OK, back of envelope calculation (including wild guesses marked by []):a.wayne said:
Noticeable distortion means 50% drop [1], so 3 secs means a time constant of 4.3 s.
Assume normal listening level means 10% of full output current draw [2] so at full output would be 0.43 s.
In 10 ms this would give a droop of 2.3%.
So 'play unplugged for 3 secs' is just another way of requiring no more than 2.3% droop between charging pulses. You can adjust this figure up or down a bit but we seem to be talking about a few percent ripple at full output.
The basic UcD invention of Bruno's has a half-bridge output. However it's fairly straightforward to connect two of them as a full bridge and synchronize the outputs.
As it is intrinsically a variable-frequency design, the issue of heterodyning and audible artifacts of same in multiple amp systems comes up as a potential problem, although any such behavior using Hypex modules is hotly denied by the company. Philips (the patent owner) will let you use the topology as long as you use NXP semiconductors by the way. I think it is an excellent approach if class D is dictated by considerations of efficiency.
Brad
Undoubtly.
They estimated the max ripple amplitude at wich the amps s NFB
would be efficient enough to bring it down to the amps s intrinsical output noise.
At 5% ripple , about 2.5V Pk generaly , an amp need 100dB PSRR
to bring it below its own output noise.
At the time this could be easily accomplshed with an RC filter for the
front end and get the remaining rejection with 60 to 80dB NFB , easily
available at 100hz or so...
All that Audio Research junk.What would you have favored/listened to, if they hadn't fckd up your mind in France ?
And Ampzilla, maybe a Dynaco if i was slumming. Yep, they screwed me up bad, those damn French.
Cool! Was that 6-8 volts on the -12dBFS signal after you had set your levels?
All that Audio Research junk.
Actually, I know from experience that Audio Research's killer markets in Europe were France and Italy.
But Ampzilla ... hmmm ... yummy! James's baby which, just like a real baby, seems to grow over time. It matures.
And James is a member of my small Trio Amp Club, with John here and Matti Otala in Finland.
Have blown fuses , haven't hurt any amplfiers yet , the weak ones had their arms raised before the first note, anyway No protection necessary , just make sure output stage bigger than PSU ..
Sounds good to me ...
Damn .. you love NFB, use it like Geritol ..
Nah , RMS meter direct voltage from amplifier ...
I can chk it by your method later , i had left my warble tone disc , so just went ahead and measured direct voltage being used ...
Oh yes it does! Those who do not learn from history are doomed to repeat it.
It's not how it's set for, it's the idea behind it, and ideas never get old. Not good ones, anyway. They evolve and live on.
Besides, as usual, you are evading the key point. The key idea was that even in those dark, old days of audio, evaluating power supplies was a very much "in" idea, albeit in a very roundabout way. The bigger the dufference between your nominal RMS power and your dynamic power, the better PSU you had to have to make it happen.
Well said . The American V8 lives on because it does the job better ( best use of parts and relativity compact , good ( ish ) dynamic balance ) . It is not perfect and perhaps that makes it even better ? I was speculating that the side valve V8 of 1927 would probably make a good diesel engine if fitted with a suitably tough crank ( diesel's are to F1 standards in that ) . We will never know because no one will invest to show us . It would be a dream for the DIY mechanic . Overhead cam and a belt for a diesel engine . Give me push rods and gear drive please with hydraulically adjusted followers ( oil filter on the engine front ) .
One protection system perhaps worth trying is a constant current source . If below constant current it does nothing ( looses 0.3 V ) . The difference is it can come on at the critical moment . I find the usual protection systems come on gradually . That sound's as if it is a good idea ? Maybe less good as it's effects are earlier . Also with a constant current PSU the amp can sit like that all day into a short . A fan can come on or a trip if wanting to be sensible with heat sinks . On the face of it CCS should be much the same as it is the same transistor curves we are using . In practice I find it isn't as pulling down the input is not the same as just switching on a CCS . I have to admit I have only my PSU experience to go by . I never design amps with protection so might have a limited knowledge ? The CCS type class A has it's advantages in this ( JLH ) . You abuse it and it clips . Douglas Self says it well . If using an over bias class B you can have the best class A . The amp simply goes into AB if abused . He then say's the amp will attempt to supply current regardless unlike conventional class A and protection will be required . A CCS PSU and over biased B will provide a better amp than either of the alternatives . Also if using my 4 section PSU ( 25 V x 4 ) it can be configured as A or B . Use something like an octal valve base with either series or parallel PSU . The octal plug also has the required CCS resistor and bias . That's 20 W class A , 100 W class AB .
One thing which slightly amuses me is people speak of progress here . With the diminishing transistor types we will be in luck just to stay still . I think a fresh look at Quads current dumping is worth doing . If it was hybridized with class G it might be excellent ( it is G , what I mean is dumpers come in at higher levels ) . The reason is feed-forward amps can use very slow dumper transistors with less high frequency distortion compared with feedback ( Quad uses both ) . I saw some 100 V germanium's in TO3 , I fancied to buy some just to see how slow one can go . I think we owe it to the DIY Audio Forum to look at class G . The 5 V levels was a good piece of research . I would sanction switch mode PSU's ( just seen some nice 48 V ones at Rapid UK ) . With that the 70 % efficiency barrier can be met . The dumpers being collector connected for maximum efficiency and performance . Use conventional PSU's once efficiency established . I think we then can measure the actual power consumption over one hour . One side class G to other D in a stereo system playing mono . I suspect there will be very little difference . Heat sink size also . Miniscule or small , who cares as small will do nicely .
Someone said about Hypex in half bridge . I think that's hype . If it wasn't it would be SE . I was glad hetrodyning was mentioned . I remember putting a class D amp on my scope for the first time . I couldn't believe what I saw . Sure my spectrum analyzer said it was OK ( ish ) in the audio band . Still couldn't believe it could work at all . If I made a conventional amp with problems like that I would doubt it could ever sound good . I would also be too ashamed to market it . Class D is the diesel of amplifiers . All very impressive . It is no jet engine .
Nige, the effects are gradual because the industry does not want your amp to cut out when you overload it, they prefer to limit but not stop power delivery.
It's also gradual because many mass produced amps are in fact highly strung in their output stages (power transistors are expensive, need matching, etc) and use the cheapest transistors available to the manufacturer, not the best for the purpose.
Lastly, their PSUs are also highly strung, they are below what they should be because that starts to be expensive very soon, and customers in the salon do not see the PSUs, but they sure see rows of LEDs, switches they have no idea about, etc, in short, "features".
Here's a specific example. I have always thought of Onkyo as at least as serious as the best of mass manufacturers, if not better. Imagine my hearbrake when I discovered that their M-282 power amp, rated at 100 WPC into 8 Ohms, uses just one pair of Toshiba's 2SC3281/2SA1302. One pair!
For the love of Christ, those are nominally 150W devices, squeezing all of 100 W from them in a continuous mode should be declared a genocide! So, what do you think will happen when that amp encounters some more difficult speakers? Not to even mention Wayne's 1 Ohm monsters. It'll do all it can - it'll try, do a little and then promptly shut down.
And THIS is the other side of the "how much uF" discussion - what use many uF of capacitors, when you have a grossly undersized output stage, so the parallel and equal question should be: how many of which output devices do we need?
It's also gradual because many mass produced amps are in fact highly strung in their output stages (power transistors are expensive, need matching, etc) and use the cheapest transistors available to the manufacturer, not the best for the purpose.
Lastly, their PSUs are also highly strung, they are below what they should be because that starts to be expensive very soon, and customers in the salon do not see the PSUs, but they sure see rows of LEDs, switches they have no idea about, etc, in short, "features".
Here's a specific example. I have always thought of Onkyo as at least as serious as the best of mass manufacturers, if not better. Imagine my hearbrake when I discovered that their M-282 power amp, rated at 100 WPC into 8 Ohms, uses just one pair of Toshiba's 2SC3281/2SA1302. One pair!
For the love of Christ, those are nominally 150W devices, squeezing all of 100 W from them in a continuous mode should be declared a genocide! So, what do you think will happen when that amp encounters some more difficult speakers? Not to even mention Wayne's 1 Ohm monsters. It'll do all it can - it'll try, do a little and then promptly shut down.
And THIS is the other side of the "how much uF" discussion - what use many uF of capacitors, when you have a grossly undersized output stage, so the parallel and equal question should be: how many of which output devices do we need?
BTW, the American V8, and historically it IS American, Ford made it first way back in the twenties, is a rapidly dying breed in these days of consumption consciousness. It was, after all, the Americans who called them "Gas guzzlers".
These days, even the Americans who pay for their petrol a lot less than us Europeans, are fast becoming consumption conscious.
In Europe, this has already become a cult. And, truth be told, the latest generation of "bonsai" engines, 1 to 1.6 litres, do have turbo chargers, making them much more efficient. The Germans are in fact in the lead so far that I doubt anyone can catch up with them any more; I was amazed to learn that a VW Golf using a 1.4 litre turbo blown engine with 122 HP drinks just 7 litres of fuel in the city! 5 litres on the highway. And trust me on this, it lacks NOTHING in terms of power, but it is a family car, not a racer, for racing get an R32, which has a turbo blown 2 litre engine delivering 270 HP.
At that engine efficiency (HP/Litre), the Chevy Corvette entry level 6.2 litre V would have some 837 HP, power enough for at least two trucks.
Just a side bar, no more car talk here.
These days, even the Americans who pay for their petrol a lot less than us Europeans, are fast becoming consumption conscious.
In Europe, this has already become a cult. And, truth be told, the latest generation of "bonsai" engines, 1 to 1.6 litres, do have turbo chargers, making them much more efficient. The Germans are in fact in the lead so far that I doubt anyone can catch up with them any more; I was amazed to learn that a VW Golf using a 1.4 litre turbo blown engine with 122 HP drinks just 7 litres of fuel in the city! 5 litres on the highway. And trust me on this, it lacks NOTHING in terms of power, but it is a family car, not a racer, for racing get an R32, which has a turbo blown 2 litre engine delivering 270 HP.
At that engine efficiency (HP/Litre), the Chevy Corvette entry level 6.2 litre V would have some 837 HP, power enough for at least two trucks.
Just a side bar, no more car talk here.
"History doesn't repeat itself but it often rhymes."
"History is just one damn thing after another."
1971 was over 40 years ago. In the world of electronics, that's three lifetimes. Yes, industry groups put some standards together back then which MAY have been slightly less useless than the "music power" or "peak power" advertising numbers that were in vogue during the '60s. And yes, if one prefers to live in 1971 and not recognize that the landscape is totally different now, that's fine.
- Status
- Not open for further replies.