|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
66.177.177.194
Don't know if any of you read my previous post about the ton of Hammond iron I aquired to experiment with. I also have some 2-300b's, 4-GM70's and 6bl7's along with some Rod Coleman boards for 300b and transmitting tubes.
I already have a really nice 2A3 I built 5 or 6 years ago and want to step it up and use some higher voltage with more output for another listening room I have. That being said, I have modeled several diagrams over the past 6 months. These are a culmination of my studies and experimentation. What do you think, here's 3 of them just to get started.
I have another half dozen or more, many with fixed bias but thought I would start out with keeping the power supply simple(well almost). I haven't adjusted all the caps yet on these diagrams....
The last one is just to step up the output since I have the giant Hammond 1642SE 75w OPT.
Has anyone tried something similar, maybe with a pentode driver? I do have some EL34's as well but don't know where to start with these.
I'm just an amateur so be easy.....
Follow Ups:
Seems to me, you'd be better off using a nice solid state amp if you need more power, for less efficient speakers, and do well just to sell all the miscellaneous parts on eBay.
The bigger the SE amp is, the more compromises you must make !!!
Jeff Medwin
I'd be leery about parallel O/P tubes, unless "fixed" or combination bias was employed. Give yourself the ability to adjust the individual "idle" currents and dodge the current hogging bullet.
Eli D.
besides, I live in Florida and would need to add another AC to cool the room.....
Started thinking about all the power supplies as well just to heat a quad set.
A FL summer should convince you to try the near Class "B" setup Jeff Yourison and I worked on. L_RD knows 6Y6s are cheap enough.
Eli D.
Nice work, that's about 200 less watts of plate dissipation than what I had in mind.
How do you like the pentode sound?
"How do you like the pentode sound?" PM Jeff, as he executed the concept. Everything I've seen from him, to date, is positive. Don't forget that I own a "Deuce" that I'm VERY fond of.
"Nice work, that's about 200 less watts of plate dissipation than what I had in mind." Near Class "B" operating points can be very useful in warm climates. Tweak the "idle" current to obtain a response "free" of notching/cross over distortion. A pair of near Class "B" GM70s would produce plenty of power. You could jack things up even more by employing some positive grid current action (B2).
Eli D.
"I'd be leery about parallel O/P tubes, unless "fixed" or combination bias was employed."
I wonder if this might be a good application for "enhanced" cathode bias. I'm referring to the use of an IC regulator to fix the DC voltage at the cathodes. It's very cheap to implement, probably less than $5-6 to bias both tubes individually, and no negative supply required.
--------------------------
Buy Chinese. Bury freedom.
Current, not voltage, is what needs to be set. Even carefully matched tubes will exhibit some differences.
When a bias supply and trim pot. are employed, the exact total bias voltage is not known. What is known, via Ohm's Law, is the cathode current.
IMO, combination bias is "best", when PSE O/P tubes are employed.
Eli D.
"Current, not voltage, is what needs to be set. Even carefully matched tubes will exhibit some differences."
Of course, but those differences will be very small if the tubes have indeed been matched, and they are not likely to be problematic in this configuration. I suggested this because the OP seemed averse to adding a negative bias supply, and this is probably the next best alternative. It's certainly better than biasing the tubes by means of fixed cathode resistors, even in the best-case scenario of separating the cathodes. Worst-case, with the cathodes tied directly together, voltage regulation at the common point will eliminate the current hogging that you mentioned.
--------------------------
Buy Chinese. Bury freedom.
Can I achieve this without a negative supply? I've seen combination bias on transistors but never tried modeling on a tube.
I tried a pentode but I have something wrong, I created some sort of notch filter below 20HZ which thru the coupling cap makes for a crappy F3.
I'm also limited with the Hammond 5K IT at 30ma
What Dave said.
And why do you want to run 1mA across the interstage primary? That's not helping anything.
And can the Hammond really handle 600V from primary to secondary?
Unless the goal is to by back bass from the LC resonance, I don't see the need for the cap since the transformer can handle the DC current. (depending on which unit it is). I had to lower the pentode plate load to 5K due to the limited inductance but the idea is clear.
Good point on the DC breakdown. The 126 series is bifilar which means bandwidth out the wazoo on the top end at the cost of lower potential breakdown voltage. I don't recall hammond ever giving a spec for their breakdown. In the referenced schematic, I'm not sure of the goal of the transformer attached to B+ but the 1ma of current may not be such a bad idea.
dave
You are losing your bass because the 44hy's of inductance is not enough for your 12K-ish source impedance. This is partially being masked by the CL resonant bump which is at around 40hz.
dave
I need to crack open that RCA handbook again, probably not a good idea using IT with that tube....
It would probably take some serious custom iron. I'll play around with the pentode more but I like the looks of
driving the GM70 with the 300B. Just need to find a tube to drive the 300B with about 40volts or so
(prefer a little headroom plus a hint of A2).
Hey,
A2 isn't an easy proposition and it isn't going to happen with that coupling cap in place. (for me grid chokes and cathode followers need not apply) Going to a 1:1 will help a bit but the DCR of the windings is added to the source Z again limiting the A2 possibilities. The next step in the design logic places a IT with a slight stepdown to keep a reasonably low impedance drive but I find the price in bandwidth and gain do not help the sonic situation. Maybe the best option is to simply direct couple the 300B to the GM-70 and then use the 1:1 to add a 5687 or 6H30 in front of the 300B to finish things off. The 1:1 can also serve plate choke duties on the 300B if you parallel the windings. 44hy is plenty of inductance but the 30ma is a bit light on the current.
dave
I paralleled the 126B and used as a plate choke, then DC coupled the 300B to the GM.
Changed bias scheme slightly
yes.... kinda....
when you parallel the wires of a transformer you do not double the inductance or current carrying capability so you still end up with a 44hy 30ma device.
dave
but why wouldn't I end up with 60ma? Wouldnt each winding be rated at 30ma x 2?
Either way, that's pretty slick, never knew these could double as a plate choke. Thanks for the info!
Edits: 06/04/15
The Bdc (DC flux) in a transformer is directly related to the amp-turns of current in the windings. When you parallel two identical windings, you do not change this relationship since you still have the same amount of current going through the same number of turns of wire. The DCR does typically get cut in half but that is about it.
dave
is above my pay grade....
I've made several attempts with spice but cant seem to get the basics of DC.
Would you still do a fixed bias with DC?
I did a IT with stepdown but I'm thinking twice about it since I get the hint this has already been tried and didn't sound too good.
You can use a current mirror to have one tube be the master, and the other the slave (see Tube CAD about this).
If we, for example, say that a single tube would want 100 Ohms nominal cathode resistance, you could use a double deck 100K pot wired as two variable resistors where rotating the knob in a given direction increases resistance on one deck and decreases it in another, then wire that in parallel with the 100 Ohm resistor. It would probably help to use a little resistance in series with the pot to prevent it from presenting 0 Ohms resistance under the filaments, and to limit the power seen through the pot. One would just need to put a meter between each filament connection and adjust the pot to get 0V DC. (You'll need separate filament windings for each tube regardless of what you do)
"Can I achieve this without a negative supply?"
I'm not 100% sure, but I think not. Low VA toroidal power trafos from AnTek are inexpensive. Even if the "tallest" RMS voltage you can buy is too low, voltage multiplying supplies are easy enough to construct.
Combination bias is a sweet deal. :> ) You get a stabilized operating point from the cathode bias resistor, excellent control of "idle" current via the bias supply/trim pot., and (if you choose) local current NFB by omitting a cathode bias resistor bypass capacitor. As DHTs are known to be runaway prone, stabilization of the operating point is (IMO) quite important.
Eli D.
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: