I've been considering building the "Einstein" push pull 6B4G amplifier that is on Tubecad here:
The real Einstein amplifier
I have transformers on hand that would be perfect for the project.
I did have a question - would the input and driver section of the amplifier have enough oomph to properly drive a push pull pair of 300B tubes (with higher B+ to the 300B's of course)?
Can anyone help a guy out? Opinions?
The real Einstein amplifier
I have transformers on hand that would be perfect for the project.
I did have a question - would the input and driver section of the amplifier have enough oomph to properly drive a push pull pair of 300B tubes (with higher B+ to the 300B's of course)?
Can anyone help a guy out? Opinions?
Hi,
If you going to pay through the nose for 4 300B's,
simply build an amplifier designed for them.
rgds, sreten.
If you going to pay through the nose for 4 300B's,
simply build an amplifier designed for them.
rgds, sreten.
Fair enough but there aren't many designs out there for push pull 300B's. 300B single ended amps a-plenty but not many push pull.
Anyone else care to look at the schematic and offer an opinion yea or nay?
Anyone else care to look at the schematic and offer an opinion yea or nay?
Lynn Olsen's "Amity" may be what the Dr. ordered. Expensive? Yes, but unlike far too many offerings, "Amity" may be a genuine value.
FWIW, I think the 300B generates excessive 2nd order HD. As Harvey Rosenberg said, the 300B puts on a show. As a push/pull setup cancels internally generated even order harmonic distortion products, 300Bs in P/P should be very nice.
FWIW, I think the 300B generates excessive 2nd order HD. As Harvey Rosenberg said, the 300B puts on a show. As a push/pull setup cancels internally generated even order harmonic distortion products, 300Bs in P/P should be very nice.
Seeing as how the 6B4G and 300B both have the same gm at 60 mA, and the near same Mu, it shouldn't be too hard to get the "Einstein" circuit to work for 300B tubes. More B+ will be needed for the extra Watts, and the plate load and bias resistors for the up front tubes will need a little work to stay within tube and R Watt ratings, plus provide another 40% signal gain. Better biasing circuitry for the 300B needed.
That "Einstein" design is really just the bog standard approach. Broskie's little mod to the splitter should be good to go, no need to put another time constant in the thing like it was setup. Without any global feedback, should be a breeze. It may not sound exactly like a gold plated, silver wired, dark energy powered ......audio elitist amp without all their expensive components. A good, well filtered, power supply will make a big difference. A good OT always recommended, but class A operation puts less demands on the OT.
Put in some TV Sweep tube outputs, in triode configuration, and you probably won't hear much difference. Just 50X cheaper.
That "Einstein" design is really just the bog standard approach. Broskie's little mod to the splitter should be good to go, no need to put another time constant in the thing like it was setup. Without any global feedback, should be a breeze. It may not sound exactly like a gold plated, silver wired, dark energy powered ......audio elitist amp without all their expensive components. A good, well filtered, power supply will make a big difference. A good OT always recommended, but class A operation puts less demands on the OT.
Put in some TV Sweep tube outputs, in triode configuration, and you probably won't hear much difference. Just 50X cheaper.
Last edited:
Thanks a lot for that.
The reason I'm asking about the 300B version is while I have the right transformers, chokes and output transformers for the 6B4G version and I plan to build it as on the Tubecad site I also have a pair of transformers and chokes with higher ratings (455-0-455 at 140mA) that looks like it would be perfect for a 300B version if the driver circuitry is up to the task.
The reason I'm asking about the 300B version is while I have the right transformers, chokes and output transformers for the 6B4G version and I plan to build it as on the Tubecad site I also have a pair of transformers and chokes with higher ratings (455-0-455 at 140mA) that looks like it would be perfect for a 300B version if the driver circuitry is up to the task.
I would check some of the 300B SE designs around and see if they are using the 6SN7 for a driver. If it works well there, then it will work for P-P well too. The 6SN7 is only rated 300 V B+, so it could be a little close on V specs. You can always switch to whatever tube they are using in the SE designs. Something with a Mu of around 30 to 35 would get you the extra gain needed, but probably better to put that in for the 1st input tube. Don't get in a hurry with punching holes in the chassis.
Last edited:
Eli, I've thought about building an Amity but I wanted to dip my toes in the water with something less expensive first. There are some push pull interstage transformers from Edcor that are really reasonable and it might be interesting to build a version of the Amity with less expensive iron.
I've also thought about running a 45 into a Hammond 124B as phase splitter/driver into a pair of 300B's. I thought of perhaps using a 6A6 as the input tube to drive the 45. I know I like the sound of a 45 - I'm currently running single ended 45 amp into a pair of Sachikos with FE206ESR drivers. I love the hell out of them.
I'm planning on putting the Fostex's into single horns, moving them into my listening room downstairs, running four FE208E sigma BIB's as subwoofers in the corners and putting corner horns into the upstairs setup. The 45 amp will go downstairs as well.
Hence the need for more amplifiers.
I've also thought about running a 45 into a Hammond 124B as phase splitter/driver into a pair of 300B's. I thought of perhaps using a 6A6 as the input tube to drive the 45. I know I like the sound of a 45 - I'm currently running single ended 45 amp into a pair of Sachikos with FE206ESR drivers. I love the hell out of them.
I'm planning on putting the Fostex's into single horns, moving them into my listening room downstairs, running four FE208E sigma BIB's as subwoofers in the corners and putting corner horns into the upstairs setup. The 45 amp will go downstairs as well.
Hence the need for more amplifiers.
That 6A6 has the higher Mu of 35 for using in the input stage position, but it doesn't look all that linear. I'm not seeing that many octal twin triodes or even single triodes with Mu around 30 to 35. There is always the 6SL7 at Mu 70, but it's still 300 V rated and rather high Rp. You may have to go with a 9 pinner.
6J6 is Mu 38, octal, but common cathode. Probably just parallel the sections. Still only 300 V rated.
6BL7 is only Mu 15, octal, but 500 V rated. Could use this for the driver stage, and put more gain into the 1st tube, like Mu 40 then.
6J6 is Mu 38, octal, but common cathode. Probably just parallel the sections. Still only 300 V rated.
6BL7 is only Mu 15, octal, but 500 V rated. Could use this for the driver stage, and put more gain into the 1st tube, like Mu 40 then.
Last edited:
Thomas Mayer did a test of the 6A6 here (a little ways down the page):
VinylSavor: 6A6
The tube curves appear to be surprisingly linear compared to the datasheet. He also recommends its use as a driver for the 45 and 2A3. I'm not religiously married to the idea of octals, however. It just seemed like a solution that would work without going the whole Amity route. 6A6 into a single 45 transformer coupled into Hammond 124B into push pull 300B would be quite a bit less expensive than spendy Lundahl iron.
I'm going to build the push pull 6B4G amp first as I have all the iron and a workable chassis all ready to go but I like to plan ahead.
VinylSavor: 6A6
The tube curves appear to be surprisingly linear compared to the datasheet. He also recommends its use as a driver for the 45 and 2A3. I'm not religiously married to the idea of octals, however. It just seemed like a solution that would work without going the whole Amity route. 6A6 into a single 45 transformer coupled into Hammond 124B into push pull 300B would be quite a bit less expensive than spendy Lundahl iron.
I'm going to build the push pull 6B4G amp first as I have all the iron and a workable chassis all ready to go but I like to plan ahead.
Woops, that 6J6 is not octal, its a 7 pin.
Anyway, could use something like 6J6 for the input tube, and 6BL7 for the driver.
The 6BL7 looks fairly linear on its datasheet. The lower Mu of 15 for the 6BL7 needs something like Mu 40 for the 1st input tube to get the drive level up for the 300B.
You could use one of the dissimilar triode Vertical amp tubes for the input and splitter. Like 6EA7 or 6EM7, 6GL7 (octals). Small triode Mu of 66 and large triode Mu only 5.5, but the gm is 6000 to 7200 for splitter use. Will be a little overkill for splitter use with a 10 Watt diss.
Anyway, could use something like 6J6 for the input tube, and 6BL7 for the driver.
The 6BL7 looks fairly linear on its datasheet. The lower Mu of 15 for the 6BL7 needs something like Mu 40 for the 1st input tube to get the drive level up for the 300B.
You could use one of the dissimilar triode Vertical amp tubes for the input and splitter. Like 6EA7 or 6EM7, 6GL7 (octals). Small triode Mu of 66 and large triode Mu only 5.5, but the gm is 6000 to 7200 for splitter use. Will be a little overkill for splitter use with a 10 Watt diss.
Last edited:
Oops, you're right. But it's a 6A6 not 6J6. But yes, the 6A6 is a seven pin.
I'll have to look into the 6BL7. Another thought I had was a 6V6 in triode mode.
I'll have to look into the 6BL7. Another thought I had was a 6V6 in triode mode.
You could use one of the dissimilar triode Vertical amp tubes for the input and splitter. Like 6EA7 or 6EM7, 6GL7 (octals). Small triode Mu of 66 and large triode Mu only 5.5, but the gm is 6000 to 7200 for splitter use. Will be a little overkill for splitter use with a 10 Watt diss. Hmmm, that low Mu of 5.5 may be a problem for a splitter, need to check on that.
Darn, it is a problem. Just use a Mosfet or 6SN7 for the splitter.
6V6 in triode has only a Mu of 10 and a somewhat lowish gm of 4000. It does have a 350 V rating.
Darn, it is a problem. Just use a Mosfet or 6SN7 for the splitter.
6V6 in triode has only a Mu of 10 and a somewhat lowish gm of 4000. It does have a 350 V rating.
Last edited:
Someone needs to simulate a splitter using a 5.5 Mu tube, it -could- be working I guess.
Just seems so counter-intuitive when it can only move the plate by 5.5 times the differential grid drive on top of the cathode swing. But 5.5X is still more than twice the bottom swing, which is the minimum to overcome it. That would mean the CF gain to the bottom side would be like 20% less than unity, which I'm not sure if that causes a linearity issue here.
Usually, the high gm is very helpful for linearity, but the upper plate V motion is reducing it considerably through the low Mu. That might actually mean the Concertina splitter is very linear here with that much internal neg. FDBK. Usually I think of lower CF gain as a problem sign. It's an interesting question in any case.
-----
OK, I guess it works alright. The plate Rp reduces at the same rate that gm increases with current, so it's a wash linearity-wise whether the NFdbk comes form the cathode or the plate (assuming complementary swings). I think I would still simulate this low Mu splitter before building it. Could be a 1st using a dissimilar Vertical amp tube this way.
Miller capacitance may be the real issue, that may be why low Mu isn't used for splitters.
Which puts us back to a small Mosfet being the best splitter. High gm, and bootstrapped gate capacitance in follower mode. Minimal drain/plate Miller capacitance.
Just seems so counter-intuitive when it can only move the plate by 5.5 times the differential grid drive on top of the cathode swing. But 5.5X is still more than twice the bottom swing, which is the minimum to overcome it. That would mean the CF gain to the bottom side would be like 20% less than unity, which I'm not sure if that causes a linearity issue here.
Usually, the high gm is very helpful for linearity, but the upper plate V motion is reducing it considerably through the low Mu. That might actually mean the Concertina splitter is very linear here with that much internal neg. FDBK. Usually I think of lower CF gain as a problem sign. It's an interesting question in any case.
-----
OK, I guess it works alright. The plate Rp reduces at the same rate that gm increases with current, so it's a wash linearity-wise whether the NFdbk comes form the cathode or the plate (assuming complementary swings). I think I would still simulate this low Mu splitter before building it. Could be a 1st using a dissimilar Vertical amp tube this way.
Miller capacitance may be the real issue, that may be why low Mu isn't used for splitters.
Which puts us back to a small Mosfet being the best splitter. High gm, and bootstrapped gate capacitance in follower mode. Minimal drain/plate Miller capacitance.
Last edited:
As pointed out already, linearity is good. Why not put that or 6N7 in the differential amp?
Sheldon
There is no differential stage in that amp.Cathodyne fasesplitter followed by an triode-amp for every fase.The two triodes have the same cathode resistance simply to avoid a decoupling capacitor.
And the Va max is 450V for the 6SN7 not 300V.Making it more useful for driving a 300A.
Mona
And the Va max is 450V for the 6SN7 not 300V.Making it more useful for driving a 300A.
Mona
"And the Va max is 450V for the 6SN7 not 300V.Making it more useful for driving a 300A."
OH, I didn't see the GTA/GTB had a higher rating. Then a 6SN7GTB driver for sure. Just need to up the input stage gain 50% or so then for the 300B outputs. 6A6 or equivalents should work then.
OH, I didn't see the GTA/GTB had a higher rating. Then a 6SN7GTB driver for sure. Just need to up the input stage gain 50% or so then for the 300B outputs. 6A6 or equivalents should work then.
Last edited:
How is not two triodes working antiphase with shared cathode not a differential stage?
If I were to adapt the Einstein amp to serious use, I would at least use bigger cathode bypass caps and coupling caps. The low end bandwidth is not very good.
CCS tails would be an obvious and significant performance improvement. Of course not applicable if this is a novelty build.
Thanks for the ideas guys. 🙂 I've learned about a couple of tubes I've never seen before.
Modeling the power transformer and choke in PSUD gives me a B+ voltage at the first cap after the choke of about 480V at the full 140mA circuit draw capacity.
I'm a little concerned that a push pull 300B amp might put too much strain on the vintage iron amperage wise.
I'm still intrigued by the idea of a 6A6 input into a 45 driving a Hammond 124B interstage/phase splitter into the two 300B's. I think that would put me well over the line amperage wise with this iron. 🙁
So is the consensus that the triodes in a 6SN7 could drive the 300B's with more B+ and a higher mu input tube?
Modeling the power transformer and choke in PSUD gives me a B+ voltage at the first cap after the choke of about 480V at the full 140mA circuit draw capacity.
I'm a little concerned that a push pull 300B amp might put too much strain on the vintage iron amperage wise.
I'm still intrigued by the idea of a 6A6 input into a 45 driving a Hammond 124B interstage/phase splitter into the two 300B's. I think that would put me well over the line amperage wise with this iron. 🙁
So is the consensus that the triodes in a 6SN7 could drive the 300B's with more B+ and a higher mu input tube?
- Status
- Not open for further replies.