Thanks, the whole thread reads like some James Joyce novel really.........
Yes, but not the speed of sound - that is a compression wave. At issue is a transverse or flex wave, which is far slower. A cantilever is about as flexible as an unstrung guitar D string - ie with no tension. If it were as long as a D string, it would not support its own weight in the same way. So one might imagine a guitar D string laid out on a table, driven at one end, with transverse vibrations propagating along it - that's about how slow. In a cantilever, lowest resonance happens when phase delay between ends is either 90 or 180 deg, which typically lies between 12kHz and 30kHz in practice.
That was the backup power for the IBM 360 at UCLA, circa 1980, giving things time to shut down in an orderly fashion in the event of a power failure.Motor-generators with a big flywheel in the garage. You know it makes sense
Yes one definitely needs the C across the transformer (or R-C). There are at least two things going on, the actual low-frequency diode conduction and the transformer leakage inductance interacting, and the much faster reverse recovery diode behavior, which of course is much better-controlled these days.
Noise and noise: when a friend got one of Hagerman's tube phono preamp kits and built it, he said he had a loud buzz. I assumed he meant in the electrical output, and he brought the unit for me to inspect. It was loud all right---acoustical emission from the ceramic disc cap across the HV secondary! I don't know how Steve failed to notice this---maybe someone substituted a different part and it was not auditioned.
I think we should start marketing them to audiots as the ultimate in clean power... And danger.... And mechanical noise.
Although the V10 diesel outside the office in my first job was seriously noisy. That was an old site (used to make propellers there in the 40s) and parts were pre national grid so about 12 generators on site, at least 3 of them manual start. By the time they had them running the mains was back on... Ah those were the days. not.
Thanks, the whole thread reads like some James Joyce novel really.........
Yes, but not the speed of sound -
Thanks, I didn't do any ME so I defer to your expertise. Confess now did you actually ever read Ulysses in its entirety.
In most of the mainframe installations of the period I encountered it was not backup power. The power was converted to 400 Hz perhaps to reduce the size of the transformers? Back then the disk drives needed to be sequenced on powerup since they drew so much power on startup. The printers were also very big and scary when operating. I think an Apple watch has more real computing power today.
The other virtue was the ride through when the grid got wonky. A backup generator would need to be 50 KW for one of those installations.
I would actually want to sell these to neurotic audiophiles: http://www.emersonnetworkpower.com/...s/discontinued product documents/sl-29300.pdf but they have been discontinued so missed another money making scheme
Ha ! Maybe nobody ever has ? I had a good try when I spent some time in Dublin, but didn't make it.
Batteries can do a LOT better than that !
Here's a paper which examines output impedance of batteries, coincidentally over the audio band, which illustrates both the resistive and reactive parts of output impedance are pretty much negligible in the application at hand on this thread. It also published a measurement method which could equally be applied to caps.
http://epg.eng.ox.ac.uk/sites/default/files/Howey/OnVehicleEIS_IEEEVTS_for_publication.pdf
No but great to sell to audiophools " Using F1 technology developed by ex-aerospace engineers" .
Actually quite nice.
The example cell is rated at 2.5 AH @ 3.3 Volts. Recharges to .8C at 10 amps in 12 minutes! Normal charge time is 1 hour. Intended for high current drain as high as 50 amps. Unit looks like it will take 10,000 charge discharge cycles in audio use. So 10 of them would power even some of J.C.'s preamps for a day of listening and recharge during bathroom breaks! Price is reasonable at $8.15 ea.
Only down side is the usual assumption cell impedance translates into noise level. That needs to be verified.
At .01 Ohms thats -218 dBV, below the level of NIST's auto-correlating amplifier which would be difficult to better. From what I've seen the real part of the impedance would be the low limit but some battery types show excess noise.
In any case you're not going to build an amplifier with 0 PSRR.
Last edited:
Actually folks do! A ratiometric switching amplifier does. That is why folks are putting global feedback into them. (Which brings up the issue of feedback's effect on noise.)
Depends on the application--ostensibly you can push more braking energy into the flywheel than you can through a generator (pound for pound, at least), which means in stop-and-go situations, they're (again, ostensibly) have a higher ceiling for efficiency. But, yes, they're definitely going to be useful only in a hybrid architecture as the overall energy storage just isn't there.
When the earliest digital-modulator-based power amps were shown to Harman, the by-far-largest and most complex parts were the power supplies. The commercialization (deeply resented by his former classmates) by Lars Risbo, the company TacT iirc, had a beautiful volume control on the front panel, which controlled the voltage rails. It was on ball bearings, and at one show people used to sneak over to it and give it a spin to see how long it would take to wind down and stop.
When the marketing geniuses decided the Harman-Kardon receivers had to have "digital" amplifiers, the one outfit in Texas sold them a bill of goods and worked closely to make sure their modules would be incorporated in the new line. The performance was mediocre, and h/k actually watered down the specs on the existing line of class AB amplifiers to make the specification comparisons look not as bad.
The interior of the new units was...interesting. One-half of the internal volume was the power supply. Another quarter was the amplifier. The last quarter was for everything else.
But hey it all had to be better. It was Digital! I think sales were still dismal.
The first sign of marketing digititis at Harman I remember was a powered speaker system that was said to have a "digital power supply". Turned out it was an offline switcher based on a Power Integrations chip.
Not huge energy storage and no significant banking to play with. You wouldn't want it necessarily in a range rover. Although only wiliams stuck with the kinetic recovery it is very widely used in lemans, so is proven in extreme environments.
The Chrysler patriot concept (gas turbine an 100kRPM Carbon fibre flywheels) had it run would have possibly had enough gyroscopic effect to have some odd handling.
The Chrysler patriot concept (gas turbine an 100kRPM Carbon fibre flywheels) had it run would have possibly had enough gyroscopic effect to have some odd handling.
- Status
- Not open for further replies.