I fixed the RMS bug which is only in my DSO. 
:?:
I modified some thing in menu.c like this:
</s>case RMS:
if(Meter[i].Track == TRACK1){
if (_1_source == HIDE) Tmp=0;
else {
Tmp =(Ka2[_A_Range]*Int_sqrt(a_Ssq/bag_max_buf)+ 512)/1024; // use bag_max_buf as average divider
if(Tmp <= 4) Tmp = 0; // round of precision error
Tmp *= Y_Attr[_A_Range].SCALE;
if (_1_source == CH_X10) Tmp=Tmp*10;
}
}
if(Meter[i].Track == TRACK2){
if (_2_source == HIDE) Tmp=0;
else {
Tmp =(Kb2[_B_Range]*Int_sqrt(b_Ssq/bag_max_buf)+ 512)/1024; // use bag_max_buf as average divider
if(Tmp <= 4) Tmp = 0; // round of precision error
Tmp *= Y_Attr[_B_Range].SCALE;
if (_2_source == CH_X10) Tmp=Tmp*10;
}
}<e>
Maybe is a little strange,
but in my situation, I find that when Vin=0, a_Ssq and b_Ssq are 0. If I add a Ka1[_A_Range] in Tmp, the RMS is bigger than actual.
Can I debug it directly with a ST-link?
Thanks,