Audible bandsaw tension gauge

[Phil Thien]

So I took what I learned in that poll of plucked bandsaw blades and I created a web site where I allow the user to take a few measurements of their blade and saw, and determine the optimum frequency of their plucked blade based upon their desired tension.

You can use a $13 chromatic tuner (like this one (http://www.amazon.com/Korg-CA-30-Solo-Chromatic-Tuner/dp/B000BBRZ3S/ref=sr_1_3?ie=UTF8&s=musical-instruments&qid=1243035171&sr=8-3)) to check your bandsaw's blade and see how close you are.

It should be pretty darn accurate. Probably more accurate than any clamp-on gauge.

If you're interested in taking a look, the page is here:

http://www.cgallery.com/jpthien/tg.htm

Let me know what you think.

[John Stevens]

Amazing, Phil. You really have a knack for seeing things in a new way. Just for fun, you may want to know that the physics and engineering principles you used for gauging bandsaw tension by vibration have been applied to measure power output for bicycle racers by measuring the frequency of the vibration of the chain via a magnetic device similar to the "pickup" on an electric bass.

Regards,

John

[Tim Put]

I agree this should be much more accurate. The problem with the clamp on gauges (besides unnecessary expense) is that they measure in parallel not series. They don't measure stress (the tension per unit area) but rather strain (the stretching caused by stress) and so can't even infer absolute stress, but only change in stress. It assumes as zero reference the tension of the blade when clamped, but either that's not zero and the meter will underdisplay in proportion, or the blade almost certainly won't be clamped straight and the meter will grossly overdisplay.

[Philip Berman]

Adagio in G For Flute, Clarinet and Bandsaw, Op. 1

This is a great idea (tensioning the bandsaw, not the Adagio!)

[Jamie Buxton]

This is an interesting idea. I'll try it on my saw, although I'm going to have to run down a scale appropriate to the blade's weight.

[Jim O'Dell]

Phil, do we need to take into account the size of the blade in length between contact point on the wheels, as well as width of the blade? ie, a fatter piano or guitar string is used for lower notes. Bigger doesn't vibrate as fast.
Very interesting idea. Jim.

[Phil Thien]

Phil, do we need to take into account the size of the blade in length between contact point on the wheels, as well as width of the blade? ie, a fatter piano or guitar string is used for lower notes. Bigger doesn't vibrate as fast.
Very interesting idea. Jim.

When you provide the wheel separation measurement in step #5, you're also providing your "blade in length between contact point on the wheels." That is, the number is the same.

For blade width, I get that in step #4, but I need the blade width w/o the teeth. I take the teeth into account later in mass only, when I compute the mass of the segment in step #5 for the purpose of computing velocity.

[Jim O'Dell]

Cool. Obviously I haven't gone to the site you linked to. I will tomorrow. Jim.

[george wilson]

Why not just learn how to tension the saw? I just go by how much the blade deflects when I pull it sideways,and how it cuts. With some sensitivity to the bandsaw it isn't that hard.

[Phil Thien]

Why not just learn how to tension the saw? I just go by how much the blade deflects when I pull it sideways,and how it cuts. With some sensitivity to the bandsaw it isn't that hard.

Hi George!

I'm not going to argue with any method of tensioning the saw that yields satisfactory results. But there has always been that mysterious 20,000 (or a little more or less) value out there. While this manufacturer suggested value may be completely arbitrary (depending on who you talk to), I've always wondered what a blade that is tensioned to 20,000 PSI feels/sounds like.

It is further complicated by the fact that it has seemed real difficult to find 20k without a load-cell based solution. The $100-$300 gauges have been reported to be less than accurate. Whether you agree or not, this type of dilemma makes the gears in my head start turning. It is simply how I'm wired.

Like I've said elsewhere (and here, I think), I'm not suggesting people use this method for everyday use. But some others will undoubtedly wonder what 20k is like on their saw. Maybe they'll ultimately feel like it is too much. Maybe they'll say, "hey, my scale already gets me to 20k, and I like this PSI" But at least they'll know what a blade tensioned to manufacturer specs "feels" like, and they can move forward form there.

It may simply boil down to a curiosity factor for many.

[Don Eddard]

Why not just learn how to tension the saw? I just go by how much the blade deflects when I pull it sideways,and how it cuts. With some sensitivity to the bandsaw it isn't that hard.

That approach works if you have someone to show you what "right" is supposed to feel like. There are a lot of guys who have never had any in-person mentoring with woodworking tools. having learned only from books, video, and the Internet. A procedure like Phil is testing is a quantifiable method by which neophytes can find the ballpark, so they can know when they are in it. It also allows the curious among the old pros to confirm they've been in the ballpark all along. ;)

[Mike Cutler]

It is further complicated by the fact that it has seemed real difficult to find 20k without a load-cell based solution. The $100-$300 gauges have been reported to be less than accurate. Whether you agree or not, this type of dilemma makes the gears in my head start turning. It is simply how I'm wired.

Phil

Cool idea:cool: I wish you luck with it. I don't need it myself, but if it helps others I'm all for it.

Concerning the above statement;

I tested 5 different clamp on bandsaw tension gauges for Mark Duginske a few years ago. 3 were by Iturra, one by Lennox and one by Starrett. All had their individual shortcomings. To post the results of the tests would be pretty laborious and read like an engineering White paper. (For our folks that are in to such things, all of the devices were tested with NIST traceable mechanical standards).
So far the "best" device I have tested is the Carter digital readout that is compatible with the Delta and Jet 14" bandsaws. It measures out in lbs/wheel force, but can be converted to pressure using Young's Modulus and the cross section of the blade width. It is more closely similar to a load cell than a strain gauge. Force acting on a piezoelectric crystal.
All of my research has indicated that the mystical 15,000 psi comes from the original standards for the Delta Bandsaw when used for metal cutting. The Lennox Tension gauge has the "green" area at 30,000 psi. I suspect this is also a metal cutting value. The hydraulically tensioned Marvel at work tension out to about 28,000psi, with a 1 1/4" metal cutting blade when the cups are set, but this an interpolated value that can't take into account the friction from rolling the blade 90 degrees twice in it's rotational path.
I have also played around a little bit with a DoAll and Kalamazoo metal cutting bandsaw.

I think that one of the areas that folks tend to overlook is the spring. The stock spring on my 14" jet was only capable of tensioning the blade to about 9800 psi. I converted it to a Cobra coil for greater range.
The heart of the band saw tension system is the spring. No spring equals no bandsaw. No matter what system is used for tensioning, if there isn't space between the coils you risk breaking something.

For the record. Mark Duginske's summation that the installed tension gauge on a 14" band saw is adequate is a valid statement. After testing the devices and coming up with a way to accurately repeat and measure strain. When the formula was worked backward the tension gauge on the back of my Jet was pretty close up to a 1/2" blade.

[Phil Thien]

Phil

Cool idea:cool: I wish you luck with it. I don't need it myself, but if it helps others I'm all for it.



Thanks for the enlightenment, Mike.

Do you still have a Carter electronic gauge? I was hoping to find someone that had access to one, that would be willing to correlate the results of my method w/ the results from the Carter.

I think there are many users out there that have no idea if their stock spring is sufficient. I also imagine there are plenty of new bandsaw users that come along that are just curious if their new saw's tension gauge is accurate. And many don't want to throw down the cash for a Carter (or can't use the Carter due to their saw's design).

My hope is to write some code that will play the derived FF, with the harmonics you'd expect from a steel string added. So all I'd need someone to do is tension 2-3 different blades to various values and see if the output from my app sounds like the blade set to a known tension via the Carter.

[Stephen Edwards]

Hi George!

I'm not going to argue with any method of tensioning the saw that yields satisfactory results. But there has always been that mysterious 20,000 (or a little more or less) value out there. While this manufacturer suggested value may be completely arbitrary (depending on who you talk to), I've always wondered what a blade that is tensioned to 20,000 PSI feels/sounds like.

It is further complicated by the fact that it has seemed real difficult to find 20k without a load-cell based solution. The $100-$300 gauges have been reported to be less than accurate. Whether you agree or not, this type of dilemma makes the gears in my head start turning. It is simply how I'm wired.

Like I've said elsewhere (and here, I think), I'm not suggesting people use this method for everyday use. But some others will undoubtedly wonder what 20k is like on their saw. Maybe they'll ultimately feel like it is too much. Maybe they'll say, "hey, my scale already gets me to 20k, and I like this PSI" But at least they'll know what a blade tensioned to manufacturer specs "feels" like, and they can move forward form there.

It may simply boil down to a curiosity factor for many.

I've thought this post from you all morning, Phil. Though the physics involved in your experiment are way over my feeble mind, and would probably not be a method that I'd ever use to test the tension on BS blades, I do appreciate YOUR interest in this area and I especially appreciate the quote where you say:

"Whether you agree or not, this type of dilemma makes the gears in my head start turning. It is simply how I'm wired."

THAT statement, in other areas of interest to me, is something that I CAN relate to. So, my hat's off to you for riding whatever horse it is that trips your trigger and leads you down the trail to seek the answers that your "wiring" leads you to pursue. I UNDERSTAND that wiring bidnez! It just IS.......

Respectfully,

[Jacob Reverb]

I just checked mine.

I have a 18" Rikon 10-345 bandsaw, currently with a 1/4" blade on it.

I tensioned the blade with the quick-tensioning lever, checked the gauge (it was on) and the blade was too loose to make a recognizable tone (and I play guitar).

That said, it cuts just fine.

Sorry.

[Phil Thien]

I just checked mine.

I have a 18" Rikon 10-345 bandsaw, currently with a 1/4" blade on it.

I tensioned the blade with the quick-tensioning lever, checked the gauge (it was on) and the blade was too loose to make a recognizable tone (and I play guitar).

That said, it cuts just fine.

Sorry.

Don't be sorry, I think that is outstanding feedback. I would appreciate it if, next time you have a larger blade installed if you'd perform the same test?

[Peter Quinn]

Wow, cool post Phil. My method has always been to add tension until I break either the blade or the saw, then back off 1/4 turn and continue cutting.:D Equally scientific but less theoretical and more applied research!:eek: Do I smell a handheld shop device in the making?

[Phil Thien]

Wow, cool post Phil. My method has always been to add tension until I break either the blade or the saw, then back off 1/4 turn and continue cutting.:D Equally scientific but less theoretical and more applied research!:eek: Do I smell a handheld shop device in the making?

How about an iphone app? I can imagine the ridicule I'd endure for that. :p

[george wilson]

No one showed me how to tension a bandsaw. I've just been careful in observing how the saw cuts,and have learned it myself.

[Barry Vabeach]

Phil, I just did the measurements for three of my blades and then did the calculations. I don't have the electronic device to check the tuning, but wondered if I messed something up - the results called for anywhere from 38.6 hz to 44 hz for both of my 1/2" blades ( Ex Lennox 1/2 6/10 bimetall on a Grizzly go513x2 - the blade is 131.25 long, weighs 185 grams, the distance between centers is 39.09 inches, the width is .408 and the thickness is .024 ) I would have guessed a much higher note - and am pretty sure I won't be able to hear 38 hz. Any idea what I screwed up?

[Phil Thien]

Any idea what I screwed up?

Maybe nothing. I plugged the #'s in and got the same results. So it is just that low, or my math is wrong (the second is just as possible as the first).

Let me ask you a couple of questions. First, the wheel separation (in step #5), you are measuring from the center point (axle) of the top wheel to the center point (axle) of the bottom wheel?

Second, if you tension the blade according to the saw's tension gauge, and then pluck it on the side opposite the guides, how does it sound? Can you find a string on the guitar tuner (below) that is close?

http://www.guitarforbeginners.com:80/onlinetuner.html

Another edit: I found a site w/ the specs for MIDI files. So I created a MIDI file with six guitar (steel string) plucks in this order: 36.7, 38.9, 43.7, 36.7, 38.9, 43.7-Hz. So three frequencies, then repeat.

The first two (36.7 and 38.9) are almost undiscernible unless you listen very carefully.

This file is called test.mid, and is zipped into test.zip. You can open test.mid in MediaPlayer or any number of other audio players. So if you don't mind, compare your blade tensioned according to the saw's instructions to this MIDI file (you can play it in MS Media Player if you use Windows). If your band doesn't pluck anything like the MIDI file I've sent, I'd also like you to try the guitar tuner above. Please let me know your observations.

Thanks for all your help!

[Alan Schaffter]

Phil

I think that one of the areas that folks tend to overlook is the spring. The stock spring on my 14" jet was only capable of tensioning the blade to about 9800 psi. I converted it to a Cobra coil for greater range.
The heart of the band saw tension system is the spring. No spring equals no bandsaw. No matter what system is used for tensioning, if there isn't space between the coils you risk breaking something.

For the record. Mark Duginske's summation that the installed tension gauge on a 14" band saw is adequate is a valid statement. After testing the devices and coming up with a way to accurately repeat and measure strain. When the formula was worked backward the tension gauge on the back of my Jet was pretty close up to a 1/2" blade.

In a response to Phil's parallel thread on another forum, I had two questions not answered.

First, after-market springs with different strength- If you change the spring, you will no longer be able to use the manufacturer's marks, correct? So what do you use? Experience?

I pointed out to Phil that a precise tension may not be needed. The procedure should be to set the tension to the mark and then adjust it for a good cut with the blade being used on the wood (thickness, hardness, tightness of turns) being used. In my opinion precision tension only comes to play when, at some later time, you want to repeat the setting when cutting the same wood under the same conditions.

During the big flap a few years ago, I asked Mark D. how blade or saw manufactures determined recommended tension. He didn't answer. If I were a betting man, I would guess it was done experimentally- have a couple of guys go into the shop and test various tensions on a given blade against a selection of wood (thickness and hardness) then pick a mark or tension that provides the best cut. How else could it be done? If this is true, precision tensioning is a meaningless term. A ball park setting followed by adjustment for the best cut is what is needed.

[Mike Cutler]

In a response to Phil's parallel thread on another forum, I had two questions not answered.

First, after-market springs with different strength- If you change the spring, you will no longer be able to use the manufacturer's marks, correct? So what do you use? Experience?

A ball park setting followed by adjustment for the best cut is what is needed.

Alan

Concerning the first statement: The amount of actual "stretch" a band saw blade encounters from fully relaxed to tensioned, and even over tensioned, is a very small amount measured in thousandths of an inch. While the tension mechanism on the back of a 14" saw is a pretty crude measurement device,it is a repeatable linear scale representing the upward travel of the upper band saw wheel. In the case of the Carter Cobra Coil I installed, which was physically longer than the OEM spring and a stronger spring in itself, it actually made the scale usable for me. Springs are fairly predictable devices and not really subject to wear and tear, so once one is installed, the gauge markings may have to be altered initially. Prior to it's installation I used to have the setting at 3/4" for a 1/2" blade.
There is nothing that says that a person couldn't scribe their own marks. It's just a soft aluminum plate.

I absolutely agree with your second statement. That is the way it's done. It doesn't matter how accurately the blade is tensioned if it isn't cutting properly, something needs to be changed.
In a PM I sent to Phil I stated that I "personally" didn't see the big mystique associated with band saw tuning. I think setting the blades on a planer/joiner is harder to do, but for some it's child's play. Not me though.:o All throughout Jr. High and High School metal and wood shops, 6 years, the band saw was always my favorite machine. I could easily convert to a table saw less shop. But that is me, I really like band saws.
I see the biggest benefit in having accurate tensioning gauges not so much to actually set the blade tension ,but to aid a person in defining and correcting a tuning issue that may actually not be a blade tension issue at all, but something else entirely.

[Alan Schaffter]

How about an iphone app? I can imagine the ridicule I'd endure for that. :p

Now you are on to something! :D

[Barry Vabeach]

Phil, thanks for your help. I couldn't hear anything with the midi file, not sure why. My length measurement is pretty much on, I even confirmed it at OWWM - using the diameter of the wheel and the height between centers it calculates the length of the blade and it is on the money. I did set it to midway on the tension gauge ( The Grizzly just has numbers and they don't correlate to blade widths) and it was between an E and the A. I guess my question is that according to this site, http://garagebandtheory.com/files/chaptersamplesforweb/001%20Note%20Names/Pages%20from%20001%20Note%20Names%20pgs%2013-15%20%206-05-07.pdf
low E on a guitar is 82.4 Hz, which is much higher than what I am shooting for ( unless we are supposed to be using harmonics )

[Phil Thien]

Phil, thanks for your help. I couldn't hear anything with the midi file, not sure why. My length measurement is pretty much on, I even confirmed it at OWWM - using the diameter of the wheel and the height between centers it calculates the length of the blade and it is on the money. I did set it to midway on the tension gauge ( The Grizzly just has numbers and they don't correlate to blade widths) and it was between an E and the A. I guess my question is that according to this site, http://garagebandtheory.com/files/chaptersamplesforweb/001%20Note%20Names/Pages%20from%20001%20Note%20Names%20pgs%2013-15%20%206-05-07.pdf
low E on a guitar is 82.4 Hz, which is much higher than what I am shooting for ( unless we are supposed to be using harmonics )

That's where it gets interesting. Assuming a fixed PSI (like 20,000), then the further apart you move your wheels, the lower the frequency your plucked blade should emit. That is because the unsupported segment of the blade (the section we're treating like a guitar string) is longer.

So it is possible that 82.4-Hz is too high for the tension you're trying to achieve. In fact, if my math is correct (a BIG IF at this point), and the #'s you're providing are accurate, then getting to 82.4-Hz may require nearly 70,000-PSI! :eek:

So what model bandsaw do you have? And is it stock? No risers or anything?

[Phil Thien]

I have updated the page so I now have approx. 40 MIDI files which will give you a good indication of what I believe your properly tensioned blade should sound like.

Working through the worksheet, you will be given a freq. that a blade tensioned to the PSI you're shooting for should generate when plucked ON YOUR SAW. You can then listen to a sample file of ten plucks that you can use to compare to your actual results.

This largely negates the need for a chromatic tuner. All you need now is a good digital kitchen scale. And if you don't have one of those, you're eating too many pretzels in one sitting. :D

[Chris Friesen]

While the calculation may be exact, I think you may be getting overly complicated.

The frequency of a vibrating string is 1/(2*L) * sqrt(*T/u) where L is the length, T is the tension in units of force, and u is the linear mass.

To a first approximation, the mass of the teeth can be ignored. This lets us consider a simple rectangular band of material. The density of steel used for bandsaw blades can be treated basically as a constant.

If we specify tension in units of pressure, this results in the tensile force varying linearly with blade cross-sectional area, in exactly the same way that the linear mass of the band does.

If we combine the above simplifications, then to a first approximation the frequency will depend only on the desired tension (in units of pressure) and the length between the axles (with the blade tensioned on the machine).

f = 1/(2*L) sqrt(Tp/rho)

where Tp is the tension in units of pressure, and rho is the density of steel in units of mass/volume.

[Barry Vabeach]

Phil, I have the Grizzly GO513x2 - it is a 17 inch saw and a large number of guys here have it. I have done some mods, but none relate to the distance between the centers.

[Phil Thien]

...To a first approximation, the mass of the teeth can be ignored. This lets us consider a simple rectangular band of material. The density of steel used for bandsaw blades can be treated basically as a constant.

Hi Chris!

The reason I compute the mass of the "string" (rather than using a constant) is that I'd like to accommodate bands with applied teeth or carbide grit. I may need that computed mass.

The other carp you see is the mine field left from me removing a feature that allowed me to derive tension from freq. That feature may return, I can't decide. I was going to figure that out today (and do clean-up) but spent much of the day deciphering MIDI and writing an app to create those sample files. It took longer than I expected.

[Chris Friesen]

The reason I compute the mass of the "string" (rather than using a constant) is that I'd like to accommodate bands with applied teeth or carbide grit. I may need that computed mass.

The other carp you see is the mine field left from me removing a feature that allowed me to derive tension from freq.

Right. That's why I said to a first approximation. I checked a few of my own blades, and depending on the particular blade it seems like ignoring the teeth could give up to about 15% error in the results. (That was with a 4tpi, 1/4" blade, where the teeth are large compared to the rest of the blade. My other blades would give less error.)

Since the workable tension range seems to be quite large, I suspect even a 15% error wouldn't be a problem. The advantage of the approximate result is the ease of calculation--you don't need to measure or weigh the blade at all, only the distance between the axles.

I'm not suggesting that you need to change your webpage (nicely done, by the way), just pointing out to those reading this that it's probably possible to get "close enough" using a much simpler formula.