AX1600i - A couple of questions

[*]Is there really no way to prevent the automatic (15 seconds long) fan self-test at startup?

The AX1600i doesn't do this.

[*]Is the fan still performing an obstruction test every few seconds, as was the case with the AX1200i?

The AX1600i doesn't do this.

[*]Does it remember user-defined fan curves during normal system restarts, if no Corsair software is running?

As long as AC power is maintained, user settings are maintained. But the AXi doesn't use non-volatile memory, so once you cut the AC power, those settings are lost.

That's great news! But then I don't understand why you wrote the following not too long ago:

Context. AX1600i does not do this. I don't think the AX1500i does it either. Last PSU to do this might have been the AX1200i.

The PSU has arrived and apart from the fact that you flat-out lied to me about the fan self-test, which does indeed happen at every (warm-)start, I am very happy with the PSU so far.

Well.. I didn't intentionally lie to you. I even checked with one of the engineers. I was told the fan only did the light spin like all of the other Zero RPM Corsair PSUs. Not that it went full tilt for calibration like the older AXi's.

The MCU (Silicon Lab C8051F380) in this power supply comes with several KB of flash memory that supports non-volatile data storage.

That may be the case, but it's not supported. When the power is removed, C-Link/iCUE settings are reset. I know this for a fact.

The fan goes to max speed immediately and stays there for the whole startup process. I honestly do not see how this could be related to any kind of calibration process.

How could you NOT see that's related to a calibration process? What do YOU think it is?

The original design from Flextronics was to spin the fan up to full PWM to see what the RPM was at that state, and then reprogram the fan controller's fan curve to ensure the proper PWM was applied to result in the correct RPM. i.e. Every fan would spin at the exact RPM regardless of lot code, etc.

Corsair had told them that the +/- 2 to 3% variance seen in fan speed was far from critical, so they asked for it to be removed and to just use a pre-determined table for fan RPM as 100 RPM one way or another doesn't make much difference.

I just could not believe that a) this is necessary and b) this must take 15 seconds. Sorry for freaking out about it.

Any chance that this could be fixed with a firmware update?

 

On an aside, why can't I set the FAN RPM any lower than 40% (960 RPM) via Corsair Link?

No. Afraid not. 40% is the lowest.

So, I'm talking to the EE that told me that it was removed and he's looking for the ECN (engineering change notice) that requested the change. Sounds like they updated the firmware for one reason or another and put the fan test back in.

This is kind of sad, because it basically makes the custom fan curve feature useless for anyone who wants quiet operation. The fan spins at about 960 RPM when set to 40% PWM, which would only be reached at a power consumption of about 1300W under normal circumstances.

I'm not going to disagree.

If the fact that Corsair Link shows a firmware version on the AX1600i means that firmware updates are available to end users

It does not mean that there are firmware updates available to end users.

AX1500i and AX1600i are different in that we offered "overclocker edition" PSUs to LN2 professional overclockers that had the OCP disabled by default via a special firmware.

Firmware updating for PSUs is not allowed because critical housekeeping tasks are done by the same MCU that is doing the reporting and monitoring. A botched firmware flash that corrupts this housekeeping could not only damage the PSU, but nearly everything the PSU is powering.