Multiwire proportional chamber

Georges Charpak's electronic readout avoided film scanning and full spark chamber discharge.^[ref]

Figure 1.
From gas ionization to an electronic wire address
. Original functional schematic; dimensions, pulse shapes and routing are illustrative. Wire cross-sections face the viewer. **Details**.

Opposite polarities identify the avalanche wire. Positive ion motion dominates integrated signal; detection precedes complete collection. Calibrated Electron travel times enable drift chambers.^[ref]

Table of contents

Multiwire proportional chamber experiment (1968)

 0  0

The original apparatus used

40 μ m

stainless-steel wires between meshes

7.5 mm

away, flowing argon with organic vapor, negative mesh voltage and individual electronic amplifiers. A

5.9 keV

X-ray source calibrated pulse height while voltage was varied.^[ref]

A coincidence circuit defined a narrow proton or pion beam. Scanning it across wires tested localization; the response edge matched the beam width, limiting the inference. Distant counts were attributed to accidental coincidences. Separate Electron scans measured pulse delay versus position, comparing gas mixtures and pulse-height-dependent electronics.^[ref]

Thresholds rejected small pulses, reducing detection efficiency. Wire-independence tests did not establish unlimited separation of overlapping tracks.^[ref]

Wire pitch ( $p$ )

 0  0

The centre-to-centre spacing

p

of adjacent wires. A wire address alone specifies a cell, rather than a continuous position.

Wire-address position model ( $σ_{x} = p / 12$ )

 0  0

Consider an original idealized track reconstruction model. Let equally spaced wires have positions

x_{k} = k p

and assign each event to the nearest wire, reporting

x = x_{k}

. Assume exactly one correct wire address, perfect detection efficiency, exact wire positions and a uniform distribution of true positions within each cell. These assumptions explicitly exclude missed pulses, false addresses and several unresolved tracks.

Writing

u = x - x_{k}

, the measurement error is

e = x - x = - u

, with

u

uniform on

[- p /2, p /2]

. Symmetry gives zero expected value. Its variance and standard deviation follow by integration:

E [e] = 0, σ_{x}^{2} = E [e^{2}] = \frac{1}{p} \int_{- p /2}^{p /2} u^{2} d u = \frac{p ^{2}}{12}, σ_{x} = \frac{p}{12} .

(1)

For chosen wire pitches

p = 2 mm

and

1 mm

this gives

0.577 mm

and

0.289 mm

respectively. These are model standard deviations, not measured chamber spatial resolutions. The largest possible absolute error in this model is

p /2

, which is a different quantity.

A useful subtlety is that the formula depends on the illumination. For an unresolved pencil beam fixed at

x = x_{k} + 0.4 p

, every reconstructed position is

x_{k}

. The repeated estimates have zero variance, but every error is

- 0.4 p

. Small repeat-to-repeat spread can coexist with a biased position. The uniform-cell result therefore cannot be transferred to an arbitrary beam profile or detector selection.

The wire address supplies one transverse coordinate in this model. Combining several separated measurements supports track reconstruction; a single address does not specify a complete trajectory. If two charged particles enter the same cell during one accepted interval, an address-only record can be identical to that from one particle. No reconstruction of that record can recover the lost multiplicity without additional information.

Figure 1.
Chosen wire-address errors and a calibrated two-wire response model
. Original teaching calculations, not measured data. Left: uniform-cell address error. Right: stipulated linear charge response with independent small channel noise; the noise estimate is a first-order approximation, not a chamber performance claim. **Details**.

Calibrated two-wire charge centroid ( $x = p q_{R} / (q_{L} + q_{R})$ )

 0  0

Can two readout values locate a hit between wires? Only with a specified response model. Place two readout wires at

0

and

p

, put

u = x / p \in [0, 1]

, and stipulate calibrated, noiseless electric charges

q_{L} = Q (1 - u), q_{R} = Q u, Q > 0.

(1)

These are assumptions for a toy sensor response, not the natural response law of adjacent anode wires. Under these assumptions the centroid

x = \frac{0 q _{L} + p q _{R}}{q _{L} + q _{R}} = p u = x

(2)

recovers position exactly. The apparent gain in precision comes from adding calibrated analogue information. Taking the mean of two fired wire addresses without their electric charges would instead give the fixed midpoint

p /2

For a chosen noise model, replace the two electric charges by

q_{L} + n_{L}

and

q_{R} + n_{R}

, with independent random variables

n_{L}, n_{R}

of zero expected value and common standard deviation

s

. Assume

s / Q ≪ 1

and no threshold censoring. First-order expansion of the ratio gives

δ x ≃ \frac{p}{Q} [(1 - u) n_{R} - u n_{L}], σ_{lin} (u) = \frac{p s}{Q} (1 - u)^{2} + u^{2} .

(3)

For stipulated

p = 2 mm

and

s / Q = 0.02

, this approximation gives

0.0283 mm

at the midpoint and

0.0400 mm

at either endpoint. The endpoint values describe uncensored signed channel readouts, including small negative values after baseline subtraction. They do not assume that both channels would pass a positive threshold. A ratio with a fluctuating denominator also need not have exactly zero bias; this is a local small-noise approximation.

Three original counterexamples show why calibration and selection matter:

If the right channel retains a relative gain $g = 1.10$ , then $x / p = gu / (1 - u + gu)$ . At $u = 1/2$ , the bias is $p /42 = 0.0476 mm$ for the chosen pitch, larger than the midpoint noise estimate above.
Suppose symmetric electronic leakage transforms the normalized right signal to $v = ϵ + (1 - 2 ϵ) u$ . Treating $v$ as $u$ gives a position bias $p ϵ (1 - 2 u)$ . A known $ϵ \neq = 1/2$ permits inversion; $ϵ = 1/2$ destroys position information entirely. This is a stipulated crosstalk model, not a measurement.
Requiring both noiseless channels to exceed $0.1 Q$ accepts only $0.1 < u < 0.9$ : $80%$ of uniformly distributed hits. Computing a position spread after that cut describes a selected sample. It does not establish improved spatial resolution for all incident hits.