LimeSuite Source calibration fails

When I use the LimeSuite Source through GNU radio the calibration fails with the following error: (italized in the trace):

INFO: device_handler::calibrate(): Rx calibration: MCU error 5 (Loopback signal weak: not connected/insufficient gain?)

This is the rest of the trace
INFO: device_handler::set_chip_mode(): SISO mode set for device number 0.
INFO: device_handler::set_rf_freq(): RF frequency set [RX]: 1419.6 MHz.
INFO: device_handler::set_samp_rate(): set [RX] sampling rate: 15 MS/s.
INFO: device_handler::set_analog_filter(): RX LPF configured
Configured bandwidth: 4.5 MHz.
INFO: device_handler::set_digital_filter(): digital filter channel 0 [RX]: 4.5 MHz.
INFO: device_handler::set_antenna(): channel 0 antenna set [RX]: LNAL.
INFO: device_handler::set_gain(): set gain [RX]: 60 dB.
INFO: device_handler::calibrate(): Rx calibration: MCU error 5 (Loopback signal weak: not connected/insufficient gain?)
INFO: source_impl::init_stream(): source channel 0 (device nr. 0) stream setup done.

gr::buffer::allocate_buffer: warning: tried to allocate
27 items of size 2400. Due to alignment requirements
128 were allocated. If this isn’t OK, consider padding
your structure to a power-of-two bytes.
On this platform, our allocation granularity is 4096 bytes.
gr::buffer::allocate_buffer: warning: tried to allocate
54 items of size 1200. Due to alignment requirements
256 were allocated. If this isn’t OK, consider padding

The LimeSuite Source parameters I thought might be relevant
Chip mode: SISO
Ch0:A (B also fails)
RF frequency: 1.14196G
Sample rate: 15M
Oversample: Default
NCO frequency: 0
CMIX mode: UPCONVERT
Calibration:On
LNA path:L
Analog filter: On
Digital filter: On

Calibrating through LimeSuite works, trying the calibration in LimeSDR USB Signal issue - #17 by jslatten
We need to calibrate using gnuradio because it sets up DSP filters.
I am running Ubuntu 16.04.4 LTS

Hi, lagrassa

Do you get the same error when you use maximum gain values in both blocks? Are you using only Source block?

No, also using other blocks as well. I tried the maximum gain value for the source block (70), and now it works with higher probability (Rx calibration succeeded) but sometimes still fails.