Paul, It's been nice to have some time today to devote to learning about programing
in assembly language. As I've read over some recent posts, I have a couple of
questions. I am pretty much convinced that capacitor discharge is superior to coil
ignition for the high rpm we are dealing with, and I'm pretty sure that's the way I'd
like to go.
1. But, am I correct in assuming that issues of dwell time would not really have to
be accounted for in the software design if capacitor discharge is used (thereby
simplifying the programing needed)?
2. Second, would it be simpler, from a software perspective, to use a magnet fixed to
the flywheel (or in my case the gear at the prop end of the belt reduction-drive)
with a hall-effect sensor, as detailed on several other posts, to count rpm and
locate tdc or another useful reference, rather than counting gear teeth? I assume
that advance has to vary depending on rpm as in recips, so all we are really looking
for is a way to locate tdc so the cpu has a reference. The design of the reduction
unit is such that I could pretty easily mount a magnet and pick-up, so I was
wondering if that might be a simpler way for me to go. Thanks,Bob
1. No doubt about it. That is the way most outboard ignitions work. Problematic is
vacuum and RPM advance while the 8051 micro computer can deal with that easily.
BTW Part of the learning of 8051 systems with USB ports effort is the ability to
leverage off relatively cheap net books and tablets to display important engine info
in the cockpit. Intrinsic to the 8051 with USB ports is the ability to talk in both
directions. Not so important with ignition systems but much more important with EFI.
2. One can determine the RPM much quicker merely timing (using the 8051) the
occurrence of a flywheel tooth passing a mag sensor. One or two instructions. Mag
sensors for anti lock braking systems are about $20. The second function is knowing
the exact angle of the E-shaft for when to inject the fuel or fire the plug.
Calculating that in software by math methods makes the program far more complicated.
To determine shaft position you just count flywheel teeth. Each tooth is worth about
3.5 degrees of e-shaft rotation. We are adding hardware to simplify the program. In
this case hardware is cheaper than software.
Paul Lamar
Paul, Interesting. It sounds like counting flywheel teeth is simpler after all. I
appreciate the information on CDI. It'll take me some time to digest it. Since the
8051 will be supplying the trigger signal, I guess advance is a software task. Is
that correct? Would that be taken care of by a software routine? Or would it be
simpler to set it up to choose an advance from a table based on rpm? Or is there a
better way to do this? I've been thinking about getting a netbook or tablet to use
for set up and diagnostics when the time comes. Would it be feasible to do the
programing work with one? I'm currently using my laptop running Windows 7 for the
work I'm doing. Thanks, Bob Blake
Yes advance is a software task. What happens typically is the computer gets a signal
from the CAS well before the time to fire the plug. Lets say 40 degrees BTDC. It then
figures out it should fire the plugs at 22 degrees. However it must turn the coil on
2 ms before it turns the coil off to actually fire the plug.
This is for inductive systems like RX8 and GM LS1 type coils. The old distributer did
this thing mechanically with a centrifugal advance mechanism.
It is also a BSFC advantage to have a vacuum advance that advances the spark timing
at lower manifold pressure that will be encountered at high altitude.
I will work on a preliminary article for the webs site going over all the stuff I
have learned so far.
One of the questions popping up is the potential for the various flash versions of
the 8051 having different ways of telling the PC how to download the software and or
re-flash. The 8051 must see one or more bytes coming in telling it a program is
coming and where to store the store the program in the proper memory locations.
Here is a bit of info on that. This is what a typical object/binary/hex file looks
like. This must be downloaded to the flash computer by the PC. We start with the asm
version of the program. Then we have a listing after the asm is assembled. Then there
is a raw hex file that must be sent to the 8051.
Paul Lamar
--
The Rotary Engine News Letter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
You Tube key word PaulLamar2
Copyright 1998-2012 All world wide rights reserved.
