Hardware
The instruments, and the numbers they are allowed to produce.
Field water testing is the one place in envelope work where a number decides an argument. We build the instruments that produce the number — the control boxes, the sensors and the leak detection — and the app that runs the procedure and writes the report.
Running today
The app that runs the test.
An ASTM E1105 water penetration test, run from an iPad: the timer, the pressure hold, the cycles, the sensor log, and the finished PDF at the end of it. It installs to the home screen, it opens with no signal, and the control box carries its own Wi-Fi so the iPad can drive it on a roof with nothing else present.
It is in use and it is not yet a public product — access is by arrangement while the first labs run it. Ask to see it on a real job.
The boxes In development
One runs the test. One watches it.
They look similar and they do opposite jobs. Which one you want depends on whether you are the agency performing the test or the person who needs to know what the test actually did.
Control box
Runs the test
Drives the blower and holds chamber pressure through an ASTM E1105 procedure — the ramp, the hold, the cycles — while the iPad logs it. The box carries its own Wi-Fi, so it works on a roof with no building network and nothing else present.
For the testing agency. With the certified sensor option it is the instrument of record for an AAMA 502 or E1105 test.
Witness box
Watches a test it is not running
A handheld that reads the same test independently and shows you what is actually happening — pressure, timing, whether the hold held. It produces an on-site preliminary report before anyone leaves, so the conversation happens while the rig is still up.
For quality control and for forensically finding where water is getting in. Informational, and labelled that way on everything it produces.
Sensor configurations
The tolerance decides it, not the price.
ASTM E1105 §6.2.3 requires the pressure instrument to read within ±2 %, or ±2.5 Pa, whichever is greater. At a 3 psf test that allowance is about ±2.9 Pa, and “whichever is greater” means a percent-of-full-scale instrument only qualifies if its full scale is small. That one sentence sorts the sensors into two groups, and it is why the control box is offered two ways rather than one way with a discount.
The sensors, in numbers
What is actually inside each box.
Published figures from the manufacturers, and our arithmetic against E1105 §6.2.3. Check any of it before you believe it.
| Standard sensor | Certified sensor | Reference gauge | |
|---|---|---|---|
| Sensor | Sensirion SDP810-500Pa | Setra 264, accuracy code F | TEC DG-1000 |
| Range | 0–500 Pa (10.4 psf) | 0–3 in. w.c. (0–747 Pa, 15.6 psf) | Full test range |
| Accuracy | 3 % of reading | ±0.25 % full scale = ±1.9 Pa | 0.9 % of reading |
| At a 3 psf test allowance ±2.9 Pa |
Outside tolerance | ±1.9 Pa — passes | ±1.3 Pa — passes |
| At a 15 psf test | Out of range entirely | Passes | ±6.5 Pa — passes |
| Interface | Digital, I²C | 0–5 V, read on a 16-bit ADC | USB, Ethernet, Wi-Fi |
| Calibration certificate | No | Yes — NIST traceable, ANSI Z540-1 | Yes |
| Recalibration | Zero check before every test | Zero before every test; 12-month traceable span | Four years |
| Use | The witness box, and any control box not used as the instrument of record | A control box used as the instrument of record for AAMA 502 and E1105 | On the same tap as the record instrument |
One number explains the whole table: the Setra’s long-term drift is about 3.7 Pa a year on this range — larger than the entire allowance at 3 psf. That is why a zero check before every test is part of the procedure and not a recommendation.
What makes it defensible
The calibration record, not the brand on the enclosure.
E1105 §6.1 permits any arrangement of equipment that performs the procedure within tolerance. So what decides whether a report survives a challenge is not who built the box — it is the paperwork behind the sensor on the day.
- A zero check before every test. Not optional. The transmitter drifts about 3.7 Pa a year, which is more than the entire allowance at 3 psf.
- Traceable span calibration every twelve months on the transmitter and the handheld.
- Six months on the spray rack, by catch box, which is the one interval E1105 does fix (§9.1.1).
- The certificate ships with the unit, and the report states the date of last calibration and where the records are kept.
Neither standard sets an interval for the manometer. The schedule above is what a competent accredited shop writes into its own quality manual, and it is what we build to.
Leak detection In development
One hub, one pad per spot, and a head to suit the place.
Water arrives differently at a joint, a deck and a sill, so the sensing end is a separate part. The pad is the same everywhere; you change the head to match what you are watching.
Leak hub
One per site
Collects from every pad on the job and reports into the same record as the test. It carries its own Wi-Fi network, so a monitored area needs no building network and no conversation with anybody’s IT.
Verified on real boards, October 2026. Case and field build in progress.
Leak pad
One per spot to watch
The battery-powered unit that sits where the water is expected. It takes any of the heads below. Sealed case, printed in ASA for UV and heat, with a moulded TPU gasket at the lid.
Verified on real boards, October 2026.
Heads — the sensing end
Rope
A ten-foot sensing rope for a joint, a transition, or any long thin place where a single point would miss it.
Rain plate
A plate for a deck or a sill, where the water could arrive anywhere across a span rather than at one point.
Copper tape
Adhesive conductive tape for following an awkward line — around a penetration, or along something that is not straight.
Options and replacement parts
| Part | What it is | Fits |
|---|---|---|
| Sensing head | Rope, rain plate or copper tape | Any pad |
| Battery | 3.7 V 1100 mAh lithium, protected cell | Any pad |
| Sealed power switch | IP67 toggle, 5 A 28 VDC, with a sealing boot | Any pad |
| Lid gasket | Moulded TPU, replaceable | Pad and hub cases |
| Case | ASA, UV and heat resistant | Pad and hub |
Those are the parts meant to be replaced in the field. A head damaged on a job should not cost you the unit.
Where it actually stands
What is running, and what is still on the bench.
- The E1105 app runs today, on real jobs, by arrangement.
- A bench control box has run a full test program from an iPad — on its own simulated chamber. The real pressure sensor and fan controller are being fitted.
- Leak pad and hub are verified on real boards. Cases and field builds are next.
- No box is for sale yet. No price, no lead time, no order page — because there is nothing to ship this week and we would rather say so.
Bring us one test.
Show us a job you are testing next month and we will show you the app running the procedure and the report that comes out of it.