Tripleableness via split equivalence relations

I have a memory that someone (Linton or Manes, maybe) proved a variation on the PTT involving split ERs. Can someone provide a reference? I thought it would be in LNM 80, but I couldn't find it there. Michael