Page 1 of 1
MOD54417 FlexBus IO Standard
Posted: Fri Jul 16, 2021 4:03 pm
by corestar
We're connecting an FPGA to a MOD54417 via FlexBus. Unfortunately, newer FPGA's (20 nm and below) do no support 3.3 V LVCMOS. They max out at 3.0 V LVCMOS and the Altera Cyclone 10, Arria 10 etc chips only have 48 of those.
Looking at the MOD54417 data sheet, it should still work so long as the FlexBus voltage FBVDD = 3.3 V. Can you confirm that is the case?
Any chance newer Netburner SOM's will support say 2.5 V LVCMOS on their FlexBus?
Re: MOD54417 FlexBus IO Standard
Posted: Mon Jul 19, 2021 3:22 pm
by pbreed
If the FPGA can handle the 3.3V drive, then 3.0 should work on the input.
Thresholds are .65*vcc for high and 0.36 Vcc for low.
So 2.145 and 1.18
Give 10% VCC tolerance...
And that's 2.25 and 1.0V Vh and Vl guaranteed
From this document
https://www.intel.com/content/dam/www/p ... /an447.pdf
looks like 3.0V LVCMOS has guaranteed high of 2.8
well above the requirements.
Re: MOD54417 FlexBus IO Standard
Posted: Tue Jul 20, 2021 6:04 pm
by corestar
Seems like it should work. I take it 0.65 Vcc = 2.145 confirms Vcc = 2.145 / 0.65 = 3.3 V for the flex bus voltage.
But the MCF54417 data sheet seems to have different numbers. It has Vin high min = 0.51 * VDD instead of 0.65 * VDD. The 0.51 and 0.42 for the Vin high min and Vin low max and did seem awfully close.
Just using the numbers in the datasheets:
Code: Select all
Netburner Cyclone C10GX
-----------------------------------------------------------
Vout low max = 0.66 Vin low max = 0.8 ok
Vout high min = 2.64 Vin high min = 1.7 ok
Vin high min = 1.68 Vout high min = 2.8 ok
Vin low max = 1.4 Vout low max = 0.2
10% tolerances could be a problem, especially on the first one but in principle I guess it will work.
Re: MOD54417 FlexBus IO Standard
Posted: Tue Jul 20, 2021 6:06 pm
by corestar
I guess my png attachment did not take.