Other Parts Discussed in Thread: TPS2061C, TPS2041C
I have deisned a four port USB hub, using the TUSB2046. For downstream power switches. I used the TPS259533, which has active low enable and active low fault indication. The hub is primarily for use as part of an in-house testing system.
The power for one of the downstream ports can be controlled by an external logic signal, which overrides the #EN signal coming from the TUSB2046. This port is typically connected to the unit under test (UUT). When I operate this disable signal, the ouput is latched off, and the #FLT output stays low. This condition can only be cleared by cycling the power, or resetting the TUSB2046. This is not what I expected. I would not expect turning the output off to result in a fault being asserted.
When I tested the short circuit fault behaviour of the TPS259533, I found that the current limit worked correctly, but no fault was asserted. This is even more worrying.
I thought that perhaps my supplier (Farnell) had shipped incorrect parts, as there are 12 variants of TPS2595. However, the chip marking (ES33) was correct.
Looking at the variants of TPS2595, there is no variant that behaves as above. The #FLT output is behaving as if it is a Quick Output Discharge (QOD), which would go low when the output is disable, and not go low on a fault. However, the only variants that have QOD instead of 8FLT are TPS259525 and TPS259535, and these would not work, as the enable is active high.
I replaced the TPS259533 chips with the TPS259573, which should work fine for me. However, the behaviour was the same.
Am I misunderstanding these devices, or maybe the TUSB2046?
