A recording pressure gauge is a pressure measurement instrument that draws a continuous trace of pressure versus time onto a paper chart. Its central component is a Bourdon tube or diaphragm element, which deflects in proportion to applied pressure and drives a linked pen arm across a clock-driven chart. The instrument exists to produce a permanent, time-stamped pressure record for processes where a snapshot reading is not enough — boiler operation, hydrostatic testing, autoclave validation. The outcome is a defensible chart that proves a pressure profile happened, the way a Foxboro 40 or Bristol Babcock recorder has done in plants since the 1920s.
Recording Pressure Gauge Interactive Calculator
Vary Bourdon tip travel and linkage geometry to see the resulting pen displacement, gain, and angularity loss on the chart.
Equation Used
The Bourdon tube tip moves by delta_tip. The pen arm multiplies that motion by R_arm / R_link. Any departure from the ideal 90 deg link angle reduces the effective travel by cos(theta_err), which is the angularity error that can make the chart trace bow.
- Bourdon tube tip deflection is proportional to applied pressure.
- Pen friction, backlash, ink drag, and temperature zero shift are ignored.
- theta_err is the deviation from the ideal 90 deg link angle.
- All length inputs use the same unit system.
Inside the Recording Pressure Gauge
The mechanism is simple in principle and unforgiving in practice. Process pressure enters a sensing element — usually a Bourdon tube, sometimes a diaphragm capsule for low ranges or a bellows for vacuum work. The element deflects by a known angular amount per psi, and a multiplying linkage converts that small deflection into a long pen-arm sweep across a circular or strip chart. A spring-driven or electric clockwork rotates the chart at a fixed rate — 24 hours per revolution is standard for boiler logs, 7 days for slower batch processes — and the inked pen leaves a continuous pressure-time trace.
Why build it this way? You want a single physical document that survives a power failure, a lawyer's subpoena, or a pharmaceutical audit. A digital data logger gives you the same information, but a paper chart from a Bristol Babcock or Foxboro recorder is admissible evidence in a way a CSV file isn't, and that's why hydrostatic test crews and ASME pressure-vessel inspectors still specify them.
If the linkage geometry is wrong the trace lies. The pen arm has to be set so the link between the Bourdon tip and the pen pivot stays close to 90° at mid-scale — get it 10° off and you introduce a sinusoidal error that reads high at one end of the chart and low at the other. The Bourdon tube must be temperature-compensated or zeroed at process temperature, because a 20°C ambient swing shifts a typical C-tube zero by 0.2-0.4% of span. Common failure modes are pen friction (drags the trace 1-2% low on rising pressure, then snaps up), dried ink starving the pen on slow chart speeds, and clock drift on mechanical movements that haven't been wound. The gauge must meet an ASME B40.100 accuracy class — typically Grade A (1% of span) or Grade 2A (0.5%) — and you verify that on a deadweight tester, not against another gauge.
Key Components
- Bourdon Tube Sensing Element: A flattened C-shaped or helical tube that uncurls slightly under internal pressure. A typical 100 mm C-tube gives 4-6 mm of tip travel at full scale, and the tube material (phosphor bronze for ≤600 psi, 316 stainless for higher ranges or aggressive media) sets both the range and the hysteresis, which should stay below 0.1% of span.
- Multiplying Linkage and Pen Arm: A pivoted lever and adjustable link that converts 4-6 mm of tube tip travel into a 100-150 mm pen sweep across the chart. The link length sets span, the pivot position sets linearity, and the geometry must be set with the link near 90° to the arm at mid-scale to keep angularity error under 0.25%.
- Clockwork Chart Drive: A spring-wound or synchronous-motor drive that rotates the circular chart at a fixed rate, commonly 24 h, 7 day, or 31 day rotations. Mechanical drives must hold ±2 minutes per 24 h to keep the time axis trustworthy on a boiler log.
- Inking Pen and Reservoir: A capillary-fed pen that maintains continuous contact with the chart at 3-8 grams pressure. Too light and the pen skips on rising pressure events, too heavy and chart drag biases the trace 1% low on the upswing.
- Paper Chart: A pre-printed circular or strip chart calibrated in the engineering unit of the gauge. Chart paper must match the gauge model — a Bristol Babcock chart in a Foxboro recorder will misalign by 1-3% because the radial scale is non-linear and instrument-specific.
- Process Connection and Snubber: A 1/4 NPT or 1/2 NPT connection with a porous-disc or piston-type snubber to damp pulsation. Without a snubber, pump pulses at 5-15 Hz will hammer the Bourdon tube and fatigue-crack it within months on a reciprocating compressor outlet.
Where the Recording Pressure Gauge Is Used
You see recording pressure gauges anywhere a regulator, an insurer, or a customer demands proof that a pressure stayed inside a window for a defined time. Process control has largely moved to digital data historians, but the paper chart survives where the record itself has legal or regulatory weight, where the environment kills electronics, or where an old plant simply works fine and nobody is going to rip out a Bristol Babcock recorder that's run for 40 years. They show up across steam systems, hydrostatic testing, gas distribution, and batch process validation.
- Power Generation: Boiler drum pressure logging on a Babcock & Wilcox industrial boiler using a 24-hour Bristol Babcock circular chart recorder for shift-handover records and insurance compliance
- Pipeline and Gas Distribution: Custody-transfer pressure recording at natural gas regulator stations using Barton 202E differential and static pressure recorders, where charts are kept for 2 years per FERC requirements
- Hydrostatic Testing: ASME Section VIII pressure vessel testing with a Crosby or Ashcroft 1188 recording gauge documenting the 1.3× MAWP hold for 30 minutes during third-party inspection
- Pharmaceutical and Sterilisation: Autoclave cycle validation on a Steris Amsco 3023 sterilizer where the chamber pressure-time trace forms part of the batch release record under 21 CFR Part 11
- Fire Protection: Standpipe and fire-pump weekly pressure tests where NFPA 25 inspectors review the recorded chart against the system design pressure
- Refrigeration and HVAC: Ammonia compressor suction and discharge logging on a Vilter VSS rotary screw compressor where the recorded pressure ratio confirms valve and seal health
The Formula Behind the Recording Pressure Gauge
The core relationship the practitioner cares about is how Bourdon-tube tip travel maps to pen displacement on the chart. Get this right and the trace is linear. Get it wrong and it bows. At the low end of the typical operating range, around 10% of span, a small zero error or pen friction dominates the reading. At full scale, the link angularity error dominates. The sweet spot — where the gauge is most accurate — is between 25% and 75% of span, which is exactly why every gauge selection guide tells you to size the gauge so normal operating pressure sits at two-thirds of full scale.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Lpen | Pen displacement at the chart radius | mm | in |
| δtip | Bourdon tube tip deflection at applied pressure | mm | in |
| Rarm | Pen arm length from pivot to pen tip | mm | in |
| Rlink | Distance from arm pivot to link attachment point | mm | in |
| θerr | Angularity error of the link from 90° at the operating point | degrees | degrees |
Worked Example: Recording Pressure Gauge in an Ashcroft 1188 hydrostatic test recorder
A small CNG fuelling station builder in Calgary is qualifying an Ashcroft 1188 recording pressure gauge for hydrostatic testing of 5,000 psi composite cylinders. The gauge has a 0-7,500 psi Bourdon tube giving 5.0 mm tip deflection at full scale, a 120 mm pen arm, and a 12 mm link radius. The technician needs to know the pen displacement at 5,000 psi (the test pressure), at 1,500 psi (the low-end approach), and at 7,200 psi (the high-end overshoot during pump shutdown), and how angularity error affects each reading.
Given
- δtip,FS = 5.0 mm at 7,500 psi
- Rarm = 120 mm
- Rlink = 12 mm
- θerr at mid-scale = 0 degrees
- θerr at full-scale = 8 degrees
Solution
Step 1 — compute the linkage multiplication ratio. The pen arm is 10× longer than the link radius, so each mm of tip travel produces 10 mm of pen tip travel before angularity correction:
Step 2 — at the nominal test pressure of 5,000 psi (66.7% of span), tip deflection scales linearly with pressure, so δtip = 5.0 × (5000/7500) = 3.33 mm. Compute pen displacement near the sweet spot of the chart where angularity error is essentially zero:
That places the pen at exactly the 5,000 psi mark on the chart — readable to about ±50 psi by eye, which is the accuracy this Grade 2A gauge is rated for.
Step 3 — at the low end of the typical test approach, 1,500 psi (20% of span), tip deflection is 1.0 mm:
At this point a 0.1 mm pen-friction step represents 1% of the displayed reading — pen drag and ink starvation dominate the error budget here, which is why nobody trusts the bottom 20% of a recording gauge for compliance work.
Step 4 — at the high-end overshoot of 7,200 psi (96% of span), the link has rotated to 8° off perpendicular, so angularity error kicks in:
That's a 0.5 mm shortfall — about 75 psi low at the top of the chart. On a hydrostatic test where you must prove the cylinder held 1.3× MAWP, that 75 psi is the difference between a passing and a borderline-failing record.
Result
The pen sits at 33. 3 mm from chart zero when the test pressure hits 5,000 psi, which the inspector reads against the printed scale to confirm the cylinder reached test pressure. At 1,500 psi the pen is at 10 mm and pen friction is the dominant error source — the trace will read 1-2% low on the rising approach. At 7,200 psi the pen reads 75 psi low because of 8° link angularity, so size your gauge so peak pressure stays below 75% of span. If your measured trace differs from the predicted curve, look at three things first: (1) a Bourdon tube zero shift from over-pressure permanent deformation will offset the entire trace by a constant amount, easily checked by venting and confirming the pen returns to zero; (2) a loose link-pin lock screw will make the trace shift unpredictably between cycles and is the single most common service issue on Ashcroft and Crosby gauges; (3) chart paper from the wrong model number — a 0-10,000 psi chart in a 0-7,500 psi instrument — will read 33% high across the entire span and is a known field error on multi-gauge test rigs.
When to Use a Recording Pressure Gauge and When Not To
A recording pressure gauge is one of three competing ways to capture a pressure-time history. The choice between a mechanical chart recorder, a digital paperless recorder, and a standalone pressure transducer with a data logger comes down to where the legal weight of the record sits, what the ambient does to electronics, and how much you're willing to pay for the front end versus the back end.
| Property | Recording Pressure Gauge | Paperless Digital Recorder | Pressure Transducer + Data Logger |
|---|---|---|---|
| Typical accuracy | 0.5-1% of span (ASME B40.100 Grade 2A/A) | 0.1-0.25% of span | 0.05-0.25% of span |
| Sample / trace resolution | Continuous analog, ~1 mm chart resolution | 0.1-1 Hz typical, configurable to 100 Hz | 1 Hz to 10 kHz depending on logger |
| Initial cost (single point) | $400-$1,200 | $1,500-$6,000 | $200-$2,000 plus host |
| Power requirement | None (spring clock) or 120 VAC | 24 VDC or 120 VAC | Loop power or battery |
| Record format | Paper chart, signed and dated | Encrypted digital file with audit trail | CSV or proprietary binary |
| Regulatory acceptance | ASME, NFPA 25, FERC, legacy 21 CFR Part 11 | 21 CFR Part 11 with validated software | Requires separate validation package |
| Service life of mechanism | 10-20 years with re-inking and clock service | 8-12 years before display/electronics obsolescence | 5-10 years sensor drift |
| Best application fit | Boiler logs, hydrostatic tests, hard-copy compliance | Continuous batch records, networked plants | R&D, transient capture, custom systems |
Frequently Asked Questions About Recording Pressure Gauge
That's classic angularity error in the multiplying linkage. The link between the Bourdon tip and the pen-arm pivot has to sit at 90° to the arm at mid-scale. If it's set at, say, 80° at mid-scale, the geometry produces a sinusoidal deviation — high at one end, low at the other, zero in the middle.
Fix it by adjusting the link length first to set span at full scale, then the pivot position to set zero, then iterate. On Ashcroft and Crosby movements there's an angularity adjustment screw specifically for this. Most field calibrators skip the iteration and live with a 1% bow.
Pick a 0-10,000 psi gauge, not 0-7,500. Two reasons. First, you want normal operating pressure between 25% and 75% of span — your 4,000 psi sits at 40% of 10,000, right in the sweet spot. Second, pulsation and spikes can hammer a Bourdon tube into permanent set if the tip routinely sees over 75% of full scale. A 7,500 psi gauge would see 6,500 psi at 87% of span on every pump start and the tube will work-harden and shift zero within a year.
The trade-off is that you lose resolution at 4,000 psi — your reading uncertainty roughly doubles. That's the price of headroom, and on a recording instrument where the chart is the legal document, headroom wins.
Almost always pen-and-chart interaction, not a real pressure event. The most common cause is a sticky pen pivot or a chart paper with a raised seam that the pen catches on once per revolution. The pen builds up a small spring force, then snaps free, then catches again — sawtooth.
Check the pen pivot for ink crust first (clean with isopropanol, not water, on a Bristol Babcock or Foxboro pen). Then inspect the chart for a seam or tape mark. If the sawtooth period matches the chart rotation period exactly, it's a chart defect. If the period is irregular, it's pen friction.
No. The mechanism's frequency response is limited by the mass of the pen arm and the inertia of the linkage — typically 1-3 Hz at best. Water hammer is a 50-500 Hz event. The chart will show a single rounded blip that under-reads the true peak by 50-90%.
If you need to capture transients, use a piezoelectric pressure transducer with a 10 kHz logger. Use the recording gauge for the steady-state record and the transducer for the spike. They serve different purposes and a working test rig usually has both.
Ambient temperature swing acting on the Bourdon tube. A phosphor-bronze C-tube has a thermal coefficient of roughly 0.02-0.04% of span per °C. On a 0-10,000 psi gauge that's 2-4 psi per °C, so a 15°C shop swing from 18°C overnight to 33°C in the afternoon sun gives you 30-60 psi of zero drift.
Three fixes, in order of cost: shade the gauge from direct sun, switch to a temperature-compensated bi-metallic movement (Ashcroft offers this on the 1188), or move to a stainless-steel diaphragm element with lower thermal coefficient. If the gauge is in an outdoor enclosure on a gas regulator station, expect the drift and design the test procedure around morning readings only.
Mechanical clockwork drift. Spring-driven 24-hour movements on older Bristol Babcock and Foxboro recorders are specified at ±2 minutes per 24 hours when fully wound, but drift to ±10-15 minutes as the spring unwinds toward the end of the cycle. If your test ran late in the chart period, the time axis compresses.
Two fixes. Wind the clock at the start of every test, not on a calendar schedule. Or replace the spring movement with a synchronous-motor drive — most recorders accept a drop-in 120 VAC clock motor that holds time to the line frequency. For pharma or custody-transfer work, the synchronous drive is effectively mandatory.
For a Grade 2A or better recording gauge used in a compliance role, you need the deadweight tester. A master gauge inherits its own uncertainty plus the comparison uncertainty, and you stack errors fast. A deadweight tester generates pressure from first principles — mass on a known piston area — at 0.025% of reading uncertainty, which is the only way to certify a 0.5%-of-span instrument with margin.
For a Grade B or lower utility gauge on a non-compliance application, master-gauge comparison is fine. The rule of thumb is the calibration standard should be 4× more accurate than the unit under test.
References & Further Reading
- Wikipedia contributors. Pressure measurement. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
