We use 1 μl cDNA template in 20 μl reaction mixture subjected to qRT-PCR experiment, for which template concentration must be strictly equal in each sample under study; then we have to determine the number of PCR amplification cycles required for the fluorescent signal to cross the threshold (i.e. exceeds background level) as the actual "raw data" in qPCR are the ct values (cycle threshold). Hence we equalize the concentration of first strand cDNA to 25ng/μl in each sample.

"1 ml cDNA template in 20 ml reaction mixture", as per your statement, here the volume of the cDNA i.e., 1ml, not going to describe that the concentration of cDNAs nor it will be equal for the different samples.

The main focus should be CONCENTRATION OF RNA from different samples or conditions, that must be equal for cDNA synthesis. Then further dilution cDNA 1/20th times, and then use equal volume let say 1ul each for qRT-PCR.

Now, there your ct values of replicates of each condition should not have the difference between

As others have said, you want your input cDNA to be in a reasonable range, to help the qPCR work optimally. It is not necessary to be extremely precise at that step. As we all know, quantification is best achieved by normalization to reference genes, not to total cDNA.

"1 μl cDNA template in 20 μl reaction mixture" is my statement...

Thanks for correcting your statement. Then here, my answer for the basis of equalizing the cDNA concentration will be relying on the CONCENTRATION OF RNA from different samples or conditions, that equal concentrations of RNA must be used for cDNA synthesis. So, what you have used you can think because any difference in RNA concentration will be going to give false-positive results.

Then further dilution cDNA i.e. in your case 1/20th times, and then there you should equal volume let say 1ul each for qRT-PCR. here, also you can think about what you have used. Any difference in sample volume will affect the result.

Now, come your ct values of endogenous controls of your choice, there the ct value of replicates (ex. Control 1A, control 1B, control 1C) of each condition should not have difference of more than