Prius RPM

[GusTheCat]

I monitor the Prius's engine RPM using Scangauge, and found that the RPMs are discrete at some fixed numbers, such as 1280, 1180, etc., instead of an arbitrary integer.

Has anybody noticed this? Is this because the algorithm regulates the engine to pre-defined RPMs, or because the RPM sensor's resolution?

[efusco]

I've been curious about the same thing for a while now. I see 5-10rpm jumps, but it seems to like 950, 1170, 1275 much more than anything in between.

[dogfriend]

Originally Posted by GusTheCat

I monitor the Prius's engine RPM using Scangauge, and found that the RPMs are discrete at some fixed numbers, such as 1280, 1180, etc., instead of an arbitrary integer.

Has anybody noticed this? Is this because the algorithm regulates the engine to pre-defined RPMs, or because the RPM sensor's resolution?

I think its the former rather than the latter because I actually do see some odd values every so often, usually when changing from one operating state to another. For example, when the ICE is running and then shuts down, every once in a while it will catch at an odd rpm like 87 rpm until it updates in another second or so and then change to 0 rpm.

[wintechsw]

I really wonder just what the RPM counter measures. Let me explain.

At enforced idle(you can achieve that), my SCII registers 992rpm and the engine tone is constant.

My normal highway speed used to be 113kph on the speedo(110 actual confirmed by GPS and our Police Radar measuring stations), and the RPM used to vary anywhere between 1700 and 2300 rpm. This is with the Cruise Control engaged at this speed on a relatively flat road. With a normal car, the engine noise varies noticeably very little when the CC is engaged. My GenII is no different in that there is no perceivable engine tone difference on level ground. Yet my SCII is telling me that the engine is anything up to +/- 300 rpm from reading to reading(2 secs I think).

My ears would recognize such a variation in rpm yet the ICE noise seems to be constant.

So my only conclusion is that the SCII RPM reading is something other than 'just the ICE rpm'.

My 2c worth.

[BAllanJ]

I always figured it was just an artifact of the number of bits used to communicate the rpm to the scangauge, but I have no real evidence of that.

[FireEngineer]

My experience has been that the RPM likes to settle on or try to find multiples of 32. So if you have a non-multiple on your ScanGage, eventually the Prius seeks to shift it to an even multiple of 32.

Wayne

[grand total]

Originally Posted by FireEngineer

My experience has been that the RPM likes to settle on or try to find multiples of 32. So if you have a non-multiple on your ScanGage, eventually the Prius seeks to shift it to an even multiple of 32.

Actually I think it is multiples of 16 RPM. My theory on why you see numbers other than multiples of 16 is that the scangauge samples engine speed periodically. If during the period that it is sampling the engine speed is steady then you see a multiple of 16 reported. If the engine speed is changing when the scanguage is sampling the engine will appear to be rotating at a speed between two 16 RPM increments.

[qbee42]

Multiples of 16 could be an artifact of the control system, or merely relate to the number of significant bits in the data reported on the CAN bus. If you drop the least significant nibble from the data it will always be in multiples of 16. Data averaging would explain the in-between values during transition.

Tom

[donee]

Hi All,

This is an artifact of the digitization of the measurement, and the limited number of bits.

Lets see if we can figure out how many bits?

There are 2^n - 1 quantum for n bits.

So, 8 bits would represent up to 4080 RPM with 16 RPM quantum. 9 bits up to 8176 RPM, 10 bits 16368.

Another possibility is the RPM has a standard representation in the computer between all the rotating components. Where one bit would be used for positive or negative. In which case 10 bits would represent a range of +/- 8176 RPM.

Based on all this I am going to gues its a 12 bit representation of +/- 32752 RPM. 12 bit is a standard Digital to Analog converter size. Say you had engine off at 120 mph, then MG1 would be at around 20000 RPM (based on 10000 RPM at 60, a coasting limit mentioned on Prius Chat). So, the 32752 RPM range appears to be required.

[grand total]

Originally Posted by donee

Hi All,

This is an artifact of the digitization of the measurement, and the limited number of bits.

Lets see if we can figure out how many bits?

There are 2^n - 1 quantum for n bits.

So, 8 bits would represent up to 4080 RPM with 16 RPM quantum. 9 bits up to 8176 RPM, 10 bits 16368.

Another possibility is the RPM has a standard representation in the computer between all the rotating components. Where one bit would be used for positive or negative. In which case 10 bits would represent a range of +/- 8176 RPM.

Based on all this I am going to gues its a 12 bit representation of +/- 32752 RPM. 12 bit is a standard Digital to Analog converter size. Say you had engine off at 120 mph, then MG1 would be at around 20000 RPM (based on 10000 RPM at 60, a coasting limit mentioned on Prius Chat). So, the 32752 RPM range appears to be required.

I'm sure that's all good stuff, but it's the speed of the ICE that we are talking about and a range of 0-8000RPM should be more than enough. Given that then 8 bits would be sufficient if the quanta were 32 RPM, but I really think I see 16 RPM intervals on my scangauge. 9 bits doesn't make much sense to me.