Straight edge my "!#*!!!!

[Andy Kev.]

I've just been planing the narrow edge of an approx. 18" long piece of American Black Walnut using a Veritas low angle jointer plane with a freshly sharpened blade. As a straight edge I used a 60 cm long aluminium ruler the long edges of which are about 1/8" and 3/32". After a few strokes with the plane I had the proverbial single shaving so I checked the edge by placing the 1/8" side of the ruler on the wood. This indicated two humps along the length. Plane, check, plane, check. Now down to one hump. Concentrate on that bit, plane until the single shaving is there. Check again. Still got a hump.

Something made me decide to check it with the other side of the ruler. The hump was now a dip! So I checked it with the second ruler (I have two so I can use them as a pair of winding sticks). One side showed that the wood was essentially flat (as expected) and the other showed a minimal dip but not in the same place as the first ruler.

Conclusion: the aluminium rulers are not straight along their edges. I find this quite astonishing as I just assumed that an aluminium ruler was going to be bang on. This means that every time I've checked an edge with these rulers, I've not been getting correct results. Is this the sort of thing which leads e.g. Veritas to go to the trouble of manufacturing guaranteed straight edges? I don't know whether to try to find some super stable tropical hardwood as recommended by Christopher Schwarz in The Anarchist's Tool Chest and have a bash at making my own or just cough up for the Veritas.

But to come back to the aluminium rulers: is it normal for tolerances to not be particulary tight in such things?

 

[Glynne]

I bought a steel straight edge (from Axminster I think) to use purely as a straight edge. It was a bit of a luxury as I thought until I tested all of my rules - and whilst some were OK, none were perfect. I also use their aluminium winding sticks as I now have a proper reference to check them with.
The experience then made me go for a surface plate to check flatness and surprisingly to me, a granite surface plate was cheaper than the straight edge!

 

[iNewbie]

Everything is made to a tolerance and you can factor in more on the price for one thats made to a better tolerance - a straight edge is exactly that,......and you can buy a SE with an accurate rule on it.


Starrett are know for their quality/accuracy.


.

 

[MIGNAL]

Probably depends on the ruler. They are rulers and not guaranteed straight edges.
Having said that my expensive Rabone Chesterman 1 mtr rule is extremely good against my guaranteed straight edge. In fact there is no need for me to use the straight edge.

 

[Cheshirechappie]

Sorry to hear of your troubles, Andy - hope you managed acceptable straightness eventually.

Sadly, it is true that most things sold ar rulers, even quite good quality ones, are not necessarily straight. I believe that Clifton, the planemakers, sometimes receive phone calls from irate customers telling them that the sole of their brand new Clifton plane isn't straight when compared to their steel rule, and having to find a diplomatic way of telling the customer that the plane is straighter than their rule.

There are two things you could do. The first is to buy a proper straightedge - which could be expensive; you'll want one that specifies a slightly better degree of straightness (so many thous per foot, or fractions of a mm per metre) than you want to work to.

The second thing you could do is make your own. Stable tropical hardwood is great if you happen to have some, but if you don't, don't worry too much about it - anything reasonably decent will do. Don't make one straightedge - make three. Here's why. If you make two, and compare them against each other, one might have a dip, and the other a hump, which may fit neatly together, fooling you that they're both straight. If you make a third, it may neatly fit the fault of the first, but it can't fit the fault of the second. The ONLY condition in which three straight edges can fit tight to each other all along, when compared A to B, A to C and B to C is when all three are straight.

You can use that trick to check things you already have - two winding sticks and one steel rule, for example. You can't correct the steel rule, but you can do something about the winding sticks, so that may give you a couple of 'good enough' straightedges that you can check every so often against each other, and correct as necessary.

 

[Phil Pascoe]

Andy - Not meaning to be sarcastic, but I would think that as there is a market for straight edges, the straightness of rulers is obviously questionable.

 

[bugbear]

Sorry to hear of your troubles, Andy - hope you managed acceptable straightness eventually.

Sadly, it is true that most things sold ar rulers, even quite good quality ones, are not necessarily straight. I believe that Clifton, the planemakers, sometimes receive phone calls from irate customers telling them that the sole of their brand new Clifton plane isn't straight when compared to their steel rule, and having to find a diplomatic way of telling the customer that the plane is straighter than their rule.

There are two things you could do. The first is to buy a proper straightedge - which could be expensive; you'll want one that specifies a slightly better degree of straightness (so many thous per foot, or fractions of a mm per metre) than you want to work to.

The second thing you could do is make your own. Stable tropical hardwood is great if you happen to have some, but if you don't, don't worry too much about it - anything reasonably decent will do. Don't make one straightedge - make three. Here's why. If you make two, and compare them against each other, one might have a dip, and the other a hump, which may fit neatly together, fooling you that they're both straight. If you make a third, it may neatly fit the fault of the first, but it can't fit the fault of the second. The ONLY condition in which three straight edges can fit tight to each other all along, when compared A to B, A to C and B to C is when all three are straight.

You can use that trick to check things you already have - two winding sticks and one steel rule, for example. You can't correct the steel rule, but you can do something about the winding sticks, so that may give you a couple of 'good enough' straightedges that you can check every so often against each other, and correct as necessary.

Or you could buy a granite surface plate. I'm not sure how the price of these works out compared
to the same tolerance straight edge.

Edit; googling shows prices varying massively. I'll keep looking.

But it's really easy (*) to correct a cheap straight edge against a surface plate. And I recommend a surface plate
as a reference when working a plane's sole.

BugBear

(*) alright, fairly easy

 

[Harbo]

I've no idea what the current prices are but I have a certified granite surface plate which is about 300mm square which I use mainly for checking metal items and an Axy 750mm steel straight edge for checking wood items.
A planer/thicknesser table or good quality cast iron top make good reference surfaces?

Rod

 

[Sheffield Tony]

I almost got caught out by this - a rough check of the sole of my Record 5 1/2 against a new teel rule from Sc*****x showed it to be miles out of flat - big gap in the middle. What to do ? After squinting down the sole and the ruler for a bit, I tried turning the ruler over. Now it appeared convex. The ruler is banana shaped - it looks like they are guillotined from a sheet of metal and warp in the process. Glad I didn't set about my plane with sandpaper before engaging brain !

The older imperial rulers in the tool cabinet were all substantially straight.

 

[bugbear]

A planer/thicknesser table or good quality cast iron top make good reference surfaces?

Key in the safe time - how would you know how flat they were?

BugBear

 

[bugbear]

I almost got caught out by this - a rough check of the sole of my Record 5 1/2 against a new teel rule from Sc*****x showed it to be miles out of flat - big gap in the middle. What to do ? After squinting down the sole and the ruler for a bit, I tried turning the ruler over. Now it appeared convex. The ruler is banana shaped - it looks like they are guillotined from a sheet of metal and warp in the process. Glad I didn't set about my plane with sandpaper before engaging brain !

The older imperial rulers in the tool cabinet were all substantially straight.

I also found (subsequent to buying a surface plate) that the 12" rule in my Moore & Wright combo
square was seriously hollow on both sides, I assume from wear.

It is now straight and parallel

BugBear

 

[Cheshirechappie]

I'd agree that a surface plate (with a calibration certificate) is a very good reference, but it's a heavy lump to find a home for, and it's something of an 'engineering' solution to a woodworking problem (though ideal for anybody wanting to refurbish plane soles to a very high standard of flatness). The only problem with 'ad-hoc' straight or flat surfaces is that they may LOOK flat or straight, but without careful checking you don't KNOW for sure that they're flat or straight - like Andy's aluminium rule.

The trio of home-made wooden straightedges can be made in the workshop with readily available woodworking tools, and checked against each other periodically, so they're a good woodworking solution to a woodworking problem. They won't be 'engineering' accurate, but they'll be plenty good enough for all but the most exacting woodworking.

 

[Jacob]

Straight edges are unnecessary most of the time. Check it by eye - if you can't see any errors then that is perfect. Though I'd use the edge of the plane as a straight edge, across the width.
Best value general purpose straight edge is a good quality long spirit level. These "precise" ones are for engineers, not woodworkers.

they may LOOK flat or straight, but without careful checking you don't KNOW for sure that they're flat or straight

if they look straight then they are straight.

If you buy calibrated surface plates etc you may find that absolutely nothing you own is straight. This way madness lies!

 

[Cheshirechappie]

Straight edges are unnecessary most of the time. Check it by eye - if you can't see any errors then that is perfect. Though I'd use the edge of the plane as a straight edge, across the width.
Best value general purpose straight edge is a good quality long spirit level. These "precise" ones are for engineers, not woodworkers.

they may LOOK flat or straight, but without careful checking you don't KNOW for sure that they're flat or straight

if they look straight then they are straight.

If you buy calibrated surface plates etc you may find that absolutely nothing you own is straight. This way madness lies!

Andy's aluminium rule looked straight - but it wasn't, and the result was grief....

PS - Absolutely NOTHING is perfectly straight or flat. What's needed is APPROPRIATE straightness or flatness. Eyeballing is plenty good enough for 'there or thereabouts' jobs, but you may want something better when jointing up board edges, for example. Holding a straightedge to an eight-foot board is easier than heaving it's mating board up onto it at times.

Horses for courses.

 

[Jacob]

Straight edges are unnecessary most of the time. Check it by eye - if you can't see any errors then that is perfect. Though I'd use the edge of the plane as a straight edge, across the width.
Best value general purpose straight edge is a good quality long spirit level. These "precise" ones are for engineers, not woodworkers.

they may LOOK flat or straight, but without careful checking you don't KNOW for sure that they're flat or straight

if they look straight then they are straight.

If you buy calibrated surface plates etc you may find that absolutely nothing you own is straight. This way madness lies!

Andy's aluminium rule looked straight - but it wasn't, and the result was grief....

Would have been no problem if he hadn't used it at all.

 

[Andy Kev.]

Thanks for all the replies, they do tend to confirm that I wasn't imagining the problem.

Cheshire Chappie: I really do appreciate that advice but I can imagine myself muttering,"ABC, BAC, CBA" as they cart me off to the loony bin in my strait jacket. Nonetheless, I might have a bash at that some time in the future.

phil.p: I appreciate you're not being sarcastic. It's obvious when you mention it. That'll teach me to make assumptions.

Bugbear and Harbo: are these granite surface plates something that you just keep there as the ultimate reference?

Jacob: I tend to agree with your idea about it being OK if it looks OK for most purposes. However, in this case I was taking a bit of trouble because I wanted the edge dead flat so that it produced as near as it could to 90° when the wood was stood on it. Hence my checking more than usual and thus making my alarming discovery.

 

[bugbear]

Bugbear and Harbo: are these granite surface plates something that you just keep there as the ultimate reference?

They're certainly not for everyday use - it's a sort of heirachy of accuracy. You check your wood against a straight edge,
but you check your straight edge against a surface plate.

But as Cheshire said, talking of "perfect" and "not perfect" is unhelpful. It's all a question of
how far from perfect is useful - be it 1/4", 1/16", 1/100" or 1/1000". How big an error
can the process you're doing tolerate before the process fails?

BugBear

 

[Phil Pascoe]

Does anyone actually know what the tolerance is on a granite worktop?

 

[bugbear]

Does anyone actually know what the tolerance is on a granite worktop?

(formally) unspecified and (empirically) very variable.

As a working support a granite worktop (sink cutout!) can be very useful (especially
if checked against a trusted straightedge for not being grossly banana shaped) but I wouldn't
use it as reference in its own right.

BugBear

 

[Phil Pascoe]

Thanks, that's much as I thought. I might try to get an offcut to try.