DasTT

Steve is making me a BAMM so why not a boost machine as well? I had gotten the AutoSpeed articles a long time ago when they were free and decided to review once again. Now I think I understand why this all works. So I thought I'd write it all down and post.

I take the analytical approach here because of my engineering background. Mostly though I just can't see me turning dials without knowing the underlying principle at play. A quick review of my understanding....

It's all about controlling the pressure signal to the waste gate since this controls how much of the exhaust is used to generate pressure on the charge air side. Higher signal (pressure) to the waste gate allows more exhaust to bypass the turbine. In the absence of any control, the charge air pressure is fed back to the waste gate diaphram. When this is the design a spring is used to counter the feedback pressure. This sets the maximum charge air pressure (boost) by opening the waste gate when the boost gets to the spring value. The boost then settles in around the spring value.

Even with electronic controls a waste gate spring is still used to set the limp mode maximum boost. For the 225 with a K04 I believe this is about .4 Bar or 6 psi. Below this boost level the engine doesn't need any control.

The control on our engines is the N75 valve. It's purpose is to bleed off the pressure signal to the waste gate. If we completely bleed off the signal (100% duty cycle from the ECU) then the turbine gets all the exhaust to generate boost on the compressor side. At the other end (0% duty cycle) the system operates like a direct feedback system and boost is limited to the waste gate spring value.

The bleed off rate of the N75 is what makes boost smooth or not (i.e. other versions of the N75 change the oriface size or effect of duty cycle on how much of the signal is bled off) and is designed to work with a particular engine/turbo combination. In general reducing the oriface size will delay boost but hold it longer. The ECU attempts to compensate but the feedback loop has a pretty long time delay as well as having the ECU predict N75 duty cycle based on other inputs. In the end, the system has to learn driving habits and make best guesses at how to control the duty cycle for the N75.

Enter the boost machine. I'll assume the parts of the machine are well known by this group. The two valves set the onset of control (the relief valve) and the maximum pressure signal (the regulator valve) available. The third control (one way valve) releases pressure in the line to the waste gate when boost is reduced (i.e. the DV operates) by dumping it back into the charge air side.

The BM is placed in-between the charge air port and the N75 valve. It's operation can now be understood quite simply. As pressure builds on the charge air side, no signal is allowed to the N75 until the relief valve setting is reached. Then the pressure signal matches the charge air pressure until the regulator setting is reached. At that time, the pressure signal to the N75 is fixed to the regulator setting. This does two things. Firstly it allows the turbo to spin up uncontrolled until a certain charge air pressure is reached. Then we are in the area of "normal" control via N75 only. Then we move into restricted control through the N75. In this restricted mode we have limited the ability of the ECU/N75 to reduce boost. As far as the N75 is concerned, the boost is at the regulator set point. The ECU however is going to eventually decide that it can no longer control boost because it is measuring a boost higher than its command boost. The algorithm is likely very flexible given the issues in controlling this kind of system but pushed to the limit the ECU will post a DTC.

In setting the regulator (clockwise increases its set point) above the expected maximum boost and the relief valve at zero (completely CCW) we are operating without interference. As we reduce the regulator pressure (moving it CCW) we will find the point where the regulator takes over control from the N75 through restricting the pressure signal. A little bit lower (CCW) and we are "adding" boost. We're allowing the turbo into un-controlled space again. The turbo of course will attempt to boost well beyond its efficiency and as has been pointed out somewhere around 22-24 lbs is where the point of deminishing returns occurs.

The relief valve can then be brought up to a point beyond the waste gate spring value but below the regualtor value to effectively become the new spring value. But the advantage is that no pressure will flow keeping the waste gate from "creeping" open just a little. The combination of the relief value and the one way valve on shifting means we'll snap the waste gate closed preserving revolutions in the turbine (all the exhaust through the turbine). This combined with release of charge air through the DV will keep the turbo spinning faster in-between shifts. Thsi is turn results in a faster on-boost when the shift is complete.

We'll anyway, that's my thoughts/analysis. Now I just have to wait for parts to show up at the ModShack so Steve can complete orders and ship 'em out. Looking forward to joining the boost machine gang.

--bob

Black BeauTTy

All's I can say is "You Da Man!"

mudd

umm....do i get like a college credit for reading this?

1T42T (aka ZZ CETT)

I hope you're right, cuz it jives pretty much
with my understanding. And the one thing you
confirmed for me was the N75 duty cycle. I had
assumed it worked like that but didn't know for
certain.

thanks!

scot

MrFrostie

I think your write-up has gotten me a lot closer to fully understanding the boost machine. Nice write-up, thanks for educating us!

TTschwing

Red hose should be here Monday. Assembly today...

And a little visual to help understand...

<img src="http://images8.fotki.com/v132/photos/7/7305/191484/turbo_schematicvicopy-vi.jpg">

MCPaudiTT - DEFCON 1

makes me want to break out the Oxy/MAP torch I bought to make my boost machine piping. Hopefully I will get to it this weekend or next...