In the early '80s I designed an 8086 based industrial control system for Honeywell UK. We tried several 8086s (for multiple sourcing) and had trouble with the Mitsubishi. The 8086 had a protocol for for handing over control of the bus to another bus master involving an input and an output signal sharing the same pin (max mode, pin 31 I think). We found that the output signal could be so distorted that it was unusable. A Japanese engineer spent an afternoon examining our system and concluded that we were using the chip correctly, the fault was with their chip. Not long afterwards it was withdrawn.
In the early '80s I designed an 8086 based industrial control system for Honeywell UK. We tried several 8086s (for multiple sourcing) and had trouble with the Mitsubishi. The 8086 had a protocol for for handing over control of the bus to another bus master involving an input and an output signal sharing the same pin (max mode, pin 31 I think). We found that the output signal could be so distorted that it was unusable. A Japanese engineer spent an afternoon examining our system and concluded that we were using the chip correctly, the fault was with their chip. Not long afterwards it was withdrawn.