I sympathise!
I have various methods, and I chop and change depending on what feels like it's helping.
In the last week or so, my method has been to write a list of 10 or so tasks that I need to get done today, major and minor. I don't think about all the other stuff that I'll have to do (after I've established that everything else can wait a day).
Then I work my way through the list. I include minor (sometimes non-work) tasks to balance things out. For example, I'll list things that will take me hours, and things that will take me only a few minutes. This makes the list seem more manageable.
This is in addition to my standard workload document, which includes:
a) what projects are active, hours remaining and deadlines;
b) what projects are coming up (again including hours and deadlines); and
c) a basic calendar noting what hours I'll be spending on what projects each day over the next few weeks.
That - in particular the calendar, which shows me what I should be expecting in the weeks to come - makes me feel more secure that I won't fall behind with anything.
Also, when I have a lot of projects on, sometimes I do a bit of work on several each day (e.g. 2 hours each on 4 projects, rather than 8 hours on 1). That reassures me that nothing is going to fall desperately behind.
Lastly, and maybe most importantly, remember that you
can say no. I'm still having to remind myself of that a lot! It's great to have a diverse client list, but it's all right to turn something down if you've already got more than enough on your plate. Particularly if you are well established with them, as they are more likely to come back with another project (I just try not to say no twice in a row to a client).
Hope that helps!