Any help with this would be much appreciated.
I've just installed a DL1. Everything seems to be working other than the RPM data feed. I'm getting strange readings i.e. quite spikey and doesn't seem to read revs above 5200 - not really what I would expect.
Its reading off an injected Hayabusa ECU. I'm not sure whether the problem is caused by the way I've wired it up or the settings I am using in the software.
I have the software settings as follows:
RPM (Advance options):
Filtering: None
Calculas: None
Maximum rate of change: 3000
Sensor type: Radians per second
Minimum 0
Maximum 12000
(the Haybusa engine cut-off is set at around 11,000)
Vehicle options:
Ignition pulses per revolution : 2
Any advice would be much appreciated,
Jon
Problem with RPM
The units you have chosen are wrong if you want RPM to read in the traditional revolutions per minute.
A radian is a more mathematical, scientific or theoretical engineering unit of angle. There are two pi (2 x 3.141) radians per rev (360 degrees).
You will observe from the above that this should make your "RPM" read high, not low.
The reason that the reading is low is because the max. r.o.c. is applying too much filtering. I imagine what you see has lots of longish straight lines at different angles and very little, if any, rounded, sweeping changes.
First, just change the units. This may be sufficient if the filtering is applied after the unit conversion. I can't remember what I've found in the past...
... but the shape of the trace sticks!
If the shape remains very similar, raise the roc. Don't pussy foot around with 50 or 100, but increase it by 1500 at a time. If you go too far, spiky will be a better description of what you see. There will probably be some normal looking RPM trace, but insufficient filtering will allow noise in the data to trick the system into thinking the revs have jumped heaps.
It might read 30,000 or more!
As the graph will probably be autoscaling, the 'good' parts of the data might be so compressed to the bottom of the graph that they are not recognisable as a rev trace.
If the data that you are examining has been taken from driving the car, especially if it was in anger around the track, there might be too many rises and falls to quickly see that you have the software settings correct.
It is usual to need some smoothing - the default 0.2 s is probably too much, once you get the roc 'in the ball park'. Initially, I'd remove the smoothing and just change the roc.
When you are more familiar with examining logged data, you might recognise the rev trace from different circuits, but some people find it hard to even recognise the speed trace.
It is best to try to set the software filtering and make sure the ign pulses per rev is correct with a log of some steady state data in the workshop (or paddock). Just hold the revs constant for five seconds at the lowest smooth idle, then a thousand rpm higher, etc. Obviously easier if you have a reliable tacho independent of the R-T hardware.
Without a tacho, it is best to look at steadier speed driving than on a hot lap, in a known gear. If you are doing this on the track use as few gears as possible and use less throttle so the speed change is gentle. This way, it is easier to see that the rev trace rises and falls in proportion to the speed trace.
HTH.
A radian is a more mathematical, scientific or theoretical engineering unit of angle. There are two pi (2 x 3.141) radians per rev (360 degrees).
You will observe from the above that this should make your "RPM" read high, not low.
The reason that the reading is low is because the max. r.o.c. is applying too much filtering. I imagine what you see has lots of longish straight lines at different angles and very little, if any, rounded, sweeping changes.
First, just change the units. This may be sufficient if the filtering is applied after the unit conversion. I can't remember what I've found in the past...
... but the shape of the trace sticks!
If the shape remains very similar, raise the roc. Don't pussy foot around with 50 or 100, but increase it by 1500 at a time. If you go too far, spiky will be a better description of what you see. There will probably be some normal looking RPM trace, but insufficient filtering will allow noise in the data to trick the system into thinking the revs have jumped heaps.
It might read 30,000 or more!
As the graph will probably be autoscaling, the 'good' parts of the data might be so compressed to the bottom of the graph that they are not recognisable as a rev trace.
If the data that you are examining has been taken from driving the car, especially if it was in anger around the track, there might be too many rises and falls to quickly see that you have the software settings correct.
It is usual to need some smoothing - the default 0.2 s is probably too much, once you get the roc 'in the ball park'. Initially, I'd remove the smoothing and just change the roc.
When you are more familiar with examining logged data, you might recognise the rev trace from different circuits, but some people find it hard to even recognise the speed trace.
It is best to try to set the software filtering and make sure the ign pulses per rev is correct with a log of some steady state data in the workshop (or paddock). Just hold the revs constant for five seconds at the lowest smooth idle, then a thousand rpm higher, etc. Obviously easier if you have a reliable tacho independent of the R-T hardware.
Without a tacho, it is best to look at steadier speed driving than on a hot lap, in a known gear. If you are doing this on the track use as few gears as possible and use less throttle so the speed change is gentle. This way, it is easier to see that the rev trace rises and falls in proportion to the speed trace.
HTH.
Thanks for that. I'm still not having much success. The readings (from a recent race) still don't look right. I pretty much know what revs it should be reading at various points on the lap but it still doesn't look right. Its still peaky, with some sectors reading around the max revs (around 11.5k) other sectors never getting above 6-7k revs. Also, there's no recognisable traces showing gear changes where I know there should be.
I've changed the sensor type to RPM
I've set the max rate of change at 8500 - although I've tried many settings but none seem to make sense.
Min is set at 0
Max is at 12000
I've changed the ignition pulses per revolutions to 4
Unfortunately I can't run some tests in the garage as the engine is not in the car now.
I give up! Any suggestions much appreciated,
Jon
I've changed the sensor type to RPM
I've set the max rate of change at 8500 - although I've tried many settings but none seem to make sense.
Min is set at 0
Max is at 12000
I've changed the ignition pulses per revolutions to 4
Unfortunately I can't run some tests in the garage as the engine is not in the car now.
I give up! Any suggestions much appreciated,
Jon
-
supermono55
- Posts: 18
- Joined: Thu Nov 16, 2006 4:19 pm
Hi Guy's
Im experiencing something similar. An RPM feed which at a certain ( not always the same) rpm seems to double.
For me the probable causeis in hardware. Im using the spark high signal connected to the minus of the coil. My idea is that the voltage "bounces"when given time. Therfore i did put an 1Mohm resistor in line in the spark high lead. Havent tested it fully, but read somewhere that is should be the answer.
Erik
Im experiencing something similar. An RPM feed which at a certain ( not always the same) rpm seems to double.
For me the probable causeis in hardware. Im using the spark high signal connected to the minus of the coil. My idea is that the voltage "bounces"when given time. Therfore i did put an 1Mohm resistor in line in the spark high lead. Havent tested it fully, but read somewhere that is should be the answer.
Erik
Return to “Installation questions”
Who is online
Users browsing this forum: No registered users and 198 guests
