I agree with the above but keep in mind that with stripping you will probably lose some of your target protein.  If it is present at low levels then it might be even harder to detect as a consequence.  Are the antibodies from the two companies different clones?  If so it may not be necessary to strip since they should detect different epitopes.  You can always strip the membrane after you have tried the new antibody and seen no signal.

Antibodies tend to adhere to the antigen pretty strongly and even when there's no apparent signal with your luminescence reagent you will have your antigen blocked by the primary and possibly the secondary as well. It might be worth your time to strip the blot. In addition consider using a powerful luminescence reagent, like pierce's femto, when my signal died out for a primary, I switched from my ECL to this femto reagent and voila, strong bands!

I agree with the above but keep in mind that with stripping you will probably lose some of your target protein.  If it is present at low levels then it might be even harder to detect as a consequence.  Are the antibodies from the two companies different clones?  If so it may not be necessary to strip since they should detect different epitopes.  You can always strip the membrane after you have tried the new antibody and seen no signal.

It is best to try the new antibody in a fresh blot first.

As suggested I would go with striping because otherwise you simple don't know what you got. I would also add another control (which we use to debug westerns). After transfer but before blocking put a drop of diluted primary Ab in one corner of your membrane and a drop of diluted secondary Ab in another corner. Let it sit for a minute and then block, and continue as usual. At the end you should see the two spots clearly, If you don't see the primary Ab you can say that the secondary Ab does not react with it, if you see either then there is a an issue with your the secondary Ab conjugate. Takes only a minute and could save many hours of frustration...

Yoram

Another suggestion: You can try to detect your secondary antibody with a (differently) labelled tertiary antibody. Building such sandwiches will increase your signals.

Thx for all your answers. I will keep it in mind.

But my supervisor had the opinion that there is no protein and I should not waste another antibody. I did another WB with some changes in buffer etc. and used the other antibody and it worked. Despite the fact, that I am not satisfied with the result :(...