Help with XL2 log data???

[Kirkinsb]

Originally Posted by RC-Monster Mike

Kirkinsb - 2 samples per second data rate explains your low numbers for sure - you are simply missing the microsecond peak currents that happen. In my experience, the peak draw will ramp down very quickly once the car is rolling - a 200+ amp spike could be down to 120 amps within the same second. 2 samples/second frame rate will not show this, though. I understand that a higher frame rate will result less log time, but it will also yield higher data resolution so you can see more detailed data, albeit for s shorter time. Besides....do you really need to log the entire run every time? Please maximize your frame rate, gather some data, and report back. You can always change the frame rate back afterwards, but I think you will be surprised at what you see.

Also, while the Baja is a larger, heavier vehicle, it also typically is used under a higher voltage and will have less traction than an overgeared Muggy(4wd) with large, heavy tires. On the surface, logic may say that a car that is half the size will draw half the current, but in truth, there are many variables that will quickly dispell this logic. Also, a throttle "blip" can be no different than a full on drag race launch, but over a shorter time period. The actual peak current happens over a very short time period(1/10 second or even less) so it does not surprise me to see 200+ peaks on a heavy, overgeared 4wd monster truck.

magicald - The highest log rate will give you the highest resolution. In reality, the average draw is typically fairly low, but there are microsecond peak currents that are very high. Personally, I have only used the highest available frame rate, as I normally am only logging data when I am trying to dial in a setup or test a component(battery, motor, etc.). The highest resolution will provide the most information in a short period of time. The beautiful thing is that you can change the frame rate! I think it is wise to maximize the frame rate occasionally to properly visualize the truest data, but there is nothing wrong with a lower frame rate to gather more general data over a longer period.


In my opinion, you simply can't draw any accurate conclusions from data logged at different frame rates - logging at a low frame rate will mask peaks.

Good info..makes sense.... I will try the highest frame rate one of these days and see what happens.... Also good info about how higher voltage and current affect each other. Sounds good. I am just glad the XL2s seem to have the bugs worked out. They perferm great. Thanks

[Pdelcast]

Originally Posted by Kirkinsb

Good info..makes sense.... I will try the highest frame rate one of these days and see what happens.... Also good info about how higher voltage and current affect each other. Sounds good. I am just glad the XL2s seem to have the bugs worked out. They perferm great. Thanks

Just for everyone's benefit--

The ICE datalogger (used in the XL/XL2) actually samples the current several times during each PWM cycle (synchronously), and then keeps a running average of the results. The samples are then written to flash memory at the programmed "sample rate."

So the sample rate is only the rate at which the averaged samples are written to memory -- not the rate at which the controller samples the current...

Make sense??

[Kirkinsb]

Originally Posted by Pdelcast

Just for everyone's benefit--

The ICE datalogger (used in the XL/XL2) actually samples the current several times during each PWM cycle (synchronously), and then keeps a running average of the results. The samples are then written to flash memory at the programmed "sample rate."

So the sample rate is only the rate at which the averaged samples are written to memory -- not the rate at which the controller samples the current...

Make sense??

Ok, so my readings at 2/second are probably fairly accurate? And I am not missing any huge amounts of data by not using the 5/second setting? Yes/No? Hmmm, interesting. Is their much benefit to having the data written 5/second as opposed to 2/second since it is averaged? thanks Patrick.

[brian015]

Originally Posted by Kirkinsb

Ok, so my readings at 2/second are probably fairly accurate? And I am not missing any huge amounts of data by not using the 5/second setting? Yes/No? Hmmm, interesting. Is their much benefit to having the data written 5/second as opposed to 2/second since it is averaged? thanks Patrick.

No, your readings are more averaged than if you had the data written at a faster rate - so the real peak of the spike is averaged with lower currents on either side of the peak giving a lower average current written to the memory. Writing data at a faster rate should include less averaged data - so truer to the actual curve.

I think.