Today I did a little more on the J&S pulleys. Previously I'd put a pilot bore in and then roughed out the outside. Today I wanted to bore the pilot bore out to 1 1/4" for some hubs that I'm making up.
As I was looking to sell my Blake co-ax today I thought it may be worthwhile doing a comparison of locating methods.
I found what I think is the true zero (at least for this exercise) by using a tenth indicating B&S indicator and ran it across the side of the hole to find a local minima. To guard against the DTI probe not being symmetrical (which it turned out it was not by 0.1mm or there abouts) around the spindle centreline, I then turned it 180 degrees and averaged the positions.
P1030039 (Medium).JPG
Firstly, the Blake. This is a dynamic method, in that you are determining your centre with the machine running. In the first photo you can see the stop bar rotated against a clamp. When the machine is going, the tip travels around the bore, displacing a bell crank which in turn moves the dial. The theory is that you manipulate the X and Y screws to minimise the needle deflection. From memory you can run up to 500rpm, but the slower the better if you don't want the needle to be a blur. The dial is graduated in 0.0005" (5/10th) but the fine print in the instructions say that is for a 4" outside sweep tool. Position wise the centre was within 1 1/2 thou of position from the interapid. The major draw back with the tool is the Z it chews up - 205mm.
P1030038 (Medium).JPG P1030037 (Medium).JPG
With this I was able to position my hole so that the range of needle deflection was 1 to 2 graduations but because of the scaling I'm not sure what that means in absolute terms
The next to be tried was my home made "Zero-it" clone with an Interapid 5/10ths dial attached. As can be seen from the photo, it only takes up around 150mm of Z. I expected this method to be the most accurate and relatively simple to use, as really all I had was an indicator on a bar whose radius once adjusted is fixed. The method was to try to adjust the X and Y axis (one at a time) by swinging the DTI through 180 degrees and getting the two readings to coincide. In reality I could get front/ back and side/ side to agree within a thou, but not at the same reading. As the bore should not be out of round and the spindle was 'cold' I can only assume this difference was due to clearances in the vertical head bearings. As it was this set up said that the real centre was 3 1/2 thou to the right and 3 1/2 thou to the front - the symmetry of those numbers makes me suspicious
P1030035 (Medium).JPG
The last to be used was the Taster. I'd made a low profile adaptor for my Taster in my lemonade thread http://ift.tt/1BR7yWx , so Z wise it only uses around 105mm. I prefer having the adaptor on this as with the Zero-it and Blake there was the added complication of using a collet chuck to hold them and another source of error perhaps? If I'd used a drill chuck it would have added another 55mm of length.
P1010831 (Medium).JPG P1010833 (Medium).JPG
One of the draw backs with the Taster is that it needs to be set up so the axis of the probe is spot on concentric with the axis of the spindle. I've has several goes at this but have never managed spot on. One day I might have to try it with a hot spindle head and see if that helps (normally it's something I try first thing in the morning so to delay annoying the neighbours as long as possible). Using the taster and comparing with my true centre, I got 0.068mm out along the X axis and 0.058mm out along the Y (2.7 and 2.3 thou out respectively)
So to summarise, all methods are within 0.125mm (5 thou) of true position but I need to investigate more why I have a difference. Next time I might bore a hole and measure it without removing the job as that may be a source of error, and doing things when the bearings are warm will probably also help. The Taster certainly uses less Z, but I'm sure there are other ways of finding position which are even easier. One day if I can get my act together I might even try optical...
Michael
As I was looking to sell my Blake co-ax today I thought it may be worthwhile doing a comparison of locating methods.
I found what I think is the true zero (at least for this exercise) by using a tenth indicating B&S indicator and ran it across the side of the hole to find a local minima. To guard against the DTI probe not being symmetrical (which it turned out it was not by 0.1mm or there abouts) around the spindle centreline, I then turned it 180 degrees and averaged the positions.
P1030039 (Medium).JPG
Firstly, the Blake. This is a dynamic method, in that you are determining your centre with the machine running. In the first photo you can see the stop bar rotated against a clamp. When the machine is going, the tip travels around the bore, displacing a bell crank which in turn moves the dial. The theory is that you manipulate the X and Y screws to minimise the needle deflection. From memory you can run up to 500rpm, but the slower the better if you don't want the needle to be a blur. The dial is graduated in 0.0005" (5/10th) but the fine print in the instructions say that is for a 4" outside sweep tool. Position wise the centre was within 1 1/2 thou of position from the interapid. The major draw back with the tool is the Z it chews up - 205mm.
P1030038 (Medium).JPG P1030037 (Medium).JPG
With this I was able to position my hole so that the range of needle deflection was 1 to 2 graduations but because of the scaling I'm not sure what that means in absolute terms
The next to be tried was my home made "Zero-it" clone with an Interapid 5/10ths dial attached. As can be seen from the photo, it only takes up around 150mm of Z. I expected this method to be the most accurate and relatively simple to use, as really all I had was an indicator on a bar whose radius once adjusted is fixed. The method was to try to adjust the X and Y axis (one at a time) by swinging the DTI through 180 degrees and getting the two readings to coincide. In reality I could get front/ back and side/ side to agree within a thou, but not at the same reading. As the bore should not be out of round and the spindle was 'cold' I can only assume this difference was due to clearances in the vertical head bearings. As it was this set up said that the real centre was 3 1/2 thou to the right and 3 1/2 thou to the front - the symmetry of those numbers makes me suspicious
P1030035 (Medium).JPG
The last to be used was the Taster. I'd made a low profile adaptor for my Taster in my lemonade thread http://ift.tt/1BR7yWx , so Z wise it only uses around 105mm. I prefer having the adaptor on this as with the Zero-it and Blake there was the added complication of using a collet chuck to hold them and another source of error perhaps? If I'd used a drill chuck it would have added another 55mm of length.
P1010831 (Medium).JPG P1010833 (Medium).JPG
One of the draw backs with the Taster is that it needs to be set up so the axis of the probe is spot on concentric with the axis of the spindle. I've has several goes at this but have never managed spot on. One day I might have to try it with a hot spindle head and see if that helps (normally it's something I try first thing in the morning so to delay annoying the neighbours as long as possible). Using the taster and comparing with my true centre, I got 0.068mm out along the X axis and 0.058mm out along the Y (2.7 and 2.3 thou out respectively)
So to summarise, all methods are within 0.125mm (5 thou) of true position but I need to investigate more why I have a difference. Next time I might bore a hole and measure it without removing the job as that may be a source of error, and doing things when the bearings are warm will probably also help. The Taster certainly uses less Z, but I'm sure there are other ways of finding position which are even easier. One day if I can get my act together I might even try optical...
Michael
0 commentaires:
Enregistrer un commentaire