having a brain fart here. help me with this excerpt from his wheel building page on an improvised wheel dishing tool:

"... You may also check dishing by placing two piles of blocks ... on a tabletop to support the rim at two places opposite one another. The distance from the tabletop to the hub locknut on the underside should be the same, no matter which side of the wheel is down."

so let's say there's more space between drive-side locknut and table compared to the nds. which side do i tighten?

if the difference is 4mm (meaning an error of 2mm), do you try to true this slight error away? how much does it normally take to move a rim 2mm to one side? a half turn on each spoke on that side? what's a reasonable length in dishing error to leave in place?

work in half turns, do the entire wheel and watch closely what you're doing. Imagine pulling the rim over the hub: tightening brings the rim more to the side you're working on, which is the same as pushing the locknut away from the surface you use.

BTW, I made my McGiver dishing tool with tumblers and a stack of 5c coins. I don't know what that says about me

so, in my example (more space between ds locknut and table), i should tighten the nds?

crafty. and you find spare change worth more outside its face value.

THere is no substitute for using the correct tools. When redishing a wheel, I loosen spokes on one side of the wheel an tighten the other side exactly the same. Half turn, quarter turn, full turn, what ever is needed. Once it's centered then I retension the wheel.

Personally, I'd do no more than 1/4 turn at a time. It's too easy to find you've overshot things by the time you get all the way around the wheel. It seems to keep other aspects of true/tension more consistent.
I find a slow, methodical approach gets me to the end of the journey faster.

NDS usually will have a greater effect due to the angle of the spokes.
Unless the DS is already at very high tension, I don't loosen it.

Tighten the side with more space, loosen the side with less space.
edit: Doh! no, as jab points out (see below) I got that wrong. Tighten the side with less space.

Yes, unless I was feeling very lazy (which is not unprecedented). Bear in mind that you may be able to correct the error by moving a 2 mm washer from one side of the axle to the other.
surprisingly little, but I don't remember.
That's a judgment call I'm going to leave in your hands. There comes a point in a long ride when I just want to sit up straight and stretch my arms out to both sides and try to relax my neck muscles for a little while, while let the bike steer itself. At this point, I like to see the front wheel perfectly centered in front of me. If I see more of one side of the wheel than the other, and have to compensate by putting a kink in my spine, I can't relax my neck; and I know i didn't dish the wheels properly.

yeah, things can go fast. I got mine centered to within a mm by working first with half turns, then quarters, and 1/8 on the final go. I used spark plug adjusters on the final run to give my ghetto rig at least the air of precision

There are various methods of checking dish. My favorite way is to reverse the wheel. Doesn't matter if it is on the bike or in a truing stand. I think tension has an impact on how many turns it takes to move it, maybe not. I started truing again after 30+ years of not doing it. Slow and easy is the trick, as mentioned. Being a bit of an engineer, I invested in a Park tension guage and found a Spin Doctor truing stand for $20. I don't trust my ear for making sure the spokes ring the same frequency to determine tension so the the tension guage works well for me.

It is intersting to see how far off spoke tension is even when a wheel is said to be true. It is from a rotational perspective but one spoke, or two can carry the load of a third one which is not tight enough.

The other trick in this kind of effort is to make sure the spokes seat. Some nipples that are rotated 1/8 or 1/4 turn are not set. You find out when you mount the wheel and there is popping for several rotations. Now the true may change. this can be avoided or minimalized by holding the rim in your hands at the 10:30 and 1:30 positions with the bottom of the wheel in your lap and use your elbows to push on the rim (Taco forces). Rotate the wheel and do it again until all the spokes seat. Check true again and adjust as needed. Repeat. This saves mounting the tire and wheel on the bike and then riding. Can you true with the tire in place? Sure, Let all the air out so the air presure doesn't add a variable.

I disagree with rhm, tighten the side with less space (nds) as you will pull the rim towards that side and increase the space. In this case, the lock nut is moving relative to the rim.

4mm is unacceptable to me.

Depends on your tension and what side you are moving. nds with less tension will move more per rotation than the ds with more tension.

Depends. You can tension only one side to center the rim but check to see that the tension isn't too high on either side when you are finished.

Less than 1mm for me.

that's what i was thinking, iab, just wanted to make sure. so when you tighten a side, you increase the space between that side's locknut and the table.

helpful remarks, all.

A real dishing tool costs all of $30. Seems like a real time saver to me, rather than blocks on a table and some jury-rigged system.

I have a dishing tool, but have found that I don't need it as my truing stand gauges both sides of the wheel at once, so (as long as I make sure that's calibrated), the wheel is always dished by the time it's true. But yeah, loosen the side with more space between the axle and the table, and tighten the side with less room between the axle and the table to move the axle to the middle.

Fwiw a dishing gauge is a pretty simple tool to make from a bit of scrap wood and about a dollar's worth of hardware.