Does anyone know what Stabila specs the flatness of the face of their levels to be? I need to get myself a long straightedge for calibrating jointer tables and was thinking a 6' Stabila would might be a 'good enough' alternative to a real 6' steel straigtedge for a fraction of the price.
Stabila Flatness
- Thread starter Aegwyn11
- Start date
- Joined
- Nov 3, 2007
- Messages
- 5,133
That's a really good question...
The Stabila levels that are longer than 4' aren't machined so they are going to have different tolerances than those up to 48".
I tried a couple of minutes ago to get the spec's but the guy I need to talk to was on the phone. I'll try again in a little while...
Tom
The Stabila levels that are longer than 4' aren't machined so they are going to have different tolerances than those up to 48".
I tried a couple of minutes ago to get the spec's but the guy I need to talk to was on the phone. I'll try again in a little while...
Tom
If you have an indicator you could do a better job. You would need to fabricate a sled with an arm on it about the length of one wing from which you could suspend the indicator. Think of it as a surface gauge with a 3 foot arm on it. You could probably get decent info with one made of MDF as long as it sat flat on the table without rocking, or you could put three brass screws in the bottom to act as a sliding tripod. Place the screws about 5 or six inches apart. As you slid from the far left towards the spindle you could be reading the parallelism of the other wing on your indicator. You could also slide front to back to check for skew. After that, it is all about the results you get on your cuts. I think Roger Salvaterri said it best, 'Good set-up hygiene dictates checking the results of your set-up.'
I tried to contact Stabila myself too...they've got to ask the engineers in Germany and said it might be a day or two to get me a response.
Greg's idea is actually an interesting one...a guy on the other FOG (Felder) has a dial indicator mounted on the end of a 6' Johnson level for this purpose. I want to understand how straight these levels actually are before I decide to go that route. Veritas makes a 50" alu straight edge that is like $80, which is almost good enough, but I'd really like a 6'. As for using a piece of MDF or wood, I'm not sure that I have the equipment necessary to ensure that the edge is straight and consistent enough for what I'm using this for. Felder uses a real 6' precision straight edge for commissioning machines, I'm just not sure that I can justify $500 for this...
Greg's idea is actually an interesting one...a guy on the other FOG (Felder) has a dial indicator mounted on the end of a 6' Johnson level for this purpose. I want to understand how straight these levels actually are before I decide to go that route. Veritas makes a 50" alu straight edge that is like $80, which is almost good enough, but I'd really like a 6'. As for using a piece of MDF or wood, I'm not sure that I have the equipment necessary to ensure that the edge is straight and consistent enough for what I'm using this for. Felder uses a real 6' precision straight edge for commissioning machines, I'm just not sure that I can justify $500 for this...
I already got the answer back...pretty quick.
They are spec'd to 0.2mm/m. So for a 6 ft. level, it would be 0.37mm, or .014 inch.
You could compare this spec to the Veritas alu straight edges..they are .003" over the length. Stabila would be .010" over 50"
They are spec'd to 0.2mm/m. So for a 6 ft. level, it would be 0.37mm, or .014 inch.
You could compare this spec to the Veritas alu straight edges..they are .003" over the length. Stabila would be .010" over 50"
- Joined
- Nov 3, 2007
- Messages
- 5,133
Nick:
Was the spec they gave you for the machined (4' and under) surface or the non-machined surface? I would have to think that the machined surface has considerably tighter tolerances.
Tom
Was the spec they gave you for the machined (4' and under) surface or the non-machined surface? I would have to think that the machined surface has considerably tighter tolerances.
Tom
- Joined
- Nov 3, 2007
- Messages
- 5,133
The Type 196 levels' vials are guaranteed accurate to .5mm/m as long as you don't bend the frame (it takes a lot to bend those frames).
I just spoke to Paul. He said that the spec' he gave you is for all lengths of levels. I still can't help but think that the tolerances are going to be better for a machined surface (4' and under, Type 196).
Tom
I just spoke to Paul. He said that the spec' he gave you is for all lengths of levels. I still can't help but think that the tolerances are going to be better for a machined surface (4' and under, Type 196).
Tom
Interesting...so the face flatness is more stringent than the level accuracy (which you would expect).
Its interesting that the machined levels don't have tighter tolerances. Right now I'm torn between the 6' level and just getting the 50" Veritas alu straightedge. I wish I could just get the Starrett 72", but I just can't justify $450 for a straightedge that will --rarely-- get used.
Its interesting that the machined levels don't have tighter tolerances. Right now I'm torn between the 6' level and just getting the 50" Veritas alu straightedge. I wish I could just get the Starrett 72", but I just can't justify $450 for a straightedge that will --rarely-- get used.
- Joined
- Nov 3, 2007
- Messages
- 5,133
Nick:
The level accuracy is an additive accuracy of the flatness plus how accurately they can clock the vials.
I spoke with Paul and he said the spec's he gave you were for all of their levels. I'd bet you that the machined faces have significantly tighter tolerances. The engineer(s) for those levels are in Germany...
If the 50" Veritas straight edge will work, would a machined face Stabila 120th anniversary, black 48" Type 196?
Tom
The level accuracy is an additive accuracy of the flatness plus how accurately they can clock the vials.
I spoke with Paul and he said the spec's he gave you were for all of their levels. I'd bet you that the machined faces have significantly tighter tolerances. The engineer(s) for those levels are in Germany...
If the 50" Veritas straight edge will work, would a machined face Stabila 120th anniversary, black 48" Type 196?
Tom
Nick,
I don't think I have explained myself well enough. The MDF arm would not ride on the bed. It would just be an arm cantilevered to the right from a small sled, maybe 6" by 6" with three brass screws in the bottom to create a three point tripod you could slide along the bed left to right. The three screws would create a register on your bed that could not rock. If you start out with the sled at the far left of the out-feed bed and zeroed the indicator as it was attached to the MDF arm on the left table just to the left of the cutter, you could then slide to the right and watch the indicator for movement. If the tables are parallel the indicator will stay at zero. An OOP condition will be quantified and the direction will be defined by how much and by which way the needle moves from zero. With a little work, and by going in the opposite direction you could also tell whether your beds are flat and if it is both or only one. This takes some thought but has the benefit of not relying on something you think is flat but don't no for sure. Even a Starret could be OOF. If you have a quality jointer that hasn't been beat up I would guess the tables are within a few thou.
The real test is how well does the machine joint and how well it works for you. I don't believe you have stated whether you even have a problem getting good results. How well does it work for you now?
I don't think I have explained myself well enough. The MDF arm would not ride on the bed. It would just be an arm cantilevered to the right from a small sled, maybe 6" by 6" with three brass screws in the bottom to create a three point tripod you could slide along the bed left to right. The three screws would create a register on your bed that could not rock. If you start out with the sled at the far left of the out-feed bed and zeroed the indicator as it was attached to the MDF arm on the left table just to the left of the cutter, you could then slide to the right and watch the indicator for movement. If the tables are parallel the indicator will stay at zero. An OOP condition will be quantified and the direction will be defined by how much and by which way the needle moves from zero. With a little work, and by going in the opposite direction you could also tell whether your beds are flat and if it is both or only one. This takes some thought but has the benefit of not relying on something you think is flat but don't no for sure. Even a Starret could be OOF. If you have a quality jointer that hasn't been beat up I would guess the tables are within a few thou.
The real test is how well does the machine joint and how well it works for you. I don't believe you have stated whether you even have a problem getting good results. How well does it work for you now?
- Joined
- Nov 3, 2007
- Messages
- 5,133
That's an interesting jig, Greg. It sounds like you've put some thought into QC procedures in a machine shop.
I'm not sure it would just test for parallelism though. I think any OOF condition would also effect the results.
Tom
I'm not sure it would just test for parallelism though. I think any OOF condition would also effect the results.
Tom
In the end, remember that you are machining wood. It will react as soon as it gets off the tooling. How much it reacts has more to do with the tendancies of that particulare piece of stock than it does with the accuracy of the tool (jointer in this case); within a reasonable parameter.
greg mann said:
I'm learning a lot...the boys over on the Felder FOG have gotten me a long ways there. If I'm understanding it right, what you're describing is the part of the setup to ensure the edges of the two beds closest to the cutterhead are parallel to each other. For this part, the guys on the FOG all suggest using a Oneway (which I already have). I don't really see how that can help you determine if the beds are coplanar with each other, though. Thats where a long straight edge (or even better, long straight edge with a dial indicator on the end) comes in. You register the straight edge on the outfeed table and start with the indicator at the cutterblock side of the infeed, then move towards the far end of the infeed table and you can see if the infeed table slopes into or out of the cutterblock. You can see the need for your straightedge to be a good bit longer than your infeed table for this test to work well (otherwise you can't keep it registered on the outfeed table).
On my particular machine, the outfeed table is about 47" long and the infeed slightly shorter. I didn't realize they were quite this big until I measured last night. As such, the thought of the 50" Veritas or even the 48" levels is pointless. I'm not going to buy a 72" Starrett for this, so I'm back to levels. Some of the guys on the other FOG suggest a Johnson 6' level as it has machined surfaces, but I'm waiting on an email back from Johnson on their face flatness spec (who cares if the surfaces are machined if they are spec'd worse than the Stabila un-machined surfaces). Depending on their answer, I'm thinking I'll either go get me a Mastercraft 6' (green Stabila with a slightly cheaper price tag), or get me a 6' Johnson, then figure out a way to clamp a dial indicator on the end of whichever.
I understand that wood will react and move to a certain degree, but I also believe that your tooling should be much more accurately set up than whatever the level of accuracy you expect to get when actually using it. Plus remember that I'm a newbie...I've learned a lot about my tools so far by going through similar highly accurate setups (unisaw and SCMS). If I KNOW that my tools are very well set up and something doesn't fit or work right, I can focus on what I'm doing wrong instead of wondering about my tooling.
Oh, and cuts are NOT satisfactory. I got this machine used (very gently used
) and the cut quality was mediocre at best. I decide to replace the blades since I don't know what the original ones have been through, and by getting into that, I started finding other significant errors (outfeed table not parallel to the cutterhead, cutterhead too low, knives too low). There's enough issue that I decided I should just go thought a FULL calibration like I did on my unisaw and SCMS.richard.selwyn
Member
- Joined
- Jan 23, 2007
- Messages
- 635
Aegwyn11 said:
Nice to see another person completely in the thick of the FOG!
(I had my machines set up by Felder and haven't dared touch them since.)
I had a client in the shop yesterday and he saw so much Festool stuff, he thought that the Felder was a Festool as well!
He was so impressed he was on the internet today trying to buy a Festool drill. I am afraid he was pt off by the price (I didn't dare tell him how much anything cost.
Richard
Michael Kellough
Member
- Joined
- Jan 23, 2007
- Messages
- 7,101
Nick, there are two ways the beds can be out of parallel.
One is the you're thinking of (and why you want a long straight edge).
The other way is that the beds can be twisted relative to one another.
If they are twisted you don't need to bother with the first possibility.
The suggestion to test the ends of the beds close to the cutterhead
is the way to test for the twist problem. If the machine passes that test
you can apply the testing principle to the other possibility.
I think that is what Greg was describing.
You can mount a dial indicator on a base and cantilever it past the cutter head
so it touches the other bed. Slide it form one side of the bed to the other and check
to see if there is a significant difference. If there is (you and the manufacturer
have to agree on what is significant) then the beds are twisted and someone
has to adjust/shim the gibs to straighten the beds out.
If the beds aren't twisted you can move on to the overall straightness test
and you can use a similar technique as above. You just have to get creative
with the mount for the indicator. A simple way to go is to clamp one level to
the top of another level so that the upper one cantilevers out about two feet.
It doesn't matter if the levels are straight (in fact the upper arm could be a
crooked stick as long as it is fairly stiff) but it would be better if you made a
tripod rig for the lower level to sit on just in case it isn't perfectly straight.
Then, with the dial indicator mounted to the end of the upper level such that
it touches the other bed near the cutter head just slide the rig towards the cutter
head so that the indicator travels down the length of the other bed. If the beds are
co-planar there won't be a significant movement of the needle. Assuming you've
already determined that each bed is satisfactorily flat, you really don't need to
travel the full length of the bed.
And remember, if you've using the jointer to make wood ready for gluing up into
wider assemblies you don't necessarily want the boards to be dead straight.
I find it better if two assembled boards are a tiny bit concave, which allows
me to concentrate the clamps in the middle of the assembly.
One is the you're thinking of (and why you want a long straight edge).
The other way is that the beds can be twisted relative to one another.
If they are twisted you don't need to bother with the first possibility.
The suggestion to test the ends of the beds close to the cutterhead
is the way to test for the twist problem. If the machine passes that test
you can apply the testing principle to the other possibility.
I think that is what Greg was describing.
You can mount a dial indicator on a base and cantilever it past the cutter head
so it touches the other bed. Slide it form one side of the bed to the other and check
to see if there is a significant difference. If there is (you and the manufacturer
have to agree on what is significant) then the beds are twisted and someone
has to adjust/shim the gibs to straighten the beds out.
If the beds aren't twisted you can move on to the overall straightness test
and you can use a similar technique as above. You just have to get creative
with the mount for the indicator. A simple way to go is to clamp one level to
the top of another level so that the upper one cantilevers out about two feet.
It doesn't matter if the levels are straight (in fact the upper arm could be a
crooked stick as long as it is fairly stiff) but it would be better if you made a
tripod rig for the lower level to sit on just in case it isn't perfectly straight.
Then, with the dial indicator mounted to the end of the upper level such that
it touches the other bed near the cutter head just slide the rig towards the cutter
head so that the indicator travels down the length of the other bed. If the beds are
co-planar there won't be a significant movement of the needle. Assuming you've
already determined that each bed is satisfactorily flat, you really don't need to
travel the full length of the bed.
And remember, if you've using the jointer to make wood ready for gluing up into
wider assemblies you don't necessarily want the boards to be dead straight.
I find it better if two assembled boards are a tiny bit concave, which allows
me to concentrate the clamps in the middle of the assembly.
i hate to say this, but i think there is a lack of clarity here. Flatness, straightness,are not the same thing, at all. you need datums to reference. i would love a pick of your sled,i am not getting it, could be me. i am an ME and deal with these things all the time, precision is a tough mistress and like a woman, is not cheap no matter what they say. there are ways of creating geometry references that might help. i sympathize with the cost of a strait edge but but do not think of the few times you use it but as the accuracy that it brings to the rest of your shop. there is nothing like bright light of precision once it is brought into the work area.
- Joined
- Jan 16, 2007
- Messages
- 759
This is an interesting thread to read. There are so many specifics that I should know better than to try to input anything, BUT... the straightest things in my shop are the Festool guide rails. I dont know their specs but laid on sides, they give me good indication of flatness. Prior to them, I had been using a 6' Stabila to do the job. My 1 cent.
I did end up picking up one of the Masterforce levels (rebranded Stabila) and played with it this weekend. I went through the entire gamut...used the Oneway to set the outfeed relative to the cutterhead, set the knives (.001" above the outfeed table to make the slightly concave cut previously mentioned), checked that the infeed table was square to the outfeed, then parallel, then coplanar. Square and parallel were pretty easy...it was pretty much nuts on already. That just left coplanar, which was a bit of a pain, but not too bad.
To make my measurements, I clamped the steel from my combination square vertically to the end of the level (just to get a little extra height), then clamped a dial indicator that I have with a mounting tab sticking out the back. Used this to make comparitive measurements front to back, side to side.
Now the machine is within the specs given to me by Felder.
Oh, and I used the (green) Stabila level to check general table flatness on the tables. Either the level happens to be wonky in EXACTLY the same manner as the tables (which I figure is pretty highly unlikely), or the level is very flat and so are the tables. Mostly couldn't get my .0015" feeler under the level, except for a couply of extreme measurements, which were still within .005". Heck, those might even be better now that the tables are set properly.
In short, both the Oneway and the "straightedge" were invaluable in the process. I'm very happy with the results.
To make my measurements, I clamped the steel from my combination square vertically to the end of the level (just to get a little extra height), then clamped a dial indicator that I have with a mounting tab sticking out the back. Used this to make comparitive measurements front to back, side to side.
Now the machine is within the specs given to me by Felder.
Oh, and I used the (green) Stabila level to check general table flatness on the tables. Either the level happens to be wonky in EXACTLY the same manner as the tables (which I figure is pretty highly unlikely), or the level is very flat and so are the tables. Mostly couldn't get my .0015" feeler under the level, except for a couply of extreme measurements, which were still within .005". Heck, those might even be better now that the tables are set properly.
In short, both the Oneway and the "straightedge" were invaluable in the process. I'm very happy with the results.
Similar threads
