r - Step halving factor reduced below minimum in PNLS step -

i running self-starting logistic model in nlme, , have received classic error message:

error in nlme.formula(model = mass.cul ~ sslogis(gd10c, asym, xmid, scal),  :  step halving factor reduced below minimum in pnls step 

i first put in code:

shortg <- groupeddata(mass.cul ~ gd10c | expu2, data = short) nlsl <- nlslist(mass.cul ~ sslogis(gd10c, asym, xmid, scal) , data = shortg) shortg.nlme <- nlme(nlsl, random=pddiag(asym + xmid + scal ~ 1)) 

everything worked fine. went standardize variance code:

shortg.nlmeb <- update(shortg.nlme, weights = varpower()) 

this got error message. now, can fit square peg in round hole increasing pnlstol. however, after increasing first 0.01, 0.1, 1, 10, 10,000 fun, keep getting error message. leading me believe wrong apart pnls halving factor. have clue of causing error?