The best power cords I encountered so far are self-assembled.
Using:
Neotech UP-OCC Bulk AC Power Cable, cut to length: VH Audio - Wire and Cable (either NEP-3003 or NEP-3001, according to the power consumption of the client);
and
Oyaide, both Male AC plugs and IEC plugs: Oyaide, Furutech and WattGate AC Connectors (each different plugs' material gives a different tonal balance).
Ok, now when it comes to choke input. It is a GREAT way to make an all out power amp. I designed into the Gale power amp in 1976, for example. BUT it was not a good, or easy engineering solution, because it MUST be clamped, and you must run very high quiescent current to keep the choke size reasonable. (the choke not being bigger than the power transformer), and to keep the caps from being over-voltaged and destroyed.
The next best is the pi or CRC filter. Good filtering, low surge currents, but only fair regulation, fairly safe to use without a load, the best for Class AB amps with modest quiescent current. Third best, and most simple and popular is just big cap input, hoping to not exceed peak currents that will blow the input fuses, having to overrate the transformer because of I(squared)R losses, poor ripple, requiring big caps or active filtering. This is still a good enough compromise for most amps.
Let's not fight about it, different approaches for different people. That's OK.
The next best is the pi or CRC filter. Good filtering, low surge currents, but only fair regulation, fairly safe to use without a load, the best for Class AB amps with modest quiescent current. Third best, and most simple and popular is just big cap input, hoping to not exceed peak currents that will blow the input fuses, having to overrate the transformer because of I(squared)R losses, poor ripple, requiring big caps or active filtering. This is still a good enough compromise for most amps.
Let's not fight about it, different approaches for different people. That's OK.
I built my first really powerful balanced bridge power amp for Ampex Research in 1969. It was rated at 50V/ 50A , or about 2000W. I sold my boss on it, BECAUSE it could reverse a motor's drive current instantly without relays, or loss of power supply efficiency. I also had to make it current out, as it was a motor drive, so I had to use 2 separate negative feedback loops and two positive feedback loops, making my own version of a Howland Current pump. What a hassle!. But it worked. Later, they decided, when they could not figure out the circuit intuitively, for a simpler solution. As they declined to hire me on as a consultant, after I left to get back into audio, I could care less. However, I offered a version of this design to Saul Marantz and Jon Dahlquist in 1973-4, but they declined to make it. I then offered it to Mark Levinson, but he wanted a small amp, that became the JC-3--ML2.
Finally, in 1975, Ira Gale took me up, and I spent 6 mo in Switzerland and London working on it. The Gale amp was 250W/ch 4 quadrant symmetry.
Then, moving back to the USA, I got Symmetry to build a 100W/channel 4 quadrant symmetry power amp that we showed at the Tokyo Audio Fair in 1978.
Finally, I built a 250W version for VMPS in 1980 on 3 airblown heatsinks, and 64W (2A Iq per channel) that we took to CES for a number of years.
Yet, WHY does Parasound not have such an amp? Politics, really. Nobody wants to put the money and time into making a whole new product, so I stick with what we have.
The JC-1 is an IMPROVED (by every means debated here) version of the HCA-3500. I did it on purpose to show what 'subtlety' does to improve the sound, like better layout, parts, high speed diodes, better wire, better caps, etc, etc. That is the real difference in cost (and sound) between the HCA-3500 and the JC-1.
Could I do better? OF COURSE, but not without a lot of money, upfront.
Finally, in 1975, Ira Gale took me up, and I spent 6 mo in Switzerland and London working on it. The Gale amp was 250W/ch 4 quadrant symmetry.
Then, moving back to the USA, I got Symmetry to build a 100W/channel 4 quadrant symmetry power amp that we showed at the Tokyo Audio Fair in 1978.
Finally, I built a 250W version for VMPS in 1980 on 3 airblown heatsinks, and 64W (2A Iq per channel) that we took to CES for a number of years.
Yet, WHY does Parasound not have such an amp? Politics, really. Nobody wants to put the money and time into making a whole new product, so I stick with what we have.
The JC-1 is an IMPROVED (by every means debated here) version of the HCA-3500. I did it on purpose to show what 'subtlety' does to improve the sound, like better layout, parts, high speed diodes, better wire, better caps, etc, etc. That is the real difference in cost (and sound) between the HCA-3500 and the JC-1.
Could I do better? OF COURSE, but not without a lot of money, upfront.
Last edited:
Hi,
As it so happens, Yes I did. BUT it was not the current drive that destroyed the woofers. It was the motional feedback loop I used in addition letting go, which would make the Woofers voicecoils jump out of their gaps.
I eventually overcame the problem using higher power rated woofers and using a separate microphone for feedback, just like Meyer did recently on the X-10, instead of trying to derive the feedback signal from voicecoils...
The Speaker was 3-Way, with a 5" Paper Cone Studio Full range driver for a Midrange and initially the East German 1" Plastic dome tweeter, later a Motorola Piezo "Lemon Squeezer" (Voltage driven of course), Woofers where dual 8", all drivers sealed boxes.
I suspect in a PA System where you drive woofers to within an inch of their life current drive may be a problem, but even in the 80's we experimented with active systems that had limiter on each channel based on RMS Power delivered to the speaker (thermal) and sliding highpass filters on the Bass Channels that prevented excessive excursions. Actually, to be precise, they where equalising highpass filters to get EBS reflex boxes "flat" that first lost their peak and eventually slid one of the filters corners up in frequency as the signal excursion increased, all the way to 80Hz.
The only drawback was that these limiters needed a loop-though from the Amplifiers (to sense current and voltage) and needed careful setup for each changing speaker system, especially the excursion limiters.
Nowadays this kind of stuff is easily implemented in DSP and in a pure feed-forward way and I doubt you'd blow anything if such a system is set up right.
Ciao T
As it so happens, Yes I did. BUT it was not the current drive that destroyed the woofers. It was the motional feedback loop I used in addition letting go, which would make the Woofers voicecoils jump out of their gaps.
I eventually overcame the problem using higher power rated woofers and using a separate microphone for feedback, just like Meyer did recently on the X-10, instead of trying to derive the feedback signal from voicecoils...
The Speaker was 3-Way, with a 5" Paper Cone Studio Full range driver for a Midrange and initially the East German 1" Plastic dome tweeter, later a Motorola Piezo "Lemon Squeezer" (Voltage driven of course), Woofers where dual 8", all drivers sealed boxes.
I suspect in a PA System where you drive woofers to within an inch of their life current drive may be a problem, but even in the 80's we experimented with active systems that had limiter on each channel based on RMS Power delivered to the speaker (thermal) and sliding highpass filters on the Bass Channels that prevented excessive excursions. Actually, to be precise, they where equalising highpass filters to get EBS reflex boxes "flat" that first lost their peak and eventually slid one of the filters corners up in frequency as the signal excursion increased, all the way to 80Hz.
The only drawback was that these limiters needed a loop-though from the Amplifiers (to sense current and voltage) and needed careful setup for each changing speaker system, especially the excursion limiters.
Nowadays this kind of stuff is easily implemented in DSP and in a pure feed-forward way and I doubt you'd blow anything if such a system is set up right.
Ciao T
I once set the monitor speakers at my mixing console on fire - does that count?
It depends on voltage or current drive.
In the ENTEC subs I sensed the position of the voice coil to DC. It worked great. . .until you hit the excursion limit. At that point the feedback did what it was supposed to do and slammed the driver hard into its mechanical limits.
The stuff we do today with DSP and limiters makes it almost impossible to damage a driver or hear distress in a packaged system. The capabilities of $30 worth of electronics far exceeds what audiophile systems are capable of in terms of eq correction, crossovers, multi amplification and really benign limiting. And all very limited by plastic enclosures, zoomy styling and market driven sound.
The stuff we do today with DSP and limiters makes it almost impossible to damage a driver or hear distress in a packaged system. The capabilities of $30 worth of electronics far exceeds what audiophile systems are capable of in terms of eq correction, crossovers, multi amplification and really benign limiting. And all very limited by plastic enclosures, zoomy styling and market driven sound.
Hi,
Same here, but even with a rumble filter the NI drive would go to catastrophic runaway when played loud. The reasons are really obvious. Using a Microphone worked way better, despite the extra delay (it just meant crossing over to the woofer lower).
At the time I did not know the Backes & Mueller approach of sticking the microphone directly to the cone. If you do that and design the PI Regulator circuit you can stay stable to many 100Hz...
Ciao T
Same here, but even with a rumble filter the NI drive would go to catastrophic runaway when played loud. The reasons are really obvious. Using a Microphone worked way better, despite the extra delay (it just meant crossing over to the woofer lower).
At the time I did not know the Backes & Mueller approach of sticking the microphone directly to the cone. If you do that and design the PI Regulator circuit you can stay stable to many 100Hz...
Ciao T
Hi,
Yes, it is elegant. BUT it is subject to many problems.
First, we derive the back EMF using a bridge circuit with L & R.
But L in the driver varies with voice coil position and R varies with temperature. So we need to strongly limit the amount of NFB or we risk instability.
Secondly, the back EMF from the voice coil is dependent on the magnet field strength, which may be reasonably constant for the linear excursion of the driver (though you may find this belief shattered if you ever hook up a real driver to Mr. Klippels analyser), but as the voice coil progressively leaves the magnet gap it declines.
So not only is the drive to the cone reduced as the excusions become very large, so is the feedback signal, so in fact our feedback loop will overcompensate and we get into a positive feedback situation where the voice coil is pushed further and further out of the gap and the feedback signal declines more and more. Eventually suspensions and cones tear, unless the voice coil melted first. It is quite dramatic either way.
So it may seem elegant on the first blush, but the practical issues should not be underestimated...
Ciao T
Yes, it is elegant. BUT it is subject to many problems.
First, we derive the back EMF using a bridge circuit with L & R.
But L in the driver varies with voice coil position and R varies with temperature. So we need to strongly limit the amount of NFB or we risk instability.
Secondly, the back EMF from the voice coil is dependent on the magnet field strength, which may be reasonably constant for the linear excursion of the driver (though you may find this belief shattered if you ever hook up a real driver to Mr. Klippels analyser), but as the voice coil progressively leaves the magnet gap it declines.
So not only is the drive to the cone reduced as the excusions become very large, so is the feedback signal, so in fact our feedback loop will overcompensate and we get into a positive feedback situation where the voice coil is pushed further and further out of the gap and the feedback signal declines more and more. Eventually suspensions and cones tear, unless the voice coil melted first. It is quite dramatic either way.
So it may seem elegant on the first blush, but the practical issues should not be underestimated...
Ciao T
I used piezoelectric crystal with JFET buffer on the cone. Later I found that positive feedback by voltage can be used cheaper and with almost the same effect.
Last edited:
It was about 30 years ago. The biggest woofer I had then was from bass guitar cab, 15" Soviet cinema driver with resonant frequency around 60 Hz...
John,
Well, at the time for us bigger woofers where not an option.
As to positive current feedback, do you mean negative impedance?
Ciao T
Well, at the time for us bigger woofers where not an option.
As to positive current feedback, do you mean negative impedance?
Ciao T
- Status
- Not open for further replies.