Back to Pioneer and Lowther Le
The critical factor determining Le is the extent of saturation of driver's magnetic circuit. At saturation, iron permeability approaches unity, so that VC behaves as air core inductor, and it's inductance is at the minimum. If the magnetic circuit is not saturated, the VC becomes iron- core inductor whose inductance is multiplied by residual mu.
Lowther' s magnetic circuit matches the magnet and yoke's magnetic path, cross-section, and gap in such a way that yoke's iron is magnetically saturated. Using a larger, more powerful magnet would be a waste of material; using a smaller magnet would increase Le.
Apparently, this was not considered in Pioneer's design. As a result, it's magnetic circuit is not saturated, and Le is high.
Here comes the big advantage of of FC motors: saturation can be easily assured. The caveat: coil's heat dissipation capacity may be insufficient. Luckily, for most FC drivers this is not a problem.
The critical factor determining Le is the extent of saturation of driver's magnetic circuit. At saturation, iron permeability approaches unity, so that VC behaves as air core inductor, and it's inductance is at the minimum. If the magnetic circuit is not saturated, the VC becomes iron- core inductor whose inductance is multiplied by residual mu.
Lowther' s magnetic circuit matches the magnet and yoke's magnetic path, cross-section, and gap in such a way that yoke's iron is magnetically saturated. Using a larger, more powerful magnet would be a waste of material; using a smaller magnet would increase Le.
Apparently, this was not considered in Pioneer's design. As a result, it's magnetic circuit is not saturated, and Le is high.
Here comes the big advantage of of FC motors: saturation can be easily assured. The caveat: coil's heat dissipation capacity may be insufficient. Luckily, for most FC drivers this is not a problem.
bondini, I have nothing I am confident in...LOL I built a LTP from FQP1N60 and 6H6Pi for a LTP input stage long before I ever heard of SOA and it has not come apart yet...LOL
I tried this one after several real pentodes and a paranoia about g2 current wrecking the phase splitting. Curiously, I like real pentodes doing that job just as much...especially EL84's. I have some others running that I think eclipse even the EL84, but not yet measured seriously.
cheers,
Douglas
I tried this one after several real pentodes and a paranoia about g2 current wrecking the phase splitting. Curiously, I like real pentodes doing that job just as much...especially EL84's. I have some others running that I think eclipse even the EL84, but not yet measured seriously.
cheers,
Douglas
Speaker resonance impedance problem
Speaker resonance presents a problem for no-feedback pentode operation. On the upward leg of resonant impedance peak, the impedance is largely inductive, and on the downward leg it is largely capacitive. In both instances, the impedance is multiples of driver's nominal impedance. This violates two load requirements for undistorted pentode operation: load must be non-reactive and uniform across the frequency range.
I see only one solution to this problem: limit amplifier's band to keep speaker's resonance out. It makes sense because this way or that, the audio band has to be split, usually at the speaker crossover. But splitting the band at the amplifier input instead has many important advantages. First, speaker resonance problem is eliminated. Second, IM distortion in amplifier and speaker is reduced. Third, output transformer doesn't need to handle low frequencies, which greatly simplifies its design. Fourth, amplifier power is used more efficiently because amplifier now doesn't handle the energy-rich bass. Fifth, speaker-level crossover is eliminated. There are so many advantages that one should wonder why a single amplifier covering the whole audio band is so prevalent.
sser2,
Generally, you want to drive the woofer below, at, and above the woofer's resonant frequency.
Or do not use a woofer, and give up hearing the bass frequencies.
Your impedance curve drawing is good for a closed box, open baffle, etc.
A woofer in a ported enclosure has two impedance peaks. The lower side of the peaks, again is inductive; the upper side of the peaks again is capacitive, and the dip between the two peaks is resistive.
A push pull output stage often deals better with reactive loads; versus a single ended output stage deals with reactive loads.
Hint: Draw an elliptical load line on properly overlaid push and pull tube curves, and then draw the same elliptical load line on a single ended tube curve.
All Generalizations Have Exceptions.
Just my opinions.
Generally, you want to drive the woofer below, at, and above the woofer's resonant frequency.
Or do not use a woofer, and give up hearing the bass frequencies.
Your impedance curve drawing is good for a closed box, open baffle, etc.
A woofer in a ported enclosure has two impedance peaks. The lower side of the peaks, again is inductive; the upper side of the peaks again is capacitive, and the dip between the two peaks is resistive.
A push pull output stage often deals better with reactive loads; versus a single ended output stage deals with reactive loads.
Hint: Draw an elliptical load line on properly overlaid push and pull tube curves, and then draw the same elliptical load line on a single ended tube curve.
All Generalizations Have Exceptions.
Just my opinions.
Although it was not addressed fo me, I have some simulations in the other thread:
Fourier components with and without GNFB
https://www.diyaudio.com/community/...lifier-without-global-nfb.384100/post-7017330
Also the listening experience confirmed that some GNFB (12dB in my case) did greatly reduce the effect of bass resonance and reduced THD. Although this is not a pure pentode amplifier. I have some local negative feedback in the form of UL connection and cathode feedback.
I was under the impression that the distortion figures in tube datasheets (as opposed to application notes) were without negative feedback. They would be meaningless if they were with NFB and the amount of feedback wasn't quoted. This is the first time I've seen the claim that the figures include "implicit negative feedback". Simulations without NFB generally give distortion results in the same ballpark as the datasheets.
I think the only logical way of measuring a power tube THD for datasheet is a choke load in the anode, and capacitor coupling to the real resistive load (and measuring THD thereon).
I could be wrong on that. That is why I requested to see measurements of a real-world current amplifier with 807 tube.
But my main point stay - why should we need to accept higher (H3=0.8%) distortion of the best "pentode" tube 807 (actually - tetrode) implemented in a current drive amplifier (without added loudspeaker distortion), when we can have lower total (amp+loudspeaker) distortion (H3=0.2%) with a conventional solid state amp with negative feedback?
Yep, the pentode amp will distort - and I will help that along with a cheap output transformer from the stash! 🙂 As I noted earlier, low distortion solid state current amp designs are available but I don't want to make and test them - readers are welcome to do so if they wish, though the exercise might best be discussed in another forum. I will use pentodes, warts and all.
Using a single ended pentode amp, I will measure THD and Zout across the audio bandwidth at 1 Watt into a resistive load equal to the DCR of the Pioneer. The results will be risible - those who seek low Zout and/or low THD will avert their eyes, chortle, and refer me to a good psychologist. I will then drive the Pioneer (mounted in a well-damped enclosure of my choice) at 1 Watt and measure the SPL and THD of the resulting sound output across the audio bandwidth. I will be interested to see the THD result, from which I will subtract the amplifier distortion and inspect the residual, invoking the god Superposition to apportion blame to the Pioneer.
I will then repeat the procedure using a wee chip amp with low Zout and low THD (yep, it's a fixture on my test bench) again offering incantations and sacrifices to Superposition.
Will the residual THD curves be different? Where and by how much? And will the SPL curves be different?
And, yes, I will opine on how both rigs SOUND because the subjective is important to me - I didn't choose my partner on the basis of THD and SPL and I'm pretty sure that's not how she chose me. 🙂
Ideal measurement load is plain resistor (should be really big one ;-), with extra voltage to compensate for drop in it.
I don't want to give up on bass frequencies. There should be separate amplifier and transducer for that. But it is NOT going to be high NFB amplifier driving a subwoofer. Speaker distortion at low frequency is much, much bigger problem than at higher frequencies. Speaker THD of 80-90% in the 20-50 Hz range is quite common. But this should be the topic of a separate thread.
Prompted me to refresh my memory on PP pentode plate curves and load lines. Had some fun yesterday.
If the amplifiers are both HiZ out, shouldnt they "mix" simply by connecting in parallel?
I like your chart. I've always though of what you label inductive as "differentiating" and what you label capacitive as "integrating".
From the DIY point of view, unless we know by what technical means such unbelievably low speaker distortion has been achieved, it will remain a pink unicorn puking a rainbow.
But what is the advantage of such amplifier bridging? It will simply reduce Zout without any other benefit.
I am thinking of using the no-feedback pentode as low frequency amplifier - to reduce speaker distortion. But the main focus will be on improving the LF transducer.
Yes, of course, there are no real capacitors and inductors, it is a model of how amplifier output sees speaker resonance.
You guys are great!
Good discussion!
Icsaszar,
Some local negative feedback?
Pentode/beam power tubes with No negative feedback have high output impedances:
(for both single ended and push pull).
UL (local) negative feedback plus cathode (local) negative feedback: causes much lower output impedance versus No negative feedback. (for both single ended and push pull).
Bondini,
Good! Do build, measure, experiment, and listen.
Non negative feedback single ended pentode/beam power output stages have a fair amount of 2nd harmonic distortion.
Many Woofers have 2nd harmonic distortion.
Connect the amplifier and woofer one way, and the 2nd harmonic distortion of the amplifier and woofer Adds the 2nd harmonic distortion.
Now, reverse the + and - connections, and the 2nd harmonic distortion of the amplifier and woofer partially cancels the 2nd harmonic distortion.
sser2,
Regardless of whether impedance is from a real inductor, or is a woofer at frequencies a little below resonant frequency, the Voltage leads the Current.
Regardless of whether impedance is from a real capacitor, or is a woofer at frequencies a little higher than resonant frequency, the Current leads the Voltage.
In both cases, the amplifier does not “know” if it is a real inductor or a woofer below resonance; and the amplifier does not “know” if it is a real capacitor or a woofer above resonance.
All,
Now, look at the impedance curve of a mid-range or a tweeter. Many of them follow that same pattern as a woofer: DCR, inductive, resonant, capacitive, DCR, inductive.
A direct to speaker driver Bi-amp or Tri-amp sees all those impedances.
Or instead, make it even more complex by using a single amplifier to drive a passive crossover and 2 or 3 way speaker system.
All,
If you look in some of the RCA Tube Data Books, there are graphical load line plus calculations that give you the 2nd and 3rd harmonic distortion levels. That is why tube curves are so handy, they can be used to give those answers.
Many subscribers to diyAudio do Not have FFTs, spectrum analyzers, or even a THD meter.
Stay safe, healthy, and have fun!
Good discussion!
Icsaszar,
Some local negative feedback?
Pentode/beam power tubes with No negative feedback have high output impedances:
(for both single ended and push pull).
UL (local) negative feedback plus cathode (local) negative feedback: causes much lower output impedance versus No negative feedback. (for both single ended and push pull).
Bondini,
Good! Do build, measure, experiment, and listen.
Non negative feedback single ended pentode/beam power output stages have a fair amount of 2nd harmonic distortion.
Many Woofers have 2nd harmonic distortion.
Connect the amplifier and woofer one way, and the 2nd harmonic distortion of the amplifier and woofer Adds the 2nd harmonic distortion.
Now, reverse the + and - connections, and the 2nd harmonic distortion of the amplifier and woofer partially cancels the 2nd harmonic distortion.
sser2,
Regardless of whether impedance is from a real inductor, or is a woofer at frequencies a little below resonant frequency, the Voltage leads the Current.
Regardless of whether impedance is from a real capacitor, or is a woofer at frequencies a little higher than resonant frequency, the Current leads the Voltage.
In both cases, the amplifier does not “know” if it is a real inductor or a woofer below resonance; and the amplifier does not “know” if it is a real capacitor or a woofer above resonance.
All,
Now, look at the impedance curve of a mid-range or a tweeter. Many of them follow that same pattern as a woofer: DCR, inductive, resonant, capacitive, DCR, inductive.
A direct to speaker driver Bi-amp or Tri-amp sees all those impedances.
Or instead, make it even more complex by using a single amplifier to drive a passive crossover and 2 or 3 way speaker system.
All,
If you look in some of the RCA Tube Data Books, there are graphical load line plus calculations that give you the 2nd and 3rd harmonic distortion levels. That is why tube curves are so handy, they can be used to give those answers.
Many subscribers to diyAudio do Not have FFTs, spectrum analyzers, or even a THD meter.
Stay safe, healthy, and have fun!
Last edited:
? Pink unicorn puking a rainbow?
How about loudspeakers which have been available several years for DIY, like Purify PTT6.5X08 midwoofer ( https://purifi-audio.com/ ) and Bohzen CQ73 ribbon tweeter ( http://www.bzspeakers.com/index_en.html ), both used in Selah Audio Purreza loudspeakers (with "unbelievably" low distortion):
So, from DIY point of view, technical means for achieving such "unbelievably low speaker distortion" are simple:
1. Choose and buy modern, "unbelievably low distortion" drivers such as Purifi and Bohzen, which don't cost arm and leg;
2. Build speaker enclosure according to TS parameters;
3. Buy calibrated microphone and mic preamplifier;
4. Get free crossover simulation software (VituixCad, or something similar) and learn how to use it;
5. Build a crossover according to the simulation above;
6. Build or buy conventional solid state amplifier with a good amount of negative feedback;
7. Enjoy in the unbelievably low speaker+amp distortion.
or,
Fom DIY point of view, technical means for achieving high distortion (H3=0.8%) from the amplifier alone are even simpler:
1. Build "current drive" pentode/tetrode amplifier without negative fedback;
2. Good luck finding fullrange driver suitable for current drive;
3. Build loudspeaker enclosure (or infinite baffle) for the said loudspeaker;
4. Enjoy in the high distortion (H3=0.8%) from the amplifier plus unknown amount of distortion from the loudspeaker.
Last edited:
Wonderful!
But pink unicorns are too Expen$ive for some of us.
Were the 2nd order and 3rd order Intermodulation distortions tested?
But pink unicorns are too Expen$ive for some of us.
Were the 2nd order and 3rd order Intermodulation distortions tested?
Intermodulation distortion was not tested. Nor the intermodulation distortion for unicorn loudspeaker suitable for current drive, with flat frequency response, low distortion and low price.
Purifi PTT6.5X08 midwoofer costs 360 EUR, and Bozhen CQ76 tweeter costs 460 EUR (there are many equally good alternatives for the tweeter). So not unicorns and not too expensive. Both drivers were used in the "unbelievably" low distortion Selah Audio Purezza loudspeakers.
How much does it costs to build a 807 tube amplifier without feedback? How much costs loudspeaker suitable for current drive, with flat frequency response and low distortion?
Purifi PTT6.5X08 midwoofer costs 360 EUR, and Bozhen CQ76 tweeter costs 460 EUR (there are many equally good alternatives for the tweeter). So not unicorns and not too expensive. Both drivers were used in the "unbelievably" low distortion Selah Audio Purezza loudspeakers.
How much does it costs to build a 807 tube amplifier without feedback? How much costs loudspeaker suitable for current drive, with flat frequency response and low distortion?
I've read Bruno Putzeys writings about loudspeakers, and didn't find much new information. But I see his creative ways of presenting his speaker's distortion information in a way that makes unsophisticated people believe they are magic. Apples should be compared to apples. Speaker distortion should be compared at the same SPL, not at the same power input. If Mr. Putzey's provided side-by-side comparison of his drivers with similar ones from Seas, Vifa, Scan Speak, etc, I seriously doubt his would come out better.
We can look at how distortion was measured in Electrovoice DL12X and Purifi 6.5" driver. The Electrovoice was measured at 10% of its nominal power rating of 200 W (btw, this is real rating if you look at this hard-to-lift driver with the motor almost the same size as speaker itself), or 20 W. Since its sensitivity is 100 dB, the corresponding SPL was 113 dB. The Purifi was measured at 1 W input; with sensitivity of 85 dB, the SPL output was 85 dB. So, just a little bit of SPL difference, mere 28 dB, and lo - the Purifi comes out big winner!
And then this ridiculous claim that this puny 6,5" driver can continuously dissipate 200 W of heat... all you need to know about credibility of Mr. Putzeys' claims.
We can look at how distortion was measured in Electrovoice DL12X and Purifi 6.5" driver. The Electrovoice was measured at 10% of its nominal power rating of 200 W (btw, this is real rating if you look at this hard-to-lift driver with the motor almost the same size as speaker itself), or 20 W. Since its sensitivity is 100 dB, the corresponding SPL was 113 dB. The Purifi was measured at 1 W input; with sensitivity of 85 dB, the SPL output was 85 dB. So, just a little bit of SPL difference, mere 28 dB, and lo - the Purifi comes out big winner!
And then this ridiculous claim that this puny 6,5" driver can continuously dissipate 200 W of heat... all you need to know about credibility of Mr. Putzeys' claims.