|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
50.35.216.158
In Reply to: RE: Magnavox 9302 Rebuild posted by Fitero on April 04, 2014 at 22:18:10
Hi Fitero,
I also use Fulton loudspeakers (FMI100s topped by the Fulton/RTR electrostatic cubes) and usually drive them with an EL84-class amp that was originally built as a clone of the same Magnavox console amp you have. I'd like to make a few comments and suggestions about your very nice amp rebuild.
First, a couple of things I really like: the fit and finish and construction look really nice, especially the grounding scheme (large copper ground bus) and general layout. I also like that you are using a shared cathode resistor and bypass cap for each PP pair, as opposed to the original Magnavox scheme with a single resistor for all four tubes, or separate bypassed resistors for each individual tube. I tried all three cathode biasing schemes, and shared-pairs sounded best to me.
Second, one difference: I ditched the Magnavox paraphase front end for a direct-coupled concertina scheme, similar to the first two stages of a Williamson amp. I felt this was an improvement, but since you have already finished your build and like the result, I am not suggesting any changes there. I also use an ultralinear screen connection, but that is not an option with the original output transformers.
Third, some suggestions. Are those large can caps electrolytics or oils? If electrolytic, I suggest paralleling them with smaller-value film caps. Also do this with the other electrolytics in the amp, or even replace them completely with films since you seem to have enough space. The blue 35V, 22 µF film caps sold by diyhifisupply.com in Hong Kong should work well for the output bypass caps, since your bias voltage is only about 12V. Two per side should be enough to replace the electrolytics, or just one per side to supplement them.
Anyway, good luck and enjoy your amp!
-JJ
Follow Ups:
This is being worked on, lots of cool stuff to post !!
Patience please, patience.
Jeff Medwin
I have that website bookmarked, and visit every now and again, hoping to glean more info...
I am going to help Brian Beck get started on that. We have given him a TON of really great information to post. He is busy, and it looks to me like I better take the bull by the horns. I'm busy too, but we all loved RWF and I want to see it done, in honor of him.
Brian sent me his initial "History - write up" yesterday, and he is stuck on that. I can help. Give it some time and it should get populated.
Jeff Medwin
Thanks for the update Jeff.
I look forward to reading any little anecdote that you folks manage to post there. I'm not alone for sure.
JJ
I was wondering about that because even tho the 6EU7 which is a 12v 12AX7 is a great sounding tube,I'm not crazy about it as a driver.The direct-coupled concertina is a wonderful idea and I think Pete would like it as well.
What tube did you use? I actually like the Mullard pentode-triode combo even tho its implemented in a long tail pair but the DC coupled concertina is worth a shot..
Some change these over to the 12AT7 like Jim McShane suggests in single driver tube amps and that works very well indeed.
BTW,you dawg..I really miss my Fulton J's but I will gladly take those RTR6s off your hands..I have a pair as well but I needed an audio transformer for one of them..I finally found an NOS one for 30 dollars on Ebay...Are those incredible tweeters or what? Those and the Heil Air Motions should be reissued IMHO.
Honest amplification is better than excessive 2nd order distortion anytime.
Edits: 04/06/14
Mike,
Sorry, my RTR/FMI tweeters are not leaving my house until I either die or go deaf. The Fulton version only runs from 5K up but most instruments and voices image at the height of the tweeters.
The 100s run full-range since the lowered position (only few inches off floor) and limited treble dispersion of the Peerless mid-tweeters effectively low-passes the main speakers.
I use RCA 6CG7s in my PP amp for the front end. Not knowing any better, I just followed an old Williamson schematic (I will try to find it for Fitero) with 47K plate and 470R cathode resistor on the first stage, direct coupled first-stage plate to second-stage grid, 22K above and below the second stage. Then second-stage plate and cathode cap-coupled to output tube grids (no third driver stage, unlike Williamson) with 0.1 µF Dynamicaps, 220K to ground. It sounded very nice to me so I have not altered the topology since then (about 2 years ago.) No doubt it could be tweaked more. I only use 4-6 dB of feedback so it is enough gain.
As Mikey said, I would be interested in perfecting what I have built....
However, when you said "a direct-coupled concertina scheme" I looked at that like cows look at trains.....
If it is in schematic form, then I can read it, and possibly build it, otherwise I am lost, and Mike gets badgered with my pestering calls and messages.
Fitero,
You can see an example at
http://www.preservationsound.com/wp-content/uploads/2012/01/WilliamsonAmp_1947.jpg
V1 and V2 are two sections of a 6CG7 or 6SN7 twin triode (the 6CG7 will fit the noval sockets in your amp but has a completely different pinout from the 6EU7.) As mentioned in my answer to Mike's post, I used 0.1 µF caps to couple V2 to the output tubes (stage V3/V4 in the schematic deleted). The very small-value coupling caps used in the Magnavox schematic will roll off your bass badly if you reduce or eliminate the feedback.
One difference from the schematic I reference: I replaced the parallel power supply decoupling filters R2C1, R6C2 with a single filter with 3.3K, 22 µF feeding both first and second stage plates. I did not have room for two filters, and the lower power supply voltage in an EL84 amp (and higher current drawn by a 6CG7 relative to a 6EU7) mandates a smaller resistance. There is some room
for variation in the other resistances, but R3, R4, R5 and R7 should all scale by about the same proportion if you want to change them. R5 and R7 need to be high-precision or closely matched.
Good Luck,
JJ
Thank You for posting the link and taking the time to explain the pertinent info and deviations JJ. Food for thought.
Lets make your next P-P amp even simpler. Two stage, direct coupled, differential front end.On my "master" schematic, eliminate C-12, R-9, C11 and short-out R-8.
On supply, L1 and L2 is 21 Ohms or less, under two pounds. C1 is 20 uF and C2 is 50 uF.
V2 is a 12AX7 at 1 mA. of current, each section, and over 150 VDC Ea.
I heard this in a small room, with 6AH4GT Finals, and it floored me on FMI-80s.
Jeff Medwin
Edits: 04/05/14 04/06/14
Thank you for posting the schematic Jeff.
The funds aren't available at the moment to embark on another amp build.
I would like to perfect what I have built though.
Any thoughts or suggestions would be welcome.
This is the schematic that I followed...
For what it is, thats not too bad. You can't start to mess with it too much, Its sort of a proven balanced design. Leave it alone.
Some of the things I don't care for is the paraphrase ( out of time, after the fact ) phase splitter, the series cap-couple to the grids of the finals, the need for negative feedback loop, and the extra parts.
My suggestion is simpler, cleaner, on-time, zero feedback loop. My suggestion is less power, but more music. Substitute inexpensive 6AQ5s for 6BQ5s in my suggestion, and a 12AX7 front end tube. BUT, you would need a higher VAC power transformer, to get higher voltages. A whole new amp. Enjoy yours. I think you should be fine with what you have.
Jeff Medwin
Thank you for the kind comments and suggestions.
The three large caps, are oil, motor-run types, which I was told "punch above their weight".
In retrospect, I would have ordered a much larger value for the second capacitor, but the parts were already on order. I followed the M&S Mods schematic that I found.
I would like to swap the output transformers for Z565 versions when I get back from vacation next month.
Are the values of the output bypass capacitors circuit parameter/voltage based, or bypass frequency based? I don't know anything about such things obviously, and rely completely upon others' knowledge here. The original circuit had 20uf caps in this position. Mine had 100uf when I bought it, and I placed 220uf caps as per the mod schematic.
One question I do have, is regarding the driver tube plate voltages.
Specs call for 130V and 160V. I measured 185V and 190V before adjusting power supply resistance. Since I used parallel values of coupling capacitors, I assume the resulting ESR dropped by half, resulted in higher plate voltages on the driver tubes. I had to use 55K ohms resistance in the power supply to drop the voltages down to 160V, and I swear it doesn't sound as good as it did last night with the higher voltage.
Should I try to peg the voltage to the lower level, even though the 6EU7 can take 300V here, or am I missing something obvious.
Your oil cap power supply will do fine just as it is. The original console amp had much less capacitance, and just cheap electrolytics.
The Z565 output transformers should be a nice improvement and will permit you to experiment with UL operation of the output tubes, as well as reduced feedback (since those Dynaco reproduction OPTs will have better bandwidth than the Magnavox originals, they should need less or even no NFB.)
As far as I know there is no exact formula for the cathode bypass capacitance, other than that it should be large enough not to roll off the bass unacceptably. You can experiment and listen.
I don't know much about how to optimize the paraphase gain-splitter circuit. The original Magnavox parts values and operating points may reflect an effort to tailor the amplifier to the original console speaker drivers and enclosure. As long as you do not exceed the tube safety margins, I think you should just follow your ears.
Good luck,
JJ
" I also like that you are using a shared cathode resistor and bypass cap for each PP pair, as opposed to the original Magnavox scheme with a single resistor for all four tubes, or separate bypassed resistors for each individual tube. I tried all three cathode biasing schemes, and shared-pairs sounded best to me."
FWIW, I design self biased setups with RC networks shared by a PP pair. The better sound you hear may be coming from compensated differences in the individual tubes of a pair. JMO, economical and good sounding, as a combination, is HARD to argue with.
Eli D.
Eli,
As I understand it, the shared bias resistor PP pair requires reasonable tube matching, at least for idle current, so there should not be a big difference to begin with. Perhaps you mean that it helps average over differences in transconductance?
I think a lot of the benefit of sharing the bias resistor and cathode bypass cap between tubes is that you don't really hear the cap most of the time, as long as the class AB channel is running effectively as in class A, since then the current through the resistor is essentially constant (DC) and the cap is not doing much. It only charges and discharges on peaks, where the current has a large audio frequency component. With four separate caps, you would hear them more. OTOH there is little argument for using only a single cap and resistor for both channels, except as an extreme economy measure.
regards,
JJ
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: