Operating and Output Voltage of Loadcell  

RE: Operating and Output Voltage of Loadcell

I don't know if this will help you, but this is what I found after scratching around.

This is what came up when I asked Chat GPT:

The component in your photo is not an ‘IC’ or sensor chip, but rather a

➡️ foil-etched strain gauge (SG), 4-grid full bridge – exactly the type of component that is attached to a metal spring in a load cell / force or bending sensor.

Why is it almost certainly a 4-grid strain gauge?You can see four identical measuring grids (2 at the top, 2 at the bottom), typical for a full bridge for bending or force measurement. Such ‘four grid full-bridge strain gauges’ are offered exactly as shown. tst-scale.com+2tst-scale.com+2The fine, parallel conductor tracks on brown carrier foil are the classic design of a metal foil strain gauge.Four solder pads in the middle = 4 connecting wires of a Wheatstone full bridge (E+, E-, S+, S-).Meaning of the markings ‘T0’ and ‘W’

For many Asian strain gauge manufacturers (BF/BA/BM/BY, ZF, ZYM series, etc.), the printed letters/codes have a standardised meaning:

In catalogues from Zemic, Materontek & Co., codes such as T0, T2, T4, T6, T8, etc. appear as so-called ‘creep / STC codes’ – they indicate the combination of temperature and creep compensation, matching the material of the spring (steel, aluminium, etc.). zemic.co.za+2Zemic Europe+2A Chinese overview explicitly mentions:‘T0: Temperature self-compensation’,‘W: Welding Wire: Silver, Enameled Wire, Cable’ – W therefore stands for the variant with soldered connection wires. WayDigi Site Name

This fits your component perfectly: “T0” at the top and ‘W’ on the right, plus the soldered copper wires.

Exact type?

Unfortunately, I cannot read the exact type (e.g. ‘BF350-3FB-T0-W’ or similar) from the photo because:

the actual type code (BF..., BYM..., ZF...) is normally printed below/on the side – in your case, it seems to be covered by silicone and the wires,and there are several 4-grid full bridges with similar dimensions in the catalogues (backing e.g. 9.8×7.3 mm, 10.0×6.5 mm, etc.). zemic.co.za+1

However, you are certainly correct with ‘4-grid strain gauge / full-bridge strain gauge’.

How you could continue searching / replacing

If you want to find a replacement part or data sheet, search the internet for combinations such as:

four grid full bridge strain gauge T0 WBF350 4FB T0 strain gaugeZemic full bridge strain gauge W T0

and compare the backing size (~11.5 × 7.3 mm) and the grid arrangement in the catalogues with your photo. tst-scale.com+1

RE:

Prusa has published the LoveBoard schematics; see attached. The strain gauge in the load cell sensor is driven by an HX717 differential amp with a 5V supply voltage.

You will need a similar differential amp, because the resistive changes are very small, the absolute resistance values in the strain gauge will drift with temperature, and there may be electrical noise interfering with the individual measurement channels. So the bridge circuit and differential amplifier are a must. 

RE: Operating and Output Voltage of Loadcell

Posted by: @jurgen-7

and there may be electrical noise interfering with the individual measurement channels.

Yup, like the fan and/or heater PWM signal? 

RE: Operating and Output Voltage of Loadcell

Posted by: @hyiger

Yup, like the fan and/or heater PWM signal? 

Yes, it seems that the heater PWM in particular has proven to interfere with the load cell in some Core One units, even with the added robustness of the differential sensing.

In my own experiments with strain gauges (not related to 3D printers), I have found that they are great at picking up all kinds of noise. Not only do you want to use a bridge sensor and differential amplifier, but also make sure that you keep your connections short and balanced, twist them where feasible, and provide ground planes and shielding where the arrangement allows. 

RE: Operating and Output Voltage of Loadcell

Posted by: @b3n

Hi there,

does anyone know the actual operating voltage of the load cell, supplied on the Loveboard?

Further more the output voltage range of the load cell is interesting or better the correct amount of trigger force on the load cell of MK4/XL/CoreOne, etc.

I plan to convert the analog output via voltage comparator to a digital on/off signal, trying to use the load cell on boards like the Mini+ board.

Thanks in advance to the community.

The Loveboard load cell operates at 5V. Its output is a small mV signal relative to force, so a comparator would need precise calibration against your specific trigger threshold.

RE: Operating and Output Voltage of Loadcell

Thank's for the answers, I guess I have to do my own circuit (amplifier and comparator) because the HX717 only have a clocked serial data output, which makes sense because there is a testing range for the load cell which says IO or NIO.

I believe the buddy board for the Mini only has high or low signal for the PINDA sensor. 

RE: Operating and Output Voltage of Loadcell

Posted by: @b3n

Thank's for the answers, I guess I have to do my own circuit (amplifier and comparator) because the HX717 only have a clocked serial data output, which makes sense because there is a testing range for the load cell which says IO or NIO.

I believe the buddy board for the Mini only has high or low signal for the PINDA sensor. 

Even with the differential amplifier, be prepared for some drifts in the load cell signal. So a simple comparator with a fixed threshold is probably not going to cut it. If you only need to detect sudden changes (from touching the bed), an AC-coupled signal might do the trick. Or use a small microcontroller with the HX717 (or similar amp) to let you do smarter filtering, sliding averages etc. in software.

The integrated differential amplifier & ADC chip, which delivers a digital signal right away, is probably desirable -- less need to fight analog noise.

RE: Operating and Output Voltage of Loadcell

You want to replace magnet sensor with load cell sensor on the Mini?

RE: Operating and Output Voltage of Loadcell

Yes that's the plan when the CoreOne gets the INDX and the MK4S/CoreOne toolhead is replaced.

RE:

So I searched a while on different sites for this topic, like Jürgen said I think I will work with the HX717 or a complete HX711 board (both 24Bit) with stabilised voltage and all the other components ready to go.
As a microcontoller the Arduino beetle board would be nice, got some interface for serial data from HX711/HX717.
The library for the HX711 is already done by some guys with similar needs, should also work for the HX717 as it's only about receiving and filtering 24Bit serial data.
However I already have a Raspberry Pico 2W on my Mini, which doesn't have much to do, maybe that's a more hardware thrifty way.
There is already a Python program, the documentation mentioned 10 SPS.
This could lead into problems, I think Z homing and Z probing speeds are above 60mm/min (1mm/sec), therefore measurement accuracy would be >0,1mm, leads to nonsense, so Arduino will do it.