ARMAC DevNotes · pages: Overview · Design & Calculations · Throttle Pressure-Drop Tables · Build & Commissioning · Calibration
Calibration needs no loop instrumentation -- no inlet pressure sensor, no branch flow sensor. A bucket, a stopwatch, a graduated container of neat Barricade, and the existing DN50 main flow sensor cover everything. The Dosatron is the meter, so we never re-measure what it already does; we measure the gel drawn and the delivery, directly.
Two knobs, set in this order:
- Needle valve -> branch flow Q_side (target ~2000 L/H ~= 33 L/min)
- Dial -> ratio r, trimmed so delivered concentration = 1.4%
The needle valve comes first because the dial calibration meters into whatever branch flow it has set.
Step 1 -- Needle valve (bucket method)
Sets the branch flow. Measure the branch output before it rejoins the main.
- Run the system at the normal 120 L/min throttle setting.
- Suction on clean water (3-way -> flush) -- do not use gel for this step.
- Open the solenoid.
- Temporarily break the return at the check valve and route the Dosatron outlet into a bucket instead of back to the low-side tap.
- Time a known volume: 20 L in ~36 s = 2000 L/H; 10 L in ~18 s is the same. Repeat 2-3 times and average.
- Trim: open the needle valve for more flow, close for less.
- Reconnect the return.
The one correction to know
With the return open to a bucket, the branch discharges to atmosphere (~0 bar) instead of to the ~1.5 bar low side, so the test DeltaP is larger than in service and the bucket flow reads a bit high versus what you'll get reconnected.
Do not chase it with a gauge. Instead, aim slightly low on the bucket -- ~1800 L/H -- then reconnect and let Step 2 trim the dial. Because the dial is the real concentration knob, a small error in absolute branch flow washes out entirely.
Anywhere in ~1500-2500 L/H is acceptable; you compensate on the dial.
Step 2 -- Dial (drum-drawdown check)
Closes the loop against reality -- actual gel, actual delivery, including the small flow bump the branch adds. No branch sensor needed.
- Everything reconnected, running normally at the 120 L/min setting.
- Suction into a graduated container of neat Barricade.
- Run for a timed interval t (a few minutes -- long enough for a clear drawdown).
- Read the concentrate drawn DeltaV_gel off the graduations.
- Read the combined delivery Q_total off the existing main flow sensor (it reads the branch bump too, which is what we want).
- Compute:
C = DeltaV_gel / (Q_total x t)
- Trim the dial until C = 0.014 (1.4%). Re-run to confirm.
Worked example
DeltaV_gel = 5.0 L drawn t = 3 min Q_total = 130 L/min -> 390 L delivered over 3 min C = 5.0 / 390 = 0.0128 = 1.28% -> dial is low, raise it target DeltaV_gel = 0.014 x 390 = 5.46 L in 3 min so raise the dial by ~= 5.46 / 5.0 - 1 ~= +9%, then re-run to confirm
The check is self-correcting: read C, nudge the dial, re-read. Two or three passes lands it.
Step 3 -- Viscosity confirmation
During Step 2, confirm the piston primes and meters neat Barricade at ~1.7-2.0 L/min without stalling. If it labours, pre-dilute the concentrate to a pumpable stock and re-target the dial for the diluted strength (accepting the stock's shorter pot life).
Step 4 -- Flush verification
3-way -> clean water, run >= 2 min, confirm the branch and injector clear. Mandatory every shutdown -- set gel in the metering chamber is the primary failure mode for these units.
What you do not need
| Instrument | Needed? | Why |
|---|---|---|
| Branch flow sensor | No | Q_side is fixed by the needle valve; the drum-drawdown check makes its exact value irrelevant. Adds a failure point in a limb meant to be dumb and reliable. |
| Dosatron inlet pressure sensor | No (for calibration) | Concentration is verified by gel drawn / delivery, not by pressure. |
| Main flow sensor | Already fitted | Existing DN50 Hall unit supplies Q_total for Step 2. |
The design is deliberately open-loop: dial sets ratio, needle sets flow, both fixed at commissioning and left alone.
Recalibration triggers
Re-run Steps 1-2 if any of these change: the manual tap's normal setting, the sprinkler nozzle set (changes the low-side pressure and thus branch DeltaP), the Barricade product or dilution, or after any service that disturbs the needle valve or Dosatron. A quick Step-2 drawdown check alone is enough to confirm concentration between full recalibrations.