DF of more than 2-3 doesn't make much sense because voice coil impedance is always in series with amplifier output, and no feedback scheme can reduce it.
As to output transformer causing distortion in high impedance output stage (such as no-feedback pentode) due to non-linearity of magnetizing current, this is a boogeyman of no practical importance, as Crowhurst explained. At higher frequencies the magnetizing current, which is limited by transformer's primary inductance, may be so low that it can be neglected. At lower frequencies, the increase in magnetizing current can be offset by increasing primary inductance.
As to output transformer causing distortion in high impedance output stage (such as no-feedback pentode) due to non-linearity of magnetizing current, this is a boogeyman of no practical importance, as Crowhurst explained. At higher frequencies the magnetizing current, which is limited by transformer's primary inductance, may be so low that it can be neglected. At lower frequencies, the increase in magnetizing current can be offset by increasing primary inductance.
May I ask you how you desumed that? I've always found numbers one order of magnitude higher.
All too true, but you will have to convince the thousands on the forum of those facts.
Why don't you put something together to illustrate your points? 🙂
Damping improves when total resistance in voice coil loop (which is the sum of voice coil resistance and amplifier output resistance) decreases. DF of 100 is for practical purposes not better than DF of 10. With 8 Ohm speaker, total VC loop resistance is 8.08 (DF=100) and 8.8 Ohms (DF=10), respectively. As you can see, VC resistance, but not amplifier output resistance, is the dominant factor. As DF goes down, the amplifier output resistance contributes more to the loop resistance, with the inflection point at DF=1. With the example of 8 Ohm speaker, the loop resistance is 8R8 at DF=10 and 10R7 at DF=3, which is not really much of a difference.
I don't need to. There were commercial high end designs (before the dreadful onslaught of Hi-Fi amplifiers and boxed speakers) that utilized open loop (no feedback) PP pentode amplifiers. Magnavox Symphony or Belvedere consoles, for example.
Was the Magnavox Symphony console considered 'high end'? I thought those pre-fab consoles were the everyman's hi fi, while the cognoscenti went for stuff like Marantz, McIntosh, Fisher, Heath, etc. -- or even more obscure stuff like Grommes or Fairchild.
Maybe I have it all wrong. I was born about that time, so I didn't get to hear that stuff, even though my parents had a pretty cool mono console stuffed with Stromberg-Carlson electronics, a push-pull 6L6 integrated amp complete with multiple phono EQs and a really beautiful tuner. It even had a big Altec 2-way speaker, but I don't have any photos of it, Just recollections of it playing top-40 AM radio in 1966 or so. Good times. As I recall, that hi-fi sounded really good. But you know, I was a little kid.
Maybe I have it all wrong. I was born about that time, so I didn't get to hear that stuff, even though my parents had a pretty cool mono console stuffed with Stromberg-Carlson electronics, a push-pull 6L6 integrated amp complete with multiple phono EQs and a really beautiful tuner. It even had a big Altec 2-way speaker, but I don't have any photos of it, Just recollections of it playing top-40 AM radio in 1966 or so. Good times. As I recall, that hi-fi sounded really good. But you know, I was a little kid.
The 7355 output tube in some Harmon Kardon amplifiers surprises me with its 42000 Ohm Rp. It would take a lot of Neg. FDBK to get the output Z down to practical level. What was the point of this tube? Same for 6V6 with its 70000 Ohm Rp.
The voice coil resistance is designed-in by the loudspeaker designer. You don't normally tamper with it.
Since about 1953, loudspeaker designers like infinite damping because it simplifies small-box design.
That does not mean you must have infinite damping!! Everything is a compromise.
If designed for infinite damping, and you have non-infinite damping, you get a response-rise (boom) of about:
DF bump
40 - 0.2dB
10 - 1dB
5 - 2dB
3 - 2.5dB
2 - 3.5dB
1 - 6dB
A low-Qt speaker in a large box can be happy at DF near 3, like most triode power amps. The 2.5dB electrical bump can be complemented in the rest of the design to give a very flat response. If that is what the loudspeaker designer wants to do.
Since about 1953, loudspeaker designers like infinite damping because it simplifies small-box design.
That does not mean you must have infinite damping!! Everything is a compromise.
If designed for infinite damping, and you have non-infinite damping, you get a response-rise (boom) of about:
DF bump
40 - 0.2dB
10 - 1dB
5 - 2dB
3 - 2.5dB
2 - 3.5dB
1 - 6dB
A low-Qt speaker in a large box can be happy at DF near 3, like most triode power amps. The 2.5dB electrical bump can be complemented in the rest of the design to give a very flat response. If that is what the loudspeaker designer wants to do.
The NFB works against the parallel combination of tube and load. The transformed tube may be 42 ohms, in parallel with 8 ohms, is 6.7 Ohms. You need less than 9:1 of excess gain for NFB to get you to DF=10. If the tube were infinite impedance, you'd need 10:1 of excess gain for NFB to get you to DF=10; no great difference.
OK, but the high Rp of the tube will make for less optimal control of the OT, less of a voltage reference. Which I thought was the point of using local N Fdbk.
So a better OT may be required. Doesn't seem like a Mfr would want such a tube. A cheap 12GC6 or 6FW5 has very similar curves but with 1/2 the Rp. Although those won't work for ultralinear.
So a better OT may be required. Doesn't seem like a Mfr would want such a tube. A cheap 12GC6 or 6FW5 has very similar curves but with 1/2 the Rp. Although those won't work for ultralinear.
Last edited:
Mcintosh and Marantz were later, they belong to the HiFi era. The unity-coupled Mc became commercially available In 50s, and first Marantz preamp in 60s. In late 40s, the top Belvedere cost same as Chevrolet or Plymouth car. Isn't it high end enough? Magnavox later produced cheap mass market stereo consoles with 6V6 amplifiers, I think that's what you mean. The Belvedere is in totally different class.
In 40s, boxed speaker's were not yet invented (lucky them). Regular arrangement was open box or baffle. Drivers had lightweight cones and strong field coil magnets; they depended on natural (acoustic) damping and did not require low output impedance amplifier for good bass. Magnavox engineers didn't realize then that they invented current drive, which is superior to voltage drive because it lowers speaker-generated distortion. Because speaker distortion dominates overall distortion of the system, there is little rationale for chasing down amplifier distortion beyond 1% when speaker THD is 5%. But anything that lowers speaker distortion brings about significant improvement.
I'm olde enough to remember those things. They are/were called Public Address Amplifiers.
Inputs for mikes & phone turntables, Crap!👎
No. Belvedere was not PA. It was home radio for discriminating listeners (who could afford it). Before calling it crap, you should have done some Google search.
It is relevant, but look at the y axis scale: 0.1 dB vs 1 dB. I don't think 1 dB difference is readily perceptable.
Yes, the meaning of my previous post is: with the impedance of the speaker that drops to 4 (not so uncommon, some speakers go even lower), and a DF of 10, we have 1 dB difference (the minimum difference that human ear can detect). With a DF of 2-3 that difference would become higher. Is it 2 or 3 really enough? I've always read that above 20 is not that relevant, that's why I was asking about the order of magnitude.
To add on the Belvedere, this is a quote from a 1947 Magnavox promotional:
"...All the music - subtle filigree of overtones, the firm powerful bass - is so faithfully reproduced that listening becomes a new and wonderful experience... Superb musical instrument... Compare it to other radio phonographs regardless of price... once you hear it, you won't be satisfied until you own one... Equipment designed to take full advantage of coming improvements in radio and recording technique... $450, with FM section $62 extra".
"Coming improvements" at the time were FM and LP, high fidelity sources. It was not a machine designed for playing primarily AM radio and 78 rpm records. Of course, one should take advertisement claims with a grain of salt, but if we believe that 1930s WE amplifiers are unsurpassed, why not give credence to top achievements of other leading manufacturers of those days?
25 W PP 6L6 amplifier with no global negative feedback.
"...All the music - subtle filigree of overtones, the firm powerful bass - is so faithfully reproduced that listening becomes a new and wonderful experience... Superb musical instrument... Compare it to other radio phonographs regardless of price... once you hear it, you won't be satisfied until you own one... Equipment designed to take full advantage of coming improvements in radio and recording technique... $450, with FM section $62 extra".
"Coming improvements" at the time were FM and LP, high fidelity sources. It was not a machine designed for playing primarily AM radio and 78 rpm records. Of course, one should take advertisement claims with a grain of salt, but if we believe that 1930s WE amplifiers are unsurpassed, why not give credence to top achievements of other leading manufacturers of those days?
25 W PP 6L6 amplifier with no global negative feedback.
Last edited:
To add on the Belvedere, this is a quote from a 1947 Magnavox promotional:
----------------------
Anybody got a schematic of that or something similar?
There is an example of a PP 6L6 output section in Beitmans, a pre WW2 Philco receiver.
----------------------
Anybody got a schematic of that or something similar?
There is an example of a PP 6L6 output section in Beitmans, a pre WW2 Philco receiver.
Attachments
Last edited: