If you mean a technical interesting issue, I can only think about what I alluded already at.
Each half of the balanced amp sees only half of impedance of the speaker. In certain cases, this will be quite low, the output stage has to manage the current.
There is also doubled output impedance, which in my view is a no brainer.
Vladimir,
Yes, there are more of course I would use myself, however this set tells a lot.
First, one of these two simply plots the voltage drop against the current, in other words the DCR of the supply. As diodes are involved there is an unavoidable non-linearity. The other looks simply at the AC impedance of the Supply.
In an ideal supply (LCLC filtered using non-polar capacitors of very large values) any of these measurements yield disappointing results.
If you where to run the tests on the power supply of the AMR AM-77 or AM-777 you would find that DCR and AC Impedance are essentially the same, nonlinear expansive (that is PSU impedance lowers the more current is drawn) and around 50milliohm for DC to a few 100KHz for normal signal levels falling to under 5milliohm at rated power...
Arguably, this type of supply does not use chokes, instead it tries to "simulate" a near-ideal CLC supply without the L. It works rather well.
This is simply not true. Going outside audio many systems with widely varying current demand use choke input supplies (I am specifically thinking of radio transmitters) and again, they are in use precisely because they offer a lower variation of supply voltages with load.
So put in the simplest and most direct way, what you state is the opposite of the actual situation and repeating another 100 times will not change these simple facts.
We may debate the precise design approaches and their results, however to simply state "LC filter have worse dynamic charateristics" is wrong. Correctly implemented they have superior dynamic characteristics to simple C Filters. Again, you can repeat your position another 100 times and it will not be any more true.
If and when I have nothing better to do I'll find you some references. Until then, how about you look at the Electronics Book to which Anatoly kindly posted a link. It seems to have the relevant details.
I'll leave you with one more link:
LC Choke-Input Filter
To quote the critical part:
"The LC choke-input filter is used primarily in power supplies where voltage regulation is important and where the output current is relatively high and subject to varying load conditions. This filter is used in high power applications such as those found in radars and communication transmitters."
quod erat demonstratum
Ciao T
Yes, there are more of course I would use myself, however this set tells a lot.
First, one of these two simply plots the voltage drop against the current, in other words the DCR of the supply. As diodes are involved there is an unavoidable non-linearity. The other looks simply at the AC impedance of the Supply.
In an ideal supply (LCLC filtered using non-polar capacitors of very large values) any of these measurements yield disappointing results.
If you where to run the tests on the power supply of the AMR AM-77 or AM-777 you would find that DCR and AC Impedance are essentially the same, nonlinear expansive (that is PSU impedance lowers the more current is drawn) and around 50milliohm for DC to a few 100KHz for normal signal levels falling to under 5milliohm at rated power...
Arguably, this type of supply does not use chokes, instead it tries to "simulate" a near-ideal CLC supply without the L. It works rather well.
This is simply not true. Going outside audio many systems with widely varying current demand use choke input supplies (I am specifically thinking of radio transmitters) and again, they are in use precisely because they offer a lower variation of supply voltages with load.
So put in the simplest and most direct way, what you state is the opposite of the actual situation and repeating another 100 times will not change these simple facts.
We may debate the precise design approaches and their results, however to simply state "LC filter have worse dynamic charateristics" is wrong. Correctly implemented they have superior dynamic characteristics to simple C Filters. Again, you can repeat your position another 100 times and it will not be any more true.
If and when I have nothing better to do I'll find you some references. Until then, how about you look at the Electronics Book to which Anatoly kindly posted a link. It seems to have the relevant details.
I'll leave you with one more link:
LC Choke-Input Filter
To quote the critical part:
"The LC choke-input filter is used primarily in power supplies where voltage regulation is important and where the output current is relatively high and subject to varying load conditions. This filter is used in high power applications such as those found in radars and communication transmitters."
quod erat demonstratum
Ciao T
A few corrections are in order-
A cap across the power line won't increase the power consumed, a pure reactance doesn't consume power. The power consumed would be in the wires connected to the cap but not in the cap. Your power meter does a good job of measuring real power.
The power factor the cap fixes the power factor from all of those inductive motors in your house (if you still have any, possibly your refrigerator). The power factor problem is the harmonics from the rectified power supply. And I believe those harmonics will need to be controlled per the latest EU regulations.
A gas tube across the line will eat a surge just fine and promptly blow the fuse or breaker. Once the gas tube is ignited the power needs to be removed to reset it. Just use an MOV. They are benign, reliable and economical.
A cap across the power line won't increase the power consumed, a pure reactance doesn't consume power. The power consumed would be in the wires connected to the cap but not in the cap. Your power meter does a good job of measuring real power.
The power factor the cap fixes the power factor from all of those inductive motors in your house (if you still have any, possibly your refrigerator). The power factor problem is the harmonics from the rectified power supply. And I believe those harmonics will need to be controlled per the latest EU regulations.
A gas tube across the line will eat a surge just fine and promptly blow the fuse or breaker. Once the gas tube is ignited the power needs to be removed to reset it. Just use an MOV. They are benign, reliable and economical.
Hi,
The classic electromechanical meters I am familiar with use a pecial arrangement of coils to they react to voltage and current (thus measuring power), they quite happily rotate with nothing but a reactive component as load.
It is possible that modern electronic meters act differently, I have never investigated how they work and I rarely encounter them in practice.
Not just motors... Mains transformers also have an inductive component.
Yes, it is called the waveform crossing zero. I personally use MOV's, but always eye them with suspicion, they by far too much resemble two-legged fuses...
Ciao T
The classic electromechanical meters I am familiar with use a pecial arrangement of coils to they react to voltage and current (thus measuring power), they quite happily rotate with nothing but a reactive component as load.
It is possible that modern electronic meters act differently, I have never investigated how they work and I rarely encounter them in practice.
Not just motors... Mains transformers also have an inductive component.
Yes, it is called the waveform crossing zero. I personally use MOV's, but always eye them with suspicion, they by far too much resemble two-legged fuses...
Ciao T
On this scale it is probably lost in the round off, I doubt if 10uF changes your whole house power factor much, but you would need to take into account the utilities power factor penalty in the long run. We buy our power in bulk (41 households) and get an extra charge based on average power factor, because in fact it costs the provider.
T-
The mechanical meters are very good even with high power factors. The electronic ones are also good. The reading is probably a mix of real power heating your wires and the outer limits of the power factor.
The gas tube won't extinguish on a cycle to cycle basis. You need to remove power for a longer period.
More on MOV's when I have time but to keep it simple an MOV rated for twice the line voltage will probably never fail.
S-
Large power consumers do pay for power factor but individual consumers don't normally.
Sent from my T-Mobile G2 using Tapatalk
The mechanical meters are very good even with high power factors. The electronic ones are also good. The reading is probably a mix of real power heating your wires and the outer limits of the power factor.
The gas tube won't extinguish on a cycle to cycle basis. You need to remove power for a longer period.
More on MOV's when I have time but to keep it simple an MOV rated for twice the line voltage will probably never fail.
S-
Large power consumers do pay for power factor but individual consumers don't normally.
Sent from my T-Mobile G2 using Tapatalk
I also saw similar statements in some books. I do not know now what exactly authors had in mind while saying that. In that period I believed them and spent some efforts buying chokes and using them. Fortunately I soon stopped doing this, since for class A amps other solutions are even better. In my references, discarding any advantages of LC filter for AB amps, I proceed from statements of experts in audio, who did a lot of simulations and listening expertise. Also I would like to note, that chokes are very small energy stores compared to lytics. If they act in positive manner, usually it happens for rather specific and limited application conditions.
Excuse me for this dispute, but your and Anatoliy's references give nothing close to the subject.
A choke input supply with a choke which is too small (equivalently, has too low current draw) will show significant voltage variations with current. This just shows it is badly designed, perhaps by someone who does not understand choke input supplies. You can't simply add a choke at the input of an existing cap input PSU and expect to see an improvement. Chokes are like negative feedback - they have to be used properly otherwise they make things worse instead of better.VladimirK said:
The overvoltage problem when the load disappears is about 55%, so quite large but it can be controlled by adding some fat zeners which normally do nothing.
Subsonic resonance can be a problem, but occurs in both choke input and CLC supplies. If the load does not damp the resonance sufficiently, you could always add a snubber.
The 350V GDT's we used for K20 compliance fired on every cycle, thats what causes the damage (kicked back through the xDSL transformer).
- Ferraris principle energy counters (the mechanical ones) do only count real power. Also do the electronic ones.
- Newer household counters as well as industry counters are able to count four quadrants, i.e. reactive as well as real power, and forward / backward (if you have say a photovoltaic system)
- You cannot correct for power factors <>1 just with a capacitor if the reactive power is caused from switching diodes or saturated cores etc. It is only possible for mainly inductive loads like a motor eg.
- Newer household counters as well as industry counters are able to count four quadrants, i.e. reactive as well as real power, and forward / backward (if you have say a photovoltaic system)
- You cannot correct for power factors <>1 just with a capacitor if the reactive power is caused from switching diodes or saturated cores etc. It is only possible for mainly inductive loads like a motor eg.
I have a few hilarious Ed Wodenjack stories of my own. Is he still with us?
How many were direct to disk? I see The Boston Pops which probably was not? I have this one it even survives being youtubed (sort of) though not my kind of music. http://www.youtube.com/watch?v=tcWROzlO4sk
Last edited:
I have a few hilarious Ed Wodenjack stories of my own. Is he still with us?
I don't know if Ed is still around but I attended a direct-to-disc orchestral session he was producing with composer Morton Gould at Watford Town hall in the late 1970s. Engineer was Bert Whyte and John Meyer was operating the lathe. Ed was one of the largest-than-life characters I have encountered in 35 years of magazine work. Didn't he change his name in the late 1980s?
John Atkinson
Editor, Stereophile
Well, I see the evening has been busy. First, the 'Boston Pops record WAS direct disc. How do I know? I have a credit on the record jacket, and they would have told me.
This was the problem, back then, with direct disc recording: IF you screwed up the recording process, then EVERYTHING that you did to make the recording possible would be lost, and it sometimes was. I was an eyewitness to this happening at a Crystal Clear recording session here in the SF EBay, maybe 30 years ago or more. The band was ready, the equipment was set up, and POOF the cutting head was destroyed, apparently by a transient, as John Meyer would remove the cutting head protection just before a live take. Everybody had to go home and they were not happy. No wonder that they wanted a 'quality' back up recording for the Boston Pops. Think of the money already invested in getting the orchestra there! Some people think that DIGITAL even in its early conception, as it was then, sounds better than live. That is where that rumor came from.
In fact, that recording bothered Ed Wodenjak, because he did not think that it was 'forward sounding' enough to sell. He approached me at an AES before its release and had me hear it over headphones. I said that it sounded closer to a live performance than ANY other recording of an Orchestra in my experience, because that's what they sound like. My experience with orchestral music may not be as complete as some here, but I did have a long term relationship with Susan Milan, then principal flutist with the RPO in London, and then I later married a violinist, who was once a student at the IHEM or The Institute of Advanced Musical Studies, Montreux, were I worked at a faculty level for 1 1/2 years and I heard recitals virtually every evening. I have a pretty good ear for musical instruments. I KNOW that they would have put the digital credits on the record jacket, IF they had used it, as they even put my name there.
This was the problem, back then, with direct disc recording: IF you screwed up the recording process, then EVERYTHING that you did to make the recording possible would be lost, and it sometimes was. I was an eyewitness to this happening at a Crystal Clear recording session here in the SF EBay, maybe 30 years ago or more. The band was ready, the equipment was set up, and POOF the cutting head was destroyed, apparently by a transient, as John Meyer would remove the cutting head protection just before a live take. Everybody had to go home and they were not happy. No wonder that they wanted a 'quality' back up recording for the Boston Pops. Think of the money already invested in getting the orchestra there! Some people think that DIGITAL even in its early conception, as it was then, sounds better than live. That is where that rumor came from.
In fact, that recording bothered Ed Wodenjak, because he did not think that it was 'forward sounding' enough to sell. He approached me at an AES before its release and had me hear it over headphones. I said that it sounded closer to a live performance than ANY other recording of an Orchestra in my experience, because that's what they sound like. My experience with orchestral music may not be as complete as some here, but I did have a long term relationship with Susan Milan, then principal flutist with the RPO in London, and then I later married a violinist, who was once a student at the IHEM or The Institute of Advanced Musical Studies, Montreux, were I worked at a faculty level for 1 1/2 years and I heard recitals virtually every evening. I have a pretty good ear for musical instruments. I KNOW that they would have put the digital credits on the record jacket, IF they had used it, as they even put my name there.
Last edited:
- Status
- Not open for further replies.