I'm in the same situation. I'm very close to launching
http://pitcher.io ... with all the usual SaaS things - 3 tiers, free trial, maybe a free account level. I have thought long and hard about free accounts though. Here's roughly where I'm at:
1. Are free accounts actually worth having? If your service offers enough value to the users that they'll invest time and energy using it then why won't they also invent a token amount of money? Even if it's just $10/year, that's going to be worthwhile for you if there are 5,000 accounts at that level.
2. Will a fee really put people off signing up? Putting credit card in to the sign up process definitely will, so don't do that. But people will try software even knowing it's going to cost them eventually if the pain you're taking away is great enough. If you can afford to give them a free account, just give them a long term free trial instead. "3 months free" is effectively the same as a free account to someone trying a new app.
3. Remember the popular saying "If you're not paying, you're the product." that people love to quote on HN when they're admonishing Facebook. It has an element of truth to it. If you're planning to make money from users by means other than direct payment, be upfront and honest about it. Tell them that they'll see adverts unless they hand over money. Tell them that you're mining their data. Tell them that you've invented a revolutionary new business model that makes money out of nothing. And tell me. :)
4. Consider that if you don't have a free tier you won't need to subsidise your free users, so you might be able to lower prices. But also consider that the price you start with will be the highest you ever charge. Users don't like increases in prices. But you can always grandfather old price points in.
5. Ultimately, until you launch, it's all speculation really. Just get it out there and get data to make proper decisions with.
I have no tips about money collection because I'm in the UK.
Actually I do: "Don't use Paypal."