Interlaboratory Evaluation of a Material with Unequal Numbers of

The cutoff characteristics of a single section low-pass RC filter with fc = ... express our appreciation to Dr. Dieter Ziessow for his aid in certain ...
0 downloads 0 Views 97KB Size
Ar

”C Nul. Abun. DIETHYL PHTHALATE

c3.C~

I

~

4

13mm TUBE 4 0 0 sac SCAN 8192 DATA P I S . ~

~

5

CltC2

;c=

>CH2

-CH3 0

CSFB

I

H-..-..._....._..

4

1524 6

Figure 9. Natural abundance 13C Fourier transform spectrum of diethylphthalate

were taken t o optimize conditions prior t o recording the spectrum. The solution was contained in a 13-mm tube with approximately 25 C6F6added for field-frequency locking. The digitizing rate was 10 KHz with an rf pulse width of 50 psec. Pulse widths of 50 psec were generally used for 13Cnuclei since these yielded an optimum flip angle. Satellites present with some peaks are spinning sidebands. The average signal t o noise ratio for the peaks is greater than 100 t o 1. The cutoff characteristics of a single section low-pass RC filter with f c = Nyquist frequency allows down-conversion (“aliasing”) of high frequency noise. All spectra except that of natural abundance diethylphthalate (Figure 9) were determined using an RC filter. The filter used in Figure 9 was a 4-pole Tchebyscheff active filter. The active filter was constructed from two experimental components (Analog Devices N o . 9171, N o . 9172) which allowed sixfc’s to be selected under computer control in a manner identical to the rate selection in the data clock. Generally, a filter bandwith corresponding t o one-half the current digitizing rate is used. The Tchebyscheff transfer function achieves an extremely sharp roll-off per number of poles. The filter phase shift is greater than 180’ throughout the frequency range and becomes distinctly nonlinear near f c . The present method of phase correction handles such characteristics extremely well. Substitution of the active filter for the RC filter has increased S/N by a factor of two. Although we have software for forming multiple pulse sequences, a power amplifier which will yield a / 2 pulses of 10 psec or less is needed, and extensive modification of the Bruker high resolution probe is necessary to withstand the high voltages developed. Multiple pulse experiments have therefore been curtailed. However, rapid accumulation of periodic pulses is shown to give significant improvements which warrant extensive use of the pulse-Fourier system. We are currently modifying a Collins 3OL-1 linear amplifier

(500 watt) to decrease the pulse width necessary for a 90” pulse. ACKNOWLEDGMENT

We thank Mr. L. Berman for his assistance in the design of of the electronic circuitry. We should also like to express our appreciation to Dr, Dieter Ziessow for his aid in certain aspects of the computer programming. RECEIVED for review January 25, 1971. Accepted June 7, 1971. Presented at the 160th National Meeting of the American Chemical Society, Chicago, Ill., September 13-1 8, 1970. This work was supported by a National Institutes of Health Grant, RR 00356, Biotechnology Resources Branch.

CORRECTION lnterlaboratory Evaluation of a Material with Unequal Numbers of Replicates In this article by J. Mandel and R. C. Paule [ANAL.CHEM., 42, 1194 (1970)l there is a n error in the iteration formula (Eq.

ac

22). The correct formula should read -

ax

= i

The original formula gives the correct answer when convergence is obtained. In a few cases, however, divergence occurs. The above corrected formula will always converge provided that only positive values of X are permitted. This is a reasonable restriction since X is a ratio of two variances. The corrected formula also requires fewer iterations. The authors thank Mr. Lars H. Sjodahl for detection of the error. (51

- ,G)’].

ANALYTICAL CHEMISTRY, VOL. 43, NO. 10,AUGUST 1971

1287