Main Site Documentation

Tristate logic level issue


#1

Sparkfun sells a ttl http://www.sparkfun.com/products/718 usb breakout. Allows 1.8, 3.3, and 5v sources on VCCIO. One guy posted the tx/rx LEDs did not turn off at logic level 1.8v.

The LEDs are connected to sink pins on the chip and 5v usb source. So my question is, could it be that when the pins try to close the port they can’t because the chip logic is only 1.8v and the input is currently 5v so door is being overpowered from the outside and 5v keeps flowing to ground? If so, is there a name for that? tia


#2

Sounds like what you’ve got going on is the higher VCCIO voltage is treating the core voltage like a sink (current is going into the supply rail). Take a look at the first drawing. I’m just guessing what’s inside the output stage of the chip but it’s plausible, right?

Second drawing is a crude way around it. The N-FET effectively shorts the LED. No voltage drop across the LED = no light. Crude 'cause it always draws current – through either the FET or the LED.


#3

Your first pic would seem to hit the issue. Don’t see any other way that could be happening. I think I recall some discussion on this effect here, but can’t remember where. Thank you.


#4

But note that according to the FTDI spec, the led is supposed to be powered from the VCCIO rail. So if that rail is 1.8V then nothing should be connected to 5V anywhere, not even the LED…


#5

@ errol. I saw that in the spec. And that is how I hooked up my own version of the board and will see how that goes. However, the question was intriguing as to how/why it was happening in the first place. Hooking to vccio does mean the led will only see 6.7mA @ 1.8v. Probably be very dim if it even lights.
FTDI also has a bug in there docs. I think the first few Figures show hooking Leds to Vcc. Then latter on, they hook to Vccio. Maybe that is how sparkfun got the hw bug.