|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
73.181.211.24
In Reply to: RE: The question is why 200 henries? posted by Chip647 on May 30, 2016 at 13:47:11
I have done a few experiments, enough to say with confidence that "more is better" with regard to inductance of a plate choke for an output stage.
A recent, specific example is increasing the inductance for a 300B amp from 24 henries to 40 henries, both with a 3000 ohm output transformer. The bass improvement is clearly audible. (This is equivalent to changing from 80 to 133 henries for a 10K output, which is typical for a Type 211 tube.)
In an older experiment, I made a preamp with a 6BL7 parafeed into a 5K output transformer; I set it up with four different inductances plus a current source, which could be selected with a switch. I no longer recall the specific inductances, but they covered about a 2.5:1 ratio. The smallest inductance gave perfectly acceptable measured bass extension, but every increase was audibly better, and the current source was still better than the largest inductance, indicating room for more inductance.
Follow Ups:
"A recent, specific example is increasing the inductance for a 300B amp from 24 henries to 40 henries, both with a 3000 ohm output transformer. The bass improvement is clearly audible."
Paul, the rule of thumb I use regarding the shunt reactance across an AC circuit is that it should be no less than 5X the operating impedance at the lowest frequency of operation. 10X the operating impedance is better. In the example you cited, with an operating impedance of 3K ohms, XL should ideally be about 30K, although 15K is probably acceptable. The 40H choke in your example fails to meet this requirement, as it presents only about 5K. Making matters worse, it's in parallel with the inductance of the OPT, so the final value is likely to be considerably less. The difficulty in maintaining minimum required inductance is one of many reasons I've never been tempted to build parafeed amplifiers.
--------------------------
Buy Chinese. Bury freedom.
Please see my comment below, about the nearly-resistive load that is possible with parallel feed.
What about a CCS in the plate circuit?A friend just did that with a 45 and loves it.
P.S. I know I posted this before but I don't think you can get enough inductance (either plate choke/parafeed or series feed SE) to play bass properly.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 06/07/16 06/07/16 06/08/16
The problem with a CCS (aside from my solid-state allergies) is its lack of energy storage. Even ignoring that, I just don't have any reason to go that direction personally. There are many excellent SET OPTs on the market that don't require a plate choke or blocking capacitor.
--------------------------
Buy Chinese. Bury freedom.
Question, in a series feed SE OPT, how high does the inductive reactance need to be (vs. the reflected impedance) to keep the load line from being elliptical?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"Question, in a series feed SE OPT, how high does the inductive reactance need to be (vs. the reflected impedance) to keep the load line from being elliptical?". A lot higher than is usually available. 40-50henries still has the loadline looking like a smashed banana. JH
A series feed opt with 200h of primary inductance would have to have tiny primary wire to allow enough turns, then the dcr would be out the window and other parasitic losses as well. No way to have your cake and eat it too, hence the opting for different topologies in finals feeding. JH
I have been considering 211 push pull instead of se with hf filaments it wouldn't be a expensive amp to build and run. I am not sure how long the tubes last. I run my main amp 18 hours a day so running costs add up for me. Any one that has posted about a 211 push pull amp over the years(and it seems not many people do) has been positive about them.
Perhaps a Ccore would be the best option to get anywhere near enough inductance. I am thinking 160H would enough So halved across the transformer (is that what happens approximately?) that would work out in the ballpark of 80H which is what the best 211 trans have that I have been shopping around for. 211 is not the only tube I want to try but it is the highest plate resistance of the tubes I want to try. The ones that are higher I want to use in A2. Would 160H put in back into the possibility of fitting into the window of a affordable EI core like one done by heyboer?
From the Lundahl datasheet,
"We define Power Low Frequency Limit, FPL
, as the frequency where LP= R LOAD. (The reactive impedance of
the transformer equals the primary load impedance). At FPL, the output power is reduced to 50%"
I don't think that's right.
If you had a 10k single ended transformer with 80Hy of inductance and you were driving it with a 10k source then at 20Hz (where the reactance of the 80Hy is 10k) the power would be 50%. (10k//10k=5k)
But the 3300 ohm plate resistance of the 211 (for instance) would swamp that equation. So that aside,
What the 10k reactance of the 80Hy of inductance in parallel with the 10k reflected impedance will do is cause an elliptical load line.
With 160Hy the reactance would be 20k at 20Hz so 10k//20k=6.667k and that will still cause an elliptical load line.
And that elliptical load line won't just be for 20Hz.
When low frequencies are present, all the frequencies will be following that elliptical load line and elliptical load lines increase distortion.
TK's 5X or 10X rule would prevent a elliptical load line but I don't see how it can be obtained in a gapped transformer (or plate choke).
Parafeed, using a CCS for the plate load, would work (parafeed output transformer don't have air gaps and have plenty of inductance) but if a plate choke or series feed SE output transformer is used.....using them to play bass seems out of the question.
P.S. I know I didn't answer your question. I don't know how much inductance can be had without causing other problems. Whatever the answer is, I don't think it's enough.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Plan B LL9202 I have posted above a question about the LL920 am verses the silicon core. I am leaning towards going that route first. Probably a higher chance of success.Then I can give up on the parallel feed 211 and have wound a smaller choke for the a2 tubes.
Edits: 06/10/16
"No way to have your cake and eat it too, hence the opting for different topologies in finals feeding"
I agree and that was my point.
Single ended amplifiers don't/can't play bass.
I use SS for my woofers and SET for the mids and tweeters.
I think that's as close as it gets to "having your cake and eating it too".
Thanks.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"Single ended amplifiers don't/can't play bass."
This I have to disagree with. My 211s work to 20 Hz with no problem. The Magnequest 10K SE OPTs are superb.
--------------------------
Buy Chinese. Bury freedom.
I didn't know magnesquest made 10k trans. I never had a pair of magnequest trans. But the magnequest preamp trans I have are so much better then any other I have tried.
Edits: 06/11/16
The transformers I'm using are FS-100, designed for 211/845 amplifiers. I run them at 1KV and 70-75 mA in my 211 SETs. I've read that there have been a few different versions of these, including one for parafeed, all with the same model no.
--------------------------
Buy Chinese. Bury freedom.
Thanks triode I found the blog too and I found the forum. People keep referring to trans that I know nothing about. I want a pair of decent trans.. I am tired of buying shit. I will wait until mike is winding something I can use and afford. How ever long it takes.. I just hope he will wind me a pair. It looks like he is getting busier and busier. I want to try find and research the specs of all the old trans that he winds too. I think they are somewhere on line. I really want a pair of classic push pull output designs with nickel lams or part there of. I will get around too the a big A2 amp eventually. I have enough other amps to finish in the mean time. I will try a a2 3c24 I have trans for that allready. If there any good. I have my doubts. They are soft tones I bought them from japan but they look like chinese to me.
I meant technically speaking. I'm sure subjectively your amps are fine.How much inductance is there in those Magnequest 10K SE OPTs?
Do they meet your 5X/10X rule?
It would take 400Hy to be 5X (50k ohms) at 20Hz.
The load line is going to be an ellipse. I don't think there's any getting around that.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 06/10/16
I get the sense that you think the reactance of the inductance of a plate choke in a pararfeed circuit is in parallel (shunt) with the reflected impedance at the primary of the parafeed OPT.I believe that is true.
I also get the sense that you don't think the reactance of the inductance of the primary winding of a series feed OPT is in shunt with the reflected impedance.
As far as I know it is.
The only difference is the caveat that you pointed out, the reactance of the plate choke's inductance is also in parallel with the reactance of the parafeed OPT but that would be a small factor. Otherwise both situations are the same.
Bottom line, I don't think a gapped transformer (or plate choke) can have enough inductance to keep the load line from being an ellipse at the lower frequencies.
That is why I say "SET amplifiers can't play bass".
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 06/10/16
I don't bother with that. I look for transformers that perform well in terms of frequency response and full power (low frequency) distortion in my application . If a transformer can do that, and it sounds good with my specific output stage, I'm done.
--------------------------
Buy Chinese. Bury freedom.
Edits: 06/08/16
What would your guess be about the result of lowered output Z from the local FB given by the distributed loading( part of the choke in the cathode)? Seems to me this would be quite useful in reducing its tolerance of the elliptical load line inadequate inductance would deliver.
Cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
I like the idea of applying feedback to the higher mu tt tubes. they seem made for it.
I have some 65W/6.5 mA/V TT's with mu of 25 that would need some sort of FB reduce output Z a bit. Have a few ideas of implementing this, but it may just be that they'll get put in front of horns and left on their own...:) They're just too neat to leave unbuilt.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Just realized I did not address the elliptical load line issue.
In a parallel feed arrangement, the output capacitor interacts with the plate choke. This interaction can be optimized to maintain a nearly resistive load as seen by the tube, at least within the pass band and assuming a resistive speaker load. For the same inductance, parafeed can perform much better than series feed, with respect to elliptical load lines.
Good question! It has bothered me for some time. We (at Bottlehead) have over the years experimented with a few different arrangements, including a cathode winding, ultralinear, plate to cathode, plate to grid, cathode follower, etc.
At this point I personally have concluded that feedback is feedback, no matter how it's implemented. It cleans up the measured performance, while sucking some of the life from the music. That last sentence offends me as a scientist, but as an engineer I have to accept it until a more successful option is discovered.
Interestingly (to me), this happens in speakers as well - the more complex the crossover, the cleaner the sound but the less involving it becomes.
Finding a satisfactory compromise is IMHO very specific to each design - generalizations are simply not reliable enough. You still have to listen!
!
The Mind has No Firewall~ U.S. Army War College.
yea in more experience more is better but if the dcr gets to high I worry that will the life out of it too.So Paul if I ask for just that
4" stack of laminations with 2.5" tongue (i.e. the core is 7.5" x 6.25" x 4") would give 237 henries at 100mA maximum current. Weight about 35 pounds, capacitance around 1000 pF.you think the dcr will come out ok and is that capacitance acceptable? I really wouldn't know.
Actually I just remembered I wanted to try a pair of 805,s too so I need 120ma max So I guess I will just have to give them my specs and let them work it out. I probably shouldn't be messing with these sorts of voltages but I have been working with tubes for 15 years now and only been got once in the beginning. I am worried though and that's probably a good thing.
Edits: 05/30/16 05/30/16
I do NOT think the capacitance is acceptable, it was just a fast calculation to determine the likely parameters. And the air gap I assumed (20 mils) was just a guess. Best to find a good designer and ask for their recommendation. My point is that it's not a trivial effort, and it's not a small chunk of iron. Engineering is all about the compromises, and in a case like this compromises are unavoidable.
yea I not sure who to ask I am thinking of trying heyboer i haven't ordered any from them before.
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: