Thanks for your kind words. It's appreciated.

I really like that mounting and I think I'll copy it :). My saab pump looks very close, could be the same pump from a different application. As I have no climate, locks, or trans, I believe the one pump with a good sized vac res will suffice for brakes and shutdown. My only hold up right now is getting to the boneyard to pull a OM617 pump to use as a cover. Maybe Friday....although the Alfa has been getting my wrench love.

I think you'll be just fine for that app. Make sure you post some pics of your bracket.

mach4, do you currently have a vaccum reservoir other than the one in the brake booster?

Yes, there are two reservoirs on the 107 chassis. There is an "egg crate" type one in the passenger side wheel well that services the vacuum door locks and another "globe" type one tucked away (virtually inaccessible) on the drivers side between the fender well and the interior panel that services the climate system. The brake has it's own reservoir.

X2, more excellent work.

I've been pondering the "vacuum leak in the transmission" system issue, and am coming to the conclusion that this may not be as big a problem as has been speculated. Here's my current understanding -

The VCV is designed to provide a vacuum signal to the transmission which emulates the vacuum signal of a gas engine so that one transmission design can be used for both. So the VCV provides a "high" vacuum at idle (10-12" or so depending on how the spring in the VCV is adjusted) and at full throttle goes toward 0" (likely 1-2" in practice). This vacuum acts on the modulator which is used to control the firmness of the shifts.

So at idle (in a hypothetical system) the VCV would allow 12" of vacuum to enter the VCV chamber and "pull" the modulator accordingly. In a steady state, no further vacuum is needed so no further vacuum is used. However when the throttle is pressed, the 12" of vacuum bleeds down based on how far the throttle is pressed. At full throttle, all the vacuum is bled and as long as the engine stays in a full throttle state, no additional vacuum is needed. Now, when the throttle is released, assume to some "cruise" position, vacuum is replenished to say 8", and again, assuming steady-state cruise, no further vacuum is needed until the throttle is released further when the VCV replenishes the lost vacuum to 12" again assuming it goes to idle.

This is no different than the operation of a tight climate system where vacuum is used in a POD to change the position of a flap and remains steady-state until the vacuum in the POD is released to return the flap to the neutral (default) position.

So if all this is correct, assuming that the volume of vacuum represented in the VCV as essentially equal to the volume of a POD in the climate system, this should not over tax an electrical vacuum system, unless someone (maybe like my wife :) ) is constantly mashing the throttle and releasing it to maintain a speed (what I call PWM throttle control :) ). And of course, assuming the systems are all tight and leak-free.

That's my current understanding - am I missing something?

One more thing...

In looking at the vacuum routing for the VCV there is no reservoir that this is tied to so it operates directly based on existing vacuum in the feed line to the brake booster. So it would make sense to either add another dedicated vacuum reservoir (there's plenty of room) or possibly move the vacuum supply tap to upstream of the climate check valve to use the climate reservoir. And same for the shutdown vacuum supply.

I have the same understanding, so I believe you are correct. You could test your theory by "T"ing your vac gauge to the modulator line and driving down the road with the gauge held by the windshield wiper. You could see the vac condition at all times (don't text while doing this test either :D)....Rich

Making good progress on the project.

I got the control circuit bread-boarded and the code written.

Here is the test setup.


http://www.peachparts.com/shopforum/...1&d=1388798779

I'm using an Arduino micro-controller and a GM MAP sensor that gives me a perfect 1bar range from 0-30 in-Hg. I'll be building a 12v -> 5v converter to power the system off the car battery. I had hoped to be able to drive the motor directly from the micro-controller using a high-capacity logic-level MOSFET, but have opted instead to drive a Bosch relay with a power-transistor instead. The package will be mounted in an aluminum MB electronics enclosure with its pin connector and pigtail I salvaged off the 123 donor car. I just opened up the case and cut the board off the pins and will solder wire connectors to the pin stubs. The MAP sensor will be mounted on the pump bracket.

How the system works: The vacuum sensor reads the current vacuum level once a second from the MAP sensor. If the vacuum is lower than the set point, the pump turns on and stays on to a specified level above the set point. Right now it's coded to turn on at 15" and turn off at 18". Those are just variables, so changing the setup is as easy as uploading a modified program through a standard USB cable. I can adjust the amount of hysteresis as needed based on real-world testing. I've currently got two LEDs configured. The green LED simply shows the status of the pump - ON or OFF. To protect the pump, I've got it set to calculate the amount of consecutive minutes of run time and if it exceeds that it will shut down the pump for a specified "cool-down" interval and turn on a red warning LED in the cabin. Before it shuts down the pump however, it sets the red LED blinking as a warning. I'm guessing 15 minutes of run time to get to shutoff and 20 minutes of cool down before resetting. The warning LED will probably be set at 8 minutes. This will be determined and adjusted based on testing. A ver 2 mod might be to add a temperature sensor to actually read the pump temperature and handle the shutdown based on temp rather than a proxy of duty-cycle.

The LEDs will be run to the cluster and occupy an empty warning slot next to the Anti-lock brake warning light.

The electronics are pretty simple, but the code was a challenge. I haven't coded anything since the dBase days (guess that dates me :) ) so learning a new language was interesting.

My in-vehicle testing process will be to create a small pressure reservoir (I'm thinking 6" of 1" PVC pipe or something similar) to which I'll hook up the just the VCV. By using such a small reservoir I'll get a chance to estimate the amount of vacuum that the VCV uses in a real world experience by just monitoring the pump cycle. We know the door locks and likely the climate controls (assuming a tight system) are very low vacuum consumers so the big question is the VCV. By the way, I could easily move the other non-brake vacuum users to the same reservoir and estimate their aggregate usage.

As I mentioned previously, I'm considering running a dual pump system with climate, doors, VCV and shutoff on one pump and brakes with its own dedicated pump. (Just need to score another $20 pump.)

More to come.

Wow. This is excellent. Once I comprehend it all, it will probably seem even more excellent.

Out of curiosity, what do you/did you do for work?

The system is just about ready to be mounted in the car. I redid the electronics on a very small breadboard and mounted it in a relay case from the 123 donor that I gutted.

It was really tricky for a number of reasons. First the Arduino runs on 5v so I needed to build a voltage regulator to drop the cars pretty dirty 12v (really 11-14 volts) down to a clean 5v. The Arduino will actually take 12v, but not the higher charging voltages. Then I had the problem of driving the pump. The 5 volts available on the Arduino was insufficient to drive a 12v relay (and the 5v ones just wouldn't handle the amps). So I'm using a power transistor to drive the relay.

I thought I had my test vacuum reservoir figured out, but couldn't get it leak free, so that's still on the drawing board.

It all seems overly complex when a Vacuum Switch with a Relay would work to turn the Vacuum Pump on and off.
If you go to someplace like Summit Racing that is what they have in their Electric Vacuum Pump Kits.

Of course.... but what would be the fun of that :) I stated earlier in the thread that this isn't being done as a "solution", but rather a learning project and a proof of concept. Learning is always best when working on a real project...and boy have I learned a lot.

I also don't think the fairly small volume of the transmission modulator requires more flow than the pump can deliver. I suspect the flow restriction of the VCV is the main limiter in how fast vaccuum is applied to the modulator. Before adding a reservoir, try a simple test. Disconnect the tube to the modulator and turn on the vacuum pump. Measure the vacuum pump outlet pressure vs the VCV outlet at various "throttle" positions. If the pump keeps up, even at wide-open-"throttle", a reservoir wouldn't help much. 1985 cars got a "vacuum amplifier" (blue moon) to account for the VCV restriction (and to process a boost signal).

The reason for the test reservoir is to create a bit of a volume to better gauge the amount of vacuum consumed by the VCV. If I ran it without a reservoir, the pump would presumably come on almost every time there was a change in throttle position and just run for a second to replenish the vacuum in the lines. This wouldn't give any meaningful data. By having the small reservoir there, it will take a bit of running of the pump to get it to baseline vacuum level, and the elapsed time between pump runs will provide a qualitative indication of the amount of vacuum used. I'm using about a 7oz bottle as my test reservoir.

At a fixed throttle position, there should be no vacuum used. I guess we'll see.

When I first did the swap, I had the 722.4x Cali transmission in the car and I used the "blue saucer" with the transmission. When that transmission went out, I replaced it with a 722.3x unit and kept the "blue saucer" for awhile. It worked fine. Then I decided to clean up the underhood mess a bit and removed it. The transmission actually does just a tiny bit better.

I have no idea what the vacuum amplifier does - it would be nice to know.