Beware Salter Electronic Kitchen Scales
Farry
If anyone's considering buying one of those expensive Salter Electronic
Kitchen Scales, you'll probably want to be aware of a design problem
that messes up the accuracy. You can work around the problem, and still
get accurate readings, but it is a rather daft fault. This is what I've
put on Salter's feedback form, and I'm waiting for their reply:
-------------------- Begin quote --------------------
I bought one of your 1004 "Electronic kitchen scales with stainless
steel platform" about a year ago and I was very happy with it. However,
it developed a fault, and I had it replaced under guarantee with a new
one. Unfortunately, I am far from happy with the behaviour of the new
scale and I have found what I believe is a serious design fault that has
been introduced into it.
I noticed that when I was measuring powders or liquids, that I was
ending up with quantities that were visibly too large for the displayed
weight. When I then transferred the powder or liquid to some old
balance scales that were cumbersome but accurate, I found that the
quantities were as much as 20% too high.
I then investigated the problem, and found that the scale misreads if
the powder or liquid is added too slowly. I presume that this is because
a software change has been introduced into the scale's electronics to
try to compensate for drifting errors in the weight sensors.
Didn't it occur to the designers that the natural way to use a kitchen
scale is to put a bowl on the scale, zero it, pour in most of the
quantity of powder that is required, then slowly add the rest to bring
it up to the target weight. If you do this however, the scale
under-reads. The only way to get around this behaviour, and to get a
reasonably accurate reading, is to press a spoon into the powder for a
moment, when adding small quantities, to make a large change to the
weight which disables this software trick.
I have to say that this has made the scale much harder to use and is
very irritating. I'd like to return the scale for one that behaves the
way that the old scale did. Before I do however, I'd like to know if any
are available, or do all of your kitchen scales now have this misleading
drift compensation thing?
-------------------- End quote --------------------
--
Farry
Dave Fawthrop
<davefa...@OMiTTHiSyahooANDTHiS.co.uk> wrote:
| Hi.
|
| If anyone's considering buying one of those expensive Salter Electronic
| Kitchen Scales, you'll probably want to be aware of a design problem
| that messes up the accuracy. You can work around the problem, and still
| get accurate readings, but it is a rather daft fault. This is what I've
| put on Salter's feedback form, and I'm waiting for their reply:
I use the Salter electronic Aquatronic model number 3007, all plastic, and
find it everything I had hoped. I have checked the accuracy a few times
and found it fine.
Dave F
Farry
>I use the Salter electronic Aquatronic model number 3007, all plastic, and
>find it everything I had hoped. I have checked the accuracy a few times
>and found it fine.
I would also have said that, but of the scale I bought a year ago,
rather than the "compensated" scale. When did you buy yours?
--
Farry
Dave Fawthrop
Sounds as if yours has a simple mechanical fault. Perhaps crud has built up
inside, or the platform has got twisted.
I have had mine about ?2? years.
I have just tried adding sugar slowly, no problem.
Dave F
Farry
OK, the model 3007 is not displayed on their website, so I guess it's an
old model and predates the compensation trick.
As I said, it's the brand new scale that's got the problem. It's almost
certainly designed into the software because I can think of no way that
a fault could introduce behaviour like that, but I can think of why a
software designer would want to add that compensation, if he didn't
properly think through the consequences.
--
Farry
Dave Fawthrop
<davefa...@OMiTTHiSyahooANDTHiS.co.uk> wrote:
Just reread your original post.
As both an ex-engineer and a software writer, it still looks like a
mechanical problem, and not a software problem. My guess is that the
platform is dragging, catching or juddering on something.
Mechanical problems sometimes build up slowly, or suddenly appear.
Software problems are either there, or not there, all the time.
Dave F
Dave Gibson
"Farry" <davefa...@OMiTTHiSyahooANDTHiS.co.uk> wrote in message
news:bu8710173rl99dac1...@4ax.com...
The kitchen scales use strain guages as the weight transducer. Although
these are not linear in output (which is compensated), they do not exhibit
the fault you describe. Its more likely to be a fault in the weighing
platform suspension mechanism sticking (imagine a shaft going through a hole
and rubbing on the side). Probably a batch manufacturing problem. Check the
batch number of your model and take it back for exchange, making sure the
new one is from a different batch/manufacture date.
Dave
Farry
>Just reread your original post.
>
>As both an ex-engineer and a software writer, it still looks like a
>mechanical problem, and not a software problem. My guess is that the
>platform is dragging, catching or juddering on something.
>Mechanical problems sometimes build up slowly, or suddenly appear.
>Software problems are either there, or not there, all the time.
OK, I've just done this.
1. Place clean mirror on worktop to make a perfectly flat surface.
2. Make sure both sides of scale are clean, and place on mirror.
3. Zero scale.
4. Put cereal bowl on scale - 274g.
5. Add tablespoon (20g) of muesli - 294g.
6. Tip out muesli and put bowl back - 274g.
7. Add same muesli back very slowly over 1 minute - 276g.
i.e. it's somehow lost 18g.
8. Tip out muesli and put bowl back - 256g.
Still 18g down.
9. Remove bowl - and scale shows "----" (below zero).
This is quite repeatable. The scale is accurate and consistent, and does
not drift even over several minutes, PROVIDED that the weight changes
are swift.
QED?
--
Farry
Dave Fawthrop
Judder!
A well known mechanical engineering problem. Often experianced in car
brakes when the car vibrates when braking.
Friction varies with position and time, not like they teach you in school.
I have seen it in lots of things but never in scales.
Take it back and demonstrate the problem, Even a shop assistant should
understand that it does not work properly. Get a different model, it may
be a batch problem.
Dave F
Edwin Pawlowski
>> 6. Tip out muesli and put bowl back - 274g.
>
> 7. Add same muesli back very slowly over 1 minute - 276g.
> i.e. it's somehow lost 18g.
>
> 8. Tip out muesli and put bowl back - 256g.
> Still 18g down.
>
> 9. Remove bowl - and scale shows "----" (below zero).
Interesting. Some recipes call for a half dozen ingreditents and I add them
one a t a time. It can easily take a few minutes as you put one away and
get out the next. That would be a problem.
--
Ed
e...@snet.net
http://pages.cthome.net/edhome
Farry
>Judder!
>A well known mechanical engineering problem. Often experianced in car
>brakes when the car vibrates when braking.
>Friction varies with position and time, not like they teach you in school.
>I have seen it in lots of things but never in scales.
>
>Take it back and demonstrate the problem, Even a shop assistant should
>understand that it does not work properly. Get a different model, it may
>be a batch problem.
I can't take it back to the shop, because I bought the original a year
ago, and now I've just got a replacement under guarantee by post from
Salter. So I'll wait for the response from Salter to my query, first.
I appreciate that it's hard to believe that this behaviour could be
designed in. OK, next experiment.
I repeat the process of adding muesli to the bowl three times, without
zeroing the scale, and weigh the bowl each time. It starts at 274g, then
it's 254g, then 236g, then 216g. i.e. 18-20g down each time leaving it
58g low. And since I completely remove the bowl from the scale to tip
out the muesli each time, that should have sorted out any judder.
--
Farry
Farry
>Interesting. Some recipes call for a half dozen ingreditents and I add them
>one a t a time. It can easily take a few minutes as you put one away and
>get out the next. That would be a problem.
Actually, that can be a problem, but not for the reason that I've
outlined. If you leave the scale without changing the weight on it, the
display will switch off after a couple of minutes to save power, and you
can't switch it on again without zeroing the scale. You have to remember
to dab it with your finger every minute to keep it alive, and the
displayed weight will remain consistent. Just don't pour in an
ingredient too slowly.
--
Farry
Farry
>The kitchen scales use strain guages as the weight transducer. Although
>these are not linear in output (which is compensated), they do not exhibit
>the fault you describe. Its more likely to be a fault in the weighing
>platform suspension mechanism sticking (imagine a shaft going through a hole
>and rubbing on the side). Probably a batch manufacturing problem. Check the
>batch number of your model and take it back for exchange, making sure the
>new one is from a different batch/manufacture date.
Unfortunately, there's no batch number, the model number is the same,
and the appearance of the old and new scales are identical, as far as I
can remember. I believe that I've eliminated the possibility of the
mechanism sticking mechanically, so that leaves either a bizarre
electronic fault, or more likely in my opinion, an ill-thought attempt
to compensate for drift in the strain gauge.
--
Farry
Dave Fawthrop
<davefa...@OMiTTHiSyahooANDTHiS.co.uk> wrote:
| I can't take it back to the shop, because I bought the original a year
| ago, and now I've just got a replacement under guarantee by post from
| Salter. So I'll wait for the response from Salter to my query, first.
Salter offer a 10 year guarantee
http://www.salterhousewares.com/pages/products/index.asp?code=14#
and click on 1004.
Why not get it swapped for a 3007 which they still sell on
http://www.salterhousewares.com/pages/products/index.asp?code=14&offset=2#
Not as fancy but it works OK.
Better than stripping what you have and risking buggering it up.
--
Dave Fawthrop <da...@hyphenologist.co.uk>
Sick and tired of Junk Snail Mail? Register your family
surname and address with www.mpsonline.org.uk
Adrian Tupper
news:cth71055ashf7aad8...@4ax.com:
Not convinced Dave. At least it must be coupled with a software fault
if the bowl showed different weights at different times.
--
Adrian
Adrian Tupper
news:m6s710djbsegdbb6m...@4ax.com:
So, what is the advantage of digital scales over my springy ones?
--
Adrian
rmp
sticking is unlikely to be an issue.
Software issue may relate to low cost analog to digital conversion, reading
frequency, etc.
"Dave Gibson" <s...@m.less> wrote in message
news:bv0evg$age$1...@titan.btinternet.com...
June Hughes
<adrian...@totalise.co.uk> writes
the time. The only problem is that the bowl is large (not a problem)
and round (a problem if you are tipping the contents into a smaller
bowl).
--
June Hughes
Fred
> >
>
> So, what is the advantage of digital scales over my springy ones?
>
> --
> Adrian
Greater accuracy.
Fred
The Good Gourmet
http://www.thegoodgourmet.com
Edwin Pawlowski
In some cases, but not all. A good balance scale will be far more accurate
than a digital, but some spring scales are junk also. It may have changed,
but a few years ago, Ohaus tripe beam balance scales were allowed for gold
assay, but not the digital.
use...@isbd.co.uk
>
> > >
> >
> > So, what is the advantage of digital scales over my springy ones?
> >
> > --
> > Adrian
>
> Greater accuracy.
>
measurement to the electronics are just the same as for any
'mechanical' scales.
--
Chris Green
Peter Aitken
news:bv2qn8$nl5hd$1...@ID-61610.news.uni-berlin.de...
I do not believe that is true. The electronic balances that I am familiar
with use some sort of solid state pressure transducer that converts the
weight into a voltage. Spring balances use the weight of the food against a
spring to move the indicator dial. Very different processes.
Peter G. Aitken
Dave Fawthrop
wrote:
The movement caused by weight on the scales, in my Salter scales of a
different model is 1-2mm which is quite enough for friction effects to
cause problems.
Dave F
use...@isbd.co.uk
to zero set correctly. I agree that a different mechanical basis for
different sorts of scales will affect the accuracy. With electronic
scales the mechanics of the interface to the pressure transducer (if
that's what is used) will define the accuracy. It's very easy to be
fooled by a digital display into believing that all the digits it
displays represent actual accuracy.
--
Chris Green
graham
<use...@isbd.co.uk> wrote in message
news:bv3e07$o08nu$1...@ID-61610.news.uni-berlin.de...
> In uk.food+drink.misc Peter Aitken <pai...@crapnc.rr.com> wrote:
> > <use...@isbd.co.uk> wrote in message
> > news:bv2qn8$nl5hd$1...@ID-61610.news.uni-berlin.de...
> > > In uk.food+drink.misc Fred <blades...@starband.net> wrote:
> > > >
> > > > > >
> > > > >
> > > > > So, what is the advantage of digital scales over my springy ones?
> > > > >
> > > > > --
> > > > > Adrian
> > > >
> > > > Greater accuracy.
> > > >
> > > Accuracy of indication maybe but the mechanics of transferring the
> > > measurement to the electronics are just the same as for any
> > > 'mechanical' scales.
> > >
> > > --
> > > Chris Green
> >
> > I do not believe that is true. The electronic balances that I am
familiar
> > with use some sort of solid state pressure transducer that converts the
> > weight into a voltage. Spring balances use the weight of the food
against a
> > spring to move the indicator dial. Very different processes.
> >
> How about balance type mechanical scales, very accurate and very easy
> to zero set correctly.
I have some balance scales but what with parallax etc., it's difficult to
see when the pointer is exactly at the mark (at least on mine). I use my
Salter electronic all the time now and I always press the scale after adding
the last couple of grams to make certain that the reading is OK. From what
has been said here, I think that I might have to invest in some weights to
check their accuracy.
My only complaint is the cost of the batteries and, to me, their relatively
short life-span.
Graham
Adrian Tupper
$hn2.173...@twister2.starband.net:
>
>> >
>>
>> So, what is the advantage of digital scales over my springy ones?
>>
>> --
>> Adrian
>
> Greater accuracy.
Not according to the OP. And just how accurate does one need to be?
All the recipes I know have quantities given in whole oz or sometimes
half-oz. My scales are accurate enough for that.
--
Adrian
Farry
>Fred wrote:
>
>>> So, what is the advantage of digital scales over my springy ones?
>>> --
>>> Adrian
>>
>> Greater accuracy.
>
>In some cases, but not all. A good balance scale will be far more accurate
>than a digital, but some spring scales are junk also. It may have changed,
>but a few years ago, Ohaus tripe beam balance scales were allowed for gold
>assay, but not the digital.
Balance scales are very accurate, but they are a bit cumbersome. I must
admit that even having to work around the problem that I've described
with the electronic scales, they are still easier to use than my old
balance scales.
--
Farry
Mark Thorson
> Not according to the OP. And just how accurate does one need
> to be? All the recipes I know have quantities given in whole oz
> or sometimes half-oz. My scales are accurate enough for that.
Maybe he's using the scale to weigh illegal drugs,
in which case he certainly might care about fractions
of an ounce. He couldn't just say that's what he's doing
in a Usenet newsgroup.
Farry
>Maybe he's using the scale to weigh illegal drugs,
>in which case he certainly might care about fractions
>of an ounce. He couldn't just say that's what he's doing
>in a Usenet newsgroup.
Fol de rol.
-L.
> Hi.
>
> If anyone's considering buying one of those expensive Salter Electronic
> Kitchen Scales, you'll probably want to be aware of a design problem
> that messes up the accuracy. You can work around the problem, and still
> get accurate readings, but it is a rather daft fault. This is what I've
> put on Salter's feedback form, and I'm waiting for their reply:
<snip>
Just FYI. The proper way to weigh solids is to add an excess to the
scale and then *remove* the material until you reach the desired
weight.
-L.
Mark Willstatter
Jumping in late here, I think if you go back and look at the data you
posted from your experiment, all you've really shown is that you get
different results if the weight is added suddenly (whether bowl or
muesli) or gradually. By far the easiest and most likely explanation
of this is friction: there's something "sticky" in the mechanics
between bowl and strain gauge and the friction is not overcome when
weight is added very slowly. I may have used different words but in
effect, I'm with Dave here.
- Mark W.
Dave Fawthrop
I must disagree with that, it is absolutely forbidden in Chemistry because
there putting anything back in the jar may cause contamination. In
cooking it must be a bad thing.
Dave F
Mark Willstatter
> Usually strain gaged loadcells have very small deflections. So movement
> sticking is unlikely to be an issue.
>
> Software issue may relate to low cost analog to digital conversion, reading
> frequency, etc.
>
and anyway the deflection on *my* digital scale is large enough to be
visible. If you go back and look at the data from the original
poster's experiment, the only real "mystery" here is why the results
are different between adding weight quickly and very slowly. The
easiest (and therefore most likely, IMO) way to explain that result
is, as Dave has said a number of different ways, by the effects of
friction: there is something "sticky", if you like, in the mechanism
between bowl and strain gauge; that stickiness is overcome if you add
(or withdraw) weight quickly but not if you sneak it on slowly. Any
other explanation is a stretch. The other possibilities that have
been mentioned - software, strain gauge/gage, a/d conversion, and so
on, are unlikely to vary with time. In other words, I'm with Dave.
- Mark W.
David Wilkinson
>The other possibilities that have
> been mentioned - software, strain gauge/gage, a/d conversion, and so
> on, are unlikely to vary with time. In other words, I'm with Dave.
But the removal and replacement of the bowl (a large sudden change,
which should overcome any inherent "stickiness") clearly demonstrates
that the bowl has "lost weight". i.e. the scales have changed their
concept of the "zero point".
It would be interested to see if the experiment could be repeated a
number of times until the weight of the bowl registered "zero".
Based on the evidence to date, I could well believe that the scales are
adjusting their "zero point" over time.
Dave W.
Farry
>But the removal and replacement of the bowl (a large sudden change,
>which should overcome any inherent "stickiness") clearly demonstrates
>that the bowl has "lost weight". i.e. the scales have changed their
>concept of the "zero point".
>
>It would be interested to see if the experiment could be repeated a
>number of times until the weight of the bowl registered "zero".
>
>Based on the evidence to date, I could well believe that the scales are
>adjusting their "zero point" over time.
Yes indeed. But if anybody's still doubting, look at this:
1. Weigh a spoon, a cup, and a bowl in turn - 30g, 194g, and 274g.
2. Slowly add 20g of powder to bowl and take combined weight - 278g.
16g lost this time.
3. Without zeroing, weigh spoon, cup, and bowl again - 14g, 176g, 258g.
Each is 16g or 18g lower (the scale increments in 2g steps).
So the zero point has shifted by about -17g.
--
Farry
Adrian Tupper
Ah. A sensible explanation!
--
Adrian
Adrian Tupper
Nearly everything weighed in this manner is dry and I wouldn't hesitate
putting sugar, raisins or even butter back in the packet.
--
Adrian
Elaine Jones
posted on 27 Jan 2004 by Adrian Tupper
I would like to add:
There's an electronic scales for re-loading ammunition, weighing powder,
calibrated in grains - can't work that out in grammes/ounces at this time
of night.
--
...ElaineJ... Home Pages and FAQ of uk.food+drink.indian can be viewed at
...Kinetic... http://www.users.zetnet.co.uk/ejones/ufdi/index.html
..StrongArm.. Under construction, FAQ, recipes, tips, booklist, links
...RISC PC... Questions and suggestions please, email or to the newsgroup
pltrgyst
>>> to be? All the recipes I know have quantities given in whole oz
>>> or sometimes half-oz. My scales are accurate enough for that.
>>
>> Maybe he's using the scale to weigh illegal drugs,
>> in which case he certainly might care about fractions
>> of an ounce. He couldn't just say that's what he's doing
>> in a Usenet newsgroup.
Some of us use our scales to weigh pool cues and shafts, in 1g or 1/10
oz. increments. 8;)
-- Larry
-L.
Actually it is the recommended method by most makers of analytical
balances. That's how they are calibrated.
-L.
Dave Fawthrop
| Dave Fawthrop <hyp...@hyphenologist.co.uk> wrote in message news:<2j7d10t7ghu3gkh6n...@4ax.com>...
| > On 27 Jan 2004 08:54:22 -0800, k3_...@yahoo.com (-L.) wrote:
| >
|
| > | Just FYI. The proper way to weigh solids is to add an excess to the
| > | scale and then *remove* the material until you reach the desired
| > | weight.
| >
| > I must disagree with that, it is absolutely forbidden in Chemistry because
| > there putting anything back in the jar may cause contamination. In
| > cooking it must be a bad thing.
| Actually it is the recommended method by most makers of analytical
| balances. That's how they are calibrated.
Very Interesting!
It means that the analytical balance people are well aware of the friction
problems and take steps the minimise them.
In the kitchen where we have a choice between absolute accuracy and
possible contamination of our stocks. For myself I would prefer to avoid
contamination.
Dave F
occupant
I have a Soehnle digit scale made in Switzerland I bought at the
Canadian Tire chain store for about CDN $50. I couldn't figure out how
the sale took readings until I just read in this news group how it is
done. Easy switch from pounds to grams. Had it about a month so don't
know the battery usage pattern but easily wired to wall outlet if
battery life is short. I have been looking for years for an affordable
digital scale. This is it. While it has a million uses, I particularly
like the ability to pour the flour for my bread machine recipes onto the
scale without having to tap the flour to settle it for measure.
The scale is a good teaching tool, too. After you have weighted out
something like chicken breasts portions a few times, you can almost
cut "meat" up perfectly to a specific weight accuracy checking randomly.
graham
"occupant" <electronicm...@telust.net> wrote in message
news:401770BD...@telust.net...
> graham wrote:
> >
> > My only complaint is the cost of the batteries and, to me, their
relatively
> > short life-span.
> > Graham
>
> I have a Soehnle digit scale made in Switzerland I bought at the
> Canadian Tire chain store for about CDN $50.
I'll take a look at these this w/e. By then, the daytime temperature is
forecast to be above -30ºC :-((
Graham
-L.
It probably depends on what it is. I don't think re-introducing
staples like sugar or flour is an issue. Anything with poterntial
quick spoilage may be an issue, though.
-L.
Joe Doe
<hyp...@hyphenologist.co.uk> wrote:
> Very Interesting!
> It means that the analytical balance people are well aware of the friction
> problems and take steps the minimise them.
> Dave F
I work in laboratory and routinely use numerous electronic balances whose
accuracy range from 0.01 gram to 0.00001 gram. We do not see any of the
problems use describe. I do not think this is a design feature of
electronic balances in general. Our balances cost many thousands of
dollars.
I have a cheap electronic balance at home that is more temperamental
(does not tare more than once) but probably cost less than $50-100.
Roland
Abso
> Dave Fawthrop <hyp...@hyphenologist.co.uk> wrote:
>
> > Judder!
> > A well known mechanical engineering problem. Often experianced in
> > car brakes when the car vibrates when braking.
> > Friction varies with position and time, not like they teach you in
> > school. I have seen it in lots of things but never in scales.
> >
> > Take it back and demonstrate the problem, Even a shop assistant
> > should understand that it does not work properly. Get a
> > different model, it may be a batch problem.
>
> I can't take it back to the shop, because I bought the original a year
> ago, and now I've just got a replacement under guarantee by post from
> Salter. So I'll wait for the response from Salter to my query, first.
>
> I appreciate that it's hard to believe that this behaviour could be
> designed in. OK, next experiment.
>
> I repeat the process of adding muesli to the bowl three times, without
> zeroing the scale, and weigh the bowl each time. It starts at 274g,
> then it's 254g, then 236g, then 216g. i.e. 18-20g down each time
> leaving it 58g low. And since I completely remove the bowl from the
> scale to tip out the muesli each time, that should have sorted out
> any judder.
How about an experiment in which something is added *really* slowly but
over a long period of time such that the final amount of the item is
significant enough that you can rule out judder? I'm originally
thought of water dripping at a set rate but of course evaporation will
skew the results. Maybe a slowly dispensed powder?
--
Abso [at] ukrm [dot] net - Ignore header email address
The uk.people.consumers.ebay FAQ is at www.upce.org.uk
In order to maintain secrecy, this posting will self-destruct
in five seconds. Memorize it, then eat your computer.
Farry
>How about an experiment in which something is added *really* slowly but
>over a long period of time such that the final amount of the item is
>significant enough that you can rule out judder? I'm originally
>thought of water dripping at a set rate but of course evaporation will
>skew the results. Maybe a slowly dispensed powder?
The trouble with that is that the display times out after 2 minutes. I
think that I've shown that the shift is too consistent for judder. (See
the post about shifting the zero point.)
--
Farry
Farry
I've found that I'm 10 times more likely to spill something trying to
put it back in the packet, compared with transferring it to a bowl. If
it's an ingredient that comes in a cardboard box with a plastic liner,
then the powder can get dropped between the two where it will spoil, and
if I later tip the packet into the bowl then out comes the spoiled
powder.
--
Farry
Dave Fawthrop
Did you fail to use the 10 year Guarantee which the Salter web site
advertises for your particular scale? Are there hidden problems in this
guarantee? At worst you should be able to get your money back or get it
replaced with a different product.
Dave F
Farry
scale problem, and one of them mentioned that he had a Salter bathroom
scale. He said that if he weighed himself twice with a gap of about 1
hour between weighings, then the scale would show almost 1kG difference
(that's 2.2lbs for American readers). We asked if he might have done
anything in between that might have altered his weight, but he insisted
not. That seems worse than the consistency of mechanical bathroom
scales.
--
Farry
Farry
>Did you fail to use the 10 year Guarantee which the Salter web site
>advertises for your particular scale? Are there hidden problems in this
>guarantee? At worst you should be able to get your money back or get it
>replaced with a different product.
I did use the 10 year guarantee to get the current scale. I sent a
feedback form message to Salter enquiring if they had any scales without
this problem, but they've not replied yet. It's a bit pointless asking
for a replacement, and risking losing the scale in the post, if it's
going to be the same. I hesitate to phone them because the chances of
one of their customer support people knowing about the problem is zero.
If they don't reply in within a few days, I'll consider phoning them
anyway.
--
Farry
Dave Gibson
I think 'just under 1kg change' in an hour is neither here nor there ..... a
pint of beer is just over 1/2 kg, having a good slash will loose 1/2 kg, a
good doohdah could be 1 kg, removing shoes, taking money out of pocket ....
even a cup of tea will account for 1/3 kg .... PLUS you can get that sort of
change if your C of G changes, ie try leaning forwards or backwards while
standing on the scale and see the effect!
Dave
Farry
yesterday, and it's behaviour is even more surprising than the Salter
electronic kitchen scale.
This scale belongs to a friend of mine who showed me how it works. You
tap it with your foot to wake it up and wait for it to zero, then you
stand on it, and after a few moments it works out your weight and
displays that on an LCD. You can't then change that be shifting your
weight because the display remains fixed until you step right off it.
The display has a 0.1kG resolution.
My friend demonstrated that the scale was very consistent from one
moment to the next by stepping on it several times, and each time the
displayed weight was exactly the same. He then said that if we waited
for a period before stepping on it then the display would show a
different weight; he wasn't sure how long, exactly. This demonstration
took a surprising turn when his wife weighed herself on the scale, and
then he weighed himself again, and then the display showed 0.7kG
difference from his previous weight.
After a lot of experimenting, we worked out what was going on. In actual
fact, the scale was less consistent than most mechanical bathroom
scales. The displayed weight varied over a range of about 1kG despite
having a display resolution of 0.1kG. However this inconsistency is
hidden by a filter in the electronics.
The scale tries to work out if the person that had stepped on the scale
previously was stepping on the scale again, and will then display
exactly the same weight. It does this by remembering the previous
person's weight, and if the weight of the next person to stand on the
scale seems to differ by less than about 1kG, then it's probably the
same person, so it displays exactly the same weight as previously. You
can defeat this simply by having somebody else stand on the scale for a
moment so that the scale forgets the previous person's weight.
We probed this behaviour by stepping on the scale, noting the display,
and then stepping on the scale again while holding a weight in the hand,
and found that if the weight that we held was less than ~1kG, then the
displayed weight would be the same as before, but if it was greater than
~1kG, the displayed weight would change. The exact threshold weight was
a bit difficult to determine because of the inconsistency. We knew that
the scale would eventually forget the stored weight, if we waited long
enough, but we haven't worked out how long.
Presumably, the whole idea of this is to prevent the consumer from
discovering how inconsistent the Salter electronic bathroom scales are.
If people immediately discovered that the expensive scale that they'd
bought was less consistent than a mechanical scale, they might take it
straight back to the shop. Also, anybody doing a review of the scale for
a publication would probably think to check how consistent the scale
was, but they'd probably not discover the stored weight trick in the
time they had available.
I'm tempted to have a word with Trading Standards to see what they think
about this.
--
Farry
Mark Willstatter
Mea culpa - it was a case of me reading what I wanted to read, not
what was written. I was just *sure* there was a re-zero in there
somewhere ;^) So let me propose and alternate explanation, one that
has been suggested before although not in quite the same way:
software. The only way I can expain this is if the designers of this
instrument felt they needed to compensate for possible drift - in the
strain gauge bridge circuit, amplifier, A/D converter, wherever. IOW,
they knew users would be disconcerted about watching the displayed
weight change by itself as the electronics drifted and so included
provisions in the software to distinguish between slow changes that
would presumably be drift and fast changes that would presumably be
caused by weight being added or removed by a human. When the weight
was added very slowly, it was interpreted as drift and so the zero
was, in fact, changed. How's that sound? My advice: don't do that!
Congratulate yourself on having successfully simulating electronic
drift in this experiment. If I'm right (a big "if", I'll grant you)
then the scale will be fine in normal use.
- Mark W.
Farry
>Mea culpa - it was a case of me reading what I wanted to read, not
>what was written. I was just *sure* there was a re-zero in there
>somewhere ;^) So let me propose and alternate explanation, one that
>has been suggested before although not in quite the same way:
>software. The only way I can expain this is if the designers of this
>instrument felt they needed to compensate for possible drift - in the
>strain gauge bridge circuit, amplifier, A/D converter, wherever. IOW,
>they knew users would be disconcerted about watching the displayed
>weight change by itself as the electronics drifted and so included
>provisions in the software to distinguish between slow changes that
>would presumably be drift and fast changes that would presumably be
>caused by weight being added or removed by a human. When the weight
>was added very slowly, it was interpreted as drift and so the zero
>was, in fact, changed. How's that sound? My advice: don't do that!
>Congratulate yourself on having successfully simulating electronic
>drift in this experiment. If I'm right (a big "if", I'll grant you)
>then the scale will be fine in normal use.
I guess you missed the first post in this thread:
http://groups.google.com/groups?threadm=8g5710dtfmirt9hupk5s247ofktfvpvgd1%404ax.com
--
Farry
Mark Willstatter
Actually, I didn't - but it's been so long, I forgot! So I guess we
wasted a lot of time here. What's the problem?
Ron Bean
Farry <davefa...@OMiTTHiSyahooANDTHiS.co.uk> writes:
>If people immediately discovered that the expensive scale that they'd
>bought was less consistent than a mechanical scale, they might take it
>straight back to the shop.
I don't know what's "expensive" for a scale, so I took a quick
look at an industrial catalog. It shows small mechanical scales
for $80, electronic scales for $200, and "counting scales" (which
require high repeatability for counting parts) for $900.
It would be interesting to know what the difference is between
the $200 scale and the $900 scale. Do these scales have any
moving parts, or is it all in the electronics?