Cyclone IV JTAG pins connected to VCCA (2.5V) with IO bank of 3.3V

[HAlex]

Hello

I am using Cyclone IV GX (EP4CGX150CF23C8N).

JTAG and configuration device (AS - EPCQ128ASI16N) connection done as described on figure 8-29 (page 220 of Cyclone IV device handbook Volume 1).

All VCCIO in my project are 3.3V.

I am wondering is it correct to connect TMS, TDI (with pull-up resistors) and JTAG connector pin 4 (for USB/BYTE blaster) to VCCA (2.5V) while VCCIO = 3.3V?

In page 210 of handbook it is also mentioned:

"All the JTAG input pins are powered by the VCCIO pin and support the LVTTL I/O
standard only."

Thank you

[ak6dn · Accepted solution]

For 3.3V LVTTL, the spec for VIH(min) is 1.7V and VIH(max) is 3.6V.  So a pullup to a voltage between these limits (2.5V) is OK.

Normally tho one would connect the pullup to the same voltage source as drives the VCCIO for those I/O pins.

The only concern would be what are the powerup characteristics of the 2.5V and 3.3V rails in your system?

Do they powerup simultaneously, or is there a significant time differential in their ramp rates?

For example, if the 2.5V rail turns on significantly later, you could get into the situation where the pullup to 2.5V is actually a pulldown.

Personally I would connect the pullups to the same voltage source that drives the VCCIO pin for those I/Os.

[HAlex]

Hello

My 2.5V turns ON after 3.3V (DC-DC regulation from 3.3V to 2.5V) with S.S setup of few mSec (less than 5mSec).

Standard POR time in Cyclone is up to 50mSec - so it should be ok.

But just to be on the safe side I decided to do also an alternative connection for VCCIO (of 3.3V - Bank 9).

By default I will assembly resistors only to VCCA (2.5V).

Thank you very much

Alex