Fisher X-200/KX-200 Design Notes

dcgillespie

Fisher SA-100 Clone
Subscriber
I have read with interest many of the threads on these two amplifiers (primarily the KX-200), and the various modification efforts to restore them. The closest I've ever come to owning one was decades ago when I received a KX-200 that was so mangled as to be nearly unidentifiable. That ultimately ended up in the thread I posted here:

http://www.audiokarma.org/forums/showthread.php?t=553403

But recently, I had one here for evaluation that turned up a number of interesting observations. With the recent efforts of tube-a-lou and others to unravel the mysteries of this amplifier, I thought I would share what I learned from that process.

GENERAL: These amplifiers include the biggest and best 7591 iron Fisher ever produced, being identical with that used in the X-202B and likely even the X-202C as well, although I have never had a C version of that amplifier here to verify this. The X-200 also almost surely represents the end of Fisher's production of the fabled 200 series, as identified by three basic observations:

1. The X-200 is a scaled down version of the X-202C, which itself is a VERY scaled down version of the X-202B. With the introduction of the C, the active feedback tone controls are gone, as well as the remote control option, which allowed the tube count to be reduced by two. The indicator lights are gone, as are various EQ settings, the dual tapped volume control, and other circuit details. All of this down scaling carries over into the X-200 as well, with these further reductions noted of that model:

2. With the introduction of the X-200, Fisher finally gave up the ghost on its SpaceXpander feature: Gone are the Reverb In/Out jacks! Good riddance!

3. Also with the X-200, the permanent LF filter I have mentioned so many times is gone: With newer changers having so much less rumble than the older ones, and heavy use of cassettes in play at the time, the filter became an anachronism. In fact, even the switchable LF filter itself disappeared on the scaled down X-200.


It would be natural that the KX-200 would already be a simplified version of Fisher built designs, to keep build issues by home constructors to a minimum. So in essence then, the Fisher built 200 series came in for a soft landing at the KX-200 level. The on-board bias meter of the KX-200 (to simplify the process of setting those controls for home constructors) simply became the final new feature of Fisher built 200 series amplifiers. As a result, these amplifiers are in fact so close as to certainly be cousins, and by all appearances, even kissing cousins. But there are some significant differences that set these two amplifiers apart. In terms of end goals, the X-200 might be considered to be a refined version of the KX-200. After all, the final Fisher built 200 series amplifier can't be outdone by a home built kit!! Undoubtedly however, they were both an effort to bring the whole 200 series of amplifiers (kit and manufacturer produced) and its inventory to a close, with the SS revolution that was upon them.

The detail differences will follow in the next installment.

Dave
 
Hiya Dave,

As a KX-200 owner this ought to be a rather interesting read.

Frannie
 
Great information as always Dave. I have a KX-200 that I use daily and it's a wonderful amplifier. Very, very quiet on all inputs. You don't even know it is powered up until the music starts.
 
Thanks Dave I'll be looking forward for your info on this amp, since I've worked on
the amp I've been using it daily for about a week now and sounding better and better
each day, it's really opening up with detail and sound.

Tube
 
I like the kx-200 so much that I rendered it inoperable so I don't blow it up before updating.

attachment.php
 
Nice ! I have a pair of X200s.
I'm interested with what you find. Also the nuance differences between the X and the KX
 
X-200/KX-200 Circuit Differences

The circuit differences between the X-200/KX-200 amplifiers are all confined to the power amplifier and line/tone sections, as the phono preamp and power supply sections are virtually identical between them, save the X-200 using a dual AC power switch and a noise bypass cap on the power transformer's HV secondary, versus the KX-200 which only uses a spst AC switch and no cap. Beyond that however, it's a toss up.

STRATEGIC DIFFERENCES: The X-200 and KX-200 are rated for the same power output, distortion, and sensitivity levels between them, but achieve the specifications for the latter two performance parameters with very different approaches:

The X-200 employs a rather typical Fisher 7591 power amplifier section, as used in their integrated amplifier lines. As such, it uses double the amount of NFB of similar designs in typical Fisher receiver and console only units, to target the higher performance level that a dedicated amplifier would be expected to deliver over that of the other more encompassing units. But it also uniquely uses a very high output stage quiescent current to achieve even lower overall distortion levels than prior 200 series amplifiers produce.

The line and tone amp stages employ a slightly different passive tone control design from the previous long standing passive design that Fisher classically used. Otherwise however, this section is very typical of such Fisher circuits, complimenting the sensitivity of the power amp section to produce a rated line level input sensitivity of the 350 mv to achieve full power output.

In the KX-200, a nearly identical power amp section is used, except that the very high output stage quiescent current levels are not quite as high (which raises the distortion of the stage), while the global NFB level has been uniquely increased even further to the highest levels of any Fisher 7591 product, to account for the increased output stage distortion. However, the increased NFB also reduces the sensitivity of the power amp section as well, so the gain of the line and tone amp stages is increased in this unit to compensate for the reduced power amplifier sensitivity, which then ultimately produces the same rated line level input sensitivity as that of the X-200. Both units employ the same passive tone control design.

The upshot of the two different approaches is that power, sensitivity and distortion remains basically unchanged between the two units for all intents and purposes. But in the process, the X-200 is even harder on the output tubes than the KX-200 is, which itself is well known to be a tube killer as designed. More on that later. Specific differences between the units include:

1. POWER AMP SECTION: The feedback resistor in the KX-200 (680 ohms) is 57% smaller than in the X-200 (1200 ohms), resulting in a 157% increase in the NFB signal applied around the power amp section. Also, because the output tubes operate at such a very high quiescent current level in the X-200, the grid resistors -- being of the traditional Fisher value of 330K in the KX-200 -- are lowered to 82K in the X-200, but the coupling cap values remain the same between both units. This means that the X-200 will have a higher degree of LF stability over that of the KX-200, which was already pretty good. Finally, a couple of small caps have been added to the X-200's driver stage to help curtail the HF response of that section after feedback is applied. These improve the HF transient performance of the design, by helping to minimize any ringing on HF transient signals. It is these enhanced stability measures that the X-200 has over that of the KX-200, that makes it a more refined design.

2. LINE/TONE AMP SECTION: In this section, the plate resistors in the X-200 (150K) have been reduced from the 220K values used in the KX-200. Additionally, the cathode resistors for these stages have been increased in the X-200 (1K and 1.5K) over that used in the KX-200 (both 820 ohms). These differences cause the line/tone stages of the X-200 to produce less gain than their counterparts in the KX-200, allowing this section in each amplifier then to compliment the needs of its respective power amplifier section, and produce a common sensitivity specification between them. In the X-200, this also creates more bias for the line/tone stages, allowing greater headroom against overload on strong input signals. This can be important, as like most Fisher designs, the line/tone amp stages operate flat out at full signal levels all the time, since they both appear before the volume control in the signal chain. Finally, the coupling cap from the line amp stage in the X-200 is 1/2 the value (.01 uF) of that used in the KX-200 (.022 uF). There are other very minor differences in these stages between the two units, but they are so small as to be of little consequence.

OTHER OBSERVATIONS:

1. Frequency Response:

The X-200 unit I had here had previously been serviced by a well known tech, who recapped the signal circuits using caps of double the original value at each location in the line/tone and power amp sections. Therefore, it will make the LF response readings I made of somewhat dubious value. However, the HF results should be largely unchanged from that produced by the stock values.

Aux 2 Input: Relative to 0 db = 1 kHz, tone controls centered, and volume control at max, HF response shows a gradual rise peaking at +1.5 db at 9 kHz, before leveling off at +0.75 db at 20 kHz. LF response showed a gradual rise of +1.75 db at 20 Hz. These are both significant, since the typical Fisher response curve usually shows a roll off at the frequency extremes. The frequency response specification for both units is given as +/- 1 db, 20 Hz to 20 kHz. With a 2 db spread then, the unit was technically in compliance, but with a response rise noted at the frequency extremes, this definitely departs from the Fisher norm. The numbers given are an average of the two channels, with the worst case deviation being at 20 Hz in the right channel, which displayed a +2.0 db rise at that frequency.

Tape Monitor Input: To see whether the response issues originated in the line stage or the tone controls themselves, response was then checked through the Tape Monitor inputs. Signals applied to these inputs bypass the line stage, and are applied directly to the tone control network. Here, LF response was unchanged, indicating that the rise in LF response is in fact generated in the tone control network itself.

On the HF end however, Fisher's habit of using the response of one stage to compensate for that of another is clearly in play. Here, response at 20 Khz was now down significantly at -2.6 db. Compared to the response at this frequency displayed through the Aux 2 inputs, that means that the input line stage displays a rising response of over 3.3 db at 20 kHz all by itself to achieve the over all response of + 0.75 db through the Aux 2 inputs. This shows that the first stage is clearly over compensating for the response loss through the tone controls, and that any signals taken from the Recording Output jacks will have an elevated response at high frequencies. Note too that these response levels were all produced with the volume control at the maximum setting. With the volume control at a typical user setting, HF response would be notably lower.

2. Output Stage Quiescent Current: Here, god love 'em, the Fisher engineers just fell completely off the bandwagon. But published efforts to address the matter simply exchange one problem for another. That, and some final comments will be the subject of the last installment.

Dave
 
The On-board Bias Meter

This is such a cool feature of these amplifiers, and the only integrated Fisher amplifiers to use them. No need to install individual output tube cathode resistors: it's already part of the design. And no need to add test points either: who needs them when a meter and switching facilities are built right into the unit?

Now coming back to the real world, the meter lives in a fairly harsh environment with regards to heat. The one in the X-200 here was DOA. But removing, cleaning, touching up a loose internal connection, and gluing the meter back plate back in place brought it back to life. When operating properly, the needle should sit over the "BAL" mark with the unit turned off, and should easily show free movement if the unit or meter itself is wiggled from side to side. With an ohmmeter, the meter movement should measure just over 1000 ohms out of circuit. If the needle appears frozen, then first check that the backing plate is still securely mounted. It's possible that as with the one here, even when (if) the meter movement is freed up, the backing plate glue may have become compromised with time and the heat, so that it has come lose to the point that it contacts the needle, and won't allow it to move. If that's the case, then simply re-glue it back in place using some silicone adhesive.

With the meter fully repaired in the unit here, I then took measurements to see if these amplifiers are the serial tube killers they are made out to be. Based on at least the Fisher schematic of these units, it certainly appears they are:

The schematic of the X-200 indicates that each cathode should operate with a cathode voltage of 0.54 vdc. Now with the meter in the off (open circuit) position and individual 10 ohm cathode resistors, this equates to 54 ma of quiescent current flow per tube. Allowing that about 3 ma of that is for the screen grid, that means that 51 ma of current is flowing to the plate. The circuit also shows that there is 430 vdc on the output tube plates. Subtracting off the (basically) half a volt at the cathode, means that the plates are dissipating 429.5 volts at 51 ma each, or......wait for it........21.9 watts per tube at just the plate! With a Design Maximum rating of 19 watts at the plate for 7591 tubes, that means that each tube is being asked to continuously dissipate 115% of it's DM rating. What the heck were these people thinking? On the other hand, this is all based off of schematic information, so it's gotta be a misprint or something, right?

Enter the newly repaired meter of the X-200 here. With a precision 390 ohm resistor placed in series with the meter movement, it takes 0.51 vdc to move the needle precisely over the "Bias" mark. Now when this meter is switched into the Bias position for either channel, the switch connects the two cathodes of a given channel together in parallel, for an effective cathode resistance of 5 ohms. But there's more. The meter movement itself must also be counted in the equation as well. With a 390 ohm series resistor, this amounts to 1400 ohms. On top of that however, in (only) the X-200, Fisher wanted to raise the quiescent current even further, and so placed a 47 ohm shunt resistor across each 5 ohm cathode resistance when the meter is switched into the circuit. Doing the math with all of these resistive elements in parallel, that means that each "5 ohm" cathode resistor is really only 4.505 ohms when the meter switch is turned to either Bias position. So when the bias controls are adjusted to cause the meter to deflect to exactly the center of the bias mark, then .51 vdc is being developed across a 4.505 ohm resistor. That means that total output stage current flow at that point is 113 ma, or 56.5 ma per tube. Subtract out the 3 ma of screen current leaves the plate receiving a current flow of 53.5 ma, or within 5% of what the figures on the schematic indicated. Close enough to throw out the possibility of a misprint. At a schematic plate voltage of 430 volts, that means that this particular little meter -- which appears to be operating properly in every way -- is causing the tubes to dissipate 23 watts each at the plate -- and god help us if the AC line voltage is high!!! In fact, even with a CL-80 installed in this particular unit, plate voltage from an actual line voltage of 121 vac is 450-455 vdc. So this time, the question appears to be very real........ What the heck were these people thinking???????

The KX-200 is little different, only omitting the 47 ohm shunt resistor from the design. But still, with it removed, the effective cathode resistance with the meter set to either Bias position is 4.982 ohms. Using the same process as above, this still yields a plate current flow of 48 ma per tube. With a (clearly low) schematic plate voltage indication of 400 vdc, we end up with 399.5 vdc at 48 ma, for 19.2 watts of plate dissipation. If the more accurate plate voltage of the X-200 schematic is used, then it becomes 429.5 vdc at 48 ma, for 20.6 watts of plate dissipation, which is still operating at > 108% of the DM rating -- and we all know that B+ voltages today are even higher than that.

Clearly then, both of these amplifiers are in fact the serial tube killers that their history shows them to be -- and sadly, by design, too. With the transistor revolution upon them, and no succeeding model to correct the issue in, there was little motivation to change the production run of what was certainly a low production model by the time the X-200 appeared. Now enter a popular modern resolve for the issue.

That resolve recommends replacing the 390 ohm series resistor for the meter with a 150 ohm resistor. With a resistor of this value, the meter circuit now takes exactly half of the voltage originally required to produce a Bias mark indication, or .255 vdc. With the 47 ohm shunt resistor removed (ala KX-200), such a resistor has the output tubes now drawing only 25.5 ma each when the meter reaches the Bias mark. With about 23 ma of this reaching the plate, and even a realistic 450 vdc on the plate with today's line voltages, that equates to 10.35 watts of plate dissipation (54.5% of rating), which will send distortion figures to the moon in this design -- and would also clearly be audible. Why such a resistor value was recommended escapes me. While it is a good value for tube life, the compromises it makes on amplifier performance are significant. Such a value sends full power 1 kHz THD to 4.0% on average, with notch distortion being visible, and maximum power output being reduced as well.

A much better value to use would be one of 240 ohms -- and in the X-200 -- this value, plus removing the 47 ohm shunt resistor as well. Such a configuration causes the output tubes to draw about 35.5 ma each when the on-board meter reaches the Bias mark. This causes the output tube plates to dissipate just over 15 watts each, or about 80% of their rating. This is an entirely satisfactorily level to operate the tubes at, continuing to produce very good life expectancy, and has full power (at the onset of clipping) 1 kHz THD -- in the X-200 -- registering around .6% -- nearly meeting the specification for the amplifier. In the KX-200, it should in fact continue to meet the specification due to the increased NFB level used in that model.

So with that issue addressed, the only other comments I would make include:

1. Both of these models definitely need Screen Stability resistors installed, whether the power supply is recapped with new caps or not. The conditions are such as to warrant them regardless of that effort. New manufacture, NOS tubes, it doesn't matter. They need to be installed.

2. For the more adventurous diyer, a worthwhile improvement might be to convert the High Filter switch into a Tone Control Bypass or Tone Control Enable switch. With the response burps and bumps noted, it would certainly help to flatten things out if desired. For ideas on how to approach this goal, you might refer to my "Improving the Fisher 400" thread, which includes such a project.

These amplifiers may be scaled down versions of the high water mark that the X-202B represents, but they still have all the goods that count, including that nifty little meter, such that these units can be just as good as any other model in the 200 line of amplifiers.

Happy Listening!

Dave
 
Yes, what were they thinking??? Over and over I learn of another Fisher model that ran their output tubes way over the limit. I know this is also the case in my 20A, and 30A, 6BQ5 amps. Such that new production tubes don't (or can't) enjoy a very long life. Was this common in the industry at the time, since tubes were cheap (relatively) and plentiful? Or, was it a Fisher thing. The situation doesn't get any better with modern wall voltages. How much compromise of performance (if any) is to be had by sticking within the design parameters of the tubes? Are there not distortions produced BECAUSE a tube set is over taxed? Or, do they just sing along happily till failure?
 
Hiya,

Dave you now have me looking square at my KX-200 and pondering its new life. Thanks so much for this detailed article.

I kind of think that Fisher like a few people were victims of the burgeoning "Marketing" induced "Watt Wars" and the shift of music from Classical , Jazz , Big Band to louder Rock and Roll.

Louder Rock and Roll meaning the volume people listen to it. Earlier music was as far as I know not blasted at the level that Rock and Roll lovers turned it up to.

Mind you I sound like an Old Bat saying this. Trust me I love that volume setting of 11 :)

Frannie
 
Hiya,

That's just because you're a hard of hearing Old Bat. ;)

Yah you need volume to overcome the tinnitus .. no little girl with a acoustic guitar music in my world. I can't hear it!!!

Frannie
 
Last edited:
.....
Louder Rock and Roll meaning the volume people listen to it. Earlier music was as far as I know not blasted at the level that Rock and Roll lovers turned it up to.
.....

Louder rock and roll, even before the "loudness wars" of recent years, was always more compressed than classical or jazz, so the average level would likely be higher even before cranking the volume to "11".
 
Thanks Dave for your informative going over of the 200 series amps, I'm going
to install the 240 ohm resistor and other upgrades when I get finished with my
other project. I did hear a change in sound when I had the 150 ohm resistor
in the bias circuit a much harder sound which I since removed for better sound
and now with this new info it will be better all around.

Thanks again
Tube
 
Hi all,
I change the 390 and put in with Dave's above advice a 240 ohm resistor
I got the readings about 35mv and with that the meter when set for that is
just above the red line but still in the red square, which is pretty much a bullseye
in my tube book. Thanks again to Dave for all his time getting into these great
amps.

Tube
 
Dave, What would be a good screen resistor value for the KX-200? Also, Could you give more insight on using the the high filter switch as a tone in or out switch? Thanks,Cory
 
Back
Top Bottom