Please explain tongue/groove technique...

[Rick Cicciarelli]

So watching NYW the last couple of weeks I've seen Norm use the same technique for making sure the groove is centered....run it through the table saw once in one direction, and flip the board around and run it through in the other direction...perfectly centered groove. Now I understand how that works, but what I seem to be missing is the idea that if I want an exactly 1/4" groove...if I don't have the fence set up perfectly in the first place, then when I go and run the board through twice like the approach mentioned above, I have the potential for actually ending up with a larger groove than I want don't I?? Sure it will be centered, but if the measuring isn't set up perfectly for one run through, when I run it through that second time it will shave off just a little bit more from that other edge of a groove resulting in a wider than 1/4" groove. Or am I missing something?

[Mike Cutler]

Nope. You're not missing anything.
Having not seen Norm do this, I will speculate that he is doing a T&G joint on a tablesaw for a panel glueup. In this instance the width of the groove is not as important as the fact that it is as close to centered as possible for alignment.
I would imaging that the tongue was cut in similar fashion and then finished with a shoulder plane.
I'm willing to bet that Norm's skill with a TS is good enough that you would need a vernier to measure the deviation from centered, and a nominal 1/4" width though.

If the width of the groove is the critical dimension, a quality T&G router bit set would be an easier and quicker alternative.

[frank shic]

i'd rather set up either a 1/4" slot cutter or a tongue/groove cutter like mike mentioned on a router table or a 1/4" dado stack on the table saw instead of fiddling around with trying to get the setting just right with a 1/8" table saw blade. there's less chance of error if you make just one pass instead of two. i don't use tongue and groove for panel glue-ups either.

[Rick Cicciarelli]

Yeah, I've seen him use this same technique for a few things, but last night I was watching him make a cd drawer/case. The sides were made up of panels with rails and stiles with a 1/4" panel in the center. If I recall, he used a stacked dado for the 1/4" groove using the method I described above.

[Richard Niemiec]

I have in the past and just yesterday used the technique you mention to groove some rails and stiles to make some flat panel doors. From my experience, assuming a 1/8 kerf blade, 3/4 stock and a 1/4 slot, set your fence to 1/4, set depth of blade to your liking, and run the stock both ways on some scrap and then get out your ruler and check the results before you run the good stuff through. Now if you have a so-so fence this may be an issue, but with a bies or in my case a unifence, it works like a charm for me.

[Tim Sproul]

Or am I missing something?

Why does the groove have to be dead nuts at 1/4 inch wide?

Make the grooves. If the groove is .260 inches or .240 inches doesn't matter. Make all the grooves first.

Now cut the tongues to fit the groove.

When milling tongue and grooves on longer stock, I find using a router more accurate as it rides any bit of warp and keeps the groove and tongue aligned with the reference face. When using a table saw or shaper or router table with long stock, I find I need a lot of pressure from hold downs and featherboards to keep the work tight against the fence and table as I feed it. I have yet to mill a 6 or 8 or 10 foot long piece of stock and have it come out perfectly 4-square - hence the need for featherboards and such when using table and fence based machining.

Routing a lot of tongue and groove will let you know if you purchased a quality router or not.