|
I'll have to politely disagree with Paul's comment about it being logical that in the case of Lightspeed since it took a lot of development to come up with an advance curve it is logically "proprietary." Once the first unit was sold the advance curve can be measured, so it makes no sense to keep it secret - all Klaus is doing is frustrating some customers and driving some potential customers away. From data posted by Scott it looks like the curve is linear from 0 advance at 30 inches to 14 degrees at 10 inches regardless of rpm. It also looks like there is no rpm advance (except for some way to switch from zero advance during cranking to the base advance). I'd like to see more data to get a more accurate picture. Say, do a ground runup and check the advance at WOT/max rpm. Then fly at 5,000 ft or less and measure the advance at 2300, 2500 and 2700 rpm at manifold pressures of 15, 17, 19, 21, 23 and 15 inches. This would give me a good picture of the curve. From what I see so far the curve looks logical, but I would prefer to have zero advance at 25 inches and have it limited to something more like 10 degrees. The advance at 25 inches means that leaning to best power at 25 inches might get one close to the detonation limit. Having more advance that would be needed at high altitude cruise means that if the MAP sensor failed to a low value the advance could go beyond what could be tolerated by the engine.
Paul is correct in that a digital (microprocessor-based) system is often more complex, but only in the software - an analog system is often far more mechanically complex (more components) than the digital approach. Which is more prone to failure? Not sure, but Klaus' approach to keeping in analog is a good one. Maybe not the only one, though. A good microprocessor-based system should be just as reliable and would be capable of more complex algorithms.
Gary I had nothing to do with Atlas or any other rockets (Yes, I did - I paid for them!) and while I've stayed at a Holiday Inn Express it wasn't last night. So I guess my opinions are only worth what the reader thinks. On Jan 7, 2008, at 3:01 AM, Lancair Mailing List wrote:
Date: January 6, 2008 9:29:54 PM PST Subject: PRISM
Prism! Prism! Prism! 'Sorta reminds me of a friend who won't buy a new computer because he's so sure that as soon as he buys one, someone else will come along with a computer for sale that's faster, has more memory, and costs less. PLASMA uses MAP and rpm to determine timing, and Klaus has taken a very conservative approach, even 'tho he's a liberal, to avoid any timing that could result in detonation - pre-ignition. This timing curve is a result of his many hours of experimentation, and is proprietary, so he's obviously not going to give it out! This timing may not be the absolute best obtained by measuring individual cylinder pressures and adjusting individual cylinder ignition timing to get right up next to this very-best-of-all-worlds timing where LPP, location of peak pressure, is at 14 deg ATDC, but consider how much more complex and costly that approach must be; more sensors to go bad, more wiring to break. Let's see. If you lose one sensor, and the other 3 or 5 are at peak, and this one is at something less than peak, will the engine still run smoothly? Will having every cylinder pressure sensor working before takeoff be a requirement, and how will it be tested on run-up? The PLASMA is a simple but very good analog EI, and a lot of it is due to its NOT being digital! Go on the VANS forum, and read all of the postings about a very popular ignition system that's up to V25 or V26 of its software. Would you like to keep sending in your ignition system for updating every time a new software bug is discovered? Has your home computer ever gone off to Never-Never Land? I've designed digital with DCTL, RTL, TTL, ECL, and CMOS, and I've designed analog circuits from DC to X-band, and I know the pluses and minuses of each. I worked on four different computers that were used for the guidance of the Atlas ICBM, Titan range safety, guidance of the Atlas used for re-entry vehicle testing, and guidance of the Atlas Space Booster - the Burroughs J3, IBM 7088 and 7090, and Harris /4. I've always pointed out to my digital-only compatriots that digital is just analog driven to saturation and cut-off! And when they get into the high speed stuff, they'd better know about analog stuff like transmission lines and reflections and capacitive loading and wire inductance. So, what is the latest release date of PRISM?
|