At  least part of that should be attributable to the length.  In protein expression, initiation is generally limiting, not elongation.  So if all the initiation signals are the same, then the long protein should produce the same moles = more mass, than the short protein, in direct proportion to the length in amino-acids.  What is the relative length difference? 

If the effect is more than can be explained by length difference alone, then I guess the truncation must be making your protein unstable.  For example that could be happening by disrupting the folding, causing it to precipitate out of solution, or exposing protease recognition sites for degradation.  This is expectable as well, and there's probably going to be relatively little you can do about it. 

At  least part of that should be attributable to the length.  In protein expression, initiation is generally limiting, not elongation.  So if all the initiation signals are the same, then the long protein should produce the same moles = more mass, than the short protein, in direct proportion to the length in amino-acids.  What is the relative length difference? 

If the effect is more than can be explained by length difference alone, then I guess the truncation must be making your protein unstable.  For example that could be happening by disrupting the folding, causing it to precipitate out of solution, or exposing protease recognition sites for degradation.  This is expectable as well, and there's probably going to be relatively little you can do about it. 

Also there is variation between transformants. If you transform your plasmid and pick 5 independent colonies to grow up, you are likely to observe different expression levels between them, especially if you let them age a bit. So before trying anything complicated, I would just retransform the weak expressing clone and test a few more colonies.

Transformed bacteria can kick out the plasmid if the concentration of beta-lactamase is high. this can happen if the cells have been incubated on plates for an extended period. I would check the pET manual for tips on maintaining plasmids like the use of carbanecillin instead of amp.

Agree with repeating transformation and check the new different colonies for expression of both gene.  Could be useful to check the mRNA expression before protein analysis.  If you have similar result and assuming your proteins are extracellular, then the problems of lower expression  might be due to uncorrect  folding suitable for protein excretion or lack of  recognition site/domain (in the shorter protein)  required for extracellular  transport

Try to check mRNA level for both genes and their stability. Then make comparative predictive analyze of possible secondary structures of the molecules. For example, stem loops in translation initiation region may inhibit translation initiation. Thus you can exclude possible problems on transcriptional and posttranscriptional stage.

If mRNA levels, structures and stability are similar for the both molecules, the problem may be in correct protein folding and its quick degradation. Try to use different expression strains or manipulate with expression conditions.  

Hi

A smaller size protein are relatively less stable especially in  a heterologous system. To overcome this problem, you can tag your proteins with glutathione-s-transferase, maltose binding system, chitin binding protein. The tag can be removed by either site specific protease or chemically. 

All the best.