Adjusting Brake Booster Characteristics

I've been unhappy with my power brakes for awhile. I believe that the boost is coming in too early, assisting too much at low input force, and running out of steam before maximum braking is achieved. Relevant discussion of boost ratios and helping force can be found HERE. Here's me talking myself through how the booster poppet valve works:

Here is my sketch of the brake booster. 1 is input force, 2 is atmospheric pressure, 3 is metered vacuum supplied through the poppet valve, 4 is full engine vacuum, and 5 is output force at the master. 

Here is a poppet valve:

You can imagine that orifice 2 is atmospheric pressure, orifice 3 is the pressure inside the #3 section shown above. Therefore the poppet valve self regulates. The input force must overcome the spring on the left hand side that is trying to close the valve. The input force is aided by the pressure differential between 2&3. As pressure in #3 rises, the assistance from the pressure differential drops, requiring more input force to overcome the left side spring. This cycle continues until input force has risen to equal the spring pressure, at which point the "helping force" referenced by previous posts has been maxed out. 

 

So maximum helping force is solely a function of diaphragm size, while the ratio is determined by spring pressure, and likely the size of the diaphragm in the poppet valve. So 2 different 8" boosters may have different boost ratios, but the same maximum helping force. This means that the higher ratio booster hits the maximum helping force with less effort, but does not generate more maximum force.

After reading a few more threads around the internet and looking at some diagrams, I headed to the garage to disassemble my 1994 single diaphragm miata booster. The output seal comes out easily, and I was hoping to access the reaction disc. The reaction disc is a rubber disc that acts as the "left side spring" in the valve illustration above. My plan was to shim this disc, or determine a way to increase its spring rate. However the output plunger seems to be captured by the return spring on this particular booster.

No worries, I flipped it over and decided to remove the input plunger. I had read somewhere that it could be removed by pulling on it with a large amount of force.... Turns out that this does not apply to the miata booster. Instead I managed to snap the plastic tube that houses the poppet into 2 pieces. Here it can be easily seen that the input plunger is captured by a key, that must be removed. Unfortunately, I don't see a way to remove the key without opening the booster body. 

This booster is now toast, but I will disassemble it completely to update this thread and gain some education. A new booster is on order.

Seems like it might be easier to adjust your curve with the pedal linkage. 

It seems also that while you may improve the rate at which assistance happens, nothing here will change the "running out of steam before maximum braking."
Unless your complaint isn't that it's not enough assistance at max braking but simply that assistance stops increasing too far before maximum braking...

I'm guessing you're one step ahead of me there, but I wanted to sanity check what you're trying to achieve.

You're more hoping to change the rate of assistance so that it applies more evenly across the range of braking you're using?

I hadn't really thought about it, but I wonder if it's one of those things like the OEM understeer; we want to help you apply the brakes easily but we also don't want to help too much when you're panic stomping? Do big, grippy tires effectively want to you to run assist further up the slope if not to a greater total assistance?

Is there such a thing as a "performance" booster which is more linear across its range? Or have I finally worked my way back to the thing you're trying to achieve which doesn't exist any other way?

I need to dig up a couple references before I reply in any detail. Generally speaking, IIRC, the profile of the reaction disc influences the initial apply characteristic aka the "jump in" ie how much helping force is delivered at onset of braking. The ratio disc determines the force multiplication from jump in to runout. Again IIRC, reducing boost ratio will raise the runout point. I'll be back later, possibly not til tomorrow, but I promise I'll be back.