Real-time PCR low efficiency - (Aug/30/2007 )

Hi everyone!
I am new here… Hope someone can help me!

I am doing quantitative real-time PCR to determine mitochondrial DNA genomic copy number (for the ones that are not familiar with mtDNA: copy number varies from cell to cell and with different conditions). I am using SYBR green and standard curve quantification and have already done this with success in the past. However, I changed to another lab and am now having problems.

I am doing exactly the same I was doing, but with a different machine (past: Corbett; present: BioRad) and different master mix/SYBR reagent (past: Quantace; present: BioRad). The reactions look nice: high r square (99%), no contaminations, and nice melt-curves. However, efficiencies are always low (75-85%). Indeed, from one standard to the next (10-fold dilution series) the Ct always increases 4 units (which is too high). It seems that I have inhibitors in my reactions. However, the samples I use to create the standards are isolated through base-column assay. I have already tried different primers concentrations and annealing temperatures and times. At last, I increased the extension times and got 2 reactions with 95% efficiencies. But next reactions got wrong again!

Should I try different primers? If so, which primer-design program should I use? Is the one from Invitrogen available online good? Or should I use a purchase-one?

Thank you in advance. And sorry for being so long…