There's nothing like writing a big long thing like a dissertation to focus you in on your process for getting the thing done. My process has followed the same basic pattern, once I figured out what worked best for me. This is, I think, the hardest thing to teach people, that there is no one right way to go about writing. It's taken me a long time to get away from that idea myself. I would find out that some writer I respected wrote a certain way and decide, "Oh, that's how I should be doing it." When it didn't work out for me, I'd blame myself rather than realizing that that method just wasn't working for me.
So here's the method I've developed for myself. First, I must do the reading, thinking, notetaking part. I treat this just as I would writing and I do it every morning first thing. At some point, I will kick it into overdrive and start doing this in the evening and maybe do a marathon session on the weekend of working out what I might want to write. Then I begin to write. Usually, the reading and notetaking doesn't stop as I find I need to fill in gaps. I write for an hour every morning. I don't look back at what I wrote before. I used to do this and I found it paralyzing. Often, at the end of a writing session, I'll jot down a couple of notes about what comes next.
Eventually, these one-hour writing sessions end with a completed chapter, paper, whatever. Then I start reading and revising. I will either do this as a contintuation of the one-hour sessions or, preferably, take a chunk of time on the weekend to read and revise. Sometimes I'll print out my document. Sometimes I'll get Mr. Geeky to read it. It depends on how I'm feeling about it. Then I send it off. This whole process generally takes 1-2 months, depending on the complexity of the project.
Between chapters or projects, I find I need about a week of downtime. I might do a little library research to find new sources for the next chapter, but I don't do much reading and writing. And then I gear back up again.
It's been interesting to see this pattern develop over the last year. I used to worry about that week off, for example, but then I realized I just need to rest and that I would get started again. It'll be interesting to see if this pattern continues to work for me, or if I find a new pattern or different patterns for different projects. We'll see.