From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Spoke Tensiometers:  Which brand is better?
Date: 28 Jun 2001 20:38:13 GMT

Karl Frisch writes:


> Jobst Brandt wrote:

>> For this reason I designed a precision gauge that measures from the
>> support side and uses a dial micrometer so that small deflections can
>> be used.  It also zeroes on the spoke and, because it uses such a
>> small deflection, bending stiffness over a 100mm span is small enough
>> to make the difference between 1.6 and 2.0mm diameter spokes
>> insignificant.  The instrument is shown in "the Bicycle Wheel" and was
>> in production at Avocet but shops were unwilling to pay $300 for this
>> precision calibrated gauge.  100 of them were subsequently sold by DT
>> in Europe.

> You know I had one of those fine instruments on the bench at
> PAB......anyone see it recently? I understand a junior mechanic
> snatched it. A truly irreplaceable toy for those who care about such
> things. By the way one of the bicycle magazines ran an article on
> "one of the worlds finest mechanics," in one picture he is holding
> an instrument that looks identical to the Brandt/Avocet device...
> the caption was something to like "a custom tensiometer built by his
> brother" just plain weird.

Wheelsmith has my first one and a second one while Tim Parker,
probably the man in the article, had his brother build about ten of
them in his own machine shop from my drawings.  Unfortunately the
person who got your gauge probable doesn't use it but keeps it as a
trophy trinket.

I have offered the drawings to several bicycle tool folks but they
don't think there is a market for it.  My impression was that they
didn't believe there was any benefit from it in the first place, from
the questions they asked.  I use mine with every wheel I build to
sample tension level.  I pluck spokes to make the wheel uniform
because I can do that more easily.

Jobst Brandt    <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Spoke Tensiometers:  Which brand is better?
Date: 28 Jun 2001 23:10:59 GMT

Dan Goldenberg writes:

> I use mine with every wheel I build to sample tension level.

> I am assuming this is to arrive at a tension value determined previously
> by the method described in your book. What value does this typically
> come out to for an MA-2 rim, 36 spokes?

Yes, but then my tensiometer was never calibrated and I just use the
dial readings to characterize tension.  0.013 right side rear, 0.025
front wheel.  That doesn't help you much but it works great for me.

The Avocet units had metric dials by Starrett.  Mine has an inch dial
by Mitutoyo.

Jobst Brandt    <jbrandt@hpl.hp.com>

From: jbrandt@hpl.hp.com (Jobst Brandt)
Newsgroups: rec.bicycles.tech
Subject: Re: Spoke Tensiometers:  Which brand is better?
Date: 28 Jun 2001 23:27:29 GMT

Tim McNamara writes:

>> The problem with all tensiometers that I have seen offered is that
>> they measure across the spoke so part of the reading is spoke
>> thickness and none of these instruments allows zeroing the gauge on
>> the spoke, tare, so to speak.  Besides, Hozan in particular and
>> others I've seen use too large a test load, one that will affect
>> tension of the spoke being measured.

> You can zero the Wheelsmith tool on a spoke that is not under
> tension.  I'm not sure if that addresses your concern, however.

I don't think I made that clear.  My instrument has a dial which is
set to zero, on the spoke before applying the test load.  Subsequently
the reading on the dial gives the deflection in the units of the dial
micrometer.  The Wheelsmith instrument has no dial it uses a vernier
scale that is rigidly attached to the two scissor image halves.
Besides having a friction pivot, it has no bearing between spoke and
contacts so that the cosine error binds the instrument.  The Avocet
unit used two ball bearings as spoke support so there was essentially
no friction involved in the reading.

> I hadn't thought of the test load affecting spoke tension
> significantly; this seems like it could cause a falsely high reading?

With Hozan, that is appreciable and on top of that, one presses down
on the spoke with the hand grip so that all the application force is
resisted by the spoke.  It's a BAD instrument.  DT's new one has a lot
of potential but its gauge length is too short, and it measures across
the wire, and it has friction contact, and it uses too large a test
load, and it isn't zeroed.  It's electronics are great and I would
love to have had them apply that to the Avocet/DT tensiometer but I
suspect they didn't appreciate what they had.  I would have bought one
just for the micro processor abilities.  They haven't talked to me
since I panned their revolution spoke years ago.  Having used Berg and
Redaelli 1.5mm spokes years ago and had them fail in torsion, I know
how bad windup failures are.  DT's 2.0mm ends cause more torsion than
the ones I mentioned at 1.8mm.  It's the friction, not the thread ramp
that causes all the problem and that increases with thread diameter as
it does with 2.34mm diameter spokes.

Jobst Brandt    <jbrandt@hpl.hp.com>

From: jobst.brandt@stanfordalumni.org
Subject: Re: safe spoke elongtion?
Newsgroups: rec.bicycles.tech
Message-ID: <OQ9G8.10591$lX2.61460@typhoon.sonic.net>
Date: Mon, 20 May 2002 16:51:58 GMT

Seb who? writes:

> I'm designing a tensiometer, and have one issue left to nail down.
> Depending on the dimensions I use, measuring the tension will
> require stretching the spoke a variable amount.  In reasonable
> designs, this varies from .5 to 3 mm.

What is it you are trying to do that isn't accomplished by the
tensiometer you saw in "the Bicycle Wheel"?  The tensiometer shown
there has features that others do not.

1. Spoke thickness does not affect readings, support and measurement
   being made from the same side.

2. By using a precision micrometer, deflections produced by a small
   test load can be accurately detected.  For this a "zero on the fly"
   must be made to cancel slight irregularities something this
   instrument does.

3. A light test load has minimal effect on existing tension that would
   otherwise be compromised as is common with tensiometers like the
   Hozan.

4. Ball bearing support at the end of the test section prevents
   friction from retarding spoke deflection (cosine error)

5. The four inch test span is reasonably large to permit high accuracy
   with small micrometer read deflections.

I don't know what you propose to do in "stretching".  This is not a
reasonable the way of measuring tension in wires.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: safe spoke elongtion?
Newsgroups: rec.bicycles.tech
Message-ID: <cyyG8.378$3w2.1554@typhoon.sonic.net>
Date: Tue, 21 May 2002 20:58:48 GMT

Seb who? writes:

By the way, why so secretive.  This reminds me of speaking to a spy
with dark glasses and hat drawn low in Casablanca.

>> What is it you are trying to do that isn't accomplished by the
>> tensiometer you saw in "the Bicycle Wheel"?

> Actually, I'm not sure, since theres no specific dimensions or
> forces given for the tensiometer in "the Bicycle Wheel".

> In essence, I am attempting to apply "appropriate technology" design
> to the tensiometer.  I don't think the one in "the Bicycle Wheel"
> could be built cheaply at home by a hobbiest with no special tools,
> which conflicts directly with the philosophy of "appropriate
> technology", and (less directly) with the spirit of "Do It
> Yourself".

You say that as though it were trivial measurement while in fact it is
not.  Deflecting the spoke from its straight path causes a length
change which if large will change its tension.  The instrument shown
deflects the spoke 0.013" in a 4" span, which gives an angle of 0.37
degrees.  Spoke length change for that angle is 0.000021" and bending
is kept small enough so that it does not materially affect readings
for spoke thicknesses of 1.6 to 2.0mm diameter wire.

> I'm trying to design a reasonably accurate tensiometer that can be
> built for less than $10 from items available in any decent hardware
> store, or mass produced for less than $1.  I'm trying to make this
> design and the method of its use freely available information.

If you look at the Wheelsmith instrument, you'll see that it has two
pieces of stamped sheet metal, a spring and a pivot pin.  That's hard
to beat in simplicity but it is not as accurate as the instrument in
the book because it has friction and measures across the wire.  It
also cannot be checked for proper zero or be adjusted for kinks in
spokes.

> Since "It is not necessary to know the actual tension but only that
> the tension has reached the desired mark", I figure an inexpensive
> tensiometer that gives precise, scaled, and repeatable (if not
> accurate) readings would be all a hobbiest wheelbuilder needs to
> eliminate the problem that "final tension of a wheel is affected by
> the mood of its builder."

I think you have a contradiction there.  Knowing you are at the
desired mark IS determining the actual tension.  I don't see how else
you can define that.

> Granted, I'm "cheating" by using a torque wrench to do the actual
> measuring, but I figure most people would rather buy a nice,
> generally useful torque wrench (which costs maybe $30) and make a
> cheap accessory tensiometer than buy a tensiometer (which costs at
> the least I've seen $90) that they only use when building (or maybe
> truing) wheels.

The torque wrench is not free and it is not sensitive enough to give
you what you want.  Fine instrument torque wrenches (1-50 in-oz) are
at least as expensive as a good tensiometer.

>> The tensiometer shown there has features that others do not.

> I appreciate its features, but feel I would have no need to
> duplicate them, or benefit from doing so.  Your design's features
> and functions do not seem duplicable without precision machining
> and / or specialized parts, and it has moving parts which complicate
> home construction.  My design can be made with a hand drill and a
> brazing kit, and has no moving parts.

That is the crux of the matter.  Why do you think tensiometers cost
as much as they do.  The instrument must be largely non invasive and
that requires precision.  What you seem to be looking for is far
easier to achieve by building a wheel as described in the book and to
know what tone a spoke of the appropriate tension makes.  You'll
notice that tension specifications are not offered with most rims, so
that is in itself your problem to solve.

>>1. Spoke thickness does not affect readings, support and measurement
>>   being made from the same side.

> Your design still calls for a table to calculate absolute tension
> based on the measurements, correct?

That's only because its production was too small (less than 200) to
have a special gauge dial, something that could be done electronically
today.

> How much extra work would it be to make a separate table for each of
> a few spoke thicknesses (and maybe brands)?  That's all that would
> be required to compensate for spoke thickness when using my design.
> Given a formula for spoke elasticity and the appropriate data for
> various spokes, I could set up a spreadsheet to produce all the
> required tables, I think.  It could be verified with measurement- a
> precision tensiometer such as your design would be very helpful
> there!

Brands of spokes makes no difference.  All the ones of interest are
steel and and its modulus of elasticity is well defined.  Thickness
only enters into the problem with more than the customary range of up
to 2.0mm diameter.  Flat spokes do not change the readings.

>>2. By using a precision micrometer, deflections produced by a small
>>   test load can be accurately detected.  For this a "zero on the
>>   fly" must be made to cancel slight irregularities something this
>>   instrument does.

> Since I don't know what it really means, I can't say its something I
> ever wanted or not.  I'll grant its not a feature of my design.
> Will that negatively affect a hobbiest wheel builder who desires to
> use a tensiometer primarily to ascertain if his new wheel has enough
> or to much tension?

I don't think you understand the problem or you wouldn't say what you
do.  The hobbyist needs to get a good reading at least as much as the
professional wheel builder.

>>3. A light test load has minimal effect on existing tension that
>>   would otherwise be compromised as is common with tensiometers
>>   like the Hozan.

> Actually, by posting this thread, I was trying to figure out how
> much "test load" is ideal for a design like mine.  How much is
> "light"?  What do you suggest?

About 2kgf gives a deflection of around 0.4mm, a useful deflection yet
one that does not change existing tension because the length change it
causes is diminishingly small.

> In this case, a light test load would produces changes that are hard
> to ascertain accurately without precise alignments or measurements.
> I could conceivably reconfigure my design to use a very lightweight
> torque wrench and very small deflections, but I was shooting for
> something a hobbiest could build cheaply and quickly, and use with a
> minimum of operator error.

Try building a cheap microscope.

>>4. Ball bearing support at the end of the test section prevents
>>   friction from retarding spoke deflection (cosine error)

> Would smooth round posts with lubrication at all 3 points of contact
> (2 loaded, 1 unloaded) in my design do nearly as well?

No but your method would not cause any length change in the
measurement length so friction is not a problem.  There are a few
problems in your design, one of them being that the three pegs cannot
be in a straight line because deflections need to be kept to size that
is far smaller than what a pin that could support the load could take.
The concept is good but it defies execution in the range of forces
involved.

>>5. The four inch test span is reasonably large to permit high
>>   accuracy with small micrometer read deflections.

> I was thinking of a total 6.6mm (1/4 inch) deflection over 10 cm (4
> inches, simply because that fits well on most spokes).  Is that far
> off from what your rig does?  I can't say what my design
> accomplishes relative to yours without knowing the dimensions and
> forces you use.

There is the killer.  Deflections must be kept to less than 0.5mm
(0.02") or the measurement will be too intrusive.  That is to say, the
measurement result will be rim and spoke length and number dependent.

>> I don't know what you propose to do in "stretching".  This is not a
>> reasonable the way of measuring tension in wires.

> "a tensiometer measures the deflection of a spoke over a given span
> in response to a standard load".  Assuming the endpoints do no move,
> is it possible to deflect a span without elongating it?  To me, it
> seems is impossible to measure spoke tension using a tensiometer
> without "stretching" the spokes at least a bit- if not the spoke
> being "deflected", then all the other spokes in the wheel, so as to
> move the endpoints of the "deflected" spoke!

I think the above paragraph was dealt with earlier.  Cosine error can
be so small that it has no stretching effect on the spoke.

> However, it may very well be the rig you used, and perhaps other
> tensiometers, cause less elongation in the deflected spoke than I
> was planning on.  That is exactly why I asked how much is
> reasonable.  If the answer is "none", what is the mechanism used?

> Perhaps I could ascertain the folly of my plans myself if you would
> be so kind as to point out where / how I could get some precise,
> accurate spoke elongation data?  I trust you are as reasonable as I
> if not much more so.  Still, I'm quite bull headed and would rather
> not take your word for the infeasability of my design without doing
> my own analysis and experiments.

Give the concept some more thought.  Maybe you can devise a
configuration that can be made to work.  The concept is not bad but it
seems to lie beyond implementation in the constraints involved.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: safe spoke elongtion?
Newsgroups: rec.bicycles.tech
Message-ID: <0iEG8.549$3w2.2664@typhoon.sonic.net>
Date: Wed, 22 May 2002 03:31:08 GMT

Seb who? writes:

>> By the way, why so secretive.  This reminds me of speaking to a spy
>> with dark glasses and hat drawn low in Casablanca.

> What value would my last name add to the conversation?  If I was a
> published author or otherwise felt a need to reference my
> credentials, it might be pertinent.

Just the contrary.  You can readily use your name because you have
nothing to hide or have said previously.  Besides, it's rude to talk
to people with the visor of your armor closed, incognito.  This is
part of common courtesy that doesn't hurt.  You must not understand why
most contributors here do not hide behind "handles".

>> You say that as though it were trivial measurement while in fact it
>> is not.

> No, I had not assumed it was.  I felt it went without saying.  I
> think the fact that I felt the need to design a device for the
> purpose, and had some admitted design quandaries, showed the
> non-triviality.

Your suggestion that you have a simple way of measuring tension
implies that you think it is simple at best, or that those who make
them are unclear on the concept at worst.

>> Deflecting the spoke from its straight path causes a length change
>> which if large will change its tension.  The instrument shown
>> deflects the spoke 0.013" in a 4" span, which gives an angle of
>> 0.37 degrees.  Spoke length change for that angle is 0.000021" and
>> bending is kept small enough so that it does not materially affect
>> readings for spoke thicknesses of 1.6 to 2.0mm diameter wire.

> OK, that's exactly the sort of info I was asking for.  Thank you.
> That is indeed very difficult to replicate, or even approximate,
> with a home built tool.

You would do well to build a prototype and investigate the effects
before proposing what you have.  You are asking the net to do your
R&D, at least in the theoretical realm, before you have an idea of
how to accomplish this beyond sketchy ideas.

>>> Since "It is not necessary to know the actual tension but only
>>> that the tension has reached the desired mark", I figure an
>>> inexpensive tensiometer that gives precise, scaled, and repeatable
>>> (if not accurate) readings would be all a hobbiest wheelbuilder
>>> needs to eliminate the problem that "final tension of a wheel is
>>> affected by the mood of its builder."

>> I think you have a contradiction there.  Knowing you are at the
>> desired mark IS determining the actual tension.  I don't see how
>> else you can define that.

> The words in quotes are from your own book (p. 118 3ed), and contain
> the entire contradiction.  How would you define the difference?

Vague insinuations will get you nowhere.

>>> Granted, I'm "cheating" by using a torque wrench to do the actual
>>> measuring, but I figure most people would rather buy a nice,
>>> generally useful torque wrench (which costs maybe $30) and make a
>>> cheap accessory tensiometer than buy a tensiometer (which costs at
>>> the least I've seen $90) that they only use when building (or
>>> maybe truing) wheels.

>> The torque wrench is not free and it is not sensitive enough to
>> give you what you want.  Fine instrument torque wrenches (1-50
>> in-oz) are at least as expensive as a good tensiometer.

> Yes, so I am finding.  They can be built fairly easily and
> calibrated with a test mass, but that still isn't free, and might be
> an equal waste of time.

Oh oh.  Here we go again.  Next I see that you want to wind your own
springs and make steel shapes.

>> That is the crux of the matter.  Why do you think tensiometers cost
>> as much as they do.  The instrument must be largely non invasive
>> and that requires precision.  What you seem to be looking for is
>> far easier to achieve by building a wheel as described in the book
>> and to know what tone a spoke of the appropriate tension makes.
>> You'll notice that tension specifications are not offered with most
>> rims, so that is in itself your problem to solve.

> I had no sure indication that is was not possible to design an
> inexpensive tensiometer, and it didn't seem harmful to try.

In that case I suggest you do that and don't use this forum as a
scratch pad for Idle conjecture.  It's not as though you proposed
specific questions but rather you made a grand "open system" proposal.

>> I don't think you understand the problem or you wouldn't say what you
>> do.  The hobbyist needs to get a good reading at least as much as the
>> professional wheel builder.

> Well, its certainly not your responsibility to explain the problem.
> I do wonder, if the Wheelsmith tensiometer "can't be properly
> zeroed", how does it "zero on the fly", and why is it so useful?

Try to imagine a way to do it and you'll see it cannot be done on any
other tensiometer because the measurement is generally made by the
depressing anvil.  That is why my instrument starts out with an
undeflected spoke against which the dial of the micrometer is zeroed
(rotated) with the thumb before applying the load.  For such small
deflections otherwise undetectable kinks in spokes can be nulled, and
they occur as is evident from use.

>> About 2kgf gives a deflection of around 0.4mm, a useful deflection
>> yet one that does not change existing tension because the length
>> change it causes is diminishingly small.

> If tension/length is changed, by a known amount, why can't the
> additional tension be calculated and canceled out?  Is it likely to
> be harmful to the spoke?

Because it depends on spoke length, diameter, number, and rim
stiffness, all basically unknown effects.  Therefore, making any
length changes is an invasion on what you want to measure.

>>> Would smooth round posts with lubrication at all 3 points of
>>> contact (2 loaded, 1 unloaded) in my design do nearly as well?

>> No but your method would not cause any length change in the
>> measurement length so friction is not a problem.  There are a few
>> problems in your design, one of them being that the three pegs
>> cannot be in a straight line because deflections need to be kept to
>> size that is far smaller than what a pin that could support the
>> load could take.  The concept is good but it defies execution in
>> the range of forces involved.

> That's exactly what I was aiming to figure out from the "safe
> elongation" of a spoke, or compute using spoke elasticity data.

Let me emphasize that paragraph.  There would be no length change in
the gauge span and therefore no need for anti-friction bearings.  The
size of the test load would determine the amount of length change
(cosine error).

>> There is the killer.  Deflections must be kept to less than 0.5mm
>> (0.02") or the measurement will be too intrusive.  That is to say,
>> the measurement result will be rim and spoke length and number
>> dependent.

> A flimsier rim would cause a different reading, I hadn't considered
> that.  Nuts.

>> Give the concept some more thought.  Maybe you can devise a
>> configuration that can be made to work.  The concept is not bad but
>> it seems to lie beyond implementation in the constraints involved.

> I'd considered some alternate configurations, but they aren't
> fundamentally different in construction from Wheelsmith's design, and
> require similar tolerances.  With access to a machine shop, building
> one might save me money, but that wasn't my purpose.

That design was made by Norm Ogle, a clever engineer and bike rider
who died of degenerative causes at an early age.  I am surprised that the
instrument is as expensive as it is considering that it has so few and
simple parts.  The spring requiring the most accuracy.

The instrument shown in "the Bicycle Wheel" was sold for about $250 at
the time it was on the market, the dial gauge being the main expense.
DT sold most of them in Europe, there being little interest here in
the USA at the time.

Jobst Brandt  <jobst.brandt@stanfordalumni.org>  Palo Alto CA

From: jobst.brandt@stanfordalumni.org
Subject: Re: Use aircraft cable tensiometer for spoke tension measurements?
Newsgroups: rec.bicycles.tech
Message-ID: <WMS%b.4300$_3.67807@typhoon.sonic.net>
Date: Sat, 28 Feb 2004 02:11:02 GMT

Michael P. Bassler writes:

> I use one like this where I work. It has Three riser blocks. that
> allow you to change, depending on cable diameter. If you can, you
> would want to kind of calibrate it for a known tension, say on a set
> of wheels built by someone who used a known good tensiometer. When
> they are calibrated they come back with a correction table that
> corresponds to temp and cable size not really designed for spokes. I
> believe you could use it.

When I first looked for a tensiometer, I came across this kind of
instrument and wondered why they chose to measure across the wire.
The ones I found were too complicated to put on a spoke, the lever
being on the far side of the gauge with respect to the wire (spoke),
sort of inside the wheel.

That's why I designed a one sided tensiometer that measures from the
same side as the spoke support and can be zeroed on the fly.  100 of
these instruments were sold through DT a few years ago but it seems
that the features of the instrument, low test load and high precision
gauge, zero-on-the-spoke, and one sided measurement were not
understood.  Today there are only spoke tensiometers that use a high
test load, measuring across the spoke.  None can be zeroed.

"I taught him everything I know and he still knows nothing!"

Jobst Brandt
jobst.brandt@stanfordalumni.org

From: jobst.brandt@stanfordalumni.org
Subject: Re: Spoke Tensiometer questions
Newsgroups: rec.bicycles.tech
Message-ID: <HRs0c.4639$_3.72775@typhoon.sonic.net>
Date: Sun, 29 Feb 2004 21:30:15 GMT

V Copelan writes:

> After years of wheel building without a spoke tensiometer, I'm
> considering buying one.  I've been looking at three different spoke
> tensiometers.  The Park tool at under $50, the Wheelsmith tool at
> around $120, and the DT Proline dial tensiometer at $250.  I've read
> the critical comments about the Park tool from Jobst Brandt. The
> Wheelsmith and Park tool appear similar.  Which tool should I
> consider?

If you do this much, the Wheelsmith tensiometer is the easiest and
quickest to use.  It isn't easily readable and convertible but it is
accurate and fast (handy).  It was invented by Norm Ogle who was a
great talent in such things.  The simplicity and durability of the
instrument should receive design awards.

That said, it was too bad that its measurement includes the spoke
thickness, but then I can't suggest a way of getting past that with
the design at hand.

Jobst Brandt
jobst.brandt@stanfordalumni.org

From: jobst.brandt@stanfordalumni.org
Subject: Re: Rim lateral flex question
Newsgroups: alt.mountain-bike,rec.bicycles.tech
Date: Sun, 09 Dec 2001 02:19:39 GMT

Jose Rizal writes:

> I've had the wheels tensioned at my LBS.  The guy who did it used a
> DT Swiss tensiometer (fancy-looking red gizmo with dial gauge) and
> consulted a DT chart for recommended tension readings.  I actually
> saw him test and adjust all spokes, but I don't know if the tension
> in each is at the maximum the spokes can take.  Assuming, of course,
> that DT recommended spoke tension isn't the maximum for the spoke
> material.  Should have asked...

Don't worry about spokes being near their limit.  Unless there are
fewer than 20 spokes in the wheel, rim collapse will most likely limit
spoke tension before spokes are threatened.  Meanwhile, tensiometers
are best used to sample overall tension by reading a couple of spokes
of each side of the wheel.  Final adjustment should be done by ear so
that spokes of the same side (rear wheels) are all the same as well as
possible.  This is not done by a tensiometer because they don't have
sufficient resolution for task.

Jobst Brandt    <jobst.brandt@stanfordalumni.org>