If you save your file as a 'xlsx' file, you could take advantage of the 'Range' option in readtable.
Without it, you may have to be a little more creative. The approach I took was to disable a lot of the autodetection readtable uses. This approach means your files must all have the same format. Some of the 'header' values look incorrect because this approach treats them all as numeric values. Those that aren't get read in as NaN.
file = 'https://www.mathworks.com/matlabcentral/answers/uploaded_files/1076420/LDB100B.CSV';
data = readtable(file,'ReadRowNames',true,'ExpectedNumVariables',5,'PartialFieldRule','fill','numheaderlines',0)
data =
Var2 Var3 Var4 Var5
_________ __________ ______ ______
Header Size 15 NaN NaN NaN
Model Name NaN NaN NaN NaN
Comment NaN NaN NaN NaN
BlockNumber 1 NaN NaN NaN
TraceName NaN NaN NaN NaN
BlockSize 41955 41955 41955 41955
VUnit NaN NaN NaN NaN
SampleRate 5 5 5 5
HResolution 0.2 0.2 0.2 0.2
HOffset -18018 -18018 -18018 -18018
HUnit NaN NaN NaN NaN
DisplayPointNo. 1 1 1 1
PhaseShift 1 1 1 1
Date NaN NaN NaN NaN
Time NaN NaN NaN NaN
Row16 0.0001725 0.00017167 1.6 1.6417
hRes = data{'HResolution',1}
vals = data{(data{'Header Size',1}+1):end,:}
vals =
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.6000 1.6417
0.0002 0.0002 1.5996 1.6417
