|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.6.199.104
A few weeks ago Nelson Pass came out with a white paper on using current source amps to drive fullrange speakers such as Lowthers and Fostexes.This got me very intrigued since I have Lowther speakers and his ideas seemed to make sense. Now Nelson is making a solid state transconductance amp, but being a tube kind of guy right now, It immediately struck me that pentodes would be a perfect match for this. Then I realized that my little BDT power amp was already a very good transconductance amp, but it wasn't working like it needed to for this application because the OPT ratio was too high, by designing the OPT and operatin point for an 8 ohm load it produced a CCS that didn't have enough voltage compliance to handle the higher impedance ranges of the speaker. After thinking about it for a while I realized all I had to do to get this to work was use the 16 ohm tap on the OPT instead of the 4 ohm tap. (I was using a 8k to 4/8/16 OPT set to the 4 ohm setting to get 16k on the primary) This shoul increase the voltage compliance enough that the amp could start to take care of the higher impedance of the driver in the bass and treble areas.
So tonight I rwired the secondaries and turned it on, WOOOHW, this was no subtle change, the bass was WAY extended, I was getting floor shaking from pipe organ recordings, these speakers have hinted at the bass power but never actually shook the floor, now the DO! Its like adding a subwoofer. The high end also improved significantly. The Lowthers drop off firly steeply above 12KHz, but with this change they go to at least 15KHz before falling off. The wonderful you are thereness stayed, the beautiful midrange stayed, but now it has BALLS!
There is "air" a plenty now with incredible sparkling highs and deep bass that is not boomy, just there. This was the single biggest transformation I have ever heard in my system. And all it took was changing the tap on the OPT!!!! Every apect of sound you can think of is better, including the air banjo factor. I had a VERY hard time tearing myself away long enough to write this mesage.According to Nelson this is only going to work well with the high eficiency single driver speakers, so don't expect miracles out of 85dB 3 way complex crossover designs.
My thoughts are that pentodes should also work very well for this, but it has to be straight pentodes, not triode mode or ultralinear mode configurations. The big difference is setting the OPT so that the most efficient power transfer happens at 40ohms or so, NOT 8 ohms. Then when used with a single driver there will be enough voltage range to maintain good transcunductance when the speaker is at 40ohms.
I think this is why some people have said that they wind up using 100 watt PP pentode amps even though they have 105db efficient speakers, such an amp will have enough voltage on the output to handle the speaker even though its got an OPT designed for 8 ohms. So if you design an amp for this specific use, you could get away with a 5 watt SE pentode amp and it would probably sound aweome.
This opens up a whole new realm of experimentation for the tube guys, straight pentode mode just might make a comeback.
I can state here that this is not a crackpot idea, it really does work and the results are phenomenal.
BTW the BDT amp sounded very good as it was originally designed, better than my 45 SET in many ways, but after this simple change to the OPTs, it was, pardon the pun, transformed.
One nice advantage of doing the amp this way is I don't have to use such insanely high ratio OPTs, I can use much more readily available ones.
Anyway, thats all the time I have now, Bach organ music is calling me!
John S.
Follow Ups:
I got my zero NFB OTL a while back and have been in heaven ever since. The output impedance is about 50ohms and I was worried about the bass performance. However....WOW it is great!!!!!!!!!!Granted, I use closed speaker boxes since I think ports slow and mush the bass. So the bass from the speakers is naturally tight.
The real problem with the high output impedance is the crossover in a multi driver speaker like mine. I have a three-way with xover at 300 and 3500Hz. The output impedance of the amp is in series with the impedance of the speaker drivers, so the crossover must be modified for approx 55ohms, and not 8ohms. This is good because as the impedance of the speaker drivers vary with frequency, it's variance is now relatively much less. An 8ohm speaker with peaks to say 25ohms is a load that varies 300%, while with the amps 50ohms the variance is from nominal 58ohm to 83ohms, or 'only' 30%. That's better for both the amp and the speaker drivers. The crossover sees a more constant impedance over frequency.
I had planned to add feedback to lower the output impedance, but the amp sits there night after night just playing music, and since my speaker crossovers has been fixed, I have no hurry to add NFB. Ideally the speaker could have two crossovers, one for transistor muck and designed for low source impedance, and one for higher impedance amps, and a switch giving the option of the two.
Anyhow, I dont get impressed anymore by amps that boast super high figures in dampening factor.
Regards, Nick
Ever since playing with IC-based current amps, I've been thinking about tube-based current amps. A pentode is naturally a current device, as it is supposed to deliver current in proportion to grid voltage at any plate voltage, which means it has a near-infinite Rp. In such a situation, the main purpose of an output transformer is not to lower Zout but to increase Iout to levels that can drive speakers. There are also interesting possibilities for no-feedback OTLs, especially low-power versions using a SE 6550 for example, or perhaps even better, differential output stage PP current amps. A PP EL34 current amp with a Gary Pimm CCS in the cathode would be a true "New Wave" amplifier design.It's not as if people haven't gone this way before. Susumu Sakuma has done some interesting NFB SE pentode amps, driven in full pentode mode with a fully separate power supply for the screen.
JohnGood-on-ya, mate! (as they say in Oz)
Back around 1960 some amps (pentode) had variable feedback circuits to tune for the optimum output impedance. A higher output impedance would more effectively transfer power to the high impedance frequencies in the speaker, fattening up the bass resonance, and anything else that wasn't tied down (like the crossover frequencies). Anyway it was tweakable. OTL amps really like a 100 Ohm load better than an 8 Ohm load, and will fatten up the bass a lot unless you dial in a bunch of feedback.
Any resistance in the LF circuit of a speaker raises the Q at resonance and a high DCR choke will require resetting the bass reflex tuning when you assume a voltage source for transistor amps. I used to do that stuff all the time before I got into tubes; now I'm way to wierd, but I remember those simple days...
So if you have Lowthers, that really don't want much more than a watt, especially in the bass, the fact that you are LESS efficiently transferring power to the lower impedance sections of your passband (midrange) won't be bad at all. The fact that you get MORE BASS will be way cool, and the extension above 12K will be nice, too. Basically, you'll put a "loudness contour" EQ on them, which is just what they are screaming for (so to speak).
If I read you right you are implying that there is sort of an auto-adjust to this where the production of speaker output power is inherently more linear with the system set up this way. I don't have enough insight to know if something that cool is happening or if it is just a fortuitous special-case. At any rate, it sure seems like a good setting in which to "dial in your octane" (if you remember Sunoco 260).
Now I have read the article. I don't quite get how a change of taps to 16 ohm get your power amp into a current amp, but I hope somebody explain it in the future. (And I don't get why pentodes should be better)
Anyway I do have a pair of Lowthers for which I ordered a pair of Gizmos yesterday to tame them a bit, but by reading the firstwatt text it seems like the same results could be obtained by two resistors and a cap in parallel with the speaker. Have you played with this network?
The quick answer is that it already WAS a current source amp, it just didn't have enough voltage range to handle the high reflected impedance from the speaker. Changing the tap decreased the voltage ratio, so given the amount of voltage swing available on the tube, the voltage range available at the tube doubled.Here's a little more detail, I hope I can explain this properly. Tubes in general operate in whats called a transconductance mode, that means that the current through the tube is proportional to the voltage on the grid. With the grid voltage held constant if you change the impedance of the load the tubes varies the voltage across the load so it can maintain the current set by the grid voltage. As the impedance keeps on rising the voltage also has to keep on rising to mainain the set current. At some point the tube runs out of voltage, either the B+ is not high enough or the mu of the tube does not allow enough gain.
The problem is that most amps are designed so the voltage across the load is proportional to the grid voltage, not the current. For the particular type of speakers Nelson is talking about they seem to sound better when the current is proportional to the input voltage, thus "current drive".
As to the pentodes, they more closely approximate a true transconductance amp than do triodes (as long as you don't apply various forms of negative fedback designed to turn them into better voltage amplifiers!). The BDT amp is already a very good transconductance amp as long as the gain is kept within the rather low mu range of the tube.
As to the impedance of the OPT, most amps are designed so that the maximum voltage swing of the amp gets transformed into whats appropriate for an 8 ohm speaker. This means that in the region that the speaker is at high impedance the amp tries to output more voltage to the speaker to maintain the current, but it can't because its already maxed out. By going to a 4 times higher tap on the transformer the voltage step down ratio was also cut in half. Thus the maximum voltage swing of the amp became twice as much voltage swing at the speaker which allowed the current to be maintained at higher impedances than with the lower impedance tap.
I hope that makes sense!
Humm, I think my knowledge guides me to start reading posts again instead of posting my stupid questions.Normally I play with this amp (http://db.audioasylum.com/cgi/m.mpl?forum=tubediy&n=42896&highlight=pufff&r=&session=) which also have a current source at the power tubes, which I think should be a bit similar to what your BDT amp (If I look at a bdt tube as a kind of triode with a current source buildt in) But my amp is in process of being rebuild, and right now I'm using a SS amp, which, as Nelson Pass writes, make the Lowthers sound like a transistor radio. And this transistor sound really got me interested in having some possibilities for adjusting the sound also when my tube amp is back in service.
So my last question before I step out of this thead and wait to see others results are: Would my amp could be used as a transconductance amp, if I connect the output tap to the 16 ohm tap?
Your amp will transfer more power into the higher impedance portion of your speaker's passband (bass resonance and high treble) if attached to the 16 Ohm tap, and it will supply less power to the midrange.
If you have 8 Ohm Lowthers you might get 1 dB more bass AND 1 dB less midrange with this adjustment. You will get a warmer, fatter sound, but it may also be fuzzier.
If you have Hammond output transformers, they can usually be configured for 20 Ohm output impedance, which will do this even more.
John and current-source, push-pull afficianadosLook at the 9/1/04 thread on push-pull spud amps that ends with my "Dumb and Dumberer" re-re-evaluation of the weird characteristics of the 6N6D "Direct Coupled Power Amplifier" in a bottle.
This thing makes 4-6.5 WPC into a 10K load P-P, with 300V B+, 51 mA, total current draw, and a plate resistance of 24K, though they consider it a dual-triode with the output valve biased into positive grid. The entire current (9mA) of the signal valve passes THROUGH the grid of the output valve before meeting the internal cathode resistor to ground. Drive a P-P pair of these with the BDT preamp into the Lowthers!
Hmm, do you have a link to that Nelson Pass article? My guess is that more than me is interessted.
Also did you ever publish a schematic of your power amp? I have only a copy of your preamp. (And parts on its way ;-))Anyway it sounds a bit like it is a kind of diff amp you have build, and they do have better slam and bass than SE amps.
Regards pufff
http://www.firstwatt.com/current_source_amps_1.htm
To infinity and beyond!!!
no text, reading the link ;-)
This post is made possible by the generous support of people like you and our sponsors: