Most testosterone in the circulation is bound to protein; about 60% to SHBG and 38% to albumin. Only about 2% is unbound. The answer to your question depends on definitions: total testosterone is the sum of all three fractions; free testosterone is the unbound fraction, and "bioavailable testosterone" is the testosterone not bound to SHBG.

More useful might be FAI (free androgen index) which is the ratio of total testosterone to SHBG, and can be anywhere between 30 and 150.

But the Endocrine Society has long since agreed that the FAI should not be used. Yes, your definitions are EXACTLY how I am asking the question. Shea et al., 2014 & Vesper et al., 2014 would agree. 

Again, I am talking about measured, NOT calculated or estimated values. Thank you. 

I was only pointing out that if you calculate FAI then you might be reassured that the levels your are seeing are indeed within the reference range.

You can measure the free testosterone o the relationship between total testosterone and SHBG.  There is normally with age an increase in SHBG and thus the normal situation is that you can have "normal levels " of  total testosterone  but low levels of free testosterone. Since the activity is related to the free ( or also the not so intense binding of testosterone with albumin= bioavailable testosterone) the expected situation should be more in this direction. However if by any reason SHBG should be reduced it is possible to have low total Testosterone and normal free T levels  

It's clear that with low SHBG a low total Testosterone can be associated with a normal free Testosterone level.  It's also clear that reliable free Testosterone assays are not always available, some are very poor.  I strongly recommend, instead of the flawed if easy FAI approach - with outlying SHBG values it's very misleading, instead using a well validated algorithm such as that of Vermeulen & Kaufman which are readily accessible on line.