Table Saw Setup! - Cut pieces have a concave bow! - Help!

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bp122

Expert at Jibber-Jabber
Joined
20 Aug 2019
Messages
1,123
Reaction score
664
Location
Haddenham
Hi All

I'm quite new to woodworking but now have a garage space to turn into a workshop.

Bought an Axminster Table Saw (their smallest one) last month.
I thought I had set it up alright.

But When I am cutting ply or mdf on the TS, the edges seem to have a concave bow (up to 1 - 1.2 mm in the middle of the edge) along the length. When i started noticing it, I started to look at the fence when I am pushing the sheet through, the front corner of the sheet edge which is touching the fence moves away from the fence marginally as I am feeding the sheet and then the back edge comes back in contact with it.

I tried to maintain a more constant pressure sideways while cutting, it still happens.

I loosened the fence adjustment screws and tried to adjust it when it was locked in contact with the blade, still have this issue. It is not all the time though, only some times.

After I adjusted the fence screws etc, I checked the distance between the cutting edge tip on the blade and the fence - The near side (front) on the blade is about 0.6-0.8mm to my right than the far (rear) side on the blade - effectively proving that either the fence or the blade (or maybe both) are off parallel. I feel this skewed setup is what is pulling the sheet away from the fence and causing the bow. Am I right to assume this?

How do I correct this?

Is there an obvious bit of setup I am missing here? how else is the best way to setup the fence?

Please help, as I have a few projects coming up.

Best regards
bp122

P.S- I can't access the forum at work, so please forgive my late response(s)
 
You say that you have set the machine up. How did you do this and what procedure did you follow? If you have set it up correctly it should perform as intended.
 
I think my word "setup" just included a check with a square, it looked square at first.
As you can imagine, it is my first table saw and I was foolishly excited to cut something on it :D
Now I can only see the error!

Elaborating on the "setup":
1. While everything was disconnected from mains, gave the blade a spin, I didn't see any wobble in it.
2. Set the fence on various positions to see if it had any flex - all was okay.
3. Checked the fence against the mitre slot - again, they both seemed parallel
4. NEVER checked the distance from the fence to both front AND back of the blade, only checked the front as I saw no wobble before.
5. NEVER checked if the fence was truly perpendicular to the edge of the cast iron top.

I hope this explains the initial blunders.

After getting a few bows, I checked - readjusted - checked again - readjusted again the four screws to set the fence, no matter how careful I adjust, the fence is 0.6-0.8mm out with teh blade (dimension between the fence and the rear of the blade is larger compared to the front side of the blade)
 
bp122":da08ii07 said:
The front corner of the sheet edge which is touching the fence moves away from the fence marginally as I am feeding the sheet and then the back edge comes back in contact with it".
You can't cut against a reference edge that's not straight and expect the result edge to be straight without first "correcting" it, however good the table saw.

Have you assumed the edge of the plywood e.g. the uncut edge is absolutely straight before your first cut, it might not be? Or, have you continued to cut more plywood using a cut edge that's (now) not absolutely straight before another cut?

Have you checked there isn't a bow in the fence and you are simply pressing the plywood into it with the first plywood corner? The result might be variable if you are comparing a long cut edge with a short one...
 
If the edge that you are cutting from is straight I don't see why it's not following that and producing a straight cut. The fence should if anything give slight clearance as material passes the blade. I take it the fence is not one of those full length ones?
 
One thought - try using push sticks? Push the work from as close to the blade as possible - by that I mean on a line with the blade. If you push from close to the fence, you will force the work away from the fence, towards the blade, which will make a concave cut. Pushing in line with the blade makes the work want to turn towards the fence, which will keep it automatically snug and straight.

Obviously, you don't want your hands running into the blade, so you use push sticks. One to push the work forward from the trailing edge, (and this is the one that should be virtually in line with the blade) and the other forwards, closer to the blade but still before it, to give a little pressure downwards and into the fence. Every cut, every time, use push sticks, not because it is safe, but because you get the pressure in the right place, and get a clean cut. You also get to keep your fingers, which is nice.

(Apologies if I sound like I am an expert - I am four months ahead of you in knowledge, with the same saw, which might be why I know about the numpty mistake - I did that, too. Not saying for a minute you are doing it wrong - this is just what I did wrong).
 
Thank you all for the responses.

I decided to re check the distances between the mitre slot and the blade, and then between the mitre slot and the fence.

The blade is about 0.25mm out of square, which apart from unscrewing and adjusting the cast table top, I don't see another way to adjust. Not even sure if there is against built into it.

The fence is straight, however, was out by 2mm towards the far end, toeing out, over a length of about 600mm.

So it turns out what you guys said was true after all, the whole setup was done by an excited silly person rather than a methodical one.

I setup the fence parallel to the mitre slot, the error now is less than 0.5mm over 600mm, which is still not perfect but something I can (or have to) live with for the time being.
Perhaps when I can afford a dial gauge (or find someone kind enough to lend me theirs :D ), I can calibrate the whole thing properly.

However, I did see the shortcomings of the little plastic push stock that came with the TS. As a temporary measure, repurposed an old 14" wooden plane into a heavy push block until I can design a better solution.

It was also my technique in feeding the stock which was faulty, with the new push block, I am getting better results. But still struggling with ripping steps of less than 25mm as there is no way for me to grip the stock well enough to get parallel strips. Need to make one of those slim push plates.

Thanks for your help.

And please treat me like the world's first silly person and please don't feel the need to sugar coat anything :D. I've got to learn thing's the right way. I'd rather gain experience than boost my ego.
 
If I had no flat thing to reference of, or no way of making something flat...
You should learn to use a plane if you haven't learned already, before you use a tablesaw in my opinion.
But thinking you could get by if cutting sheet goods ......Just for the moment :eek:
I would start by getting that blade parallel with the table using the pick a tooth method.
[youtube]LJPNlJPC6Bw[/youtube]
Setting your fence is how you prefer it, some like it parallel with the blade for tenon jigs for example.
I make rip cuts on it only at the moment, so making do with the fence a hair or two away from the back of the blade.
That method on the video should get you pretty close.
You could then make a crosscut sled and rip a sheet of ply accurately if you have hold downs incorporated.

Use a pair of push sticks made of soft ply, or look at Steve Maskery's HSE approved design.
Plenty of designs make your hand reach over the blade in close proximity so beware!.

Tom
 
Ttrees":3nde5599 said:
If I had no flat thing to reference of, or no way of making something flat...
You should learn to use a plane if you haven't learned already, before you use a tablesaw in my opinion.
But thinking you could get by if cutting sheet goods ......Just for the moment :eek:
I would start by getting that blade parallel with the table using the pick a tooth method.
[youtube]LJPNlJPC6Bw[/youtube]
Setting your fence is how you prefer it, some like it parallel with the blade for tenon jigs for example.
I make rip cuts on it only at the moment, so making do with the fence a hair or two away from the back of the blade.
That method on the video should get you pretty close.
You could then make a crosscut sled and rip a sheet of ply accurately if you have hold downs incorporated.

Use a pair of push sticks made of soft ply, or look at Steve Maskery's HSE approved design.
Plenty of designs make your hand reach over the blade in close proximity so beware!.

Tom

Just to add, most of the table saw accidents I have come across (via toutube or here) seem to involve using a block rather than push sticks. This might just be from my random sampling, but is seems that the block gives a false sense of security, band also gets your hand much closer to the blade, where it will always be at risk. YMMV.
 
Your wording about the 25 mm strips makes me think you have a 25 mm piece against the fence.
Whenever possible you have the larger piece against the fence, allowing the smaller piece to come slightly away from the blade as it is cut.

If you are cutting long thin strips then maybe a jig like this will help
https://www.youtube.com/watch?v=S9PTFizoA64&t=711s
Its the second jig half way through, but the first one is also very handy to have.
 
In that video above, he is using an unguarded blade, no riving knife, no push stick and is reaching over the blade whilst it is still spinning.
Although the jigs themselves may be very good, it is irresponsible to show this sort of behaviour, especially as it is usually beginners who search out this stuff online.
No, no, no.
 
Steve, agreed, but tell him, not us.
He is american, which explains (but not excuses) everything.
On the plus side, I have made both of those jigs and they are extremely useful.
 
Ttrees":2wnx5e8e said:
If I had no flat thing to reference of, or no way of making something flat...
You should learn to use a plane if you haven't learned already, before you use a tablesaw in my opinion.
But thinking you could get by if cutting sheet goods ......Just for the moment :eek:
I would start by getting that blade parallel with the table using the pick a tooth method.
[youtube]LJPNlJPC6Bw[/youtube]
Setting your fence is how you prefer it, some like it parallel with the blade for tenon jigs for example.
I make rip cuts on it only at the moment, so making do with the fence a hair or two away from the back of the blade.
That method on the video should get you pretty close.
You could then make a crosscut sled and rip a sheet of ply accurately if you have hold downs incorporated.

Use a pair of push sticks made of soft ply, or look at Steve Maskery's HSE approved design.
Plenty of designs make your hand reach over the blade in close proximity so beware!.

Tom

I tried this method, this is where I found the 0.25mm discrepancy from front to back.
I am now getting myself a dial gauge from work this weekend, which should help determine the exact amount.
I also tried this method in a different way, where I made the blade tooth brush the end of the combination square and checked the rest of the teeth. This is where I detected that out of 48 teeth, 17 of them touch or brush the end of the square, the remaining don't. I can see the blade move out and in in relation to the square, not by much, by about 0.2-0.3mm.

sunnybob":2wnx5e8e said:
Your wording about the 25 mm strips makes me think you have a 25 mm piece against the fence.
Whenever possible you have the larger piece against the fence, allowing the smaller piece to come slightly away from the blade as it is cut.

If you are cutting long thin strips then maybe a jig like this will help
https://www.youtube.com/watch?v=S9PTFizoA64&t=711s
Its the second jig half way through, but the first one is also very handy to have.

You assume right. I only did this because I wanted the strip from the one side of the board without a bow, so the flat reference was against the fence. But then again, I could have trued the other side of the board first and then cut a strip off it - one can tell I'm new at this! :D
 
Sounds like there might be crud on the arbor flanges, or the blade is no good.
The arbor flanges probably have a hollow profile and are in contact with the blade on the edge
so as the nut is tightened they flatten out somewhat.
Just a light lick for 20 seconds each on good 400g on a surface plate made all the difference for me.
I think you would be able to get satisfactory results after using the pick a tooth method.

Good luck
Tom
 
I also tried this method in a different way, where I made the blade tooth brush the end of the combination square and checked the rest of the teeth. This is where I detected that out of 48 teeth, 17 of them touch or brush the end of the square, the remaining don't. I can see the blade move out and in in relation to the square, not by much, by about 0.2-0.3mm.


Thats not relevant. As the blade spins, the teeth stay in the same relationship to each other, so the only effect that will have is to just cut a fractionally wider slot than the kerf measures.
Provided of course that the blade itself is not wobbling on the shaft.
 
Ttrees":q0i2v81m said:
Sounds like there might be crud on the arbor flanges, or the blade is no good.
The arbor flanges probably have a hollow profile and are in contact with the blade on the edge
so as the nut is tightened they flatten out somewhat.
Just a light lick for 20 seconds each on good 400g on a surface plate made all the difference for me.
I think you would be able to get satisfactory results after using the pick a tooth method.

Good luck
Tom

Thanks, Tom. I shall look for that tonight.

sunnybob":q0i2v81m said:
I also tried this method in a different way, where I made the blade tooth brush the end of the combination square and checked the rest of the teeth. This is where I detected that out of 48 teeth, 17 of them touch or brush the end of the square, the remaining don't. I can see the blade move out and in in relation to the square, not by much, by about 0.2-0.3mm.


Thats not relevant. As the blade spins, the teeth stay in the same relationship to each other, so the only effect that will have is to just cut a fractionally wider slot than the kerf measures.
Provided of course that the blade itself is not wobbling on the shaft.

Fair point, when I checked it didn't look like the blade was wobbling on the shaft, but I will take another look.

Thank you, sunnybob!
 
bp122":20hhw4mm said:
Ttrees":20hhw4mm said:
Sounds like there might be crud on the arbor flanges, or the blade is no good.
The arbor flanges probably have a hollow profile and are in contact with the blade on the edge
so as the nut is tightened they flatten out somewhat.
Just a light lick for 20 seconds each on good 400g on a surface plate made all the difference for me.
I think you would be able to get satisfactory results after using the pick a tooth method.

Good luck
Tom

Thanks, Tom. I shall look for that tonight.

sunnybob":20hhw4mm said:
I also tried this method in a different way, where I made the blade tooth brush the end of the combination square and checked the rest of the teeth. This is where I detected that out of 48 teeth, 17 of them touch or brush the end of the square, the remaining don't. I can see the blade move out and in in relation to the square, not by much, by about 0.2-0.3mm.


Thats not relevant. As the blade spins, the teeth stay in the same relationship to each other, so the only effect that will have is to just cut a fractionally wider slot than the kerf measures.
Provided of course that the blade itself is not wobbling on the shaft.

Fair point, when I checked it didn't look like the blade was wobbling on the shaft, but I will take another look.

Thank you, sunnybob!

For work that matters, it's good practise to make your cut a fraction wider than you need to, then plane it to the size you want (handplane not powerplane!), this will take out any discrepancies from the saw like kerf marks, burn etc etc, which can be common unless you are buying £50 precision blades.
 
I took the opportunity this morning while I was working on my first ever "successful" edge joint , to clean out the blade, the washer and the arbor of my table saw and to see if some debris was causing it to have the 0.25mm wobble I was talking about.
There wasn't anything obvious there. I couldn't have checked the blade for flatness as I don't have anything flatter thna my table saw table, which, along with its mitre slots and blade slots, doesn't have enough area to check the flatness of the blade.

However, while doing all this, I also discovered that the blade insert that came with the TS, although made out of cast Aluminium, it is very distorted and is kicking the workpieces in the air a bit. I tried playing around with the adjustment screws on the back of the plate. If I got the plate flat, it would either sit too shallow or too proud of the cast table, if I levelled it with one side of it lining up with the TS table, the other side is distorted again. It isn't very well made.

So my next project may be to make an insert out of wood.
Any thoughts?
 
If the Throat Plate (TP) is a nice fit, but just not very flat, use it as a template to make some new throat plates. DS tape it to you material (mine are MDF, 12mm IIRC) and flush trim the edges.

Drill a couple of finger holes, one at each end, and round them over nicely. These will make removal of the TP easier.

Insert short screws underneath where the location points are and adjust the screws until the TP sits flush.

Make a short cut in the far end so that the Riving Knife can come up, then glue in a tiny piece behind the RK position to prevent the TP from becoming too weak.

Lower the blade completely. Put the fence over the TP BUT NOT OVER THE BLADE. Wedge the TP down underneath the fence. Start the saw and, very carefully, raise the blade. The RK should be able to come up behind it. You now have a Zero Clearance Insert (ZCI), suitable for that blade.

It should improve the quality of your cuts, especially cross-cuts. Make a batch, so you always have a new one ready when you need it. And when it wears out, you can turn it end for end and use the other edge.
 
Thank you for that precise set of instructions, Steve. I shall try that. Then at least my thin strips won't end up inside the saw! Plus the cut quality improvement is an added bonus.
 
Back
Top