Bernard,
The information above is good, but there is another way to approach this that might be helpful.
Always start with low pressure - just enough to begin to compact the curds together, let excess whey escape. After 30 minutes, increase the pressure - just to the point where whey is seeping out gradually. After 30 minutes, increase again, again looking for whey just seeping out. By that point, you will have a good idea whether or not the curds are going to take just a small amount to knit, or whether you're really going to have to ramp it up. The further along you get, the less you will see much of any whey coming out.
As soon as possible, start flipping the cheese each time you increase the pressure - I say it that way because, with something like a cheddar, it may still be way too crumbly to handle it the first time or even two.
A couple of other points: temperature of the curd makes a difference; warm curd knits more easily than cold curd.
And super-important: cheeses that go into the press before salting (e.g., tomme, gouda, asiago, parmesan, etc.) need to be monitored carefully so that they go into the brine (or start dry salting) at the proper pH. For most cheeses, that means around 5.3. In my experience, if I follow the typical recipe to let a Gouda stay in the press overnight, I don't wind up with Gouda; I wind up with something that has gotten far too acidic, and as a result the paste is crumbly rather than flexible, and the taste is sharper than it should be. Ideally, you control the pH curve so that the cheese has time to knit well by the time it reaches the target pH.
If you want the ultimate work-out, make a Cantal - the curd is pressed, unsalted, for 24 hours, allowing it to bottom out on pH; then it is milled and salted, and then it is pressed again - now cold and low pH. Takes a huge amount of psi to get a decent knit!