Mini-splinters in support of partner

Here's a promising hand I held at the club last week, playing with Len: ♠A53 ♥KJT9 ♦7 ♣AK986.  All were vulnerable and, as the dealer, I had no difficulty in choosing an opening bid: 1♣.  After a pass on my left, partner bid 1♥.  Now, my hand has become even better.  But just how good is it?  In particular, is it sufficiently good for a bid of 4♦?  If partner has as little as ♠Kxxx ♥Qxxx ♦xxx ♣xx, game in hearts would be a pretty good bet, though perhaps not cold.  I decided therefore to bid 4♦ and partner asked about key cards.  Pretty soon, partner was installed in 6♥ and I was feeling more than a little smug.  Most probably we'd be one of very few pairs in 6♥ and it rated to be a pretty good contract.  Not everyone, I told myself, would have made a splinter bid at our second turn.

Partner turned out to have a very suitable hand: ♠QJ6 ♥AQ42 ♦A62 ♣T75.  We had a definite club loser but if clubs split (they did), partner would be able to ditch two spades on the long clubs.  With such a strong trump holding, there should be no difficulty in ruffing two diamonds in dummy.  Even if clubs were 4-2, the ♠K might be onside.  Unfortunately, in order to negotiate two diamond ruffs, set up the clubs and draw trumps, it was necessary that trumps split 3-2, which they had a 68% right so to do.

They didn't.  In the end, partner went down 2, though double-dummy play (i.e. not trying to make 6) would make 11 tricks.

I was feeling a little less smug now.  You know you're getting an absolute zero when you try for an iffy slam and it goes down.

So what went wrong?  There was nothing wrong with partner's bidding.  I think I overbid my hand.  Here's how I could have got the same message across (support with short diamonds): at my second turn, I could have called 3♦, a mini-splinter in support of partner, but forcing only to the three-level.  Obviously, with two opening hands and a fit, we're still going to reach game, but we're not going to get to the iffy slam.  Give partner ♠Kxx ♥AQxx ♦Axx ♣xxx or ♠QJx ♥AQxx ♦Axx ♣Qxx and now the slam is odds-on, even opposite my mini-splinter.

Had my minor suits been switched, the mini-splinter would not have been available, because my jump rebid would have been a game-forcing and natural jump shift.  To summarize, here are the three available mini-splinter sequences:

• 1♣ – 1♥ – 3♦;
• 1♣ – 1♠ – 3♦;
• 1♦ – 1♠ – 3♥.

Furthermore, opposite a non-passed hand, we might consider the following four sequences as mini-splinters, as opposed to jump shifts (according to partnership style and after discussion):

• 1♣ – 1♦ – 2♥;
• 1♣ – 1♦ – 2♠;
• 1♣ – 1♥ – 2♠;
• 1♦ – 1♥ – 2♠.
  

Would partner have recognized 3♦ as a mini-splinter?  It's a jump where the non-jump would have been forcing (in this case a reverse), so it must be a splinter.  However, we'd never discussed what ranges the two splinters might show.  But I think it makes total sense for the mini-splinter to be forcing only to the three-level, since we also have available the full splinter, i.e. a double-jump for a game force.