Fisher 400 Power Supply Refresh

sony6060

Super Member
I like the Fisher 400. The audio is just stunning. Rolled a few tubes, sounds great. The amp section does not need any upgrades such as boutique coupling caps. If it sounds good, do not fix it.

To ensure the Fisher 400 performs, I rebuilt the power supply. 50 year old electrolytics are a risk. I also noticed the can type capacitor next to the transformer operated very hot. Although picking up heat from the transformer, still was overly hot due to excessive leakage current.

I used a software based power supply simulator and found the factory stock power supply produced over 20 volts of ripple on the high B+. Part is due to the unbalanced voltage doubler using unmatched 200uF & 100uF in series. Fisher did this due to the first cap (200uF) does supply 1/2 wave DC to lower voltage circuits. However, the current draw is minimal. Calculating a new design power supply using much higher proposed capacitance on a power supply simulator, the peak to average current from the power transformer only showed minimal difference. Excessive high peak & short duration capacitor charging current can lead to excessive transformer heating.

The new power supply uses CDE 680uF @ 400 volt capacitors in series in the doubler circuit for effective 340uF at the high B+. Some manufactures such as Dynaco ST70 makes use of a small choke after the first filter section for a C-L-C filter feeding the power tubes. This is called a capacitor-inductor-capacitor (CLC) filter. I designed a 340uF-1HY-10uF filter. The new filter provides less than .6 volts AC ripple on the B+ (high voltage) for the power tubes. I also prefer a polyproplyene cap in the last section for best possible sonics. I used ARC manufacture vs Solen manufacture that I do not favor. The non-polarized ARC 10uF capacitor is rated 330 volts AC as viewed in picture below. To conservatively convert the AC rating to DC I calculate 330 volts AC (x) 1.4= 462 volts DC rating. 462 volts DC is well within safe operating parameters in the Fisher circuit. The choke I used is a Triad C24X rated at 240ma and 50 ohms DCR. The Triad C24X is available at Mouser for $8.77. The choke is hi-pot tested to 1500 volts. The choke does not spec maximum DC volt rating, but should be well within parameters to operate at 420 volts DC plus being wound on a plastic bobbin provides for extra insulation. The choke drops only 7 volts B+ (high voltage) with the Fisher 400 current draw.

I also rebuilt the bias supply. In stock form the bias was only supplying -16.6 volts. I added two 4700uF caps & reduced the 1K resistor that feeds the last filter section 100uF capacitor to 680 ohms. I also replaced the stock 100uF capacitor with a new capacitor rated 100uF @ 35 volts DC. I prefer to retain the factory additional bias filter section some simply remove. The simulated AC ripple from the rebuilt bias supply is approximately 1 millivolt or well within low imposed hum parameters for the sensitive 7868 high gain power tube grids.

Now the 7868 tubes enjoy -19 volts bias and operate cooler plus less stress on the power & audio transformers too. The expensive NOS RCA black plate 7868s I use will last longer. Some prefer -20 to -21 volts bias, however at 19 volts bias & 410 volts plate including 370 volts screen voltage the plate curves indicate 30ma plate current. That equals 12.3 watts plate dissipation with a 19 watt rated plate. The rest of can type capacitors (2) were replaced with 4 section 30uF @ 475 volts can type capacitors available from Antique Radio Supply. On one replacement multisection can type capacitor I changed from the factory 40-40-40-20uF sections to 30uF each for no other reason than the capacitor values was available off the shelf. The upgraded PS doubler section provides more pure DC to this new capacitor. In other words less capacitance did not exceed the Fisher 400 factory stock AC ripple content to other circuits. The other can type capacitor I replaced that now 'lost' the 200uF doubler capacitor section for 680uF, I simply paralleled two 30uF sections for 60uF vs the factory designed 50uF. I also changed the rectifiers to two series UF4007s in each leg now providing a conservative 2000 PIV rating. UF4007 rectifiers are low noise type diodes. I also removed the original power B+ shunt capacitor for clipping spikes and replaced it with a .01uF @ 1KV ceramic cap. Also, this power supply modification must add a C90 thermistor in the 120 volt AC line to limit start up inrush current that would be excessive with 680uF capacitors installed.

The Fisher 400 with new power supply provides the same sonic signature as original with now a slightly tighter bass response, brighter overall with the slight darkness removed and more airy presentation in a blacker background. The difference is not a 'night & day' change, but is sonically noticeable. Also interesting is the treble control does not need to be fully maxed out now. The sonics are more crisp with the power supply upgrades. The new can capacitors operate cooler & the power transformer also operates cooler with the higher bias on the 7868 tubes. The 1-2 volt higher 7868 bias from factory specs negates the slightly higher peak to average current demand of the new 680uF main capacitors. I avoid drill holes in a collectable vintage chassis, plus removing the power transformer for new mounting screws would be difficult. I simply soldered the new choke to the grounding lug on the lug type solder terminals. Although installed a little askew, the choke is hidden under the chassis. A 200 watt soldering gun or pencil is needed for this task.

Enjoy your Fisher 400 with a superior power supply. The Fisher engineering team back then could only dream about the high performance capacitors available today. :music:
 
Last edited:
Finished PS Pix

I removed the insulated can cover & attached over CDE new cap with silicon for a better finished appearance. :D
 
Last edited:
Nice.
It is refreshing to see a new approach to an old problem.
Despite our best efforts, sometimes our best "how to do" threads start and reinforce an echo chamber mentality.
I'll have to read this a number of times to fully absorb it, but I applaud your fresh approach.
There's also a difference between restore and redesign. I appreciate your chronicle of the redesign effort.

This is another example of "There's more than one way to do things".
 
Last edited:
Thanks All

I like quality & the Fisher certainly has it. The Fisher engineers must have performed critical tuning & listening- it is no accident the Fisher 400 sounds this good. A few minor changes is ok such as this new & simple power supply. Perhaps an RIAA mod (only two wires) is worthwhile, I'll think about it. The German ERO coupling caps at the 7868 tube grids are spot on. :yes:
 
Last edited:
I like the Fisher 400. The audio is just stunning. Rolled a few tubes, sounds great. The amp section does not need any upgrades such as boutique coupling caps. If it sounds good, do not fix it.

To ensure the Fisher 400 performs, I rebuilt the power supply. 50 year old electrolytics are a risk. I also noticed the can type capacitor next to the transformer operated very hot. Although picking up heat from the transformer, still was overly hot due to excessive leakage current.

I used a software based power supply simulator and found the factory stock power supply produced over 20 volts of ripple on the high B+. Part is due to the unbalanced voltage doubler using a 200uF & 100uF series capacitor doubler circuit. Fisher did this due to the first cap (200uF) does supply 1/2 wave DC to lower voltage circuits. However, the current draw is minimal. Calculating a new design power supply using much higher proposed capacitance on a power supply simulator the peak to average current from the power transformer only showed minimal difference.

The new power supply uses CDE 680uF @ 400 volt capacitors for effective 340uF at the high B+. Some manufactures such as Dynaco ST70 uses a small choke after the first filter section for a C-L-C filter feeding the power tubes. I designed a 340uF-1HY-10uF filter. The new filter provides less than .6 volts ripple. I also prefer a polyproplyene cap in the last section for best possible sonics. I used ARC manufacture vs Solen manufacture that I do not favor. The non-polarized ARC 10uF capacitor is rated 330 volts AC. 330 volts AC (x) 1.4= 462 volts DC rating and is well within safe operating parameters in the Fisher circuit. The choke is a Triad C24X rated at 240ma and 50 ohms DCR and is available at Mouser for $8.77. The choke is hi-pot tested to 1500 volts. The choke does not spec maximum DC volt rating, but should be well within parameters to operate at 420 volts DC plus wound on a plastic bobbin for extra insulation. The choke drops approximately 7 volts under the Fisher 400 current draw.

I also rebuilt the bias supply. In stock form the bias was only supplying -16.6 volts. I added two 4700uF caps & reduced the 1K resistor that feeds the last filter section 100uF capacitor with 2.2K shunt resistor to 680 ohms. I also replaced the 100uF capacitor. I prefer the factory additional bias filter section some simply remove. The simulated AC ripple is approximately 1 millivolt or well within parameters for sensitive high gain power tube grids regarding AC hum.

Now the 7868 tubes enjoy -19 volts bias and operate cooler plus less stress on the power & audio transformers too. The expensive NOS RCA black plate 7868s will last longer. Some prefer -20 to -21 volts bias, however at 19 volts bias, 410 volts plate including audio transformer voltage drop and 370 volts screen the plate curves indicate 30ma plate current. That equals 12.3 watts plate dissipation with a 19 watt rated plate. The rest of can type capacitors (2) were replaced with 4 section 30uF @ 475 volts can type capacitors available from Antique Radio Supply. On one replacement multisection can type capacitor I did change from the factory 40-40-40-20uF sections to 30uF each. However, the upgraded primary PS section supplies more pure DC to this new capacitor. The other can type capacitor that now 'lost' the 200uF section for 680uF, I simply paralleled two sections for 60uF vs the factory designed 50uF. I also changed the rectifiers to two series UF4007s in each leg now providing a 2000 PIV rating. UF4007 rectifiers are low noise type replacing the older design & more noisy 1N4007 rectifier. I removed the original power B+ shunt capacitor for clipping spikes and replaced with a .01uF @ 1KV ceramic cap. Also, this power supply modification must add a C90 thermistor in the 120 volt AC line to limit start up inrush current that would be excessive with 680uF capacitors and wearing on the power switch.

The Fisher 400 with new power supply provides the same sonic signature as original with now a slightly tighter bass response, brighter overall with the slight darkness removed and more airy presentation in a blacker background. The treble control does not need to be fully maxed out now- interesting and unexpected. The new can capacitors operate cooler & the power transformer also operates cooler with higher bias on the 7868 tubes. The 1-2 volt higher 7868 bias from factory specs negates the slightly higher peak to average current of the 680uF main caps plus also better for the higher than stock B+ with higher line voltages nowadays. I will not drill holes in a collectable vintage chassis. I simply soldered the new choke to the grounding lug on the lug type solder terminals. Although installed a little askew, the choke is hidden under the chassis.

I'll be the first to admit that I'm way over my head on this update...any chance you'll show/post the changes/mods on the schematic? I'm thinking I'd like to try this.

Many thanks :thmbsp:
 
I'll be the first to admit that I'm way over my head on this update...any chance you'll show/post the changes/mods on the schematic? I'm thinking I'd like to try this.

Many thanks :thmbsp:

I will draw up the schematic & likely post tonight. A little tricky for persons without experience. However, tracing a few wires & seperating the voltage doubler capacitors then applying the choke is not difficult. Also, soldering skills is needed for the choke mounting with a high heat pencil or solder gun. Drilling the chassis would require removal of the power transformer further complicating the upgrade.
 
Schematic

See attached. Note, the wire to choke that feeds the 7868 tubes s is shown on wrong side of choke. Feed the 7868 tubes after choke at same point as 10uF poly cap. The cap must be a poly type as 10uF in an electrolytic may experience too high ripple current in that position. An option would be an electrolytic type 47uF 450 volt cap. Nichicon EKX series handles .5 amp ripple current. About $2.20.
 
Last edited:
Kinda makes me wanna pick up a 400 :scratch2: :scratch2: :scratch2:

Do it, these will only escalate in price. I do not understand why these Fisher throw such a huge soundstage, but other tube amps sounds more closed in with the exception of my DIY 2A3 amps. Other I do not get is ceramic coupling caps in preamp sections usually have an gritty edge. Not these ceramics in my Fisher. Ceramic caps seem to last forever.
 
You could do this PS mod on your 800 - probably sound pretty stellar


Yeah, but I'd have to take apart what I only recently did. Naw.....I'd rather rip into a virgin one...I mean, well, you know what I mean :thmbsp:

I've been meaning to pick up another unit anyway :D
 
Yeah, but I'd have to take apart what I only recently did. Naw.....I'd rather rip into a virgin one...I mean, well, you know what I mean :thmbsp:

I've been meaning to pick up another unit anyway :D


I know exactly what you mean. I sell mine on trying to spread the love which gives me an excuse to get another.

Can't go wrong with a 400 :yes:
 
On the subject of ceramic caps in the Fisher, I believe Sony6060 and I might have had a skirmish on this issue, which I regret.
When he stated in another thread that he was going to remove them all, I stated that they were not in the signal chain.
What I wrote and what I intended were not aligned.
What I meant was that I do not believe that the ceramic caps are used in the signal chain where AC rides on the DC, eventually to be output as sound.

The ceramics are used in tank circuits, bandpass filters and other non-sonics affecting circuitry.
 
Corrected schematic

Note high B+ wire supplying 7868 tubes via audio transformer center tap to correct side of new choke.
 
Last edited:
Sony6060,
I see that the top 680uf is on the far side of the choke in your correction. Hopefully I have that correct.

Are that 680uf cap and the 10uf ASC cap tapping off the same spot and terminating at the same spot? Is the 10uf acting as some kind of bypass cap? I don't understand that part. Can you help please?

The top 680uf appears to be replacing C56 in your schematic? The bottom 680uf cap appears to be replacing C57c.

Then you're using a 30/30/30/30 can for C52A and C52B with each being a 60uf combo of two 30uf sections paralleled?
I'm just trying to triangulate your posts on the other page with this schematic to make sure I understand this. I'd like to try it.

Also, if I understand it correctly, you're substituting a 30/30/30/30 for C52A-C52D because the upstream ripple is now far less.

That cap that you changed to a ceramic cap before diode CR3 - what is its circuit function? I thought that was the death cap. If not, what is is?
 
Last edited:
Back
Top Bottom