Could you please let us know the full part number of the devices which you were using previously? While ordering the earlier batch of devices, did you specifically order for parts with VIO functionality?
Can you please provide the top markings of few of the newer batch of devices that are failing?
>>Is the device capable of supporting 1.8V I/O?
It is possible to place an order for parts with VIO functionality enabled. The programming can be done in such a way that devices come with VIO functionality.
If possible, can you provide the wave forms for both working parts and failing parts?
Thank you for the information.
We were not aware that we were ordering a special part previously.
The part we have been using (the part that was working for us) is:
It has markings:
The part we recently used (that does not work) is:
It has markings:
As far as we knew, we were just substituting the Automotive range part for the Industrial range part.
I'm not sure about getting waveforms. I'll have to look into that.
Can you tell us how we go about ordering the part with the VIO capability?
Thanks again for your assistance with this.
For the S70FL01GS product the VIO function is not available. The datasheet
"Note that the Versatile I/O (VIO) power supply (pin 14 on the 16-pin SO package and ball E4 on the 5x5 BGA package) is not
supported, and pin 14 and ball E4 are RFU (Reserved for Future Use) in the standard configuration of the S70FL01GS device.
Contact your local sales office to confirm availability with the VIO feature enabled."
I do not know how you received S70FS01GS product with the VIO feature enabled. The Bonding diagrams show VIO pin is not connected to the Dies. You can ask Sale Rep, if possible to enable the VIO feature.
This is quite a mystery then. We have been using these parts in our boards connected to a Zynq 7 series Xilinx FPGA with the I/O voltage at 1.8 volts. Everything has been working fine until this latest batch of boards where the CM changed the part to the Automotive range part. We approved the change, assuming the only difference was the temperature range.
It seems unlikely (but not impossible) that the parts would work without the VIO being functional. The best case (1.8V) high level out of the Zynq would be 54% of Vcc for the flash. The minimum Vih for the flash is 0.7 * Vcc. We have even programmed flash at high and low temperatures.
I will capture some waveforms to confirm, but it will be pretty surprising if the parts we have been using do not have VIO enabled.
Thanks for looking into this for us.
This is a mystery, can you confirm with waveform the previous S70FL01GS devices outputted only 1.8V for VOH. This would confirm that these units had the VIO bonded out and enabled.
You are correct. The working parts are putting out about 2.5V (clamped by the Zynq input I suspect). This means that for some reason, the Industrial range parts have been working for us with the Zynq I/O voltage at 1.8V and the flash I/O at 3.3V. Murphy got us this time.
I'm trying to understand why the industrial parts seem to work OK like this, but the Automotive range parts do not. I am lowering the flash Vcc to be closer to the Zynq I/O voltage to see if it will start working with the Automotive parts. Just in order to fully understand the issue. This wouldn't be a workable fix to the issue.
Obviously, we need to identify a part that will work for us. It would be great if we can find a drop in replacement that doesn't require a board spin. Are yo aware of anything that would work?
It is possible for Cypress to supply the S70FL01GS devices with VIO enabled. You need to contact your Cypress sales reprehensive and request the device. I don't know the cost or NRE that would be involved, it may be less expensive to do your board spin.
We have contacted sales and they are looking into it.