|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
108.73.71.36
In Reply to: RE: Pentode vs. Ultra Linear posted by Freo-1 on April 19, 2015 at 10:54:06
" He explained that when the plate pulls the top of the output transformer winding towards ground, the ultra-linear tap pushes the screen grid so low that it renders the tube unable to drive difficult loads...."
Winding an OPT with the tap in a reasonable spot takes care of that on amps with HP-race B+ levels. It is a matter of tube choice and OPT design...IOW, U-L amp building 101...which I am pretty sure he did well in, which means what coming from a good designer? surely not, "I screwed up the design"?
It is easier to crank up the g2 voltage on a pentode so the loadline stays where the plate curves are horizontal no matter what the speakers do.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Follow Ups:
...I first ran across it as Bob Carver was promoting his mods for the Cit II that would enable it to produce something like 100-120WPC. I think it involved conversion to pentode mode and addition of g1 clamps on finals allowing deep class AB operation. I've got details stored away somewhere but am not able to locate them at the moment. Bob soon moved on to mfg of limited edition tube amps employing these same basic concepts. Interesting coincidence?
Regardless of exact details, the bottom line is that this story comes from Carver as he was in self promoting mode so how much is fact and how much is embellishment is anybody's guess.
Phase Linear was out of Seattle. Bob Carver has always been known as a real character. Very smart though.
CITATION II MOD
It's easy to almost double the power of the Citation amp. It takes four things -
one, some new Tung-Sol KT120's. Two, a pair of 1N4148 fast small signal diodes.
Three, a dymo label maker to relabel the output transformer taps. Four, some 390
uf / 450 volt capacitors.
Here's how it works. The new KT120's are rated at 150 watts of audio output per
pair, and by installing them in place of the 6550's or KT88's we have the
potential of powers well beyond 60 watts, the Citation's official rating.
However, unless we can change the output transformer's impedance to match the
new tubes, we will not get any of this new power. Step one is to install the new
tubes. Step two is to install the two diodes and a resistor configured as the DC
restorer. This step is essential as it substantially reduces the load on the
power transformer allowing it to deliver the extra power without getting too
hot. In step three we use the dymo labeler to relabel the speaker taps - 16 ohms
becomes eight ohms, eight ohms becomes four ohms and so on. By doing this we
effectively change the turns ratio by a factor of 1.414, and since 1.414 squared
= 2, we can double the output power this way. For a variety of technical reasons
it does not quite double, but it does yield approximately 115 watts. Step four
requires paralleling the new capacitors with the existing filter caps to
increase the energy storage by about ten times. This helps the output power as
well as the DC bounce (that Citations tend to have on transients). This more
than helps the bounce, it eliminates it.
All four of these things must be done at once, otherwise the increased
performance will not fully materialize. The only thing not absolutely necessary
is the Dymo machine, in which case it will be necessary to remember which
speaker tap is which.
=========
The Stu Hegeman mod for the Citation II
Performs three important functions:
One, it adds energy storage to the power supply, increasing it by more than a
full order of magnitude.
Two, it incorporates the DC restorer, reducing distortion substantially as well
as dissipation in the plate circuit of each output tube from about 50 watts to
about 9 watts. It adds filtering to the low level stages, thereby eliminating
"DC bounce" in the front-end voltage amplifier.
Three, it replaces the ultra-linear output stage with a pure pentode one,
yielding substantially better drive for all (and especially difficult)
speakers. Compensation is changed to take full advantage of the new output stage.
The following should be made to both channels. See Carvermod.gif
Add a 200 µf/450 volt cap across C4 A and C4 B. Add a 390 µf/450 volt cap
across C3 in the power supply. Add a 1N4004 silicon diode across the existing
selenium rectifier M5. Delete C16, R35, R26, C15. Cut and remove the ultra-
linear wires that go to R27 and R34 of the output tubes. Add a network
consisting of a 600 ohm resistor and .0027 µf cap across R12. Add a network
consisting of a 43 K resistor across R36. Add a network consisting of a .012 µf
and a 300 ohm resistor across R9. connect the 270 ohm screen resistors to the
center tap of the output transformer transformer (455 volt source). Add the DC
restorer network of 91 K and two 1N4148 diodes from pin 5 of the output tubes
to -55 volts. Parallel a 300 K resistor across R24 and R31. Add a 1N4004 silicon
diode across the bias meter, and a shunt resistor of 4.7 ohms across R72. This
completes the modification.
Set the idle current such that a voltage reading of 0.32 is measured across R28
and R38, the 15 ohm cathode resistors in the output stage. Make a permanent mark
on the built-in bias meter for ease in future adjustment.
The electrolytic cap values do not have to be exactly 200 µf and 390 µf - they
can be any close or convenient value - not critical at all.
===========
Some of the false beliefs that have sprung up around these mods. It is
understandable as these mods may fly in the face of generally acceptable
methods. First the notion that by raising the power and installing Kt120's, the
power transformer may be overloaded. It actually will run much COOLER and will
regulate better. Here's why: The filaments on the KT120's draw about 13 watts
instead of ten, or 12 watts more total, and it makes the power transformer
hotter, that's bad. However the output tubes are idling at about 9 watts instead
of close or at 50 watts each, and that represents a savings of 164 watts,
allowing the transformer to run cooler. That's good. We add 12 watts and
subtract 164 watts - a good trade. As for maximum output with music; we find the
peak-to-average power ratio of music is on average about eight to one - so
delivering 120 watts rms means the heat associated with it is only 120 / 8 = 15
watts. For 60 watts we have 60 /8 = 7.5 watts. That is a total of 15 - 7.5 = 7.5
watts. So...........these mods reduce the total dissipation by +12 - 164 +15 -
7.5 = 144 watts. That is a lot, and in fact the transformer now runs relatively
cool where it used to run so hot it was hard to touch. Now it's just a nice
hand-warmer. With the added efficiency, we find that the filament voltage is
6.34 volts rms and the B+ regulation is 96% whereas it used to be 64%. Cooler
and better!
I saw a simulation of the DC restorer in operation that had been posted, and it
accurately illustrated how the DC restorer worked. Then the author said "...but
it may cause distortion on the negative drive signal swing." - or words to that
effect. A false belief. Consider that a class AB output stage drives positive
and negative and each tube delivers only half a sine wave - 67 % distortion,
pure even THD, lots of it, and the transformer combines those two haversines to
yield a full and undistorted one. Since the simulation accurately showed
distortion on the negative grid drive, we may think that it will somehow show up
at the output, but it CANNOT because the tube is normally cut off during that
time and not conducting at all!
The proof is that the measured THD as well as IM distortion is about one third
of normal WITH the DC restorer than without it. And at all levels! Substantially
lower!
Nobody ever went broke underestimating the intelligence of the American public
H. L. Mencken
Mike Samra: As you know that is not the way the tube is being used.....They are putting the KT-120 in there Audio Research amps and imaged that they have doubled the power............I have never heard of anyone making that modification before installing the KT-120.......People don't read between the lines.......Will
Don't recall Carver modding Cit II's. His original line of tube amps were the Silver 7's, IIRC, 200 watt monoblocks. The original prototypes used the huge Dyna A-440 (?) outputs with tertiary windings. These were the transformers used in the Mk VI which never really made it to the market.
IIRC, Victor Goldstein, brought a pair to Seacliff and HP simply loved them, and eventually convinced Carver to produce them. Of course, the transformers were no longer in production and definitely not available for production runs, so Carver built his own. The Dynas were rated for only about 120 watts so the originals used a pair per side , paralleled. Production run amps had one huge output....not quite the same.
He did come out with a smaller amp later, though, perhaps that was what you were thinking of?
When did Carver do this work? I remember him as an innovator with solid state, but not tube products (back in the day). I did see him selling a small amp loaded with Chinese parts on eBay about a year ago. If I recall correctly, it sold for much more than I would have ever thought it was worth.
--------------------------
Buy Chinese. Bury freedom.
...the Carver tube stuff of that era while serious efforts always struck me as a marketing gimmick intended to prove his sand products were every bit as worthy of audiophile appreciation as the newly resurgent tube designs from ARC and CJ etc. In interviews, Carver also came off as considering tube design a trivial task compared to good SS design.
Thanks, I was thinking it was probably an era that came after his SS heyday. Carver was practically a genius with solid state and had quite a few innovative patents to his name. I've never associated him with vacuum tube design, though. Funny that he would dismiss the challenges of tube design, but maybe I agree. It's easier to create a high quality amplifier when you start with better basic components. :)
--------------------------
Buy Chinese. Bury freedom.
the Silver 7 prototypes used paralleled pair of A-451 outputs.
the Mk VI's used a model specific transformer that was never
offered as a catalog item. And it's primary impedance was
nomially 1650 ohms CT. As opposed to the A451's nim pri impedance
of 2200 ohms CT. The A-440 (and A-441) had a nom pri impedance of
4300 ohms CT and it is the 440 series that had the tertiary winding...
There was another Dyna output with a tertiary winding... though it never
went into production... the original tranneys developed for prototyping
the ST-70 amp had an ouput model A-460--- which had a 4300 ohm pri CT with separate screen windings.
MSL
Builder of MagneQuest & Peerless transformers since 1989
but the transformer sizes were similar hence my thinking they were 440's.
However if they indeed were custom and used only for the Mk VI and used a UL design, it would explain why they went ballistic so often . GE 8417's screens can not take more than 400volts continuously: oh they my work for a while but eventually they go up in smoke
I'm getting senile.I have what I thought are 441's with supposedly tertiary windings but IIRC the seller told me the primaries had a lower impedance of about 1650. Nice light blue potting cases, but no other id. Luckily I bought them before the Silver 7's came out.
Thanks for the info and crash course in dyna iron.
Funny that Dyna came out with those tertiary winding iron so late. I thought the whole idea of UL was to cut manufacturing costs by eliminating the fixed screen circuit as in Altec 340's, 350's and Allen organ amps which regulator tubes like OD-3's and such to power the screens.
UL transformers certainly paid off for Keroes and Hafler.
Edits: 04/20/15
That tertiary winding is also quite useful for CFB topology. That, with a fixed screen supply gives the same effective g2 U-L but with the CFB added to the mix. You can bring g2 to the cathode( with the extra winding attached to the g2's, and cathode grounded), or bring the cathode to the screen( using the extra winding tied to the cathodes and its CT grounded).
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Carver wasn't selling modded Cit IIs, he was promoting his latest tube amp design concepts that included the g1 clamp, true pentode operation and deep AB biasing. He kinda demonstrated these concepts by marking up the Cit II schematic with them and slyly inviting DIYers to have at it. This was around 2011-2013. At the same time he was offering very limited edition, hand built custom amps utilizing these same concepts on eBay of all places. Not much later he began producing a range of tube amps under the Carver brand with names like "Bob Carver Cherry", Bob Carver Black Magic" and so on. These amps appeared in the 2013 Music Direct catalog among other places. No relationship to the much earlier Silver Seven flagship tube amp that became the sonic basis/reference of his "Transfer Function" marketing campaign for a series of sand amps.
The Silver 7's were designed to show he could design a good tube amp and were a product of his assertion that by altering the "transfer" function, he could make his solid state amps sound like any competing amp available. Didn't work apparently....
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: