Hey, this post is like 5 years old at this point in time.  

Some additional items have come to the surface so I will share them here and let you ignore the picures and such past this point that show different things.

First, hose size and filter nipple IDs -- stay with 3/8 and only go as small as 5/16th at the very smallest.  Why?  3/8" is good because the hose is big enough that in the bends it never goes below 5/16ths effective inner diameter.  The ID of the 3/8" filter nipples never restrict anything to less than the stock petcock and carburetor nipples already do.   Your normal small screw type hose clamps can easily make 3/8" hose grip the petcock and carburetor nipples so that the rest of the system is known NOT TO BE THE FLOW RESTRICTION.

Next, the orginal filter style show in the original pictures isn't readily available any more.   Your clear sided filters are either FRAM G2 or FRAM G3 filters or else they are the oriental rip-offs of that design type.  Note that these FRAM types have a lot more pleats (more square inches) of filter paper than the original filters had.

    20 pleats of paper on a side (40 total)
FRAM

   14 pleats of paper on a side (28 total)
oriental rip off of a FRAM

There are some filters out there with only 7 pleats per side (14 total) so count your pleats before you buy the filter.

If you go with the oriental rip off "FRAM" filter, please do a flow through test using a syphon hose on the intake side maintaining only 2-3 inches of vertical drop (gas surface to filter inlet)  in the suction system.   Some oriental filter paper can be very restrictive, so you are looking for syphon amounts in the order of a cup a minute or better once you get the system flowing at full rate at the designated 2-3 inches of drop.   Most clear in line filters of this size did a lot better than this minimum cup a minute rate in Charon's testing.

Or, you can do a in-place test on the bike with the hose to the carburetor detached and aimed at the catch cup.  You will need to do this when you are well down in the reserve portion of the tank.  The pass point is a cup full of gas inside 60 seconds.

(a little bit more of a functional test when done on the bike, I think, although you will have to separate out your vac petcock as a potential failure cause if you chose to do it this way and it fails).

=====================================================



(note that extra quick clamps were purchased to go over all fuel connections for vibrational security)

As you view this picture you will see the stock Suzuki fuel line tubing measures 0.2675" on the inside diameter.  This has led some to purchase quarter inch I.D. (0.250") (BAD IDEA) fuel hose, including the example shown.  This hose is a tight fit to all the nipples used in this system and is not "easily installed" on the metric carburetor nipple and thus the 0.250" hose is a BAD IDEA so please don't do that to yourself.

Others have purchased 5/16 I.D. (0.3125") fuel hose (BETTER IDEA) and have found they got acceptable grip on the nipple fittings and no fuel leakage.  You may need to purchase your clear filter first and then consider fitting the hose to the nipples you have on the clear filter, with the slack being taken up at the slightly larger brass carburetor nipple barb.  You need to replace the stock Suzuki hose with your new hose to make the run to the clear filter, then use the new hose to go from filter exit to carburetor nipple.  Note the use of a long broad loop to make the return run to the carburetor -- this way the hose is never temped to pinch itself shut due to a "too tight" bend anywhere in the curved return run.

Clear fuel filters of the type shown in the pictures can be purchased at Wal-Mart in the automotive section (Fram has two clear fuel filters currently hanging up on the wall in the Wal-Mart auto fuel filter section).    Autozone no longer sells the exact same looking clear fuel filter shown in the first picture in this thread.

3/8" filter nipples and 3/8" diameter rubber hose is the BEST IDEA because it absolutely guarantees the stock nipples on petcock and carburetor are definitely the "as designed in" restriction points in the fuel delivery system.  

Use of effective hose clamps is required at the petcock nipple and the carburetor nipple when using the larger 3/8" ID rubber hose.


=============================

"Small filters" referred to by some folks as their preferred solutions generally come from Briggs and Stratton type fuel filters (lawnmower section) and from dedicated motorcycle shops.  Watch your fuel line sizes as most of these are quarter inch nipple sizes and may require quarter inch fuel hose.  Watch out though, as some are 3/8" nipple sizes and won't work for you without downsizing a hose somewhere.

The small fuel filters either have a metal or a nylon mesh screen as their active element.  This is the exact same thing that exists on your petcock, so don't expect to see any finer filtration or water removal.

True micron level filtration requires PAPER PLEATS and lots of them, so look for clear plastic filters in the auto area that have some decent size to them.

=============================

Discussion of the rusting of the insides of Savage gas tanks and the need for the larger clear sided 10 micron filter paper fuel filters seems to center around the increasing use of alcohol in cheaper grades of regular gas.  This inclusion of alcohol used to be unusual and used to be always stated at the pump, but now this is no longer a requirement in some states.  Alcohol is accepted as part of the anti-knock additive packages that may be used in standard grades of regular gas, especially at the low cost distribution points.

What is new and potentially frightening to you and your bike is a creeping new technology that is also part of the "anti-knock additive package" classification as it rides right along with the use of alcohol that is used to carry it into the gasoline mixture.  Note this is not a new patent, and it makes reference to even older existing patents in the fuel industry.  

(you need to scroll down several pages to get to the actual percentage of H2O that this technology will support -- yes, that was up to 22% -- you did read it correctly)

http://www.freepatentsonline.com/4158551.html

And now you know why cheap gas can seem a bit cloudy when you put it in a clear sided container.  

And also you know why it may be a good idea to put a sizeable clear 10 micron pleated paper filter on your fuel line as the growing intentional inclusion of alcohol as a fuel additive has potential downside consequences for your 1980's designed carburetor system (and your 1980's old style uncoated steel fuel tank).

While some automotive filters may incorporate a water trap, or have one in series with a paper element, I don't think the paper itself can sift out water molecules as you imply.  Water molecules are on the order of angstroms (a few ten thousandths of a micron) and a good paper filter only filters things larger than about 10 microns.  I doubt the Autozone cheapys are even that good. So unless there's some other voodoo going on with ions or hydrogen bonds or something, paper is not going to hold back H2O for you.

Sintered brass filters typically provide 35-40 micron filtration.  That's better than a petcock screen and good enough for a carbureted motorbike.

That makes sense Oldfeller, I had not thought that clusters of water molecules, in solution, might also have their own surface tension, sans water/air interface.  And it is those microscopic droplets that cannot pass through the filter until alcohol comes along to grease the skids.. interesting.  

So moisture will not hang around in the filter thanks to the gasohol that we're stuck with these days. The engine has to pass X amount of water that is in the gas.  Is it better to have a paper element in case of a watery non-gasohol tankful, or do you think the filter ought to just pass it?  I think the paper one might be better in that case, so's to avoid passing a sudden "slug" of water that settled to the bottom while the bike was sitting.  But then if that same slug was sitting in the filter, it seems that would stop up the fuel delivery also.

Yep, a quarter inch of water in your gas is probably better to have in your filter than your carb bowl.    1) because if the filter is clear, you can see it  and 2) that amount can sit there until you get some gasohol to dissolve it, whereas in the carb bowl it'd probably be too much for your engine to start.  

I'm shocked and apalled there's deliberate water in gas?! Bastards.

Add the clear filter guys. You want proof on other vehicles? My 69 Dodge runs the large clear Fram. I change it once a month. The water and alcohol in modern fuels speeds the corrosion of gasoline storage tanks. I ALWAYS have "sand" in the filter. Or "chalk" dust. Or pieces of dissolved fuel line. My 49 Dodge has a fuel filter sediment bowl at the carb tocatch whatever makes it past the fuel pump. Which also has a sediment bowl. Both are metal mesh screened stock. Both converted to paper replacement elements. Had the 49 running 17 years ago when I first started seeing E10. Bowls were always full of crud. With the sediment screen too much debris got passed into the carb. It is now a paperweight. Literally, the crud E10 forms blocked every passage and hardened solid. And carb cleaner only gets so much of it out.

Here's what mine looks like right now.


Another example of a correct installation.

Does anyone have any bad experiences with the fram g3 filter, which I believe has 3/8 inlet and outlet? How often do you guys replace them and can I do the same routing as the fellow with the ryca setup using 3/8 fuel line?