 1K ohm and 1.2K ohm resistors in the negative feedback that the problem goes away. So this is somehow related to the negative feedback loop. The impedance in this area is low since the origin points are the 4 ohm and 16 ohm taps of the transformer secondary. There is a 220 ohm resistor to ground and the two 1K and 1.2K ohm resistors in series with the 220 ohm resistor to ground, so I doubt that hum pick-up is the issue.

I need to recheck the values I used. The capacitor that couples back the energy from the 4 ohm tap is supposed to be a 2400pFd (or .0024mFd). Perhaps I am off a factor of 10?

I should add that if I swap the plate leads, then I really have positive feedback and a very strong oscillation, so the issue does not seem to be the phase of the feedback.

More checking ahead.

Joe

 

[dcgillespie]

Joe -- I would strongly suggest that you have the NFB leads reversed between the channels -- that is, feedback from the left channel is inserted into the right channel, and visa versa. That would let the unit run fine with one (or either) channel's tubes installed, but not both.

As for the drive within each channel to the output tubes, the plates of the phase inverters (pin #1) should drive the output tube with the blue OPT plate lead, and the cathode terminal of the inverters (pin #3) should drive the output tube with the blu-wht lead. Note that this orientation is only correct if the OPT secondary connections are as according to the modified schematic, where the black (common) lead of the OPT is now grounded. If that modification has not been made yet (i.e., the brown lead is still grounded), then connecting the drive to the output tubes as described will result in oscillation.

Keep us posted!

Dave

 

[audmod01]

Dave;

Thanks for that input. I had not thought of that possibility. I will take a look soon. The common lead of each secondary is definitely grounded right near where they exit the transformers at the same terminal strip where the cloth covered leads go to the speaker switch up front. The amp is so close now to working correctly, so I still have hope.

My thoughts were turning to using the scope and looking at phase of signals at the grids of the driver, inverters and outputs versus the secondaries of the transformer. Since the scope has two channels I could easily compare phase of signals from input to output of each stage.

Joe

 

[audmod01]

Secondary Leads of Xfmrs Swapped!

Late yesterday I looked at the output circuitry closer and discovered that I had managed to swap the secondary leads at the speaker terminals left for right!. I took time to move them back to where they should have been and re-tried the amp. It still has the loud hum/buzz.

I think I will temporarily disconnect the speaker/earphone wiring and ground the output common speaker terminals to the chassis like the speaker common wires of the output transformers. Then with the amplifier separated from the preamp stages by removing the jumper cables, I can troubleshoot just the driver, phase inverter and output stages. Once I solve the problem I can always go back and work on the speaker selector switch wiring.

What I have verified so far is:
V5 pin 1 output is definitely coupled to V7 pin 6
V5 pin 3 output is definitely coupled to V8 pin 6
V6 pin 1 output is definitely coupled to V9 pin 6
V6 pin 3 output is definitely coupled to V10 pin 6

I swapped the feedback leads left to right late yesterday and there was substantially worse oscillation going on. I moved the leads back to their initial positions. I may need to work on one channel at a time and use load resistors to handle the output instead of listening to my speakers. After I ground the speaker output terminal, I will recheck the feedback wiring once more.

The general approach will be to use my signal generator to inject a test signal, either square wave or sine wave and trace through the driver, phase splitter then output stage. I will do one thing at a time to the unit, such as the grounding of the output speaker terminal and disconnecting the speaker/earphone selector switch and do a quick check to see what if any change has been made in the outputs.

The driver, phase inverter stages are supplied B+ by power supply source A. The output tubes are supplied power by the direct B+ developed at C25. If I were to use the preamp stages there could be a possible internal leakage section to section in C27, so by only using the output stages of the amp I can eliminate that possibility. Section to section shorts or leakage in electrolytic can capacitors is rare but can happen. Isolating the output stages from the preamp stages will help narrow the search.

Joe

 

[dcgillespie]

Joe -- then IF:

1. V7 pin 3 serves a blue output lead, and

2. V8 pin 3 serves a blu-wht output lead, and

3. V9 pin 3 serves a blue output lead, and

4. V-10 pin 3 serves a blu-wht output lead

AND

5. The OPT Black secondary leads of both OPTs are grounded, and

6. The NFB connections are made to the Brn 4 ohm and Yel 16 ohm taps, and

7. The NFB loops are verified to serve their own channel (no intermixing)

....and you still have oscillation issues, then I would be highly suspicious of the OPTs themselves. A quick and easy way to see if the oscillation is originating from the NFB circuits is to temporarily ground the high side of resistors R54 and R58. That will remove all NFB from both channels to aid in your troubleshooting.

Let us know!

Dave

 

[audmod01]

Results of Grounded Feedback Test

Dave;

First I want to say a big thank you for your help!

I did find that I actually had two problems. I did have a loud hum that turned out to be a poor ground on an audio input cable to the tuner input jack on the left channel. This was somewhat masking the problem with feedback issues.

I corrected to poor ground at the input jack of the tuner input. Then I just had the buzz/howl coming from both channels.

The connections of the plates of V7, V8, V9 and V10 were as you described and match what we show in the schematic we developed. The common leads of both output transformers are grounded and I also disconnected the speaker switch wiring and grounded the speaker common terminals direct to chassis and even ran a stranded wire from the two common speaker terminals all the way back to the grounding point of the common wires of the output transformers. That way I eliminated all the speaker/earphone switch wiring as a possible source of trouble.

I rechecked the 4 ohm and 16 ohm leads from each channel back to its driver stage. The left channel feedback wires definitely go to the cathode circuit of V5. The right channel feedback wires definitely go to the cathode circuit of V6, so that part of the wiring is correct. I also verified that I did order the correct values of capacitors (2400pFd and 620pFd) and used 1200 ohm 1/4 watt resistors in addition to the existing 1K ohm resistors that were already there. The 100pFd and 10K ohm parts are at pin 2 of each phase splitter as indicated in the schematic.

With the feedback circuits active, there is a loud howl/oscillation in both channels. If I ground R54 and R58 the howl/oscillation goes away and the audio is quite decent although there is likely residual distortion, but fairly low. It is quite listenable like that.

The unit came with Fisher output transformers, but of course it is possible that they are bad. I have no way to know exactly what abuse they may have been subjected to in the past. From documentation that came with the unit, one transformer was replaced at some point with one supplied by the Fisher Doctor. The part number is slightly different from the other one which is still OEM.

If the transformers are bad, what would the likely failure mode be?

Joe

 

[dcgillespie]

Joe -- The transformers may not be bad per se, but I recall looking at the pics of your unit and thinking that the wires coming from the transformers didn't appear to be the typical wire that was used on Fisher transformers. That's why I wondered if the leads had been extended -- and possibly improperly so.

At this point, I would verify the relationship of all wires to one another -- at this point assuming that the ONLY thing you know for sure is that you know which wires are related to the primary, and which are for the secondary. Beyond that, you can take nothing for granted. Might even be helpful if the end bell where the wires exit could be removed to see for your self any "extension" work done inside the end bell.

My worst fear would be if some one previously had the transformers rewound in an effort to repair them. I know there are services that do this, but I have NEVER heard of a good outcome. Without popping an end bell however, you have no idea if the leads are merely extended, or the product of a rewind. I'm sure of course that you have verified the part number of the transformer to that called out by Fisher for the model.

In any event, I really think this is where things are leading you to. The test of killing the NFB system proves that the problem is one of phase. You have verified the phase of the signal all the way through the power amplifier circuits, meaning that the transformers themselves are the only thing NOT verified. With the look of the wiring coming out of them, they are the next stop.

Dave

 

[audmod01]

Output Transformer Wires

Dave;

One thing that I did do was to ohm out the secondaries with my Fluke 8600A digital meter. Here is what I found;
Right Channel Transformer
Black to Brown=0.37 ohms (4 ohm tap)
Black to Green= 0.55 ohms (8 ohm tap)
black to Yellow= 0.74 ohms (16 ohm tap)
Left Channel Transformer
Black to Brown= 0.42 ohms (4 ohm tap)
Black to Green= 0.58 ohms (8 ohm tap)
Black to Yellow=0.79 ohms (16 ohm tap)
I did extend one of the yellow wires myself as the lead was cut short. I covered it with some yellow heat shrink tubing. When I had the end bells off to paint them I could tell that the original wire colors were correct when comparing one to the other. External to the bells at least the OEM transformer's lead colors had changed some due to age.

The colors of the wires only showed discoloration from outside the bell to inside the bell, so I think they have been there a good number of years. As to whether they might have been rewound or not, I am not sure I could tell.

If the phase of the secondary windings are incorrect, is it likely that swapping the leads and using the yellow as common, the green as 4 ohms and the brown as 8 ohms and the Black as 16 ohms, would correct the problem? Or is it more likely that the phase is something other than 180 degrees opposite of the primary and just off enough to create feedback issues?

Would Heyboer be a good potential rewinder or total new build? Or how about Triode?

Joe

 

[dcgillespie]

The next thing I would do is try swapping either the drive leads to the output stage from the phase inverter -- OR -- swap the primary plate leads, but not both.

I think you may have tried this once before, but I believe that other things may have changed since that time, so it would be worth another try.

You do not want to swap the secondary leads. The Common and 16 ohm leads can swap, and the 4 ohm tap is the center tap of the winding, so it represents that impedance from either the Common or 16 ohm end of the winding. But the 8 ohm tap only works relative to the actual Common lead -- not the 16 ohm lead.

Dave

 

[audmod01]

Feedback Issues

Dave;

I will try swapping the plate leads of the output transformers again to see if it helps. One thing that did change was that I swapped the transformer secondaries left for right channel as I discovered that I had mistakenly cross-connected them when I reinstalled them after cleaning and repainting. It is worth a try.

Nice to know the way the secondary windings were done. I had thought the secondary windings were all in series. Perhaps Fisher wanted the best match to 8 ohm speakers that could be provided. I see now why you chose the 4 ohm and 16 ohm taps as feedback source connections.

I found this site that discusses the design and manufacturing of audio output transformers:
http://education.lenardaudio.com/en/14_valve_amps_5.html
I knew that different methods of winding audio output transformers had been developed. This particular site is pretty thorough in covering various methods.

I will take a look at the primary winding resistances to see if there is any notable difference there.

Joe

 

[audmod01]

Plate Leads Swapped - It Works!

Dave;

This morning I measured the resistances of the output transformer's primary windings. This is what I found:
Right Xfmr.
CT to Blu/Whi=113.3 ohms, Fluke 8600A
CT to Blu=123.0 ohms, Fluke 8600A
Blu/Whi to Blu=~240 ohms with multimeter
Blu/Whi to Blu=.23 ohms with Fluke 8600A
Left Xfmr.
CT to Blu/Whi=112.5 ohms, Fluke 8600A
CT to Blu=121.0 ohms, Fluke 8600A
Blu/Whi to Blu=~250 ohms with multimeter
Blu/Whi to Blu=.23 ohms, Fluke 8600A
The discrepancy between the end to end measurements with the Fluke versus the RS Micronta 50K ohms/V multimeter must have something to do with the way the Fluke measures resistance internally.

I decided to try a last ditch effort and swapped the Blu/Whi leads with the Blue leads for each channel at the output tube sockets. I made sure the negative feedback circuitry was in place and not grounded for each channel. I turned the unit on waiting to quickly turn it off if I heard oscillation. Instead the amplifier came up working correctly and sounded quite good. So somehow the way these transformers were wired caused them to be the reverse of the normal wiring. Either that or someone did go into the transformers and changed the connections at some point in time. Who would have thought this would be the case?

I will go back now and reconnect the speaker/earphone wiring and recheck. That should not change anything. If it does the only thing that would explain it would be that the left and right channel leads got reversed in the speaker switch wiring. That should be easy to correct.

What a relief!!!!!

Joe

 

[audmod01]

Speaker/Earphone Switch Wiring Recheck

After I completed reattaching the corrected output transformer plate leads to the tube sockets, redressed the wires and secured them with cable ties I took a look at the speaker/earphone selector wiring. I carefully reattached the wires as I had in the beginning.

I had a red wire connected to what will be the right channel of the earphone jack. I had a white wire connected to the left channel earphone jack. The red wire attaches to the right channel 4 ohm speaker terminal lug. The left channel white wire attaches to the left channel speaker terminal lug. I removed the temporary direct grounds on the speaker terminal lugs and reattached the speaker gounding wires (the cloth covered ones). The darker colored one connects to the right channel speaker terminal lug and the lighter colored one to the left channel speaker terminal lug.

A quick recheck of operation showed that everything was still working correctly.

In retrospect, I believe I had two problems. I had wrongly swapped the left and right channel secondary wires at the speaker terminal strips. Although I had the feedback wires going to the correct driver stages, the cross coupling of the two outputs caused some of the feedback issues I was experiencing. Then it turned out the output transformer plate leads needed to be swapped to the altenate plates on each channel in order to get rid of positive feedback. There was still the possibility I could have somehow cross-connected the two channels at the speaker/earphone selector switch. Taking the speaker/earphone selector switch out of the speaker wiring and grounding the common leads of the speaker terminal lugs allowed me to begin eliminating possibilities.

I believe some of the difficulties I encountered were related to the non-standard rewiring that I found when I first began the restoration.

What a relief it is to have all that behind me now. I can set the final output tube's current and do the AC signal balancing to the output tube grids. Currently I have the AC balance controls set to the center of their adjustment range.

Joe

 

[audmod01]

Output Bias & AC Balancing

I finished cleaning up the lead dress of wires in the output tube area, securing them with cable ties again.

I set the negative grid bias to provide 32mA per tube, which I thought was a good operating point. I then adjusted the AC balance using a 1kHz tone fed into the amplifier. I used my Fluke 8600A set to measure AC voltage and adjusted the AC balance controls until I had equal AC measurements to either grid on each pair of output tubes. The ground lead of the meter was connected to the filtered DC voltage going to the grids. The other lead was connected directly to the feed going to each grid. The final AC measurements were within .01VAC of each other which I believe is about as good as I can expect.

Listening tests confirm good frequency response while listening to my favorite FM station. I heard one recording with pipe organ pedal notes accompanying an orchestra playing Respigi's Festival Overture. Everything sounds quite clear and balanced left to right with the balance control set to mid position. Tone controls work well with no scratchiness. The slide switches I installed are working flawlessly. The volume control has no scratchiness either. The Mode Switch works like it should and there is no popping when changing modes (make before break contacts).

Overall results are quite pleasing. I might replace the PEC tone control circuits with discrete parts, but that can wait a while.

This Fisher bird sings once again!

Joe

 

[dcgillespie]

What a super job, Joe! That unit would have been lucky just to be a parts unit, let alone one restored to the superb level you have taken it! And your documentation of your work is second to none. I know you've particularly got to enjoy it now seeing the final results from all your work. We need some final pics!

One suggestion to offer if the proper equipment is available, is instead of measuring and equalizing the AC drive voltage supplied to the output tube grids, do the following:

1. Connect a proper dummy load and scope to the amplifier. Connect your DVM between the two cathode terminals (pin 5) of the output tubes in one channel. Set DVM for low DC volts.

2. Using 1 kHz drive signal, drive the channel to about 1 db below maximum power output (or about 78% of max power).

3. Adjust Phase Inverter adjustment for minimum voltage difference between cathode terminals.

Where as adjusting the phase inverter drive for equality to the output stage is ideally correct, it rarely produces minimum distortion, as it doesn't take into account the individual characteristics of the output tubes, or the AC balance of the output transformer. By adjusting the inverter drive so that the current draw between the output tubes is equalized under high power dynamic conditions, these variables are then accounted for, producing a more accurate minimum distortion setting.

Congrats on a super job, very well done!!

Dave