Okay... as promised in another thread, I removed the pressure differential switch from my 70 Vette and disassembled it (again). I decided to start a new post so I wouldn't be hi-jacking somebody elses.

As a side note...
One vendor has a C3 switch break down as follows-
1969
1970-1973
1974-1977
1978-1982

So far, here is what I found (for the second time)...

All metal parts are brass except for the switch pin and two internal springs.
The switch pin does not move on this test part.
When it's centered, it is sitting in fluid, not touching anything. When the piston pushes the dog-bone to one side, the dog-bone contacts the switch pin. The dog bone (besides touching the grounded case) is touching the pistons, which touch the springs which touch the inverted flare inserts, which touch the switch housing which touches ground. That's when the light comes on. It will stay on until the dog-bone is re-centerd and the connection is broken (open circuit).

The switch DOES close off the offending port(s) when activated (so let's please stop debating whether or not the switch closes off the port).

But there is a caveat...

The bore is measuring at .375 diameter while the pistons are at .341 and the dog-bone in the center is at .342.

The difference in diameter between the pistons and the bore is .374 - .342 or about .032. This means there is .016 clearance between the piston walls and the bore wall (probably a little tighter on a new switch).

Meaning.... even though the pistons move to cover up the fluid input ports, there is sufficient clearance for fluid to leak past the piston and still get access to the output ports.

I reassembled the switch and stuck a rod through the front right brake line output port (bottom end as mounted) and forced the pistons and dog-bone to one side as far as it could go (hit the limiting switch pin), simulating a rear system failure. The top piston moved sufficiently far enough to block off the rear fluid input port.

Then I did the converse... pushed the internals the opposite way... and the other input port was blocked.

This a "visual".... I can see the piston move to block the port... good for me... the doubting Thomas.

The amount of movement is governed by the dog-bone in the center of the switch. The switch pin is .256 in diameter... the inside distance between the ends of the dog-bone is .386.

At rest, the pistons are right at the edge of the input ports. The springs keep the pistons and dog-bone "stacked up" with no clearance/endplay between them.

The test done several years ago.... with a simulated popped left front brake hose (steel line routed back to M/C), the car stopped, but not very well.

I am taking measurements of the switch etc. so I can draw a cutaway view.

Next, I am going to try to actually measure the affect on flow using a regulated water supply and a bunch of fittings I will have to make up. I will get a flow rate with the switch un-activated, and then a flow rate with the switch pushed to one side.

The pressure will be low, but the % difference should be an indication of how it will work under real braking situations..... such as the test I did a few years ago.

Stay tuned.

 

This is close to what you have and will explain in more detail how it works. Hope it helps.

http://patimg1.uspto.gov/.piw?docid=...B33F0&HomeUrl=

 

Thanks...

I think I quoted a few lines from that URL a few posts ago.

I was surprised when google came up with a patent URL.

On a personal note-
I tend not to depend on things that are written such as "white papers".

The work I'm doing now involves patents and white papers & such and I (we) have even found our own research to be mis-quoted and wrongly applied to other "white papers", so the errors just propogate.
We can't believe everything we read.

So... I tend to do my own "research" and put the scholars to the test.... see if their mechanical philosophies actually work in the real world.

I'm trying to get as accurate dimensions as possible on this "switch" before I put it in CAD. The inside corners are all radiused making it difficult to measure.

I'll be back.

 

Testing continued…

This switch is the original, soft brass, and is pretty well worn… so the numbers may be slightly different with a new switch with closer tolerances. The information stamped on the integral (swedged) bracket is:
PATENT
3369090
3374322
It also has a hexagon with what looks like a W in the center.

November 1981 GM parts manual group 4.690 calls for
65-66 #3878944 ($126.00) W/H.D. Brakes
67-68 #3904303 ($25.75)
70-73 #3980797 same as in the 1970 AIM ($24.70)
74-77 #339780 ($39.25)
78-82 #1257208 ($43.75)

The switch pin which controls the dash light also limits the dog-bone in the center of the housing. The dog-bone moves toward the leaky brake circuit, driven by the o-ringed pistons. The single o-ring on each piston is offset toward the center of the housing. The switch pin diameter is .254. The total dog-bone inside clearance is .376. This leaves .. (376-.254=) .122 unoccupied space between the inside edges of the dog-bone. This means the dog-bone can only travel ½ of that…. or .061 when the piston is forced all the way to one side as a leak occurs. At rest, each piston is behind its respective fluid input port by several thousandths… i.e.- the pistons are not right on top of the port (hole).

The piston diameter is about .341. The inside bore of the housing ..342. However, the housing bore diameter increases to .374 even further back away from the fluid port- about 3/16 from the inside edge of the port. This is why the piston o-ring is offset toward the center of the housing… to keep it from slipping into the wider .374 diameter part of the bore.

This means when a circuit leaks, the piston can only travel .061, and must first make up the distance it is behind the port, and then cover up the port to shut off the flow. Since the FRONT port is 3/32 diameter (.093), the piston would have to travel that far to close off the port entirely…. which it cannot…. it only moves .061 max (about 2/3 of the FRONT input port gets covered). This is difficult to see when looking through the fluid input port hole, even with 3.00 reading glasses and a flashlight. At first glance, it appears as if the entire port gets closed off.

This coupled with the increase in piston to bore clearance from .001 to .015 at the port means the piston cannot entirely close off the leaky circuit.

This is what I found by testing with water, less viscous than brake fluid, but good enough for this purpose.

I found a maximum of about a 20% drop in water flow when the switch was activated.

I only tested the FRONT circuit…. the REAR input port is a larger diameter (5/32) and piston movement of .061 would cover less of this port than the FRONT port.

Bottom line… the piston does slide in front of the fluid input ports, but doesn’t entirely close off the fluid.

They could easily have shaved .060 off the piston, resulting in no port closure at all, and the switch would still have functioned to turn the dash light on & off… in my opinion… the 20% figure (+/-) was by design.

The next step is to bench bleed the M/C, test each chamber/port separately to make sure they both independently develop maximum back pressure when the other one is “dead” (fluid ported back to the chamber), and re-install the switch. Actually, I should get a couple of gages for this to measure each port independently and then together.

Then I can re-do the stopping test.

This is getting expensive…. my digital camera died and I have to get a new one to take the rest of the pics.

 

Next phase of testing complete...

I have two steel lines going from the master cylinder output ports directly back into the master cylinder fluid chambers.

Bled the master cylinder... both chambers at the same time.

During bleeding, the pedal goes to the floor... as we all know & expect.

Removed the rear bleeding tube, and installed a port plug.
Bleeding tube still in front.
Pedal went to floor.

Now switch 'em-
Removed rear port plug, installed bleeding tube.
Re-bled both chambers.
Removed front bleeding tube, installed port plug.
Pedal went to floor.


Removed rear bleeding tube... installed a port plug.
Now have port plugs in both ports.
Pedal hard as a rock.
Master cylinder is working.

So much for the "dual master cylinder" save your life philosophy.

More on this later.

Next, re-install the switch and re-test.