You do not insist on 2H anymore?
What has changed your mind??But even 100mH is a nonsense, dear Vladimir.
Right. And "Real World" test would mean load of the amp on a load that is meant to be used with this particular amp. For example, load of 5W SE amp on a single high efficiency full range speaker, load of 100W P-P amp on 3-way speakers with crossovers, and so on. Some speakers are meant to be driven by amps with low output resistance, some need higher output resistance, some need to supply current (theoretically infinite output resistance), some even want negative output resistance for proper bass response; there are no amps for all speakers, and no speakers for all amps, and they compliment each other.
It is useful to check how well the amp would drive real speakers meant to be used with it, Dependence of distortions on power gives a good picture; dependence of distortions on load resistance gives a good useful picture. But measuring of frequency response of an amp loaded on complex load is meaningless and confusing, though sounds quite scientifically-objective: amps with lower output resistance would show flatter curve. But it means nothing because we want flatter curve by sound pressure, also we want lower distortions. Some speakers distort less when driven by amps with current output, while other speakers distort less when driven by voltage outputs.
Thanks John, I have - tho with smaller values. Just wondered if you might be doing the "two caps to ground" one from hot, one from neutral trick, or just one across the line. Cheers.
Was also wondering if when getting up into these cap values it's starting to work as power factor correction?
Was also wondering if when getting up into these cap values it's starting to work as power factor correction?
Unloaded peak voltage is pretty darn high, compared to loaded average, and what about 'shock ringing' on top? You know the resonance of the load cap and the choke? IF I were you, I would do a transient analysis of a choke input power supply like I did with ECAP on an IBM mainframe in 1966, yes, 45 years ago. Just switch off the current and see what happens.
Pano, next to me is a 10uF cap plugged into the wall outlet. Been there for years. I personally prefer oil filled polypropylene with an extra high voltage rating in a metal can. However, the guys in Taiwan put in something slightly different in the JC-3. 10uF is the 'magic' value, we worked this out over 10 years ago.
Last edited:
In honor of the Blu-ray release of "Meet Me in St. Louis".
"It will take me at least a week to dig all my dolls up from the cemetary".
Hi,
I'm not John, but I also do this kind of thing.
Some notes. The Cap's draw a blind current which will spin your meter. In the US 10uF is around 250mA (around 30W/H clocking the meter) in Europe it's more like halve an Amp (around 120W/H clocking the meter). Not the end of the world, but for example I just moved house and did the place with massive amounts of LED spot lights and LED Lightbulbs, so with ALL lights on the new place draws less than 100W, new "high efficiency" fridge as well as Aircon too... So a single 10uF cap can throw a spanner into being "green" (well, so do Class A Tube Amp's)...
Second, as the mains have substantial inductance it is necessary to "snubber" the capacitor directly connected across life & neutral, of ringing can become a problem, especially if gear is plugged in right next to it (as the rectification supplies ample stimulus for the filter to "ring")...
Third, be safe. Capacitors across Life & Neutral of the mains have strict requirements for flammability and many other ratings. They are hence marked with Symbols like UL and CE and TUV and all are shown as "X.." on the case. Failure to use these may void your insurance if your place ever burns down (even if the wrong cap's where not the cause) as well posing a risk, possibly to ones family.
Ciao T
I'm not John, but I also do this kind of thing.
Some notes. The Cap's draw a blind current which will spin your meter. In the US 10uF is around 250mA (around 30W/H clocking the meter) in Europe it's more like halve an Amp (around 120W/H clocking the meter). Not the end of the world, but for example I just moved house and did the place with massive amounts of LED spot lights and LED Lightbulbs, so with ALL lights on the new place draws less than 100W, new "high efficiency" fridge as well as Aircon too... So a single 10uF cap can throw a spanner into being "green" (well, so do Class A Tube Amp's)...
Second, as the mains have substantial inductance it is necessary to "snubber" the capacitor directly connected across life & neutral, of ringing can become a problem, especially if gear is plugged in right next to it (as the rectification supplies ample stimulus for the filter to "ring")...
Third, be safe. Capacitors across Life & Neutral of the mains have strict requirements for flammability and many other ratings. They are hence marked with Symbols like UL and CE and TUV and all are shown as "X.." on the case. Failure to use these may void your insurance if your place ever burns down (even if the wrong cap's where not the cause) as well posing a risk, possibly to ones family.
Ciao T
Well you want to point out that when you need a special power cord to make your gear work better, then the gear could have been designed better, but you are willing to put up with a loudspeaker that isn't designed to work from a voltage source!!!!
Yes vacuum tube amplifiers used to be designed to deliver maximum power to the load and followed the INCORRECT maxim that maximum efficiency occurs when the source impedance matches the load impedance.
Of course maximum efficiency occurs when the source impedance is zero! When they match the source must dissipated half of the total power. That is actually a problem Edison had. He though his power generation should have the same impedance as the load. Westinghouse and Tesla got it right.
Once you understand the difference between maximum efficiency and maximum power transfer things cool down quite a bit. That is why a properly designed loudspeaker expects a low source impedance.
John,
What about it? In my commercial implementations the frequency is two decades below the ripple frequency.
I use choke input supplies of my own design almost every day. I know the voltages and I sure as hell analyse the resulting systems for resonances. I mean I would be a complete and total moron if I did not. I even tend to hang EMI sniffer coils around things.
It is really not all that difficult, in tube circles we are really, really au fait on Choke Input Supplies. I cannot give all the secrets away, but there are many ways to get what with chokes.
Who says I don't? I actually prefer to look at the supplies step response to load change, but in the end you see the same effects at work.
Incidentally, if I switch the current off it is on the mains side. The PSU cap happily absorbs remaining stored charge of the choke until the whole system drains.
Ciao T
What about it? In my commercial implementations the frequency is two decades below the ripple frequency.
I use choke input supplies of my own design almost every day. I know the voltages and I sure as hell analyse the resulting systems for resonances. I mean I would be a complete and total moron if I did not. I even tend to hang EMI sniffer coils around things.
It is really not all that difficult, in tube circles we are really, really au fait on Choke Input Supplies. I cannot give all the secrets away, but there are many ways to get what with chokes.
Who says I don't? I actually prefer to look at the supplies step response to load change, but in the end you see the same effects at work.
Incidentally, if I switch the current off it is on the mains side. The PSU cap happily absorbs remaining stored charge of the choke until the whole system drains.
Ciao T
Ed,
Actually, given that electrodynamic speakers rect to current, only current and nothing but, this may be a good thing...
And Tube Amp's where almost never designed for matched load, but usually a ratio of 3:1 (rare exceptions exist), for example the 2A3 with it's 800 Ohm Anode impedance was specified for 2.5K load SE and 5K Push-Pull, giving a 3.125 Ratio...
Ciao T
Actually, given that electrodynamic speakers rect to current, only current and nothing but, this may be a good thing...
And Tube Amp's where almost never designed for matched load, but usually a ratio of 3:1 (rare exceptions exist), for example the 2A3 with it's 800 Ohm Anode impedance was specified for 2.5K load SE and 5K Push-Pull, giving a 3.125 Ratio...
Ciao T
- Status
- Not open for further replies.