bill@vrdxhq.verdix.com (William Spencer) (11/29/90)
My Fisher is the "The Fisher" 500-B. Cost was $100 without wooden cover.
Not a great deal but par. This amp originally was not designed as an
audiophile amp, the emphasis was on features and getting the job done.
Power is 22.5 watts per channel. Tuner is decent and features the nifty
tuning eye.
It's a good amp for experimentive type modifications because of the
terminal strip, point-to-point wiring scheme. You can easily make a small
change, listen to the difference, and then make another change.
One might prefer a more powerful amp for modifications, though, because
it's not that much more work. The following instructions should help
modify other amps also.
My modifications have resulted in a very impressive sound with full bandwidth
and definition. Bass is still a little bloated (response is flat to 30 Hz
with some distortion near the bottom). My modifications are mostly
design changes; parts quality is good but not great. Improving parts quality
and retaining the original design is not recommended.
The unmodified amp sounds pretty awful. Any cheap modern reciever will
beat it. Bass is bloated and and highs are rolled off. Inspection of the
circuit reveals many large series resistances with small ceramic capacitors
(10 pF range) in parallel with them to try to balance high frequency losses.
Mostly they add gritty highs while detail is still lost. Also, the signal
passes through about 9 feet of cable (phono, less for AUX). The preamp
section consists of two single tube inverting amp stages in series divided
seperated by channel reverse, tape monitor and passive tone control circuits.
The cabling, high impedances, and low-tech preamp stages are the reciever's
biggest weaknesses.
The power amp stage is suprisingly good. The modification with the best
work/improvement ratio is to bypass the preamp. The input to the power stages
are the shielded cables coming into the driver stages (12AX7's near the
2 output transformers) from the volume control. If you don't need volume
control, just run cables (hot and shield) from there to a pair of the RCA
jacks. Add 100K resistors between hot and ground. If you use the volume
control, you immediately have a problem, however, because the volume control
pot is 500K, enough to cause a R-C high frequency rolloff at some settings.
The gain of power stages only is enough for moderate volume play from
a CD output level, however phono is usually lower in level.
*Note: tube amps contain LETHAL voltages when operating and for some time
after. If you can't take precautions, don't modify.*
My Fisher still contained many original preamp tubes (20 years old!).
I immediately replaced them of course. Also all connectors and tube sockets
were cleaned with Cramolin, along with the incredibly dirty selector switch
contacts.
If you want to make more comprehensive mods you will need to trace portions
of the circuit. Ignore the tuner section. The preamp and phono stages are in
the front.
My mods consist of first making the controls and cabling more normal.
The preamp section was rebuilt for a better gain stage. Detailed
instructions for power stage modification are not given in this article.
My mods went as follows:
1. Hardwire the phono/tape head amp for phono. Hardwire the input selector
switching and the EQ switching (taps on ceramic resistor/capacitor
encapsulated networks). Switch the input jacks to the tape head ones so
metal shielded plugs on the back don't short their grounds together when
AUX is used (a
RF rejecting treatment is used for tape head/phono grounding).
Also, eliminate the cable runs to the backpanel phono level control (not
needed) and run short leads directly to the main section of the source selector
switch.
2. Cut out the components linking the "tape" position on the selector switch
to the tape monitor lead. This provided a "kamikaze" type tape monitor function
(if used on record feedback could occur).
3. Connect the tape out line to the output of the source selector switch.
I used what was the low filter switch as a tape send switch (turn off to avoid
loading). Originally the tape monitor came from the first preamp stage --
inverted (bad) and boosted (and rolled off in the highs).
4. Move the tape monitor switch to immediately following the source selector.
The tape record/monitor arrangement is now conventional.
5. The preamp circuit I am using is essentially a passive preamp (select,
balance, and volume driven by the source) followed by a fixed gain boost
(a PAS type preamp stage). You can eliminate the gain if you don't need it.
I needed it because of phono and also because I reduced power stage gain.
Unlike seperate passive preamps, loading is known and cabling to the
amp is short.
6. The original balance control is a shunt to ground type between the preamp
and volume control. A 47K series resistor drives it so the shunt can attenuate
the signal. I changed it to 5.6K for less loss (including capacitive). The
balance is less sensitive (good) and loads the input when set far to one side
(who cares?). It is the first function following the tape monitor switch.
7. Next function is volume. 500K is too high a value for the control.
I chose 100K. Since the original control was also power switching and
is hard to replace I relocated the volume to the opposite side of the front
panel. Many quality controls can not be used because the shafts are too short
(if you want to keep the front panel anyway).
I used the Radio Shack stereo pot and a press-on type knob reamed out enough
to jam on the little bit of shaft sticking out the front.
8. Strip out the preamp stages and figure out how to arrange the parts.
The preamp gain stage was adapted from the PAS design (seen in _Audio Amateur_).
This is two tube stages in series with feedback to the cathode of the first.
Gain is approx. 11X.
power for other channel
|
--------------*--------------*---*---R9---<original preamp power
| | C4 C5
R2 R6 *GND*
| |
*-----C1- *----C2-*--C3---*--->To power amp section
| | | | |
>--R1--T1A *---T1B | R8
| | | | |
*---- | R5 | GND
| | R4 | |
R3 | | | |
| | GND-- |
GND | |
-------R7-------
R1=1K, R2=220K, R3=5.6K, R4,R8=470K, R7=68K (all 1/4 WATT or more)
R6=100K 1/2 W, R9=1K 1 W
C1-C3 = .1 uF, 400V, C4=.047, 600V, C5=80 uF, 450V
pin #'s: 1 6
2-T1A 7-T1B
3 8
use T2 for other channel. Heater connects (4,5,9) remain.
Gain is set by R7/R3. Note R7 is low enough to load down tube stage.
C3 is needed because output is not exactly 0 V otherwise.
9. I rebuilt the power stage using hybrid cascode driver. Based on Curcio
ST-70 circuit in _Glass Audio_ #1 except using 500V, 1A MOSFETS for the
upper tube sections. Also biased different -- 12AX7's don't work down to as
low of voltages as the 6DJ8. Different transistorized current source also.
As they say, "solution is left to the reader."
10. Also, power stage feedback is taken from the same tap as used for the
speaker instead of allowing multi taps. This trick is borrowed from VTL
but they most likely get less advantage because their transformers are better.
Gain is reduced for increased negative feedback ratio.
---------------------
Hints
-----
1. If you change the power stage feedback make sure the time constant of
the stabilization is the same. A compensating capacitor is in parallel with
the feedback resistor leading from the output side of the output transformer.
This rolls off the amp below the transformer resonance. The other small
capacitor in the power stage fights oscillation due to delay between
the push and pull sides of the power stage. It has an equivalent in
different driver topologies.
BTW, the popular differential "long tailed pair" driver configuration
doesn't work too well. Gain (therefore feedback) is reduced to half.
Bandwidth is higher though.
2. BI-AMPING. Note that reducing C3 and R8 in the preamp gain stage above
results in a high pass filter. Add a low pass hanging off the gain stage
output (that doesn't load the stage too much) upstream of C3 with its own
blocking cap and you have a first order
electronic crossover. The nonzero output impedance of the gain stage
complicates matters slightly. I have been using this setup with a
cheap reciever for bass. A Radio Shack equalizer provides a known
constant load to the passive low pass section (RS includes schematics.
The reciever was found to have a varying input impedance due to the
volume control circuit.). The sound is good. Even a cheap amplifier
for bass can outperform a good amp working through a passive crossover.
Tubes are a natural for the high pass amp since the transformer has trouble
with bass, although highs may be limited also.
3. The original phono stage outperforms the phono stage of many cheap
recievers. However, the highs are a bit rolled off and the lows are
tubbier than they should be. I presently am using a seperate phono preamp.
I may go to the seperate preamp and bypass the preamp section entirely.
Will be an interesting comparison. (Other considerations are gutting
the reciever section and building a more involved power section using
more tubes in the driver and more power supply filtering. There is
no more space in the amp as is.)
4. Note additional power supply capacitor C5. The amp was unstable until
I added it. It was okay until my air conditioner clicked on. The amp
then went into oscillation around 1 Hz due to interaction of the power section,
the power supply and the preamp. Since C5 is near the preamp R9 prevents
the possibility of high current across the chassis.
I also added a diode of proper rating in series with the resistor
driving the main preamp supply to reduce turn off thump. The high pass
for bi-amping eliminated the thump, so there must have been some near
instabilty remaining.
DO NOT increase the blocking capacitances above the values shown
or C5 may not be enough to assure stability.
5. Don't try to desolder the original components from the terminal strips,
you'll destroy the strips trying.
Just cut them off and remove excess. The terminals mainly provide a fixed
mounting for components instead of providing primary connection.
6. If you use a 'scope to look at signals, remember the capacitance and
resistance is not negligible in high impedance circuits. Be sure the
probes and inputs can handle the voltage. Watch out for the ground ring
around the probe that's exposed when not using a clip hood! Generally,
most scopes can't handle the full swing of the output tube plates but
can handle everything else. Don't 'scope directly to the grid (input)
of a tube, go upstream of the grid resistor, otherwise it may go unstable.
Be careful! It's neat though looking at the stuff -- the circuits are
simple yet work very well.
------------
So are these modifications worthwhile? Depends. I have enjoyed some
excellent sound for the last 6 months or so, but there are other methods to
achieve that. Also, I find myself wanting more power (even with the
biamplification). But I have learned much. For those with enough interest
or more time and energy than income this type of modified component is
something to consider.
bill S.
Disclaimer: I can not promise to stay out of the audio profession all my life.
AND, ONCE AGAIN:
*Tube amps contain LETHAL voltages when operating and for some time
after. If you can't take precautions, don't modify.*