What is the relation between amplicon size and melting temp? - (May/23/2020 )

I am running HRM these days. One thing I never understood is the relation between the length of amplicon and melting temperature. I mean, will temperature increase if length is longer? I am guessing that the answer is negative. Probably melting temp does not work that way.

I have a set of primers that amplify a 120 bp amplicon. Then I have another set that amplify a 300 of the same gene. How different their melt peak should be? Definitely, it does not mean that the 300 bp amplicon must show a higher melting peak than that of the 120 bp? Correct? If the 120 bp product shows a peak at 81 C, it does not mean that the 300 bp product must be above 81 C? I have problem understanding this.

Yes and no.
If you try to imagine DNA like a Velcro, the longer Velcro you have the more force is needed to separate it. So it is legth dependent.
But it is also sequence dependent and mismatch dependent, of course.

But, legth only matters to some extent, DNA can be very long yet the Tm never goes much over 80. It is because longer DNA is not melted as "one" but as several separate domains (with different meting temps), if you have very long Velcro it is like several people trying to separate it. For that reason HRM recommended limit is around those 300 for SNPs. 
I would expect the Tm to be maybe slightly higher, but doesn't need to be. Also melt peak differences are not only in their negative derivative maximum but also in the shape of the curve (I think in HRM shape matters even more)

Hi Trof,
I knew you would be the one to reply because you are very good at qPCR. Thank you.

Your explanation makes sense. I think the Velcro is a great example. I just need to look up online to find a page o an article about the shape of the curve.