This is a job for a programmable thermostat, not a timer switch
The heater in question, a Cadet RM162, is a 240VAC unit, hence the two pole switch (which was inadvertently disassembled by the OP). The timer switch the OP has, unfortunately, is a single pole, 120VAC unit. While a relay could be used to get around this limitation, it's cheaper and easier to use a line-voltage programmable thermostat instead, such as the Honeywell TL8230A1003.
You may want to pause to learn a little more about switches.
The first concept is that powered things need power. Power moves in loops, you you need two wires. It's perfectly possible to set up a power loop so it is hot-to-hot, in which case you really need to switch both wires. A great example is an air conditioning condenser in North America.
However that is more expensive, so neutral is a concept we established to make wiring easier. We set up neutral so it is near ground potential, so it's considered not hazardous. And so there is no need seen to switch it.
On an old fashioned light, the only thing that takes power is the light. So the switch is a simple mechanical human-powered device, and it only needs to interrupt (break) one of the wires. It's safer to interrupt the hot wire.
On a new fangled smart switch, the switch also takes power. So now you have two powered devices, and both of them need a full loop of both hot and neutral.
- The smart-switch needs always-hot and neutral so it can be always-ready.
- The lamp proper needs switched-hot and neutral.
Note that switches did not need neutral before, and suddenly now they do. That's a real problem for a lot of old in-house wiring. Sometimes you just can't install the smart switch.
There are also smart switches which do not require neutral because they use "the incandescent cheat", exploiting a fact about incandescent bulbs where they act like nearly a dead short until enough current flows to get them hot. The smart switch (on an incandescent light) can place itself in series with a turned-off bulb, and leak enough current to function itself. However the power supplies in CFLs and LEDs are designed as voltage devices not current devices, and respond to "the incandescent cheat" in strange and unpredictable ways.
There's one more concept that comes up a lot. Why doesn't neutral have a circuit breaker?
That concept entirely relies on the notion that hots and neutrals should be married monogamously. That is, every hot should have 1 partner neutral and use only that neutral to return current. The neutral is the same wire size as the hot. If this is done properly, the neutral cannot overload unless the hot is also overloaded in exact proportion. Since the hot has overcurrent protection, that will protect the neutral also. That's the concept.
Of course that does not account for people who don't respect that concept. I see circuits all the time where a device is wired with the intended hot, but any "neutral of convenience". The easiest way to detect this is by adding a GFCI/RCD to the circuit and putting a load on every outlet. The GFCI compares current flow on "hot" to "neutral" and trips if they are not equal.
Best Answer
So your proposed wiring scheme looks great when it's in PIR mode.
However, throw the switch to "other thing" mode, and what do you see?
The other problem with PIRs is you really want to power them 24x7. They also have light sensors, and they need 24 hours to see the maximum and minimum absolute lux in the location, so they can distinguish night from day in adverse conditions.
So I would flip it around. I would make sure that PIR supply (brown) always sees hot, and PIR neutral (blue) always sees neutral. I would run PIR switched-hot to one side of the switch, and always-hot to the other side of the switch. The pump goes to the common.
Now the PIR is always up, but it either has a load, or does not.