How does the unaspirated /t/ differ from /d/?
For example, how is 'stop' pronounced differently from 'sdop', in terms of tongue position and other aspects?
Solution 1:
The contrast between /t/ and /d/ is neutralized for English speakers after an /s/ in the same syllable. In some ways, the plosive in word-initial sC clusters is like a (phonologically) voiced, or lenis plosive, and in other ways it is like a (phonologically) voiceless, or fortis plosive. There isn't any possible contrast. As long as the /s/ is phonetically fully voiceless, "sdop" would sound like "stop".
For example, using another pair of voiced-voiceless consonants (k and g) we can see that many English speakers find "disgust" and "discussed" sound basically the same. Similiarly, "disdain" to me starts out about the same as "distinguish". If you pronounced it as "disdinguish", I doubt I would notice (while I certainly would notice if you pronounced it as "dis-tinguish" with an aspirated t).
Relevant blog post by Geoff Lindsey: Sblended (exdended) (english speech services)
Relevant paper: The phonological status of English oral stops after tautosyllabic /s/: evidence from speakers’ classificatory behaviour, by José Antonio Monpeán González.
Of especial relevance is González's discussion on page 88-89 of experiments by Ohala (1983) and Sawusch and Juscyk (1981) where subjects interpreted the same phonetic sound as a voiced stop when not preceded by [s], and as a voiceless stop when it was preceded by [s].
Solution 2:
/d/ is more weakly articulated than /t/, and it causes lengthening of a preceding vowel in the same syllable. These differences persist even when /d/ is devoiced for some reason and they are general to the other oral stops b, g. The /d/ in a hypothetical */sdap/ would be weakly articulated. Even though word final /d/ is devoiced in some dialects of English, "pod" still sounds different than "pot", because the /d/ of "pod" is lenis and it lengthens the preceding vowel.