processor speed in standalone computer really matters ?
#1
Member
Thread Starter
Join Date: Apr 2001
Location: Baghdad
Posts: 1,429
Likes: 0
Received 0 Likes
on
0 Posts
processor speed in standalone computer really matters ?
im thinking to get the new autronic PnP and i read the spec , it says that the processor is intel 16bit running 16mhz.
i know that motec , haltech and microtech running 32mhz , but dont worry bout motec since i wont be able to afford it.
so im deciding whether to get autronic / haltech / microtech
thanks a bunch guys
i know that motec , haltech and microtech running 32mhz , but dont worry bout motec since i wont be able to afford it.
so im deciding whether to get autronic / haltech / microtech
thanks a bunch guys
#3
Honda-Tech Member
Join Date: Dec 2001
Location: The Swamp, FL
Posts: 7,139
Likes: 0
Received 0 Likes
on
0 Posts
Re: processor speed in standalone computer really matters ? (turbozxi)
as far as clock speed, im not sure, but i doubt it will be an issue.
Ive seen quite a few cars running autronic and they are beasts.
Ive seen quite a few cars running autronic and they are beasts.
#5
Re: processor speed in standalone computer really matters ? (Day Walker)
I would say the clock speed is not the most important thing. The faster one may recalculate the engine condition 20,000 times in a second while the slower one only does it 10,000 times. But does this really matter?
I think more important is the accuracy and resolution of the analog/digital converters and how often they deliver a new value.
It doesn't help if the processor cycles through your loop a million time per second if the processor can only read one updated sensor value per second.
Most microcontrollers only have a limited count of inputs, so the developer has to multiplex the sensors. This means you can not read all sensors at one time. In one cycle you can read the TPS, than the CPS etc.
Next thing is the resolution. If you have a 8-bit converter, you only get a value from 0 to 255 for the throttle position. This might be enough, but with a value from 0 to 65535 the processor could do a way more accuracy calculation.
But there is an other point that i am missing in all standalones so far. The ECU gets a sync for the crank position every 15 degree of crank rotation. That means the ECU always have to intercalate (calculate) the current postion till it gets a new sync. If you are now using an ECU from a B16 and use it with a B18 the r/s ratio and piston speed changes. This means the formula to calculate the current crank position is no longer accurate and your ignition/injection may be up to 2 degrees off.
I have to say this is the point of view of a software developer (me) and not an engine builder. I might be wrong on some points.
I think more important is the accuracy and resolution of the analog/digital converters and how often they deliver a new value.
It doesn't help if the processor cycles through your loop a million time per second if the processor can only read one updated sensor value per second.
Most microcontrollers only have a limited count of inputs, so the developer has to multiplex the sensors. This means you can not read all sensors at one time. In one cycle you can read the TPS, than the CPS etc.
Next thing is the resolution. If you have a 8-bit converter, you only get a value from 0 to 255 for the throttle position. This might be enough, but with a value from 0 to 65535 the processor could do a way more accuracy calculation.
But there is an other point that i am missing in all standalones so far. The ECU gets a sync for the crank position every 15 degree of crank rotation. That means the ECU always have to intercalate (calculate) the current postion till it gets a new sync. If you are now using an ECU from a B16 and use it with a B18 the r/s ratio and piston speed changes. This means the formula to calculate the current crank position is no longer accurate and your ignition/injection may be up to 2 degrees off.
I have to say this is the point of view of a software developer (me) and not an engine builder. I might be wrong on some points.
#6
Re: processor speed in standalone computer really matters ? (Andre)
Pretty much spot-on, Andre, you're obviously a good software developer, but I'll argue this:
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Andre »</TD></TR><TR><TD CLASS="quote">But there is an other point that i am missing in all standalones so far. The ECU gets a sync for the crank position every 15 degree of crank rotation. That means the ECU always have to intercalate (calculate) the current postion till it gets a new sync. If you are now using an ECU from a B16 and use it with a B18 the r/s ratio and piston speed changes. This means the formula to calculate the current crank position is no longer accurate and your ignition/injection may be up to 2 degrees off.</TD></TR></TABLE>
R/S has nothing to do with crankshaft position. It's a mechanical 2:1 ratio drive between crank and cam; rod angle, piston speed, etc, are divorced from where the crank is in it's angle of rotation, and where that places the cam/distributor in it's angle of rotation.
FYI, timing belts suck... there is already +/- 2 degrees in cam/crank alignment from that sloppy ***** of a timing belt flapping around all over the place. If there were a < $300 gear drive for Honduhs, I'd buy one.
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Andre »</TD></TR><TR><TD CLASS="quote">But there is an other point that i am missing in all standalones so far. The ECU gets a sync for the crank position every 15 degree of crank rotation. That means the ECU always have to intercalate (calculate) the current postion till it gets a new sync. If you are now using an ECU from a B16 and use it with a B18 the r/s ratio and piston speed changes. This means the formula to calculate the current crank position is no longer accurate and your ignition/injection may be up to 2 degrees off.</TD></TR></TABLE>
R/S has nothing to do with crankshaft position. It's a mechanical 2:1 ratio drive between crank and cam; rod angle, piston speed, etc, are divorced from where the crank is in it's angle of rotation, and where that places the cam/distributor in it's angle of rotation.
FYI, timing belts suck... there is already +/- 2 degrees in cam/crank alignment from that sloppy ***** of a timing belt flapping around all over the place. If there were a < $300 gear drive for Honduhs, I'd buy one.
#7
Re: processor speed in standalone computer really matters ? (mekkis)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by mekkis »</TD></TR><TR><TD CLASS="quote">Pretty much spot-on, Andre, you're obviously a good software developer, but I'll argue this:
R/S has nothing to do with crankshaft position. It's a mechanical 2:1 ratio drive between crank and cam; rod angle, piston speed, etc, are divorced from where the crank is in it's angle of rotation, and where that places the cam/distributor in it's angle of rotation.</TD></TR></TABLE>
Ok, let me explain what i mean.
The piston speed and piston acceleration is a function of the crank angle. The piston accelerates and slows down on every stroke.
The stroke and rod length influence the piston speed and acceleration. The mean speed is always the same if you don't change the stroke because the way the piston has to move is always the same. But the longer the rod, the lower the top speed.
But coming back to my example with the B16 ECU and B18 motor. The B18 has a 77mm stroke and a way better r/s ratio than the B18, which results in a lower piston speed.
The OBD1 and OBD2 distributors have 24 teeth for the crank position signal. This means every 15 degree of camshaft the ECU gets a signal from the sensor. When we are dealing with ignition timing we have the crankshaft position as the reference. The crankshaft rotates twice as fast as the camshafts, which means we only have a sync signal every 30 degree of crankshaft rotation.
If we now set the ignition timing to 10 degree BTDC (Before top dead center), the last position where the ECU knows the crankshaft position is 30 degree BTDC.
If you program a microcontroller you have to set a timer at 30° BTDC and calculate when 10° is reached. Now the B16 ECU assumes 10 milliseconds for example. But our B18 piston is faster, and after 10 milliseconds we are already at 5° BTDC.
I don't know if some standalones learn the piston speed curve, but i think the stock ECU's have fixed values for sure. That's why i don't understand why the standalones don't ask for the engine basics.
Maybe someone can enlighten me?
R/S has nothing to do with crankshaft position. It's a mechanical 2:1 ratio drive between crank and cam; rod angle, piston speed, etc, are divorced from where the crank is in it's angle of rotation, and where that places the cam/distributor in it's angle of rotation.</TD></TR></TABLE>
Ok, let me explain what i mean.
The piston speed and piston acceleration is a function of the crank angle. The piston accelerates and slows down on every stroke.
The stroke and rod length influence the piston speed and acceleration. The mean speed is always the same if you don't change the stroke because the way the piston has to move is always the same. But the longer the rod, the lower the top speed.
But coming back to my example with the B16 ECU and B18 motor. The B18 has a 77mm stroke and a way better r/s ratio than the B18, which results in a lower piston speed.
The OBD1 and OBD2 distributors have 24 teeth for the crank position signal. This means every 15 degree of camshaft the ECU gets a signal from the sensor. When we are dealing with ignition timing we have the crankshaft position as the reference. The crankshaft rotates twice as fast as the camshafts, which means we only have a sync signal every 30 degree of crankshaft rotation.
If we now set the ignition timing to 10 degree BTDC (Before top dead center), the last position where the ECU knows the crankshaft position is 30 degree BTDC.
If you program a microcontroller you have to set a timer at 30° BTDC and calculate when 10° is reached. Now the B16 ECU assumes 10 milliseconds for example. But our B18 piston is faster, and after 10 milliseconds we are already at 5° BTDC.
I don't know if some standalones learn the piston speed curve, but i think the stock ECU's have fixed values for sure. That's why i don't understand why the standalones don't ask for the engine basics.
Maybe someone can enlighten me?
Trending Topics
#8
Re: processor speed in standalone computer really matters ? (Andre)
You're confusing piston speed with crankshaft speed. Piston speeds, bore size, etc, have a direct effect on what your ideal ignition timing is, but they have no effect on how that timing is determined. For 3000 rpms with a B16 and 3000 rpms for a B18C and 3000 rpms for a goped, all these crankshafts are turning the exact same speed, and traverse (engine harmonics and dynamic imbalance aside) the same number of degrees of crankshaft rotation in the same period of time.
Best way to think about this is to forget the pistons and rods even exist, think of the crank pulley as a clock's face.
Best way to think about this is to forget the pistons and rods even exist, think of the crank pulley as a clock's face.
#9
Member
Thread Starter
Join Date: Apr 2001
Location: Baghdad
Posts: 1,429
Likes: 0
Received 0 Likes
on
0 Posts
Re: processor speed in standalone computer really matters ? (mekkis)
too hard for me to swallow those informations
so a computer with 16-bit and 16mhz processor , should be way more than enough to run our boosted engine ?
so a computer with 16-bit and 16mhz processor , should be way more than enough to run our boosted engine ?
#13
Honda-Tech Member
Re: (turb)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by turb »</TD></TR><TR><TD CLASS="quote">i think the crankshaft speed should be constant regardless of the r/s ratio. </TD></TR></TABLE>
What they're talking about is crankshaft fluctuation speed. That's how obd2 ecus know when you have a misfire, they see the difference in crankshaft speed between two cylinders. The crank is not at a constant rpm, it accelerates on a powerstroke and then slows on a compression stroke. Keep in mind, we're talking minimal here, nothing you can see or feel, but can be measuered. If you only had one trigger on the crank for crank position then the ecu doesn't know what's going on for 359 degrees of crankshaft rotation. The more pickup points on the crank, the more accurately the ecu knows exact crankshaft position.
What they're talking about is crankshaft fluctuation speed. That's how obd2 ecus know when you have a misfire, they see the difference in crankshaft speed between two cylinders. The crank is not at a constant rpm, it accelerates on a powerstroke and then slows on a compression stroke. Keep in mind, we're talking minimal here, nothing you can see or feel, but can be measuered. If you only had one trigger on the crank for crank position then the ecu doesn't know what's going on for 359 degrees of crankshaft rotation. The more pickup points on the crank, the more accurately the ecu knows exact crankshaft position.
Thread
Thread Starter
Forum
Replies
Last Post
AkiraLite
Road Racing / Autocross & Time Attack
5
01-05-2003 03:17 PM
Nuts
Road Racing / Autocross & Time Attack
14
12-26-2001 08:20 AM