constraining within-group (or lowest level) error variance to equal to 1

["Siu Woon Edmond Ng" <E.Ng@ioe.ac.uk>]

Hi, 

I wonder if someone could give me a pointer for how to contrain the 
within-group error variance to a constant value, say 1. I have aggregated data 
for performing a meta-analysis.  
The data consist of effect size, d, and its variance, varofd, and the 
study-level average of a covariate, and the study identifier. I intend to use 
'lme' to perform this analysis but I have not been able to contrain the 
within-group error variance to be one as is required in the model. 

Thanks in advance for any idea.

Best wishes,
Edmond