New scratch build el84....

[jbus]

Getting ready to build a new power amp and trying to decide what circuit to use... Any thoughts or schematics you all like? I want to stick with6bq5s because that is what OTs I have...

An st35?

What do you guys like?

Josh

 

[gearhound]

RH84 amplifier

There are quite a few threads here on AK and other audio forums about that DIY amplifier.

Steve

 

[jbus]

Thanks Steve, but I forgot to mention I am planning for it to be a push-pull... I have an old mamaggie 8600 that might get converted over to the RH84 circuit though...:scratch2:

 

[vinylkid58]

What do you guys like?

I like simple cathode biased, concertina driven EL84's. With 12AT7's, you may not even need a pre, depends on the source.

I've never heard a bad sounding PP EL84 amp, regardless of topology.

jeff

 

[jbus]

I like simple cathode biased, concertina driven EL84's. With 12AT7's, you may not even need a pre, depends on the source.

I've never heard a bad sounding PP EL84 amp, regardless of topology.

jeff

Jeff I agree with never hearing a bad one... But, I don't know "concertina"... Any examples I could look at? Also, if doing a 12a*7 I would lean toward a ax7 because I have a ton of old stocks. Would that be a quality downgrade in your opinion?

 

[dcgillespie]

The concertina phase inverter circuit is also known by name as a "phase splitter", or "cathodyne" phase inverter circuit. It is the topology as used in the Dynaco SCA/ST35, Eico HF-81, and a bazillion Fisher circuits.

Dave

 

[vinylkid58]

I don't know "concertina"... Any examples I could look at? Also, if doing a 12a*7 I would lean toward a ax7 because I have a ton of old stocks.

Here's a good explanation of the "cathodyne" that coincidentally uses the 12AX7 as an example.

http://www.freewebs.com/valvewizard/cathodyne.html

jeff

 

[Shadowdog]

"concertina" examples are the Dynaco ST-70 , SCA-35/ST-35. Very popular used on later Scott 299's &222's. Sherwood & Eico HF-81 etc. and many others used it too. Concertina or Split Load which are actually derived from the Williamson.

The Paraphase was probably the first PP circuit and Magnavox used it extensively. The Floating Paraphase was a later version was. Scott first used it for example on 299 and 222 amps then switched to the Concertina. Pilot SA-232 &260 also used it. There are many that admire this circuit. There are Dynaco circuits floating around to use it for EL84/6V6 tubes from back in the day.

http://www.diyaudio.com/forums/tubes-valves/175995-have-anyone-built-6v6-amp.html

http://diyaudioprojects.com/Schemat...-6V6-6AQ5-Dynaco-A-410-Tube-Amp-Schematic.htm

The 3rd and final circuit that came out in the 50's is the popular Mullard , also called LTP or Differential. Many of the Dynaco ST-70 boards out there now use the Mullard . It is popular to add CCS to it now days'

http://www.diytube.com/phpBB2/viewtopic.php?t=4663

http://www.audiokarma.org/forums/showthread.php?t=447318

What do you have for OPT's and do they have UL?



Randy

 

[jbus]

Wow! Thanks guys! I clearly have some studying to do. But in a nut shell what is the advantage using that scheme? I would definately like to do it as a comparison with my KNight KN928 which I would assume is not...

I'm gonna try to put pencil to paper tonight and begin roughing (by roughing I mean roughing) in the schematic. I think I would like to use the st35 as a base and make it tube rectified and the driver an ax7.


Thanks for the education and feel free to keep it coming if this seems like a bad idea!

Josh

 

[Shadowdog]

I think I found the KN928 schem. Amp part anyway. Concertina or split phase. It has a balance pot like Scott used. Found the full one too.

 

[vinylkid58]

I think I found the KN928 schem.

Nice work.:thmbsp:

The KN928 output stage is virtually identical to the HK224, even the resistor and cap values are pretty much all the same. Same weensy OPT's too.:yes:

jeff

 

[BinaryMike]

I vote for using a pentode input stage, as in the Dynaco St-70, Mk-III, etc. I say this because I've recently gone through the exercise of designing a power amp with triode input stage and found that miller effect is a fundamental and severe limitation of this architecture. It might be a good thing if you need to roll off HF gain at 2~3KHz in order to avoid oscillation, but the pentode is much better in this respect. You might very well need a little HF gain reduction anyway, depending on transformer quality, but with the pentode, you have far more control over this parameter and you're able to minimize HF distortion by maintaining HF feedback near its midrange value.

 

[jaymanaa]

Are the transformers you have Z565s? If so, I used the DIY Tube board to build this little guy. I used a separate transformer for the power tube fils so I can run 10BQ5s while at the same time, unloading the already stressed out ST35 power TX. I think I've had more offers to buy this when people hear it than any of my other amps. The hot set up in it seems to be a miniwatt Mullard driver, and a RCA black 12AU7 splitter into black plate tungsol 10BQ5s. You get individual bias, and adjustable feedback with the DIY board. It's very thick and substantial and only 50 bux I think. I'm running UL right now, but have thought about sticking a triode mode switch on the back.

 

[Shadowdog]

Hi Jay,

Classy looking amp! Beautiful job. The only thing you may try on it is Dave Gillespie's EFB mod or the STD fixed bias with a 25V separate trans. if you don't want SS as it should beat even adjustable cathode bias.

Randy

 

[jbus]

Again wow! Thanks for all the effort!!

My Output trannies are Woodward Schumachers with I think an 8k... They are only center tapped. They are really beefy at least compared to any of my other amps... Probably about three times the size and weight of the Knights.

As for the Power Transformer I am not decided yet. I have several, a smallish 250v, a 360v from an organ pull and one I just got in an old Wilcox-Gay Recordio I haven't put the meter on yet, and a couple others also untested... They all have a tap for the rectifier and should have enough current for the filaments. (The organ I think powered around 40-50 tubes)...

I have a lot of reading to do guys. Thanks so much! I love karma! I will be out of town for a job through tomorrow so the schematic is on hold but I will keep researchin':thmbsp:

-Josh

 

[Kegger]

I vote for using a pentode input stage, as in the Dynaco St-70, Mk-III, etc. I say this because I've recently gone through the exercise of designing a power amp with triode input stage and found that miller effect is a fundamental and severe limitation of this architecture. It might be a good thing if you need to roll off HF gain at 2~3KHz in order to avoid oscillation, but the pentode is much better in this respect. You might very well need a little HF gain reduction anyway, depending on transformer quality, but with the pentode, you have far more control over this parameter and you're able to minimize HF distortion by maintaining HF feedback near its midrange value.

What kind of circuit are you trying to design Mike? I mean using what tubes?

As yes miller effect is severely limiting in high output z triodes trying to feed
into low z stages or used as drivers, but use a triode input stage with a nice
high transcon triode tube or keep the input impedance on the following stage
high enough and or not demanding of swing or drive and your fine.

That's where when testing you'll find tubes like the 12AX7 are quite hindered
and you need to keep them sandwiched between high impedance stages and
then they perform decently, just don't ask them to be an driver tube though.

For gain, and being able to sit before a moderately high impedance stage the
12AT7 does a pretty good job, otherwise intermediate gain tubes like an 6DJ8
or the 6BK7/6BQ7/6BZ7 are quite nice triodes with decent gain an drive.

It's the low current and high plate resistance that kills high freq performance
in a lot of triodes, unless there setup in an environment that isn't demanding.

 

[Shadowdog]

As to Mike's suggestion to use a Pentode front end I've seen that preference somewhere else.

Could you use a stock ST-70 board (Concertina like the stock ST-70) from Triode Electronics for instance as it has an in production Pentode EF86 to drive easy to drive tubes like the EL-84, 6V6, or 7591 tubes without modification? If you wanted a Mullard or LTP circuit maybe possibly the Eico HF-50/60 circuit could be used also if it's possible. Maybe you wouldn't need a preamp with all that drive, just add a volume control and input selector if needed.

Thanks!

Randy

 

[Kegger]

Now remember the pentode in those circuits is not driving the output tube,
it is driving the phase splitter which is driving the output tubes, the issues
with the Dynas ST70 circuit was limited driving capabilities from the triodes
concertina setup, it just wasn't a very beefy driver tube for the large EL34.

Drive an gain are 2 different things, some tubes may be good at one but it
may not be good at that other, an certain circuits can "help" them as well.

The new boards like the EF86 version use an different triode like the 12BH7
to do the phase splitter/driver duties and do a better job of it of driving the
larger output tubes much easier.

The original Dyna 70 driver setup would be just fine to drive 6V6, 7591, EL84.
(just drop some gain in the pentode section by using a smaller plate load)

 

[Shadowdog]

Thanks! But would the new Triode board work as well or better if you wanted to use in production affordable easy to get tubes, instead of 7199 etc? Or use the Eico HF-50/60 circuit for EL84/6V6/7591 if you wanted to do it P2P or want the Mullard circuit?

Thanks!

Randy

 

[Kegger]

Yes, in my opinion regardless of what you were driving the updated circuit on the
new EF86 version would be a better way to go then the original Dyna 70 circuit is.