gemellocattivo.com

Which means "Evil Twin". Lets see your projects where you change boring into fun or create the fun from scratch.
It is currently Thu Mar 28, 2024 5:31 pm

All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 29 posts ]  Go to page 1, 2, 3  Next
Author Message
 Post subject: Common Rail Diesel
PostPosted: Thu Apr 07, 2016 1:55 pm 
Offline

Joined: Sun Oct 11, 2015 2:07 pm
Posts: 134
Has anyone here played with common rail Diesel engines or controls?

In theory, they are much simpler than gas:
-The fuel mass controls load, instead of air mass, fuel mass is much easier to meter/control than air mass
-Air control is not required at all, except boost limiting
-Many fewer actuators, the fuel injector per cylinder replaces injector, spark, and throttle of spark ignition


In reality, these systems have much less information on the internet. There is no open-source common rail ECU, or nothing even slightly open (like Megasquirt equivalent).

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Thu Apr 07, 2016 2:28 pm 
Offline

Joined: Thu Jan 01, 2015 6:47 pm
Posts: 4251
I have not but I know I've read stuff and thought they were just using gas ecus with an external driver for the injectors because as you say it less stuff not more????? Just do a fuel vs TSP or MAP and you're good to go right?


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Thu Apr 07, 2016 2:38 pm 
Offline

Joined: Sun Oct 11, 2015 2:07 pm
Posts: 134
Technically there is no TPS on a diesel, as there is no throttle. MAP will equal baro or boost pressure, and also doesn't indicate load (it does indicate maximum load though). A pedal sensor (ETC like) could be used as 'TPS' though.

The injection timing needs to be precise, angularly, which some ECUs don't do well (e.g. our ECUs round the injection angle to the nearest tooth, because the TPU in the MPC565 isn't very flexible). But, if you could use the injection timing and injection mass tables, I guess it would work. Injection timing is used kinda like spark timing for gas engines.


I'm investigating a new Diesel engine to use for snowmobiles, but the control system is a place I have relatively little experience with.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Thu Apr 07, 2016 3:23 pm 
Offline

Joined: Thu Jan 01, 2015 6:47 pm
Posts: 4251
Yes, I meant throttle pedal not actually TPS. Do they just let the turbo spool or control boost based on desired output?...not sure which would be more efficient.

The eTPU processors do injection timing to the 0.1 degree I think it is, enginelab is at least that good. I think even MS claims to be there....although I'm not convinced they have the ability to measure crank position to that resolution.


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Thu Apr 07, 2016 7:54 pm 
Offline

Joined: Sun Oct 11, 2015 2:07 pm
Posts: 134
About boost:
Boost is generally controlled. There are a few ways to do this. The best is to use a VNT turbo and motorized actuator and leave the vanes 'normally open', closing them to build boost instead of opening them as boost is reached. This part does not worry me, it's very similar to gas boost control.

About eTPU:
eTPU can interpolate a crank wheel to any intermediate resolution, within some limits. Normally it's used with 1/10th or 1/16th deg per bit and with 60-2 can definitely hold tolerance less than 1/2 degree which is certainly enough for common rail. The outputs (and inputs) are scheduled (or captured) in hardware, and there are two pulse events in hardware, so there isn't any interrupt latency if there is enough time between edges for the microcode to reschedule the hardware.

The older TPU is not as fortunate. In the modules we use, some fuel pulses are rounded off to the nearest tooth so the hardware timers can get good duration precision and handle multiple pulses properly. Newer Woodward/Motohawk modules use eTPU and can use the angle clock for high resolution angle and duration timing of 16 pulses per cycle (vs 2 in the TPU modules).


I definitely wouldn't trust Megasquirt to run anything important, but MS3 can probable time to a half degree on a 60-2 wheel. The's MS3 S12X is considered absolute bottom line in the automotive industry and reserved for very low function systems (think carb replacements).

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Fri Apr 08, 2016 4:24 am 
Offline

Joined: Thu Jan 01, 2015 6:47 pm
Posts: 4251
The eTPU based controllers and I assume other modern designs are designed for a WORST CASE error of 0.1 degree, normally they are under 0.01 degree error. This is where the mis-fire detection comes from, the ability to know the crank tooth did not arrive when expected.

MS....it's not the processor that troubles me although it is a 30 year old design. I bought a hall sensor from the MS distributor to use for testing my o5e project. I had all kinds of crank sync problems and after spending weeks checking code, changing settings, ect I realized the problem was the sensor was awful, the tooth edge was jittering around +/-20% or 2 degrees on my 36-1 wheel. i bought another one, exactly the same. I pulled out another brand I had for a wheel speed sensor and it was unter 5% jitter, 0.5 degree and my o5e sync problem was fixed. I pulled out a VR sensor and I couldn't measure any error.....and learned the reason I've always had better luck with VR crank sensors.

Okaaaayyyy.....I spoke to the MS people and was told I'm nuts, they've sold thousands without any issue. But there was simply no way to set windows wide enough to reliably see the teeth and narrow enough to reliably find the gap on a -1 wheel. I took another tooth off the wheel and had better luck syncing with that sensor.....but MS has no problem with the sensor so I assume they are either not error checking anything or they are averaging position, either way GIGO means that with that sensor you get +/- 2 degrees, maybe +/-1 deg if they are averaging and it doesn't really matter what the chip is capable off. With good hw/setup though I assume it can do much better, but that fact it works fine with crap HW/setup and timing could be a couple degrees off is a deal breaker for me.


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Fri Apr 08, 2016 8:51 am 
Offline

Joined: Sun Oct 11, 2015 2:07 pm
Posts: 134
Some eTPU controllers set the TCR2 resolution to 0.1 deg/bit, which would mean any angle must be rounded to the neatest 0.1 degree. Motohawk uses 1/16th instead, so 0.0625 deg. I believe they are accurate to at least 1/2 degree (this is the spec for the system I worked on), the worst-case error will probably be when interpolating across the missing tooth gap as there is an 18 degree window without a physical tooth (with 60-2). Misfire uses the times of physical teeth, and doesn't rely on the crank interpolation. Of course, eTPU can only interpolate from the crank signal so a bad crank signal will always cause some timing error.

The MS code must be checking something, because we tried MS before going to Motohawk and couldn't get it to synchronize to a 12-1 trigger wheel on a single cylinder engine. This resulted in extremely poor starting, and when it started it would occasionally lose sync. With the TPU based controllers it synchronizes with no problems.


I've never particularly liked VR sensors because the interface circuit is more complicated, but I've never had any problems with them either.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Fri Apr 08, 2016 10:10 am 
Offline

Joined: Thu Jan 01, 2015 6:47 pm
Posts: 4251
apalrd wrote:
Some eTPU controllers set the TCR2 resolution to 0.1 deg/bit, which would mean any angle must be rounded to the neatest 0.1 degree. Motohawk uses 1/16th instead, so 0.0625 deg. I believe they are accurate to at least 1/2 degree (this is the spec for the system I worked on), the worst-case error will probably be when interpolating across the missing tooth gap as there is an 18 degree window without a physical tooth (with 60-2). Misfire uses the times of physical teeth, and doesn't rely on the crank interpolation. Of course, eTPU can only interpolate from the crank signal so a bad crank signal will always cause some timing error.


The eTPU code doesn't work like that...but I don't know how Motohawk presents. What you set effectively is a clock speed and need to get no more than 1024 tics between teeth at whatever RPM is your min and that give you a resolution of (degrees between teeth/1024) * (min rpm/rpm).

Yes misfire is based on physical teeth, but as I said that is based on the sensors ability to detect the teeth. All of the hall sensors I've ever checked jitter or bounce around when reading a constant rpm wheel tooth location. VR sensors, at the sensor itself, read the tooth location more accurately than I can measure error, but it's possible the conditioning circuit adds error.

apalrd wrote:
I've never particularly liked VR sensors because the interface circuit is more complicated, but I've never had any problems with them either.


Programmers and HW designers never do.....but they are much much more accurate so us MEs love them :)

apalrd wrote:
The MS code must be checking something, because we tried MS before going to Motohawk and couldn't get it to synchronize to a 12-1 trigger wheel on a single cylinder engine. This resulted in extremely poor starting, and when it started it would occasionally lose sync. With the TPU based controllers it synchronizes with no problems.


hmmm....that might be the down side or the averaging scheme, it can't react to rapid rpm changes like you see on a single during cranking?

The eTPU code has no issue with it either as long as a VR or high quality Hall pickup is used.....once I changed sensors there was basically nothing I could do to make it break sync other than drop below the threshold RPM so it could time-out between teeth. Really good.


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Fri Apr 08, 2016 11:20 am 
Offline

Joined: Sun Oct 11, 2015 2:07 pm
Posts: 134
mk e wrote:
The eTPU code doesn't work like that...but I don't know how Motohawk presents. What you set effectively is a clock speed and need to get no more than 1024 tics between teeth at whatever RPM is your min and that give you a resolution of (degrees between teeth/1024) * (min rpm/rpm).


You can set the eTPU resolution to whatever you want as long as it doesn't exceed the limits. A faster TCR1 frequency also helps. A lot of people set it to a single value, to get 0.1 or 0.0625 deg per tick. 0.1 deg/tick seems common. Then, all of the angular math operates on 0.1 deg/bit and the same fixed point unit can be used for all of the code.

_________________
"Sometimes, the elegant implementation is a function. Not a method. Not a class. Not a framework. Just a function." ~ John Carmack

"Any sufficiently advanced technology is indistinguishable from magic" ~Arthur C. Clarke


Top
 Profile Send private message  
 
 Post subject: Re: Common Rail Diesel
PostPosted: Fri Apr 08, 2016 11:59 am 
Offline

Joined: Fri Jun 12, 2015 1:13 pm
Posts: 6
Don't ya' know I love it when you talk "eTPU", Mark!!

(best to you on fixing the engine, Mark.)

Paul


Top
 Profile Send private message  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 29 posts ]  Go to page 1, 2, 3  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Group