Improving the Fisher X-101C

[dcgillespie]

Here we go again, looking to make the most of a very good Fisher amplifier, by (1) Capitalizing on any significant compromises in the original design, and (2) better adapting it to today's audio environment.

This time, it's the X-101C, in the form of Audiodon's ratrig. Currently, it has been brought up to original performance specifications as documented in his thread:

http://www.audiokarma.org/forums/showthread.php?t=491521&page=7

and is now undergoing listening tests while the project is contemplated.

In our various conversations, Don has expressed to significant concerns with the unit, that include:

1. It tends to run out of gas in the power output department, as compared to other similarly tubed Fisher units, and

2. The heater string cathode bias system used in this design limits any relaxing of output tube quiescent current to help accommodate newer tubes or extended tube life in general: If you reduce it much at all, the heaters in the small signal tubes won't come up to temp, so the unit won't work.

Both of these issues are related, so fixing the bias system can potentially yield a twofer in results. He has also asked that some of the improvements I made to his Fisher X-1000 be considered for this project as well. They include:

1. A tone control bypass option, and

2. Modifying the near useless Reverb Input/Output jacks into much more useful Preamp Output/Power Amp Input jacks.

To this, I added dumping the powered center channel output, and converting the main L and R outputs so that the Common output connection is in fact at ground level, rather than floating above ground as found in the original design.

This third modification has three distinct advantages:

1. It makes testing of the unit painless and much more effective with test equipment having dual inputs with a common ground (such as scopes), or equipment that includes a common ground path through the AC line ground system. Since testing/measuring is an area that Don is is pursuing, this will ease his effort into this new area for him.

2. It allows the use of external speaker switching systems or circuits that require a common ground output.

3. In the case of the X-101C, it will also make the unit correct with regard to absolute phase. As originally designed, the X-101C reverses phase for all inputs except that heard through the Tape Monitor setting.

The elephant in the middle of the room with this project, is that if the small signal tube heaters are no longer to be powered by output tube current draw, then an alternate source (power supply) is a must. While this will in fact involve some slight alterations to the X-101C's chassis to accomplish this, addressing this issue is very much more than worth the effort, as testing has shown that the original bias system used represents a HUGE compromise to this Fisher. In fact, the compromises are virtually identical to those I addressed in my original article on EFB(tm), so for those interested in this project, a review of that article would be of benefit in understanding why the compromise is so significant. In that article, you can virtually substitute "Fisher X-101C" for "Dynaco SCA35", and the article will still be 100% accurate with regards to the problems discussed regarding the original design, and the results produced when modifying the unit with EFB. In fact, the results that EFB can produce in the X-101C is even more significant than that achieved in the SCA/ST35 amplifiers, because the performance deterioration that occurs with both channels driven in the X-101C is even greater than that which occurred in the small Dynaco amplifiers. The original article on the introduction of EFB can be found here:

http://www.tronola.com/A_New_Look_At_An_Old_Friend.pdf

Finally, it should also be noted that the Fisher X-101C is for all intents and purposes (except for the bias system employed), the audio section of a Fisher 400 receiver. IN PARTICULAR, specifically that of an EARLY version 400 receiver, as along with the same basic topology, even though they have different part numbers, the output transformers in this unit measure EXACTLY the same electrically and physically as that of a bonafide early version Fisher 400 OPT that Don had sent me for testing last year. So, owners of the early version 400 receiver just might want to follow along with this project as well.

So, there's the project. More as it begins to unfold!

Dave

 

[audiodon]

It is interesting to note that the 400's output transformers are driving 7591s in this configuration, while in the 400 the same output transformers drive 7868s.

 

[dcgillespie]

It is easy to think that specific output transformers and tubes are married together, and while certain transformers and tubes would in fact appear to be (for example, an Acrosound TO-300 and 6L6 class tubes), tubes can in fact be operated in a variety of operating conditions, with each condition requiring a different optimum load. To that point, note that the EFB installation in my late version 400 not only provides EFB action for the output stage, but also alters the base operating point of the tubes, so that the match between them and that offered by the OPT becomes optimum. In the case at hand, the comparison between connecting the same basic OPT to either 7591s or 7868s has zero impact of performance, since these are basically the same tube, but with different base configurations.

Dave

 

[eduarsan1]

Dave .

I´m ready to help in the drawings when you need it!!!:thmbsp:

Regards

Luis

 

[dcgillespie]

Initial Thoughts For The Elephant

Since any real improvement to the output stage begins with lighting the small signal tube heaters from an alternate source, an additional power supply will be required. But in designing this power supply, other aspects of the project should be considered. This includes the request for Preamp Output/Power Amp Input jacks, and the application of EFB.

By careful consideration of all of these elements, a common Radio Shack transformer was found to be ideally suited for the job. With a 25.2 volt CT 450 ma secondary, it can easily light up the four small signal tubes (converted to two paralleled circuits of two tubes in series, ala Fisher receiver style), and with its center tapped secondary, can center this voltage as a + and - source relative to ground. That in turn allows this supply to also quite easily power a small buffer circuit required for the Preamp Out/Power Amp Input jacks -- which will also handle Miller at the power amp input stage to boot. Finally, the new supply will also allow for EFB to be applied to the output stage with (in part) an EFB control grid regulator (versus an EFB cathode regulator as used in the small Dynaco amplifiers), which after development, will then allow the whole mess to be rather easily installed in an EARLY version Fisher 400 receiver if desired (An EFB cathode regulator could have been used, but would not be as compatible with the fixed bias design of the receiver). Besides all of this, the transformer is uniquely enclosed in a metal case, providing for superior shielding, even in close proximity to the power amp driver tubes. I have used more of these little transformers than I can count over the years, with nary a single failure in a wide variety of applications.

So, thoughts are starting to jell. The pics show where the chosen transformer is to be mounted. The underside will of course have to be examined to ensure that the necessary holes for installation will not interfere with existing circuitry. If it does, an alternate plan for that circuitry will be required, as there is no alternate mounting location that is visually or physically workable for mounting the transformer. The chosen location does cover up the serial number information for this unit, but that information can be relabeled in a different location (on top of the control cover maybe), and besides, the area under where the transformer is to be mounted is severely rusted anyway. So really, it will help spruce things up!

More as it comes together.

Dave

 

[audiodon]

Well Dave,
This is a pleasant turn of events.
I do plan to copy the work when you've completed it in my Kymers X-101-C and the Little Bitty Pretty One may make a return visit this summer.

Perhaps eduarsan1 can do his X-101-D also.
Don

 

[eduarsan1]

Well Dave,
This is a pleasant turn of events.
I do plan to copy the work when you've completed it in my Kymers X-101-C and the Little Bitty Pretty One may make a return visit this summer.

Perhaps eduarsan1 can do his X-101-D also.
Don

Don

You're right!!!! I have a pair of OT from a early 400 model, and I'm considering in use those OT in my X-101-D!!!

Luis

 

[audiodon]

Doesn't it have the same OT already? I suppose it might have the later 400 style transformer.

 

[dcgillespie]

Finding Space

As usual, space is very tight in a Fisher project. The mounting screws for the new transformer come in EXACTLY under the existing terminal strip between the adjustment pots, and the driver tubes. The only resolve is to remove the strip, drill the holes for the transformer and mount it, and remount the strip using a nut as a standoff to clear the mounting hardware for the transformer.

Also, the area under the main power transformer will need to be cleared out to accommodate the new power supply components and EFB circuits.

Why in the heck could Fisher engineers not have been more accommodating, and taken into account that some goof 50+ years later was going add all this stuff???

Pics are provided of the areas to be addressed. I can work on them while I wait on my parts order just placed.

Dave

 

[eduarsan1]

I bought the x-101-d with replaced trans from a scott 299 so i'll change them!

Enviado desde mi GT-S7562L mediante Tapatalk

 

[audiodon]

Dave,
It seems that the ratings games that those manufacturers were playing at the time were not limited to increasing wattage.
Size consolidation was another rating . . . and the smaller the better.
The Sherwoods are even tighter than this X-101-C, with sunken and clamped can caps tilted at an angle and angled output tubes. Fisher seemed to get away from that after their X-202-B, though the KX-200/X-200 does have cutouts for the output transformer end bells because those transformers are laid in bells up and down and laminations on the side.

 

[dcgillespie]

Progress

In approaching the installation of the center piece of this project (modification of the output stage for improved performance), a thorough study of the space limitations and options for installation shows that there's not much of either. As a result, beyond the electronic modifications themselves, there is the significant pre-work required just to prepare the unit for installation of them. This will require:

1. Removal of the T-strip under the power transformer (currently mounted with a rivet), and relocating it to the nearby power transformer mount hardware.

2. Removal of the long T-strip that the adjustment controls are connected to. It is currently mounted down with three rivets, and will be reinstalled at the same location, but spaced above the deck of the chassis by the width of a nut, to allow space for the hardware mounting the new transformer that will be directly under this T-strip. A previous post clearly shows the strip requiring removal.

3. Removal of the adjustment controls. These will be reinstalled facing 180 degrees opposite of their original position, to allow room for the primary wires of the new transformer to enter the underside of the chassis at the point that they exit the transformer.

4. The general wiring of the area underneath the power transformer will need to be rerouted to clear the area to allow for a new circuit board to be installed (shown). Use of a circuit board is the only way that all of the required circuits can be neatly installed.

The new circuit board is now finished, and has been proof tested for this project. In conjunction with the new transformer, this board supplies well filtered heater power to the four small signal tubes originally powered by the output stage. One nice feature of this new supply, is that unlike the original design, any number of the 12AX7 tubes can be removed, and the remaining tubes will continue to operate without it. Of course before, trying such a move with even one tube would kill the entire amplifier, since the heaters of these tubes were all connected in series. With the new supply, there are no series heater strings involved, so any tube can be removed without affecting the others. The new supply will also accommodate the higher heater current needs of 5751 tubes as well.

Also on the board is an isolated +/- supply, perfect for powering a small buffer op-amp to support the desired modification of the Reverb jacks into Preamp Output/Power Amp Input jacks. With both + and - voltages available, it makes for a very simple addition of such a buffer circuit, eliminating the use of coupling caps and bias circuits for the op-amp, as was required for the buffer installation from the single source supply in my modification of the Fisher 400 receiver.

Also on the board, is a unique 1.5X voltage doubler, providing the necessary -50 vdc to power the new EFB Control Grid Regulator, that will ultimate provide bias voltage to the output tubes.

Finally, the EFB Control Grid Regulator itself is also located on the board as well, making good use of all available space on it.

The board is 1 3/4 inches by 2 3/4 inches, with the pics provided showing how well it accommodates all the circuitry on it, and its size relative to the transformer power it.

The EFB Screen Grid Regulator can be hard wired in over where the existing cathode bypass cap is currently located. The remaining proposed modifications do not have any space challenges associated with them, which is why the center piece of the project is being dealt with first.

With the board completed, effort will now turn towards making room for it, which will be the focus next week.

Dave

 

[audiodon]

Well, these kind of adaptations don't just pop out of nowhere, as you make them seem to. They're the result of experience, mindful planning and thoughtful execution.


This is about the worst-case scenario size-wise. The KX-100 is a comparable size, the KX-200 has a bit more space, and the X-202-B? Well that one has space but it's also loaded with wiring, so I'm not so sure.

Is that circuit board going to be a below-decks thing or would a different set of considerations be needed if it's placed on top like the board in the 400?

 

[dcgillespie]

It will be below the deck, mounted right under the main power transformer. Available height down there is also a consideration, which makes mounting a challenge as well.

The current plan is to secure a piece of similar sized phenolic (type) material to the chassis (to act as an insulator) with adhesive, and then secure the board to it with tie wraps, or some other semi-permanent means to secure it.

By comparison, the 400 had a grand canyon of space to work with in all dimensions, where as this unit has space -- but none of it usable for the needs of the project. Additionally, there are no easily holes already available to pass wiring through on this unit as there was with the 400, and this unit already has a surface mounted bias control, whereas the 400 had none. These factors all pushed for a top of chassis mount in the 400 project, but a below deck mount with this one. I am hopeful to have a 10K wafer to replace the current 5K wafer of the existing bias control. If so, this will allow the new control in the bias control location to adjust the EFB bias, for no change in overall control function.

I try to keep all evasive work to a minimum, but in this case, holes will be necessary to mount the new transformer and neatly route the leads -- if a professional looking job is to be done of it all.

No doubt other installation methods will work, with varying degrees of finished appearance. The circuits are not complicated, but rather, variations of circuits I have already published before, but adapted to operating conditions of this amplifier.

In looking ahead towards the other desired modifications, the most logical approach to installing a tone control bypass option is to convert the High Filter Switch to support the new function. That in turn then begs the question as to what to do with the High Filter network itself -- and the permanent Low Filter as well -- strapped to the line amplifier input stage. My suggestion would be to remove both of these filters in the process of that conversion to flatten the response of the stage, and eliminate the phase shift such filters produce -- even when they are "turned off".

Dave

 

[dcgillespie]

Meant to mention that I also tested Don's 5751s that are (now) installed in the phono preamp position of the X-101C (they were originally designated for the line/tone stage positions). These tubes tested dead on 1200 micromhos -- in both sections of both tubes -- exactly as published in the RCA tube manual for these tubes when operated under conditions for Ep = 250 vdc. I haven't tested a lot of these tubes, but such consistency is usually reserved for military application tubes, which these tubes likely were.

Dave

 

[audiodon]

Whatever you want to do is fine. These units are designed to be in cases anyway.

I agree to the suggestion of the removal of both the high and low filters and the repurposing of the switch.

Should I send some 12AX7s or are they unnecessary for what you're doing now?

 

[dcgillespie]

Other Measured Data and Observations

Before this unit goes under anesthesia, I thought I'd post a few more vital signs of base line performance.

Typical of many pentode output stage designs, the internal resistance of the X-101C's power amplifiers is 1.46 ohms on the 16 ohm tap. Plenty low enough to not contribute significantly to any speaker damping issues, but higher than most feedback UL and triode designs are capable of.

Measured frequency response was pretty rolled off, as is typical with the "Fisher Sound". Sweeping just the audio frequency bandwidth to 20 kHz:

1. The power amplifiers alone were down .75 db @ 20 kHz, and down 1.6 db with the volume control set for 12:00 noon.

2. Inputting a signal into the Tape Monitor inputs directly drives the tone stacks, and the stage immediately before the power amplifiers. Here, with the volume up full, 20 kHz response was down 2.8 db, while with the volume control at 12:00 noon, it was down 3.5 db.

3. Inputting a signal into the Aux inputs brings the filter circuits into play, and another stage of amplification. Whether it was by design or not, some compensation came into play with this stage to help the limited response of the tone stack/stage. Here, at 20 kHz, response was now down just 1.8 db with the volume full up, and down 2.2 db with the volume at 12:00 noon.

In any event, response that is this rolled off is clearly audible, producing a definite softening of HF detail, although lower volume control settings would act to improve HF response characteristics. After extended listening to the repaired amplifier, the sound is definitely pleasant to be sure. But when returning to equipment with a flat response in the audio bandwidth and the same listening material, the difference was apparent.

Visually, this is depicted in a series of 10 kHz scope shots:

1. Here is the stock X-101C power amplifiers, with the signal injected at the top of the volume control, and the volume control up full. High Frequency response is over compensated, but stability is excellent. Response is down .75 db at 20 kHz, with an identical presentation from both power amplifiers.

2. Here, the signal is injected into the Tape Monitor input (volume control full, tone controls centered), and response is down 2.8 db @ 20 kHz, with an identical presentation in both channels.

3. Here, the signal is injected into the Left Aux input (volume control full, tone controls centered), and response is now only down 1.8 db (average both channels) @ 20 kHz. You can see the improvement from the display in pic #2.

4. Here, the signal is injected into the Right Aux input (volume control full, tone controls centered). The filter circuits are clearly over achievers in this channel, with even more compensation applied than provided by the left channel circuits. This is one of the best reasons to remove such circuitry, as it is hardly needed in this day and age, and acts to introduce inconsistencies in Hf response between the two channels, which can smear the sound stage.

5. Here is a signal through the Left input, with the volume control at 12:00 noon. Note the rise time is reduced even more, and response is down 2.2 db @ 20 kHz.

Finally, Don has noted that this unit runs out of gas earlier than other 7591 units he is familiar with. Before the repairs were made, this was particularly true with the high power amplifier drive requirements and dynamic stability issues from the excessive feedback causing the amplifier to struggle. However, even after the repairs, a head to head comparison with my late model EFB modified Fisher 400 receiver was still no contest on demanding material (Telarc's recording of the 1812 Overture), where the X-101C really struggled with extended LF power requirements, where as the 400 took them in relative stride. So the potential for improvement has been noted both on the bench, and in the listening room.

With that, I'll turn on the gas, and begin to cut........

Dave

 

[audiodon]

As a parent, I've made more than one booboo's sting go away with the careful application of a princess or ninja turtle band-aid followed by a snack in the form of a roll of fruit-by-the-foot.

That parental behavior is exactly what I want to unlearn - not to forget, because there is a time when palliative care has value, but to unlearn it so I can expand upon that knowledge with facts and tests to help me move beyond my parental behavior.

I've used fancy caps, tube rolling, audio grade carbon film resistors, and other tricks and these have all masked the fundamental fact that the circuitry, as designed, requires improvement. The improvements are often predictable and repeatable. I take comfort in that as someone who is not a designer by inclination, but can repeat and improve on the designs of others, I will learn from this and be able to apply that knowledge to future work.

Those palliative treatments, like band-aids for a toddler, don't really make the problem go away. They give the required attention, settle the ego, hide the bleeding, and allow attention to be focused on comfort, but, for the amp, they do not heal the design compromises from back in the day that are ready for an update today.

I'm looking to heal the wound. This unit, in Dave's hands, after my attention to detail in the careful application of palliatives, should be healed when completed. The healing will expose when the palliatives genuinely assist and when they are band-aids that hide what's necessary to get to what we all want . . . amplification without compromise in a pretty package.

 

[fred soop]

.....
I've used fancy caps, tube rolling, audio grade carbon film resistors, and other tricks and these have all masked the fundamental fact that the circuitry, as designed, requires improvement.
.....

I was assisting another AK member looking for recap advice for a Heathkit AR-1500 receiver and looking at those places where a larger value electrolytic would benefit. But, looking at the power amplifier, I'm just cringing. Excellent design for its era but it's almost exactly what Bob Cordell starts with in one chapter, and with relatively inexpensive modifications, reduces distortion by 3 orders of magnitude.

The typical 2 transistor phono preamp is another example. In fact, one expert (don't remember which one) claims that the best fix for any preamp (phono or high level) over 20 years old is to gut and start over.

 

[dcgillespie]

Post Surgery

Everything went smoothly according to plan:

1. The AC wiring was removed, and the power transformer hardware removed so it could be swung free, allowing the 3 post terminal strip riveted underneath the power transformer to be drilled free. The screen dropping resistor connection was also removed from this strip.

2. The adjustment pots were all disconnected and removed, and finally,

3. The long 8 post terminal strip between the adjustment pots and the driver tubes was drilled free.

4. Mounting holes and holes to facilitate the leads of the new transformer were then marked and drilled.

5. The new holes were cleaned up and grommets inserted in those that the leads pass through.

6. The new transformer was then installed and mounted up tight, with its primary and secondary leads twisted and routed to the area where they will connect into the unit.

7. The 8 post strip was then remounted using one nut as a spacer at each mount point to raise the strip off the deck, to accommodate the mounting hardware of the new transformer, that comes through the chassis directly below this strip.

8. The adjustment pots were then reinstalled and reconnected, with the Channel A phase inverter control mounted 180 degrees out of its original position, to accommodate the primary leads of the new transformer.

9. The power transformer was remounted, using one of the mounting nuts and bolts to remount the 3 post terminal strip previously removed.

10. The AC wiring was reinstalled, with the primary winding of the new transformer connected in as well, and one of the CL-80 current limiters relocated to the 3 post terminal strip.

11. The screen dropping resistor has been (temporarily) relocated, and connected back into the circuit, and the adjustment controls have been reset.

This process allows the amplifier to continue to operate in stock form while I finish out some final measurements for installation of the new power supply/EFB board, yet has the chassis ready to go for physical installation of the board, and the new transformer mounted and awaiting its final connections as well.

So the patient survived the worst of the surgical proceedings, and should do nothing but get better from here on out. Pics include:

1. The surgery in progress. The controls have been removed and swung up high, the 8 post terminal strip has been drilled free, as has the 3 post terminal strip as well. The screen dropping resistor has also been disconnected from this last strip, all to allow working space in mounting the new transformer.

2. The 8 post strip has been remounted one nut thickness above the deck, and the adjustment controls reinstalled. Note the Channel A phase inverter control has been remounted 180 degrees opposite that of the other controls -- to accommodate the primary leads of the new transformer that appear directly behind it. If you look carefully, you can see the mounting bolts of the new transformer that appears directly under the 8 post terminal strip, as well as the transformer's yellow and black secondary leads, twisted, and running under the 8 post strip over to the area where they will connect into.

3. The finished look after surgery is complete. At this point, the amplifier will operate just as it did before the surgery, but a comparison with previous shots of the area will show that significant free space has now obviously been created directly under the power transformer to accommodate the new board. The goal with all such work is to make the underside look as if you were never under there, with all new wiring and rewiring looking as original as possible.

4. All of the surgery work went to create the space to allow installation of this --

5. And this.

With this phase of the work all complete now, some quick checks will be made to make sure that the amplifier is in fact all as it was electrically. Following that, the next phase will include rewiring the heater circuits of the small signal tubes to operate from the new power supply board, rewiring the cathode and grid bias circuits of the output tubes for use with the new power supply board, installation of the new board itself, and installation of the EFB Screen Grid Regulator.

And the project continues on.......

Dave