can ms3x run a common rail diesel ?

[frank_ster]

ok very good !

i'll source a list of componets, and draw something up on pcb123.
i'll build a prototype when i get back to montreal ( i'm currently in meppen germany installing a large automation system )

the L1 on the drawing is a voltage regulator ?

L2 would be the voltage dc-dc converter any other part numbers that would do the job nicely ? again for mass production unit cost will matter!

[SymTech Laboratories]

L1 and L2 are inductors.

[fixmann]

Hi

I dont know if this could be of any help in the construction of the driverstage, but i put some links here anyway.

http://www.caranddriver.com/features/te ... -explained

http://www.abaxor.de/publications/icit/ ... e_ICIT.pdf

http://downloads.hindawi.com/journals/e ... 578591.pdf

http://ieeexplore.ieee.org/xpl/freeabs_ ... er=4237776

Øyvind

[frank_ster]

^ nice read !

[frank_ster]

would you guys think less of me if i used a modified domestic power inverter ? 400 watt ?

[fixmann]

hi
I dont know how experienced you are in electronics, but a powerinverter can be a good choice.
The inverter is usually built up by tree parts,microcontroller, inverterstage and a H-bridge.
The inverterstage is run by the microcontroller with pwm, wich in turn makes half of the outputvoltage.

The full outputvoltage/"sinus" is put together by the H-bridge, wich also is driven by the microcontroller.
So if you disconnect the 2 pins on the microcontroller wich drives the H-bridge, then connects them together with a
inverting SN7404, you can use the output from the inverter to drive the injector directly.
I am trying to say, you use the megasquirt to drive the H-bridge in the inverter while the microcontroller in the inverter
runs as normal.

If you use a 110v inverter you get +- 55v to drive the injector, and +-110 if you use the 220v model.

The one thing you must check is the powerrating on the H-bridge, if to low they can be replaced.
For this solution you would need one inverter pr injector.
Øyvind

[frank_ster]

humm... so what about 1 of these 100 watt 12 vdc to 110vac 14$ power inverters modified for each injector .. .. that would be funny and interesting at the same time.

but from what I have been reading the injectors wat about 90- 150 volts so a european type would be better.

but your saying the the inverter stage comes before the H bridge. but isn't the inverter stage the H bridge its self. doesn't it just take 12vdc and switch it into ac using h bridge then use a simple transformer to raise the output voltage ?

[SymTech Laboratories]

You might be able to adapt a common power inverter, but remember that it will need to be rated for much more than 100W peak output. You may find fuses inside that prevent the injectors from drawing the current they need to open quickly.

[fixmann]

hi

In the "old days" the powerinverter had a "big" ironcore transformer, and these used a pushpull-circuit to
drive a 12v transformer in "reverse" to step up the voltage rate.
These type you can forget to use for this project, 50/60hz and to slow.

Now days a highfrequency driven ferite transformer is used to make half the rated outputvoltage.
The switchfrequency is around maby 50khz and benefit from this is that you get 95+% efficiency.
The H-bridge "makes" the +-110 which will be 50Hz 220v AC

Just take one appart and have a look, you will better see what am talking about.
And yes, if you want 110v you must use a european model.
You can probably squeeze a few extra volt out by modifying the feedback.
So if you put a good quality fast charging capacitor before the H-bridge and after the transformers rectifier, i think this could be possible.(Just make shure the H-bridge can take the peak current)

I understand that the peakcurrent and voltage are high, but i have not seen any commonraildiesel with a
powerplant under the hood, i would believe this could be done with a modified inverter.

It is not so difficult to make your own step up supply from scratch, but the inverters as you say are very cheap.
SG3525 is a nice circuit you can use to make the powersupply before the H-bridge.

Am on vacation, when i arrive home i can take a look, i have one i can try to modify.


Oyvind

[frank_ster]

i would think the 50 khz would be a problem tho. unless its rectified. and some what conditioned since my understanding of piezo electric is it needs dc +- for open and then reverse -+ to close.

as my original plan would of been inverter then rectified, and then capasitors, inductors and then h bridge for each injector.

well to get the party started i could use power supply to create the 100- 200 volts and drive H bridge on a proto board and see if that can be made to work.

off topic , i'm stuck in germany for another day ! stupid strike.

[fixmann]

No the 50khz is the frequency at the transformers, after the rectifier there is dc.

[frank_ster]

so they convert to ac two times ?

first at high switching 50 khz for more effecient transformer.

then back to dc

then convert to 50 hz.

is the above correct ?


at work i used a 220 volt single phase to 3 phase motor drive inverter. i ran it through a transformer, because i needed 480 3 phase. for some reason it didn't like that.

so i then took a 230v 460v dual voltage 3 phase motor 5 hp. ran the motor on half of its windings at 230 and then used the other 230 to power the transformer to make 460-480 volt and it was then perfectly clean ! i admit a bulky way of doing it, but thats what i do! make s*** work haha.

[dieselgeek]

Why not just use someone's existing injector driver? plenty of Ford CRD engines to be found in the junkyards here in the states.

Meanwhile, driving the injectors is the smallest of your challenges. Common rail diesel engines have very little variance in pulsewidth (there's only a few crankshaft degrees where the engine will accept the fuel) - the big trick is managing and monitoring the fuel pressure. Fuel pressure is used for increasing load/output much more than pulsewidth.

another issue is accuracy - I doubt the current megasquirt offerings can cut events down into the tenth-of-a-degree accuracy needed for proper timing control on CRD engines.

I worked on a prototype CRD standalone system that never made it to market (rumor has it that Gale Banks is working with that same system now, however). There's a lot required that megasquirt cannot do in it's current configuration.

[fixmann]

hi
Yup frank_ster, you got it.

I must agreee with dieselgeek, the timing and other stuff would be the biggest problem to overcome.

Øyvind

[frank_ster]

well there would be a few issues, i would think they would be lags.

i would think the injector drivers gates and processor are able to reproduce there injection ability.

there maybe a lag factor but that could be overcome with a few constants.

do you agree with the above ? or do you think the general hardware would would have a tolerance greater than the actual injection times?

the use of external oem injector drivers has been floating around for a while now.

it seams ford power strokes , gm duramax, honda diesel, and few others use these units some of which are known to fail.

for the fuel pressure variance it seams a pwm output to the pump inlet flow would take care of pressure. but i think the pressure variance is more fuel efficiency and not robbing power to pump up to high pressure for no reason.

[fixmann]

Yes i agree most problems could be overcome if enegy and effort is put in to solve them.
Im must say right away that am not wery experienced when it comes to commonrail diesels.

If it is like dieselgeek says that the pw is changed wery little, and the injected amount is most controlled
by the fuelpressure, then i would have try to solve this by making a circuit that converts the megasquirt injector-PW into
pump pwm.
Or even better make a pressuresensing closed loop , and then convert pw into actual pressure.

So to sum it up:
1.Make highvoltage injectordrive circuit controlled by timingtable.the injector-PW must be a number wich is corrected by
table-PW.
2.Convert table-PW into fuelpump-PWM(fuelpressure).
If i remember right, i think the fueltemp also must be taken into acount.

To solve this you must as metioned make some electronics(microcontrollerbased) or rewrite ms sourcecode.
This is just my guess, i think it could be done(not how well thoug).

Øyvind

[frank_ster]

the injector pumps already have a pwm solenoid limiting the fuel entering them so that would be as simple as driving an injector.

fuel temperature wont matter that much but even then it could easily be factored in.

so the biggest unknown i have is the piezo injector drivers.

[fixmann]

hi

When i am home from my vacation(14 days) i will see if i can help you with the driverpart.
Personally i dont see that the driverpart should be to difficult to put together, the timing of high,low and let go is easy done by a microcontroller.

Øyvind

[bobtooke]

Following this thread with interest...

You guys have lost me with the way you might get piezo injectors to work but when you do (I know you will), you have a willing tester. I've been pondering trying out a MS3 and a bigger turbo on my Landy TD4 for a while now.

People have told me MS3 on a BMW TD common rail won't work because of the complex injection cycles, so I'm itching to prove them wrong.

[dieselgeek]

People have told me MS3 on a BMW TD common rail won't work because of the complex injection cycles, so I'm itching to prove them wrong.

Do you plan on re-writing the base MS code? because that's what it's going to take.

For those interested in common rail diesel engine management, Bosch has a number of good books on the topic. Even the Bosch Automotive Handbook has a good chapter on Diesel engine management. You're not going to be doing it anytime soon with an MS3-based product, IMO.