Rather than using two pumps for a two-stage system, we can use one high-vacuum pump and two small(ish) tanks. The first stage will use a small tank, only modestly larger than the maximum amount of air pulled from under the plastic during the initial pulldown. Then we'll switch to the second (unpolluted) tank for the long, hard pull to get good detail.
As before, I'll show an idealized, simple version using a 3-ported valve. In this case, it must be a non-mixing valve; that is, it should connect one port to either of its output ports, or neither, but not both. Then I'll show how to relax that criterion, so we can use a cheap 2-port ball valve instead.
Code:
+ - - - - - - - - - - + +------------+
| platen | | first- |
+--------+ +---------+ +-----------+ stage |
| | | +---------+ (initial |
| | | | | pulldown) |
| | +-------+ +---+ +--+ | vac tank |
| +--------+ ball +-------+ 3-port | +------------+
+-----------+ valve +--+ +--+ valve |
+-------+ | | +---+ +--+ +-------------+
| | | | | second- |
| | | +--------+ stage |
| | +----------+ (hard pull) |
| | | vac tank |
| | | |
+--+ +--+ | |
| high | | |
| vac | | |
| pump | | |
+--+ +--+ | |
| | | |
exhaust +-------------+
To evacuate the tanks, we close the ball valve to isolate the vacuum system from the platen, engage the pump, and switch the 3-port valve to pull from one tank, then the other. When both tanks are evacuated, we set the valve to leave the path to the first-stage tank open.
When we are ready to form, we lower the plastic and open the ball valve. This allows air to rush into the first-stage tank, performing the pulldown, but polluting that small tank fairly seriously and weakening its vacuum.
Then we switch the 3-port valve to close off that tank and open up the second-stage tank. Since it's been sealed off, it's unpolluted, and can perform the long hard pull at full strength. (Initially, anyway... it will of course fill up with air leaked around the edge of the platen, slowly losing strength. But at least it'll start out with a good hard full-strength pull.)
One advantage of this scheme is that if the first tank is appreciably larger than the volume of air to be pulled out from under the plastic, it will pull pretty hard, even if it doesn't pull as hard as the second tank. For example, if it's 2x the maximum volume to be evacuated, and our vacuum is fairly hard, it will pull at least two or three times as hard as a (vacuum cleaner-type) high-volume centrifugal pump.
A disadvantage for a simple, manually operated system is that it requires a hand on the valve at the moment forming begins. Ideally the main ball valve would open just as soon as the plastic is down and can make a seal at the platen.
For a very simple system, where two hands are used to lower the plastic and hold it down, this may require a 2nd person to operate the valve. It's therefore more suited to a system where the plastic can be lowered and held down with one hand while the main valve is opened with the other.
There's also a subtle gotcha with this scheme, in that using a non-mixing 3-port valve may cause a loss of seal in the moment of switching tanks, when neither tank is connected, unless the plastic is held firmly down mechanically. If the 3-port valve is the mixing type, with both ports slightly open when the handle is in the in-between position, there's another problem---for that brief period, the fresh second-stage tank will suck air from the from the polluted first-stage tank, polluting itself somewhat.
That drawback, at least, can be easily fixed, and in fact the fix makes the plumbing a little cheaper. We can simply use a restricted-flow connection between the 2nd stage tank and the 3-port valve. This could be an "orifice" fitting (a pipe or tubing connector with a constriction to limit flow), or we can simply use a small hose to connect the tank.
The orifice or hose size should be large enough not to impede evacuation of the tank, and to allow it to comfortably keep up with leaks during the long hard pull, but too small to allow much air to be pulled from the first-stage tank during the time the valve handle is in the intermediate position.
That idea will also let us use a second (common and cheap) 2-port ball valve instead of a rarer and more expensive 3-port mixing valve, or a still harder to find 3-ported non-mixing valve. (Sure, you can find them, but finding large and cheap ones isn't so easy.)
The flow requirements are so different during the initial pulldown and the long hard pull that we can simply leave the paths to both tanks open during the initial pull down, and the second tank will not be terribly polluted as long as the flow is very restricted during that time (roughly one second).
This allows this schematic, using an orifice
Code:
+ - - - - - - - - - - + +------------+
| platen | +-------+ | first- |
+--------+ +---------+ +----+ ball +--+ stage |
| | | +--+ valve +--+ (initial |
| | | | +-------+ | pulldown) |
| | +-------+ | | | vac tank |
| +--------+ ball +-------+ | +------------+
+-----------+ valve +--+ +--+ |
+-------+ | | | | +-------------+
| | | | +------+ | second- |
| | | +--+ ori- +--+ stage |
| | +----+ fice +--+ (hard pull) |
| | +------+ | vac tank |
| | | |
+--+ +--+ | |
| high | | |
| vac | | |
| pump | | |
+--+ +--+ | |
| | | |
exhaust +-------------+
or this equivalent one using small, cheap tubing to connect the second-stage tank
Code:
+ - - - - - - - - - - + +------------+
| platen | +-------+ | first- |
+--------+ +---------+ +----+ ball +--+ stage |
| | | +--+ valve +--+ (initial |
| | | | +-------+ | pulldown) |
| | +-------+ | | | vac tank |
| +--------+ ball +-------+ | +------------+
+-----------+ valve +--+ +--+ |
+-------+ | | | | +-------------+
| | | | small | second- |
| | | ============== stage |
| | +-+ tubing | (hard pull) |
| | | vac tank |
| | | |
+--+ +--+ | |
| high | | |
| vac | | |
| pump | | |
+--+ +--+ | |
| | | |
exhaust +-------------+
To evacuate the tanks, we close the first (main) ball valve, and open the second one, and engage the vacuum pump.
During forming, we open the main ball valve, and after about a second, we close the second one.
One nice feature of this scheme is that evacuating the tanks leaves the the valves in the right positions for forming, by default. (Main valve closed, first stage valve open.) It's a little more complicated to operate than a one-valve system, but not much. As long as we don't need both hands to hold the plastic down at the beginning, it's pretty easy to use.
We don't need to hold the plastic down when switching from first-to-second stage operation, because the path to the 2nd stage is already open, albeit small, and suction is maintained. We will not lose vacuum (or our platen edge seal) during the switch.