ADMV1014 Vdet VS. Vctrl

Hi Folks!

I run into some weird problem with the 1014 on our application board which I couldn't reproduce with the evaluation board. I'm trying to understand what's going on, I don't know if it's an issue with our board design, the programming or something else.

First the DET_EN pin seems to work the other way around. I see voltage variations on the VDET pin if DET_EN is 1 (not 0 as it is in the datasheet: "The detector can be enabled by setting the DET_EN bit (Register 0x03, Bit 6) to 0. ")
The bigger problem is if the detector is enabled there is visible difference in the conversion gain, moreover there is a weird voltage on the Vctrl pin despite a 10kOhm pulldown on it.

Conversion gain (40GHz LO, upper sideband)

Below is a scope screenshot of the signals on VDET and VCTRL.

The pulses correspond to the network analyzer sweep, so the VDET signal looks all right to me.
However when there is no input, VDET is low, then VCTRL gets up to around 600mV, despite the 10k pulldown resistor on it. There is nothing on our board that would explain this behavior.

If DET_EN is 0 there is only DC on both lines.

I'm sure it's something simple that I'm overlooking. E.g. the interference of some other setting, but I can't figure it out.

Any tips would be appreciated.

Thank you,