A digital load cell carries a microprocessor inside the cell body. Each one has its own address, and the indicator talks to it over that address rather than reading a millivolt signal. That single change — moving the conversion from the indicator into the cell — is what fixes several problems that analog systems can only work around.
Digital against analog: six practical differences
1 The electronics are protected
An analog cell is powered by the indicator, so the bridge excitation sits at whatever the indicator puts out. In an industrial environment that leaves it exposed to leakage currents between indicator and cell — unstable data at best, a damaged cell at worst. Digital cells are laser-welded under protective gas and rated IP68, with leakage protection built into the circuit so the output stays stable.
2 The system tells you when something is wrong
The indicator polls every cell in turn while the scale is working. When one stops answering correctly, the indicator raises an alarm by itself — you do not have to deduce a fault from a strange number.
3 Every cell reports separately
Each cell returns its own value instead of being summed into one analog signal, which raises resolution and keeps the displayed weight accurate. It also makes the cells independent: if one fails, the scale can often keep working.
4 Corner adjustment stops being manual work
Because the indicator reads each cell directly, you place a load on each corner once and it calculates the correction factor per cell and applies it automatically. No repeated trips round the platform. Replacing a cell does not force a full recalibration either — once the new cell is identified over the RS-485 bus, you enter its parameters and the system runs.
5 Long cable runs, fast response
The output is already digital, so it is far harder to disturb — harder, not impossible. Transmission uses a master–slave protocol with error checking, and where the data is encrypted it is difficult for an outside party to read or tamper with.
6 Lower cost of ownership
Fault-finding takes less time, so maintenance costs less. Diagnostics run automatically, and equipment changes do not trigger recalibration.
What happens inside the cell
- The strain measurement produces a small electrical signal, which goes to the amplifier. The gauge element is only about 8 mm long, mounted so that bonding faults and surface electromagnetic effects do not corrupt it.
- The analog signal then passes through a 24-bit A/D converter.
- The CPU computes and compensates the converter output.
What the CPU handles
- RS-485 output
- Precise zero handling
- Linearity correction
- Ambient temperature compensation
- Gravity compensation
- Creep compensation
- Hysteresis compensation
- Automatic return to zero
ISN: the number the cell actually sends
A digital cell does not report kilograms. It reports an internal count called ISN, running from 0 at zero load to a full-scale value — typically 30,000. What one ISN is worth in weight depends entirely on the capacity of that cell.
Once calibrated at full load, the boundary values are stored in the cell itself and hold for at least ten years.
Worked examples — what one ISN is worth
| Cell | Rated capacity | Calculation | 1 ISN = |
| BTA-D | 30 tf | 30,000 kgf ÷ 30,000 div | 1 kgf |
| BTA-D | 40 tf | 40,000 kgf ÷ 30,000 div | 1.333333 kgf |
| BTA | 60 tf | 60,000 kgf ÷ 30,000 div | 2 kgf |
| SQC | 100 kgf | 100 kgf ÷ 30,000 div | 0.003333 kgf (3.333 g) |
The pattern is the same every time: rated capacity divided by full-scale divisions.
How the indicator reads a weight
Example: BTA-D 30 tf, unloaded 2 div, loaded 1002 div
- Full-scale ISN for this cell is 30,000 div.
- Loaded − unloaded = 1002 − 2 = 1000 div.
- (1000 ÷ 30,000) × 30,000 kgf = 1000 kgf.
Polling one cell, then many
Every cell has its own address. To read one, the indicator or PC sends a request to that address — much like dialling a phone number. The sequence is simply: request cell #1, cell #1 answers; request cell #2, cell #2 answers; and so on round the platform.
If a request arrives out of order, or interference alters the data in transit, the protocol format is strict enough that the theoretical error rate is around 10 in a billion, and a check code guards the exchange. A digital cell simply will not respond to a malformed request, however many times it is repeated.
When no data comes back within the set time, the request is sent again. If the second attempt also returns nothing, that cell’s data is treated as abnormal and the indicator raises an alarm until normal readings resume.
Where digital systems do go wrong
Digital cells fail less often than analog ones, and when they do the cause is usually the connection rather than the cell — and most often an address problem. Check addressing before suspecting the sensing element.
Load cell note
KALA Technical Notes
Digital is not automatically the right answer — on a short cable run with a stable environment, analog cells still do the job for less money. Where digital earns its cost is long runs, automatic corner correction and diagnostics on a bridge you cannot easily get under. Tell us the platform layout and we will say which suits it.
● Digital and analog load cells, indicators and RS-485 accessories for truck and platform scales.