Turbo boost control

[shender1]

John Slade,

 

From what I understand you are going to use a manual waste gate to control boost on your turbo. What I was wondering is

how the fuel mixture will be controlled as the amount of air to the engine is increased. Does the fuel injection compensate for this automatically or is it done manualy in the traditional fashion?

 

Sean

[John Slade]

Does the fuel injection compensate for this automatically or is it done manualy in the traditional fashion?

I believe Tracy is working on this, but right now it's manual - mixture control is a knob on the EC2 control panel.

[shender1]

I guess I don't understand how you are able to determine how much fuel to add or remove..do you watch the tach or the egt?

[John Slade]

I guess I don't understand how you are able to determine how much fuel to add or remove..do you watch the tach or the egt?

Good question. Tracy supplies an air/fuel ratio sensor which goes in the exhaust. This gives me an LED bar readout of the fuel ratio of the exhaust gases.

 

I wanted to control the boost to make sure the auto vacuum control system didnt take boost off when I needed at height, and put it on when I didnt want it (say on a go around). Greg Richter simply welded up his waste gate and used a pop off valve. I may do that later. I also have the pop off (currently set at 6 PSI) but wanted manual control of the wastegate as well. Once initial testing is over I dont expect to be adjusting the boost control very often, just experimenting with it to see how boost impacts various flight regimes. The EC2 has two computers, each with its own set of mixture parameters at various RPM. I may set up one for a full boost and the other for a zero or limited boost condition. This way I can switch between them and have the correct mixture for the different boost settings. This is very much an on-going experiment.

[shender1]

I was messing around with the Mooney on MS flight sim 2004 and the mixture is determined by TIT or Turbine Inlet Tempeature as opposed to an egt which I think is normally used on turbo charged engines. Do you plan on adding a sensor for this?

 

I understand why you are going with the stock turbo on the 13b. I know that Bruce Turrentine reccomends the the supercharger but then you are stuck coming up with some type of muffler. The turbo solves the problem but I guess the question is whether the stock turbo will take running at 75%n power all the time. I don't know if the Mazda engineers designed it for that.

 

I am still educating myself about the relationship between mixture settings and fuel flow control. I'ts still a type of voodoo science to me. I guess one of the downfalls about the 13b is that you are pretty much stuck custom building some type of fuel injection system.

 

I'll be really interested in the results you get from yours.

[dust]

we have had a big conversation about the duty cycle of an aviation vrs auto turbo and as it turns out, the aviation cycle is far easier on the turbo than the starts/stops/heating/parking cycle of the auto duty cycle

 

The auto duty syscl tends to cook the bearings where the aviation use does not

 

enjoy the build

 

dust

[mplafleur]

There are so many different methods of determining spark advance and mixture it can be quite confusing. Add a turbo, and it gets messier. It depends on what your mission, or goal is. Power or economy. Typically, in an auto, a change in throttle will drop you out of closed-loop operation (economy) into open-loop control. If you've been in steady state for a pre-determined amount of time, then you go back to closed-loop control. This loop is usually controlled around fuel mixture which is determined by the EGO (exhaust gas oxygen) sensor. The best mixture for fuel economy and reduced emmissions is 14.64:1, air to fuel. There is a map for spark advance and it is retarded when the knock sensor detects a knock. The length of the fuel injection pulse is calculated by the fuel pressure, fuel temperature, air pressure, air charge temperature, and coolant temperature. Without a turbo, barometric pressure is used. This can go on and on. It's been quite a few years since I wrote engine control software (Formula 1 racing), so things are a little sketchy. While we monitored turbo inlet and outlet pressure, we did not control the turbo. There was a manual (fixed) popoff valve.

 

 

John,

 

Does each PCM (ECU) control a separate rotor? Does each then have it's own sensors?

[John Slade]

Does each PCM (ECU) control a separate rotor? Does each then have it's own sensors?

There are two manifold pressure feeds to the computer, but only one air temp sensor. As I understand it the computers are largely redundant and either one controls both rotors.

the mixture is determined by TIT or Turbine Inlet Tempeature as opposed to an egt which I think is normally used on turbo charged engines. Do you plan on adding a sensor for this?

The air temp sensor for Tracy's ECU is in the inlet manifold. As I said earlier, I think Tracy is working on adding turbo boost control. When he comes out with that it will probably involve additional sensors.

it turns out, the aviation cycle is far easier on the turbo than the starts/stops/heating/parking cycle of the auto duty cycle

Yes. I'd heard that. Also the stock (single stage '91 turbo) comes with oil lubrication and water cooling. It seems to be a fairly robust unit. If it's anything like the one on my Saab it'll go forever with zero attention.

I guess one of the downfalls about the 13b is that you are pretty much stuck custom building some type of fuel injection system.

Hmmm. I didnt realize I was custom building the injection system I learned that I need 4 550 injectors, 2 in the stock primary location and 2 in bosses in the secondaries. These hook up to the ECU. That's about all I know.

I'll be really interested in the results you get from yours.

Not half as interested as I'll be.

[shender1]

John,

 

You are right about the fuel injection system. What I meant to say is that there really isnt any off the shelf standard intake system that I know of or that I have seen.

[Jim Sower]

I thought the parameters that drove Tracy's ECU were MAP, Inlet (charge) temp and RPM. With these you know mass flow which should give you the mass of fuel and therefore pulse width you need. I can see TIT being useful in lieu of EGT as an indicator for "trimming" mixture similar to EGT in a Lyc. That said, I have to wonder how useful it would be with a waste gate since I would expect TIT to be all over the park depending on wastegate position. In any event, I thought Tracy's ECU got you pretty close all by itself, turbo or NA. If I was turbo'd I would be very interested in a knock sensor inputting to advance. Does Tracy have any plans around that, or any kind of detonation protection?

 

What am I missing? .... Jim S.

[mplafleur]

That would be a speed-density system. Manifold absolute pressure (MAP) and air charge temperature (ACT), also called inlet air temperature (IAT) is used to determine the density of the air entering the system. From this, fuel can be metered at the correct ratio. Engine speed (RPM) is entered into the equation as a modifier. I'm not sure that TIT would vary that much. A knock sensor would be nice but they are expensive and so is the circuitry needed to integrate the input.

[Jim Sower]

<... I'm not sure that TIT would vary that much ...>

Hmmmm. With wastegate bypassing anywhere from NONE of the exhaust to ALL of it, I would have intuited that TIT would vary a lot. But I guess it doesn't really matter since you're only using it to trim mixture to near peak TIT for a given wastegate setting. It would still tell you what you want to know, just like EGT.

 

<... A knock sensor would be nice but they are expensive and so is the circuitry needed to integrate the input ...>

I had always thought that knock sensor was the "blunt instrument" of the inputs to an E-FI-I system. Cheaper than O2 or crank angle or TBI, and when it goes off, unit just knocks a few degrees off the advance every time he hears it (and then starts sneaking back up). Didn't know it was all that tough to buy or integrate.

 

Live and learn .... Jim S.

[mplafleur]

I guess it depends on where the sensor is actually located. I was thinking that the TIT sensor (should I reword that? ) would be at the inlet of the turbine and would see the temp of the exhaust regardless of whether the wastgate was open or not.

 

In reference to the knock sensor: I don't know what a replacement OEM knock sensor costs. I do know an EGO sensor is expensive. The heated ones that is. I was referring more to the electronics that have to go into the module. The knock sensor output is special and has a special IC that is interfaces with. At least all the ones I've dealt with have. With all the supplier competition out there now, maybe an off-the-shelf solution can be had for not too many $. And yes I agree, a knock sensor would be nice.