At FTDNA there is a Slovenian kit N110396 from Presladol. G-L293 and 100 % related through very specific STR combination to four Croats from the Mrsic et al. Croatian national reference Y-STR haplotype database (1100 haplotypes from 5 regions of Croatia tested). This combination involves dys19=14 + dys456=14 in addition to dys385=15-15. All of these values are off-modal for G-L293 and there is no chance whatsoever all 5 of these do no form some cluster of their own, especially as this cluster differs alot from others even at low number of STR's.

As Slovenian has 111 STR's I was able to estimate where under G-L293 tree he fits best though he has no close matches.
His DYS505=9 is indicative of Z6653+ and he is predicted at FTDNA as such.
His dys565=12 is indicative of Z6679- , as dys565=13 is one of important mutations at that level
His DYS570=17 + DYS576=15 combination is indicative of G-Z6692- as DYS570=18 + DYS576=16 are some of determining mutations at that level.
So seems due to STR combination a pretty basal G-L293.

With the Karachay-Balkar/Georgian/Abkhazian cluster G-Y36036 he shares three of their defining mutations, namely DYS532=13, DYS481=21, and DYS437=14. Doesn't share other 11 mutations from the Y111 set. Croats from the study all have dys437=15, but due to other STR's and proximity they are certainly related so these might be a private mutation 14->15 for the Croats.

So I guess this cluster is some basal G-Y36036 with TMRCA to Caucasians at about 4500-5000 ybp or just some G-Z6653*.

Of interest might be also one Pashtun G-L293 sharing with these dys19=14 and also having GATAH4=11 which is also indicative of G-Z6653* as one of defining mutations at G-Z6679 level is GATAH4=10<-11. Croats/Slovenian have GATAH4=10 but due to their other STR's they should be Z6679-. In the case they are not Y36036- this Pasthun might look more interesting as he doesn't fit as G-Y36036 nicely because of dys437=16.

Considering this cluster is basically confined to these 5 I can't help but wonder whether these are the genetic trace of Alanic Croats who are speculated to have existed by variousy authors. Personal name Horoathos, "assembly" Horouathon occurring in Sarmatians around Tanais has been long speculated to have etymological ties the the Croat ethnoym:
"Trubachyov tried to explain the original proto-type of the ethnonym from adjectives *xar-va(n)t (feminine, rich in women), which derives from the etymology of Sarmatians"

I might search for some references that some authors used to support their theory of Iranian connection. I couldn't find such genetic links with Serbs, though I tried, who were also hypothesized to have similar connection. Certain 10th century supposed mentions of Serbs (Sarban) and Croats (Krevatas/Krebatas) in Caucasus are definitely false, in the case of Sarbans the source refers to Shirwan, and Krebatas to Karabakh.

Obviously there are other possibilities. Among Avars there was also G-L293, but this particular haplotype does not exist in Hungarians or Romanians (but some others do). There is some J-Z1828 haplotype which seems surely Sarmatian and it occurs in 4 Croats from the same study but it also occurs in a Serb, in Hungary etc it might be of Iazygian extraction.

This cluster though seems pretty confined to Croats (and a Slovenian 30 km NW from Zagreb). Its regional presence: one from Northern Croatia (no.13), Western Croatia (no.10) and two from Central Croatia (no. 10 & 11) area.

Not found in Eastern Croatia.

Haplotype 25 from the Southern Croatia, also a G-L293 I had to exclude because he lacks dys19=14+dys456=14, also has usual dys437=16, he onlyl shares with most unusual dys385=15-15, not enough to put him inside this cluster.