Avery's "First" Maker: The TA-800

Was there very many value changes of the new components to reach this point, or did you get there on paper first? Pretty amazing stuff Dave, definitely the master at work here. :bowdown:
 
Thanks Steve. It was completely a paper effort at first for the new configuration, and for 10 of the 13 components used in its design. That circuit was then tacked in place, and the remaining three component values determined. Then it was all yanked out and properly installed and retested to ensure the desired outcome was achieved. In the end, there was only 1 component (a 25uF, 25 volt electrolytic cap) that is common to both the original and modified designs. It performs the same function in both circuits, but connects differently at the tube socket to accommodate the new tube. Since it was one of the new components installed in the original restoration, it could be reused again in the new design. And, since the points it connects to are closer together in the new design, it could be reinstalled without having to extend its previously cut-to-length leads. Of the other 12 components used in the new driver circuit, while many share similar functions with those of the original design, their component values are completely different.

Dave
 
Son, YA DID GOOD! On the separate graphs I think I see just a bare hint of of overshoot on Channel "A" compared to "B" but "B" has a hint more ripple. Am I seeing things? Actually on the overlay, they look so close that at 1st glance you could mistake them for being the same side. CAJOLE Steve into bringing up his TA-600 and you can A/B them when finished to see how well it actually does by ear. :naughty::D
 
Larry -- Your observations are correct. You would be hard pressed to find any stereo amplifier that produces square waves from each channel that are exactly identical. The two wave forms generated are so close however that if they had been produced by channels of identical design from the same stereo amplifier, nobody would bat an eyelash at it. To have one from the original design and one from the modified design get lost in each other when overlaid together then shows just how closely the two designs mimic each other. I plan to get it into the listening room later on this evening so as to hear the two channels, because Chanel B is still completely stock.

Gadget -- By topology, they are nearly identical, with the biggest difference being that the 800 includes an AC Balance control in the design, whereas the 600 has none. The values used to establish the basic operating points of the two stages are nearly identical as well -- which is all as it should be, since one of the goals was to have the modified design come off as Fisher oriented as possible. With the drive requirements between 7189 and 7591 class tubes being so similar, there's simply no reason not to use the same basic concept. It is in the feedback and HF stability arrangements that the two circuits differ, with those elements always being developed around the specific OPT used. Since the transformers are different between the two models, it means that those circuits will be different as well. Fisher changed the output transformer between the early and late versions of the 600, and of course changed from a tri-pent to a dual triode driver tube at the same time as well. When that happened, their approach to achieving HF stability also changed significantly -- but because of the transformer change, it's hard to tell how much of the HF stability changes were due to the transformer change, and how much was due to the driver tube change. So it was in that area where I was really starting from scratch, and therefore where all the development work was centered -- which I pretty well knew was going to be the case from the start. Had Fisher actually implemented a change to a 7247 driver tube with the same OPT in the 800, they may not have stabilized the design exactly as I did, but for all intents and purposes, the results would almost certainly be the same, since those characteristics were established in the 7199 version, which I mimicked in the modified design. I'll present the schematic of the modified design when the project wraps up.

Dave
 
yeah I figured the feedback stuff would come out different because of the transformers. I was mostly just curious how close in design and operating points they were. Is the AC balance implemented like the later receivers, a variable resistor in the plate circuit of the inverter?

Didn't realize the early and late TA-600 uses a different output transformer.
 
In the TA-800s original driver circuit, the AC Balance control was fit into the cathode circuit of the Phase Inverter stage. I elected to keep it there in an effort to minimize as much of the re-wiring as possible.....

Dave
 
The new power amplifier and modified tone control circuits in Channel A played completely uneventfully last night, along side the completely stock Channel B circuits. With the tone and Balance controls centered, the sound was balanced and presented equally from both channels, with a well focused sound stage and balanced timbre. I firmly believe that for anyone familiar with the sound of the 800, asking them to blindly identify which channel was stock and which was modified, would be purely a crap shoot guess. So the amplitude (i.e. sensitivity) and timbre of the modified channel presents for all intents and purposes identically to the stock one. However, there is one notable area where the channels can immediately be identified now as to which is which.

In a block diagram, for a 250 mV signal presented to the high level inputs of the stock TA-800, it can be shown that the tone control amplifiers amplify this signal by a factor of X1.6 which is then presented to the volume control. At its full setting, the volume control presents this (now) .40 vac signal to the power amplifier section, which then develops full power output.

In the modified TA, the same 250 mV signal is now amplified by a factor of X4.0 in the tone control amplifiers, which through a full setting volume control, now presents a 1.0 vac signal to the new power amplifiers, which then develop full power as before. In other words, the new design amplifies the signal 250% more in the tone control amplifiers, to make up for a 250% loss in gain in the power amplifier section. Because the net gain of the unit is identical in both cases however, this is why the two channels can play side by side each other within the unit, with no one knowing the better, even though they are quite different. However, what this also shows is that assuming a full power 8Ω output generates 14 volts at the TA's speaker terminals, it means that the gain of the stock power amplifier section is X35, where as the gain of the modified power amplifier section is now just X14. Because the power amplifiers appear after the volume control, this change has huge implications on how quiet the set is when turned fully down.

The stock TA -- even with a properly quiet 7199 -- is notably noisier at minimum volume settings than its later brethren is when using sensitive speakers, because of the higher gain the power amplifier section has in stock form. In piling on by adding insult to injury, locating these sensitive power amplifier driver tubes right next to the power supply section invites this higher gain design to produce more general noise, in addition to the extra hiss generated from the higher amplification displayed. With sensitive speakers, this can rather easily be heard in a quiet listening room from a not-so-close listening position. In other words, you can hear that it's on. Nothing to go postal over mind you, but very obvious none the less. However, if you add any kind of extra unusual noises from the 7199s, it gets to be intolerable real quick.

With the new design, the gain in the power amplifier section is notably reduced, while the gain in the tone control stages -- located well away from the power supply area -- is appropriately increased. As a result, now at a minimum volume setting, you don't even know that the unit is turned on. For all intents and purposes then, it is dead quiet now (in Channel A anyway). The total noise from the overall amplification process is the same in both designs. But in the new design, more of it is lost within the un-attenuated signals passing through the tone control amplifiers, making for a notably much quieter minimum volume setting noise level. It simply comes off as more refined.

7199 EPILOG

No doubt some may say, "what's all the fuss about converting the power amplifier section to using 7247 tubes? The 7199s in my TA are quiet." I'm sure they are. I have a precious few that are commendably quiet in my equipment as well. Therefore, if you've got quiet tubes, and they're good, then I have no doubt that all of this may in fact come off as much ado about nothing -- which is perfectly fine.

But that wasn't the case for Rob's unit. Both of the original Fisher 7199s in his unit were very bad, and more specifically to the point, it is the effort to obtain quiet 7199s today -- where you didn't have any before -- that presents such an unworkable situation -- which will only get worse in the future. All three 7199s that were the result of this effort were hardly blind internet buys. They were all obtained from quality vendors -- two of which were "JAN" tubes from Jim McShane, and one which was a NOS NIB Sylvania from Brendan at Tube World. Let me stress again, that all of the problems that these tubes displayed are the type that would not be caught by convention tube testing, whether pretested, burned in, you name it. The problems displayed are either due to the way they are used in circuit (direct coupling), or produce significant hum from the slightest H/K leakage. For vendors to cull out these bad examples -- on top of those that fail conventional test procedures -- would require the vendors to additionally plug the tubes into working circuits and monitor them, which either they don't have the time to do, or would send the price even higher. To the credit of both these vendors, in spite of both returned tubes being new and certainly testing as "Good", they immediately took them back, and gave full credit for them. I have dealt with Jim for years who imo is excellent, and it was he who recommended Brendan as a quality vendor when he couldn't replace the one I returned. Even though this was my first experience with Brendan, he took back the Sylvania 7199 no questions asked, for a full refund including shipping, and was then able to supply the two excellent GE 7247s the set will now employ. Cudos to both of these quality/class vendors for standing behind what they sell.

The final nail in the coffin of this story however is the third and last 7199 tube of this saga. It too is now starting to display intermittent noises -- periods of increased hiss and/or static, that can always be relieved for a period of time by tapping on it. It's hard to call it a new tube anymore since it was used for all the testing after the initial restoration was done, and then through out all the developmental work on the new driver circuit. But whatever slimmer of hope I had in finding any of these tubes in truly good, usable condition is now is completely gone. I have zero confidence in any remaining stock out there, as this project is 0 for 3 with these tubes. By comparison, with the new driver circuit design, the unit turns on with no muss no fuss to inaudibility, and out of the black background comes music -- exactly as Avery would have wanted it to. In today's audio environment then, unless you've got a stash of known good, quiet, tubes, conversion to the 7247 driver design will keep these particular Fishers performing -- and acting -- like a Fisher should.

Dave

One of the new driver tubes installed, up and running, makes it one down, and one to go......
SAM_2068.JPG
 
Nice work Dave, as always.

Did reworking the tone amp section to make up for the reduced power amp gain reduce the amount of cut/boost in the tone controls, or were you able to retain the stock range by altering the amount of feedback in the stage?
 
Thanks Gadget -- Because the tone controls are of a completely passive design, releasing some extra gain had no effect on their operation. In essence, the tone control section is made up of two halves of a 12AX7, each with its own plate-to-grid NFB loop, with the passive tone controls sandwiched in between the two stages. As did Fisher with the later version 600, I simply went to the first tone control amplifier stage and modified the NFB loop to obtain the extra gain needed. After the adjustment, the first stage is still operating with 12-13 db of NFB, which is plenty to linearize the stage and ensure a uniform response across the audio bandwidth.

Dave
 
12-13 vs I think you said 20 in the stock design?

Good deal. Nice to know it doesn't cause it to change how it performs in stock form.
 
Dave for what it's worth, I saw a NIB never used 7199 today on the auction site, Tubesforless is the vendor. No affiliation, hope this might help.... Al
 
Dave for what it's worth, I saw a NIB never used 7199 today on the auction site, Tubesforless is the vendor. No affiliation, hope this might help.... Al
I have one like that hiding here somewhere, almost afraid to use it in anything since it doesn't mean it's any good.
 
That's the thing -- if you've got one that you've had hidden away for a long time, there's frankly a better chance that it will be good than the ones for sale today. I honestly think that the ones out there today are just the throw backs. I'm not even saying its done intentionally by vendors, either. A vendor gets one back as a reject, maybe he retires and his entire stock is bought off by another vendor and there ya go, the tube is back out there. Of course, then there's the auction site, where all bets are off.........

Dave
 
Both of the new 7247 driver stages are installed now, with only the 1st stage of the Channel B tone control amplifier left to modify to finish up the whole driver tube issue. Once that's done, some testing to make sure all is well, and then (again), more listening tests for confirmation. A break over the weekend, and then the final lap of the project begins with the installation of EFB, which will start the first of next week.

Dave

SAM_2070.JPG
 
Looks amazing Dave! Definitely clean and tidy as well. This will keep this 800 in service for many years. Between this and the EFB, hopefully we can bring those specs up a bit. Less distortion and heat plus more power! Talk about a Fisher Hot-Rod! Thanks for all of your hard work!
 
Back
Top Bottom