Programmable Devices
CPLDs, FPGAs, SoC FPGAs, Configuration, and Transceivers
20641 Discussions

Database error for Cyclone III DIFFIO?

Altera_Forum
Honored Contributor II
1,195 Views

Has anyone seen any inconsistency with the databases for available pins, particularly DIFFIO? 

 

I am using the EP3C80F484C8 device. I am using a lot of DIFFIO (LVDS) IOs in my design just fine. However, I found two locations that are identified as DIFFIO pairs in the pin spreadsheets but Quartus does not acknowledge as valid differential IOs. 

 

J18/K18: should be DIFFIOR16p/n 

T4/T5: should be DIFFIOL35p/n 

 

If I hover over these IOs in the Pin Planner, the label says DIFFIO but the icon for the pin shows them as special purpose IOs and the tool will not accept a differential port to be placed here. It won't accept it for a cmd-line scripted flow either so it appears to be a mismatch in the database that Quartus uses versus the one that generated the pin spreadsheets and the pin labels in the Pin Planner 

 

Any thoughts? 

 

A side note: I have device migration enabled for the 55 and 120 and have confirmed that these pins are still avaiable after device migration takes its pound of flesh from the available IOs. 

 

Thanks in advance, 

A
0 Kudos
2 Replies
Altera_Forum
Honored Contributor II
355 Views

The pins are not available as differential pairs for EP3C55, if the pin-out file is correct.

0 Kudos
Altera_Forum
Honored Contributor II
355 Views

 

--- Quote Start ---  

The pins are not available as differential pairs for EP3C55, if the pin-out file is correct. 

--- Quote End ---  

 

 

Ahh yes, I think I see the problem.  

 

In the EPC380, J18/K18 are paired. However in the EP3C55, J18/K18 are both available as IO but are rebundled with other DIFFIO pairs. K18 is paired with pin K17 and J18 is only a single-ended IO (its diff pair being only available in the larger package). The T4/T5 pair is a similar situation. 

 

I see now that moving to the larger devices requires more VCC and GND pins and so sometimes the DIFFIO pairs are broken up and used for this. I wasn't expecting the pairs to get rebundled - as was the case here. 

 

I think I can conclude that the migration view within Pin Planner definitely is not wrong but it is also not complete since it showed these pins available across all three devices but does not indicate that their differential pairing may change or be lost. 

 

Thanks FvM. 

 

-A
0 Kudos
Reply