The Fisher 500C/800C Challenge (long..dialup warning)

[valveitude]

First a little prologue…

A little while back, I decided to get back into my favorite obsession…2 channel audio, more specifically, analog and T-O-O-O-B-S . The first step was finding some info online ( a 20 year hiatus left me wanting on the current state of things) to help get me up to speed. I found this group of forums , and quickly learned how little I know, and a good deal of what I thought I knew was wrong. The bench of knowledge here is a very deep one.

I was originally planning to build the uber ultra-fi system right out of the gate. A month plus into my turntable project, I realized at this pace it would be a l-o-o-n-g time before I ever spun my first record (phono-pre, line stage, power amp, and speakers still in the thought stage), so I decided on an interim step…a massaged vintage unit.

This brings me to the point of this post. I decided on the Fisher 800C receiver. I always liked the vintage Fishers I had listened to in the past, and the 800C has the added bonus of AM (I get my news in the morning while I’m getting ready for the day). I would also note that I bought my unit BEFORE I read the Stereophoole review on the 500C ( same unit minus AM).

It is my hope that this group can help develop THE definitive set of mods for this already great piece. First the ground rules:

*Anything done must be 100% reversible-no cutting-no separate supply ect. (it is a classic after all)

*With the exception of the phono stage (total caa-caa being the universal opinion of those who have reviewed it) I want it to still be a Fisher (more or less). Avery Fisher was no fool, and in lew of the test equipment available today, he engineered a very nice piece of gear through voicing and intuition (much like the group here). If you suggest something that is a radical departure from the original topology, please back it up as to why.

* All modified circuits must draw no more current than the originals. 40 year old iron aint so easy to come by, and tends to be a little on the frail side. If it’s the greatest sound in the world, it does me little good if it melts the trannies.

The full pdf schematic of the 500C is to large to attach, so I have cropped snippets to aid in the discussion.

Big problem first..the phono section…



… I’m sure with the turntable/cartridges of the day, it sounded ok. But everything I’ve read suggests that it will not due today. I intend to completely re-due this circuit. It is limited to two 9 pin sockets, and the voltages that can be derived from the stock supply. Currently, I think the evil Kuei’s circuit..

http://www.diyaudio.com/forums/attac...&postid=335624

.. would fill the bill nicely, but I am open to further suggestions.

Next up..the control section…



…hoo-boy, what a convoluted mess. I won’t need the high/low filters, but I would like to be able to switch the tone/loudness circuits in and out, I can re-purpose one of the filter switches for this. I also want to reconfigure the “reverb” loop as a true pre-out (after volume) main-in ( to PA voltage amp) loop. I am uncertain how much the 12ax7 in this circuit contributes to gain versus compensating for insertion loss of the RC circuits. Can I simply bypass from the grid of V15, and go straight to the balance control ?

On to the power amp proper…



…nice and simple. Some things I’m considering are, converting to cathode bias, regulating grid bias, getting rid of the cathode capacitor ,and reducing overall feedback to compensate for the new local feedback. It would be a big departure from original but I also thought about a small Jensen transformer for the phase splitting duties. This is the area I’m least certain about changing. The 500C/800C is famous for its low end extension and “slam”..I don’t want to lose that. I am all ears to the pros and cons of these mods, as well as any others you might suggest.

Finally the power supply…



…Basic stuff. I am most interested in diode recommendations here. The RC filters will be determined after phono section is finalized.

Well gentlemen, there it is..let the ideas/recommendations run forth.

Casey

[SY]

Were it mine, I'd have less respect for 50 year old ideas of production engineering. And sensible engineering to meet the requirements of typical users 50 years ago.

What don't we need?

- Phono equalization other than RIAA
- Tape monitoring
- Tone controls

So... were this piece mine, I'd wire up a permanent RIAA network, and change the tone control section over to a simple buffer for a preamp out. And I'd get a good volume/balance control. If I would need tone controls, I'd build in a version of Quad's tilt control, the best consumer tone controls I've ever used.

Now the power amp. There are better ways of doing a one tube input stage in 2005. A voltage amp direct coupled to a split load inverter is one of them. A small input transformer coupled to a diff amp is another. Either way, the output is an easy-to-drive pentode stage, so don't muck with cathode bias.

Regulation of the grid circuit is an OK idea; more importantly, you want to regulate the screen supply. If that's not regulated, the idle current will drift with changes in line voltage no matter how well the grid supply is regulated. Plate regulation is much less important here.

[Giaime]

Hello Casey,

I've been told that the phono preamp you've linked has error in its schematics: in fact it doesn't simulate well the RIAA curve.
http://www.videohifi.com/forum/topic...age=1&ARCHIVE=
In this thread on an italian forum you can see that RIAA network simulated on Spice, and the graphs of the response. Not good.
Lower on the first page you'll find a better phono preamp: also on my site you can find a tested design that I've just built: just 2 12AX7 for both channels.

[valveitude]

Hello Sy,

Were it mine, I'd have less respect for 50 year old ideas of production engineering. And sensible engineering to meet the requirements of typical users 50 years ago.

What I have respect for, is the ability to get good sound in spite of the constraints of 50 year old production engineering. No easy feat. I agree that ther are a multitude of better ways, which is why I started this thread.

What don't we need?

- Phono equalization other than RIAA

Very true

- Tape monitoring

Maybe..Tape-in could be used as another input, and tape-out could feed my pc for CD burning (can't spin the wax when I'm on the road)

- Tone controls

90+% of the time I couldn't agree more, just one more thing to muck up the signal, BUT that other 10-% of the time it is most useful. This is why I want to keep it, but be able to switch it out.

If I would need tone controls, I'd build in a version of Quad's tilt control, the best consumer tone controls I've ever used.

Ok, I'm listening..could you point me to a schematic, and maybe elaborate on what makes it the shiz-nitty of tone controls ?

A voltage amp direct coupled to a split load inverter is one of them. A small input transformer coupled to a diff amp is another. Either way, the output is an easy-to-drive pentode stage, so don't muck with cathode bias.

One of my weak points is associating circuitry with its name, could you point me to an example of a "split load inverter" ? I mentioned above that I was thinking of the input transformer approach, any implementation you could link to ?

Regulation of the grid circuit is an OK idea; more importantly, you want to regulate the screen supply. If that's not regulated, the idle current will drift with changes in line voltage no matter how well the grid supply is regulated. Plate regulation is much less important here.

Very good info. What solid state reg. scheme would you recommend for the screen?

Hello Giaime,

I've been told that the phono preamp you've linked has error in its schematics: in fact it doesn't simulate well the RIAA curve.

Hmm..I haven't simulated it myself, but I know more than a few people have built it and speak highly of it. I couldn't say one way or the other, maybe Kuei can step in.

In this thread on an italian forum you can see that RIAA network simulated on Spice, and the graphs of the response. Not good.

Can't speak Italian, and the pictures came up X's here, bummer.

also on my site you can find a tested design that I've just built: just 2 12AX7 for both channels.

Looks like that would work nicely in this application..thank you.

Keep em comin' guys.

Casey

[rcavictim]

Those odd output tube would be gone if that were mine and the sockets and circuit parameters if necessary rewired to accept a tube that is commonly available.

Having to rely on only one remaining manufacturer (Sovtek) in a brand called Harmonx Distortion or something (I forget, I've got a really bad headache right now) which has shown to be a totally poor quality, short lived tube in all the NIB examples I've bought so far that I won't use them again doesn't sound like an upgrade to ME. Oh, ElectroHarmonix. Well I was close.

Plunk in a set of real NOS 6CA7/EL34's from Mullard and enjoy the same set of valves for 15+ years. Tight center-to-center spacing of the sockets probably prevent being really adventurous as in perhaps KT66's or KT88's.

Install a real choke in that existing CRC HV power supply. At least Fisher didn't use a stupid voltage doubler like others did here. Sherwood comes first to mind.

Install a separate HV fuse to each o/p xfmer to save it in case someone puts EH tubes in the amp and isn't standing there watching when the tubes go into hot plate runaway.

Put a big knob on the tuner control shaft so you hand will intuitively be able to find it. Ditto for the volume knob.

Merry Christmas

[Giaime]

Sorry Casey for the pictures, on my pc they look good. Anyway they show that circuit under a Spice simulation (with all the voltages, currents etc etc) with the graphs of the response: and obviously the RIAA curve looks bad

Let's see if I can report them here:

Pic of the circuit (note: the second stage biasing has been changed from grid leak to LED bias)


The response of the circuit


Real RIAA response

As you can see there's something wrong...

[valveitude]

Hello rcavictam,

Clearly you have had som R-E-A-L-L-Y bad experience with EH. I can't comment one way or the other, as I have no experience (yet) with them. I can say though, that most everything I've read about the 7591 from them in the the Fisher has been very positive. In fact the Fisher "experts" I've read on the net claim the EH's sound better than nos!! Again, I have no personal experience, but even a blind pig can find a truffle once in a while
, maybe thats the case with the EH 7591. The fusing of the OPT's is a stellar idea though.
Nothing says I can't rewire for a different tube later if I have problems with reliability and availability. For the time being I want to keep to the tubes the OPT's were wound for.

Don't think I'm dismissing you out of hand, I'm not. I will be keeping a very close eye on the outputs.

Thanx for the heads up.

Hello again Giaime,

Don't tell anyone, but I'm posting from work . The local network filter is blocking the pictures. I'll check them out when I get home.

Thanx again,
Casey

[SY]

Just a few quick comments (I'm buried in a prototype at the moment):

Changing from 7591 to EL34 is not trivial. More filament current, higher drive requirement, not to mention form factor. I'd try the new 7591s.

You can find examples of the sort of input stage I'm talking about in Morgan Jones's "Valve Amplifiers." Check out the Bevois Valley project- that input stage could work very well in this application. While you're perusing that book, look up the Maida regulator circuit- that would be perfect for the screen supply.

There was a nice article about the Quad tilt control in Audio Amateur, then later in AudioXpress. The former was an article by Reg Williamson and Alan Watling; the latter was a tube adaption. I don't have the references handy, but if you can't find them, I'll dig them up.

[Tom Bavis]

JJ has a new 7591, same size as original. The EH are happier with a lower grid resistor (Fisher used 110% of the data sheet maximum...), and that may be true with the JJ as well. Wth 150-200K, grid drive is reduced further, but it should still be enough. Really should be a 7247 instead of a 12AX7, but those are hard to find...

Add 10 Ohm resistors as cathode fuses - gives you a point to check idle current too. As for regulating bias - DON'T unless you regulate the screen. As-is, the idle current driops as line voltage increases, since the control grid has more gain than the screen. Adjustable bias and balance is a good idea, if you can find space for the pots.

Install an inrush limiter (Keystone CL-70 or similar) in the primary and reduce the line fuse to 2.5A slo-blo. (too late in mine - line switch is already gone...) Add a 3-wire cord for safety, and delete the line-chassis caps.

800C Schematic on my page at http://www.audiophool.cjb.net/Audio.html

[Eli Duttman]

Casey,

SY made an IMPORTANT point about regulating screen grid B+. Stable g2 B+ is THE way to maximize open loop linearity in pentode mode

If you regulate the C- supply, you MUST also regulate the B+ too.

Replace the doubler diodes with a pair of Cree 1200 PIV Schottky parts. They have zero switching noise; so, the 10 nF. snubber on the rectifier winding can be disposed of. Acquire a pair of GE/Thermometrics NTC inrush current limiters from Jim McShane and install them in the lines from the rectifier winding to the rest of the B+ PSU. Both the power trafo and the O/P tubes are handled much more gently when the turn on surge is damped. Increase the size of the caps. in the doubler stack to 330 muF. That will stiffen the supply a little without increasing I^2R heating a lot.

Thorsten's phono circuit is out because it uses the 6922. Where is the 6.3 VDC for the 6922 heater going to come from? A DECENT and simple phono stage is the "classic" RCA passive equalization circuit. The weaknesses of the RCA design are poor drive capability and mediocre deep bass extension. With the phono section in the same "box" as the downstream circuitry, drive is not much of a concern. The deep bass extension can be improved by changing the 2nd gain block's biasing method. Ground the cathode and use a 20 MOhm Caddock part as the grid leak resistor.

Buffer the recording O/Ps with JFETs.

A possibility for the phase splitter/driver is a 12AT7 wired in a Schmitt circuit. 3X red LEDs would provide the necessary grid bias and a 6 mA. CCS is the path to ground. 47 KOhm load resistors would be used on the 'T7 anodes. The non-inverting I/P of the splitter provides a convenient point for the application of loop NFB.

Another possibility for the tone controls is the Baxendahl circuit.