I've been studying those ridiculous onomatopoeic words like ぱさぱさ,　がさがさ, ぎりぎり, etc (is there are a term for these?). There are a lot of them. I think it's best to mix them in gradually with the rest of my kanji-based vocab instead of trying to take them all on at once. But I've found success (for now, anyway) with assigning heisig-style mnemonics to the hiragana, then making an image for the word.
し=shibuya
ば=bar
ら=ramen

しばしば means ”frequently” because I frequently go to shibuya bars.
ばらばら means ”separate” because the place where they sell ramen is separate from the bar.

Anyone else do anything like this? Does it work with other hiragana words? I'm going to need something when I start on those words with the small つ in them: そっきり、さっぱり、がっしり。。。* Is there a "standard" set of mnemonics for the hiragana？ It might be useful to use the same ones as Heisig in Remembering the Hiragana.

I totally made up さっばり　and がっしり and it turns out they're real.

[Katsuo]

I use these kind of mnemonics quite often. E.g. I memorized がっしり by imagining a moth (が) with a solidly-built rear end (しり) doing sumo. That ignores the "っ" of course, but I find such approximations are usually enough to trigger the memory.
Here's links to a couple of recent related threads:
JLPT Giongo / Gitaigo
Heisig's RTK2 Kunyomi Mnemonic Chart