arch estimation with restrictions

[lilc]

I'm using eviews 7. I'm trying to replicate Engle's model where he assign weights to the errors in the conditional variance equation. I have tried clicking the options in eviews but the only one that appears to allow me to do so is the IGARCH model (and that only works if I specify a GARCH model and not ARCH model). I tried to modify the standard command to:
equation arc1.arch((0.3*c(6)+0.3*c(7)+0.2*c(8)+0.1*c(9)+0.1*c(10)), 0, m=1000, h) dly c ar(1) ma(1) ma(4), but, I guess it is not as straightforward. I would like some suggestions pls. Thanks.