Next: 5.3 Fabry-Pérot Overview Up: 5. Instrumental Characteristics Previous: 5.1 Introduction

5.2 Grating Performance Overview

The SWS instrumental signal-level unit is $\mu$ Vs^-1, the average rise of the voltage of the ramp during one second of time (see section 3.4 for an explanation of this). It is found that the signal response of the instrument is a linear function of the illumination over a very large range of input signals (0.5 - 100,000 Jy), when hysteresis effects are ignored. Therefore this unit is fully qualified for simple linear calculations backwards and forwards relating source flux density, instrument signal level, instrument noise level and noise equivalent flux density (NEFD, i.e., the input flux density which gives an output signal with a signal-to-noise ratio SNR = 1). The SWS grating sections produce a $\sim$ 1.3-3 $\mu$ Vs^-1 signal on a single detector element for a source flux density of 1 Jy.

Both the response and the noise of the detectors have been characterized as function of signal strength. Table 5.1 summarizes the overall properties of the detectors.

For sensitivity calculations, the noise per reset interval may be summarized as

$\begin{displaymath} N[\mu{}V/s] = \frac{\sqrt{N_R^2+(N_D^2+N_S^2S)t_r}}{t_r} \end{displaymath}$

(5.1)

with S the signal in $\mu$ V/s and t_r dimensionless (happens to be equal to the reset interval in seconds). The empirically determined noise parameters N_R (read noise, in $\mu$ V/s) N_D (dark current noise, in $\mu$ V/s), and N_S (signal shot noise, $\mu$ V^0.5/s^0.5) have been obtained by forcing a fit through this function, and are summarized in Table 5.2. For the SW grating and the FP detectors the function works well up to a $t_r \approx 8$ sec (although 8 second resets were only used for a short period at the start of the mission). For the LW grating detectors the function does not work well in describing the t_r dependency, and the resulting values are not necessarily directly to be associated with read noise and dark noise. Parameter values are chosen in such a way that they give the best description for the short reset intervals.

For Band 4, test measurements show consistently a very poor increase of SNR for t_r > 2 sec. The parameters listed apply to t_r = 1 sec and t_r = 2 sec. For longer t_r no improvement in noise should be expected, but the $\sqrt{t}$ function still applies.

**Table 5.1:** Summary of SWS detector properties for a reset interval t_r = 2 sec
detector	T	bias	minimum		int.⁴	average response
			system noise³
	(K)	(V)	( $\mu$ V/s)	(electrons)	( $\mu$ V/s)	( $\mu$ V/s/Jy)⁵
			gain=16
Gr-InSb	3	-	0.7	35	0.32	3.0
Gr-Si:Ga	3	12	1.5	75	0.32	1.6
Gr-Si:As	4.5	3.0	3.0	150	0.40²	1.5
Gr-Ge:Be	2	0.6	2.0	100	1.00	1.3
F-P-Si:Sb	10	1.1	1.1¹	70	0.40	0.007
F-P-Ge:Be	2	0.6	3.0	200	0.40	0.0037

Notes:

1.: values listed for element 2; noise = 0.7 $\mu$ V/s for element 1
2.: excess photon noise at $Signal > 1000 \mu$ V/s
3.: measured at a signal power at which photon noise is negligible; except FP-Ge:Be where still some photon-noise is suspected; estimated photon-noise free NEP₂ = 1.05 x 10^-16 W/ $\sqrt{Hz}$
4.: int. is intercept: photon noise at signal = 10 $\mu$ V/s
5.: 1 Jy = 10^-26 WHz^-1m^-2, ISO telescope area taken as 0.28 m²

**Table 5.2:** Grating noise parameters
Band	1	2	3	4
detector	InSb	Si:Ga	Si:As	Ge:Be
range ( $\mu m$ )	2.38-4.05	4.05-12.0	12.0-29.5	29.5-45.2
N_R	0.75	0.75	1.00	1.56
N_D	0.70	1.50	3.00	2.00
N_S	0.10	0.13	0.14	0.32

Next: 5.3 Fabry-Pérot Overview Up: 5. Instrumental Characteristics Previous: 5.1 Introduction

SWS Instrument & Data Manual, Issue 1.0, SAI/98-095/Dc