Load Cell Wiring

You cannot use suggested pattern since if you have 2 blocs on a diagonal and 2 load-cells on the other you measure (if the blocs allow it only) only the disbalance of the load versus the bloc diagonal.

For a platform you need at least 3 load-cells to get the whole load on it. The most common way is to use 4 load cells one at each corner.

I tried to send you a link with the way the SBH2 load cell has to be put in place but it did not work. Try google with "mettler toledo SBH2" as question and you will find all necessary informations.

Hello Nick name,

I do agree that 4 load-cells are preferable, but i just used (compulsery) this diagonale scheme for truck's platform. Of course, false load cells shoudn't be absolutely rigid, ideally their stiffness factor should be the same as for operating load-cells. Mechanical system has to be adjusted maximum accurately. Otherwise when an operable load-cells are placed on one side and false load-cells on another --- response of scale will depend on where loading is placed.

One more option these two load cells might be used for pallete scale.

I just came across the Metler-Tolledo ocassionaly as 'freelance" repair man (truck scale, hook scale) with similar set of equipment. They have "good" level of protective measures and refused to work if:

(1) their ADC control module did not get proper serial code of a proper load-cell.

(2) if factory preinstalled parameters of load-cell's bridge were drifted significantly from initial levels.

First what I did, I sorted out all load cells properly between modules, some of them appeared as quite operable. Couple of them showed "Error" message on display. I've measured level of Load cell's output and found its zero drifted some more than 10-15 mV. I made external regulating circuit (set of resistors) and corrected zero level to 0.5-1 mV. This way I could hack these two sets of load cells and modules.

Metler Tolledo dealer(at least in Kazakhstan) had nothing to suggest for fixing the problem unless to send these broken devices buck to US for crazy cost (warantee was expired).

regards, caramba

The reason why a load-cell is rejected if zero voltage is too high comes from a sensor behaviour. A high zero voltage can normally appear ONLY if the load-cell has been overloaded (provided that no electrical or strain-gage problem occurred) which means in the sensor body some trans-elastic zones appear. Such zones compromise sensor's linearity and modify the calibration (original) values. The sensor MUST be not only centered by electrical means but also re-calibrated if not results are WRONG!!!

Especially in weighing where the result is part of a commercial transaction this can be a problem either for the seller or for the buyer.

With respect to the usage of 2 load cells and 2 "blocks". The arrangement you suggest gives results very depending on the COG position even if stiffnesses are the same. Errors can be very important and are not acceptable for commercial weighing (at least for usual tolerances). It can be used ONLY if the errors mentioned above are small in comparison to the weighing tolerances accepted by the law in the country where the system is used. I doubt that ISO Norms will allow it.

Hello, Nick Name,

I do accept your point. Nevertheless, I would disagree that always zero voltage drift can be caused only load cells overloading. I've made hundreds of strain gauges(as I suppose you did too) and dozens were repaired. I've found a simple thumb rule(s):If you have only zero problems its likely caused by a transient unstable resistance in within and in outer of strain gauges' circuit. This instability could be caused the condenced moisture. After drying up the load cell and correction of zero level (by any means) there left a hope always that load cell is recovered. Otherwise if there'd been found significant change of sensitivity (as rule weak response)--- yes it's likely the mechanical overloading the cause and adhesive layer shearing problem, so load cell shouldn't be repaired (at least electrically).

In case with mentioned my experience to troubled Mettler devices I have removed from load cells, indication modules and cable connectors some a 3/4 glass of water. I was stunned that these devices were operable. Of course I checked out linearity of output by placing model static weights. All was ok and as i know remains so.

As for diagonal placement of load cells. I do agree it is far non perfect design. But I assume you would agree it possible in principle at least. Yes this mechanical system is significantly statically undetermined, so here's a risk always to get non-linearity and even dead zones in load cells summary response. But if there applied very accurate installing procedures and if there presumed to weigh masses in relatively narrow range in its variaton(+ or -20%) that pattern could be applied for(of course non commercial) industrial applications.

Last example from my specialization - belt scale. It should be banned as SCALE as there instead installed dozens of load cells (under every belt conveyor's idler) as rule is installed only the one or two. However preciously done installation make this task possible i.e. to wait more or less good accuracy performances even for commercial purposes sometimes, somewhere. It's real ordeal, but it's possible.

best regards,

caramba

I agree. With respect to moisture I noticed a strong and unstable loss of signal. If you got such a lot of water it means the gages were VERY well protected because if not you should have had a short circuit with either an over loaded supply or a zero output.

In general I found that zero shift at zero load is the consequence of load cell mishandling or misuse. I am not only a user but I am also involved in the design of special transducers and since many years.

I guessed so,i.e. that you had a great experience with this stuff just as a developer rather than user. And unfortunately, it's a bad tradition of my location's practice to misuse complex industrial equipment. Here's nothing to be proud.

So often I had been charged in here to make my work with an extremely scarce of needed tools to make this work as well done. I did a lot of hand made load cells in my projects not why i do like that so(you know it's a quite messy and long process), but by simply reason that either there was a lack of money to buy them or needed load cells were out of reach.

I've repaired also this or that way a broken branded load cells (Hottinger, Hasler, Mettler Tolledo and a lot of unknown for you made in Russia(SU))

As i said before I accept and respect your point as fair. I emphasized that what I said here in thread could be applied taking in account special emergency conditions.

Best regards, caramba.

I highly appreciate what you did. I know how difficult it is to fight against lack of means and money. It requires a lot of ingeniousity to make all from nothing! Engineers in the west usually do not appreciate what it means to call a supplier and have a delivery in 48h.

By the way I would be interested in your experience with Russians(SU) load-cells. I know that many were copied the modells being many times either cells from w-europe or from the US. Strain gages were made similar to HBM and some looked like japaneese. Since it is not a topic for all CR4 you amy use the channel of personal mails.

Regards

Nick Name

Oh it's a long story. Back to history, due to Soviet Union was quite an isolated country there was own (not too bad developed) industry which produce strain gauge. I'm sure we had good theory and engineering at this field. Do you agree we had at least quite significant achievements in such field as aviation. And you're aware that strain gauge based sensors is a key element during air planes developing. I have my hand books concerned strain gauge's and load cell design written in SU in 197x. It's very good one and work for me by now.

Nevertheless even in relatively good times we had quite poor in its quality load cells just as mass product. That's why in laboratory where i worked we've decided do not rely on unstable load cells for scales we developed and made very simple, very cheap and very reliable load cells (we made ones from used bearings).

As i know, contemporary load cells' producing in Russia oriented on internal market. Yes what they're making maybe looks like western's copies, especially i can say that about single-point, S load cells. And to be frank, their prices is not too attractive, as it was when that new companies started their business.

Take a look at this site(sorry no language option but Russian). It's more or less good known Russian producer. Eventually I'd dealt with their staff and can say it's quite decent one, but Mettler Tolledo's is better :).

regards, caramba

Open the case and see if there is a wiring diagram in "that thing you came across."

To use the Mettler summation box, which is what you are describing, you will need their Jaguar or similar operator interface. I have wired about a dozen of these, but still refer to the manual, which I have on CD, to connect the wires properly. Are you sure that two cells are dummies? They make floating, semi floating and fixed load cells for a 4 cell tank weighing system. The fixed unit is just that, quite rigid but it still can be compressed in the vertical axis. (you'll never see this movement - it's very minute) The semi floating will not pivot like the floating cells do. If you think you have 4 working cells and want to continue, get the operator interface and reply to me. I can help with the wiring. Of course, if you get the operator interface, you won't need my help, since you'll get the CD I am referring to.

Thanks for the advice...As you had presumed I do have the Hardware but no manual or any documents.

And actually I don't know how the connections to be made...

I tried to follow the markings in the connector now I have a display " O Kg."..

It seems I'd wired it wrong.

More Guide Lines?

Regarding the Blocks...It is dummy only and it is not wired.

I tried one load cell connecting direct to display bypassing the Junction Box...It was giving a display for a while then it went to "over load" as per a Mettlor Toledo instruction I find in Net. After the overload it continues to be in overload,Hope i didn't spoiled theload cell.

You hadn't spoiled the load cell this way. I'm sure [almost]. If you have seen "0kg" it's a good sign. Push buttons "zero" and then "tare" and try out to give loading upon the load cells. One by one and simultaneously. If your load cells is designed for significant loads (it could be up to 10 tons) the minimum load to be distinguished is some 20 kg (as i could recall).

Study the menu of display. There are some modes: 1st channel; 2nd channel; 1+2; (1+2)/2.

Unfortunately my desktop is broken now. I've stored there Mettler docs. So what I told here it's what I could recallect from 2003. If it'll be actual I'll try get data and send to you.

By means of multimeter you could measure excitation voltage (should be some 5V) then you can control output in mV. If mV output without load less than 2-3 mV you'll launch your stuff without problem.

Good luck