signal manipulation, delay, advance
lightsout
Posts: 22
I even had trouble coming up with a subject line let alone trying to explain so,bear with me as I try to explain:
I have a couple of project ideas that would take for instance on an·auto,motorcycle,etc. engine·if you had a timed/synced hall effect 5-8 volt square wave.
· Now (based on other inputs), lets say you want to advance or delay this signal.
·I don't know if you could advance/delay the existing signal ?
OR
use the existing signal a an input and "mimic it" with another signal from some kind of "chip?" that you could advance/delay it?
Feel free to ask if I didn't explain enough.
I have a couple of project ideas that would take for instance on an·auto,motorcycle,etc. engine·if you had a timed/synced hall effect 5-8 volt square wave.
· Now (based on other inputs), lets say you want to advance or delay this signal.
·I don't know if you could advance/delay the existing signal ?
OR
use the existing signal a an input and "mimic it" with another signal from some kind of "chip?" that you could advance/delay it?
Feel free to ask if I didn't explain enough.
Comments
·· This concept is used all the time by people who want to mess around with their Engine Timing, Fuel Injection, Air/Fuel mixture, etc.· And it's easy enough to do electronically.· But coding for it can be altogether different for two reasons.· One, you really need to have a thorough understanding of what the signals are, and what they're doing, as well as the effects of altering them.· And two, because you pretty much need a dedicated controller for this, or at least something run in assembly to meet the strict timing requirements.
·· The driving section (Altering the signals) is easy enough with the proper driver circuitry, like a power transistor, or Push-Pull driver chip.· Remember, things in an automotive environment tend to draw HUGE amounts of current compared to their·outside counterparts (Where applicable).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Chris wrote:One, you really need to have a thorough understanding of what the signals are, and what they're doing, as well as the effects of altering them.·
Lightsout:I don't think thats a problem for me as thats "most of my life".
Chris wrote:And two, because you pretty much need a dedicated controller for this, or at least something run in assembly to meet the strict timing requirements.
Lightsout:"something run in assembly"I guess I don't know what that means ??
·Chris wrote:The driving section (Altering the signals) is easy enough with the proper driver circuitry, like a power transistor, or Push-Pull driver chip.
Lightsout:Ok, so lets say I use a hall effect sensor to trigger a power transistor, now how would I advance or delay it?
·· Perhaps in a Fuel Injection system you will be monitoring inputs to determine fuel/air mixture and adjusting the fuel via PWM.· More fuel when needed, or less when not needed.· That's a couple of examples I have seen.· There are no specifics.· You will need to find out what the signals are that you need to monitor, and what signals you need to switch to adjust them.· And as I already said, be aware of the effects of doing this.
·· I recently read an interesting article from a forums member's website where he talked about his exploits messing around with an O2 sensor on his vehicle.· Made for some very interesting reading.· Not to mention the humor!·
http://www.bobblick.com/techref/projects/o2sensor/o2sensor.html
·· As for assembly, it's a very low level language for talking to the CPU, and is faster than interpreted BASIC, as is found on the BASIC Stamps.· I was thinking you might need an SX Micro to do what you want to do, but I guess that depends.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Regarding advancing or delaying the signal, that is done by having your "trigger" occur in advance of the event you want to control
For example, if you have a pulse that occurs exactly at Top Dead Center (TDC), the you can only delay the controlled event.
If you have a pulse that occurs 30 degrees before TDC, then you can advance the controlled event, or delay the event with respect to TDC.
The actual amount you can delay depends on two things:
Even if your pulse occurs 30 degrees before TDC, but the processor speed is slow, and the engine speed is really high, you might be limited to only 20 degrees advance.
Compared to a really fast processor at low engine speeds, you may have the capability of 29 degress advance. I takes a little math,knowledge of the processor, and knowledge of engine speeds to determine max advance.
Hope this helps more than it confuses.
K
Your best solution may be to purchase an adjustable timing control like http://www.msdignition.com/tc_2.htm·, knock the pot off it and use a digital pot for your control. I have done this in the past with good results. You can set your initial timing with the pot in any position thus giving you either full retard, full advance or 1/2 of each.
There seems to be a few people on the list in the past wanting to do this. What are you needing it for?
I use mine as presets for octane 87 or 93 and for 93 w/Nitrous in my Corvette.
KenM said:For example, if you have a pulse that occurs exactly at Top Dead Center (TDC), the you can only delay the controlled event.
·What I could do is··"virtually" advance it if I delay it for instance 350-359 degree's of shaft rotation.
I like Eric R's idea , altough it would be limited to certain aplications and I don't know if it would work with low voltage (5VDC) applications.
I know I don't want to directly contol it through the stamp due to trying to caculate and adjust for speed of the programmed stamp operations since timing·IS everything!
In a typical automotive 4 piston engine you would have a hall-effect sensor signal·of 0-250HZ (normally directly to) the computer. I could·"intercept" the hall-effect signal to trigger a transistor circuit·then the transistor circuit trigger the computer, then·I could read decision making inputs·with the stamp and have the stamp increase/set the amount of delay of the transistor indefinately UNTIL the stamp can compute a new amount of delay....................follow me??
Sooo, what I think I need is a transistor-like device that has a programable amount of delay (trigger delay?),that can interface with the stamp. Is there such a thing ??? I think IT would solve my problem.....Anyone...,anyone...,anyone...,Bueller??
In the mixed signal domain, think "monostable multivibrator", or "one-shot". Examples are the 74HC121 or the CD4538 or even the venerable '555 triple nickle timer. A trigger pulse at the input causes it to generate a longer pulse at the output, and if two are cascaded, a delayed pulse at the output. The timing is determined by a resistor and a capacitor. Control of these circuits with a Stamp would require a digital potentiometer in place of the resistor(s). Look for books with titles like, "IC timer cookbook".
That said, this is a natural application for a micro like the SX chip or a PIC, to handle the timing as an adjunct to the Stamp. I don't know of one that does what you want, ready made, though. You'll have to dig into SX/B yourself!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com