Tracking setup plots flatline for all channels in subbands 51 and 75 for all 500 MHz bands

[swh76]

The tracked frequency in tracking_setup plots in DSPv3 fw is a flat line for any channels which fall in subbands 51 and 75, but the channels are actually able to track (we know this because in Frankenpolar a couple of these channels are TES-coupled and we can read out TESs on them just fine). Additional details in this JIRA ticket - https://jira.slac.stanford.edu/browse/ESCRYODET-391.

Mitch looked at this some yesterday and found that even though the tracking setup plots for two channels in subband 75 look like this ;

but when he plotted all tracking setup data it looks like this (with no other channels on):

f, df, sync = S.tracking_setup(band=band,reset_rate_khz=4, lms_freq_hz=20129.4, \
                               fraction_full_scale=0.540, make_plot=True, show_plot=True, \
                               save_plot=True,channel=S.which_on(band),nsamp=2**18, \
                               feedback_start_frac=0.2,feedback_end_frac=0.98,lms_gain=6)
plt.plot(f)

so the channels are tracking, and the data is there, we must just have the channel mapping from fw to pysmurf wrong, for at least subbands 51 and 75, somewhere. @jmdewart wonders if part of the issue may be the mapping from the 416 DSP channels in fw to the 512 pysmurf channels.

[eyyoung24]

In [26]: S.which_on(2)
Out[26]: 
array([  0,  15,  16,  23,  31,  32,  47,  48,  55,  63,  64,  79,  80,
        87,  95,  96, 111, 112, 119, 127, 128, 143, 144, 151, 160, 176,
       183, 256, 271, 272, 279, 288, 303, 311, 319, 320, 335, 336, 343,
       352, 367, 375, 383, 384, 399, 400, 407, 416, 431, 471])

In [27]: sb = np.zeros_like(S.which_on(2))

In [28]: for i, ch in enumerate(S.which_on(2)):
    ...:     sb[i] = S.get_subband_from_channel(2, ch)
    ...: 
In [29]: sb
Out[29]: 
array([63, 55, 67, 51, 59, 65, 57, 69, 53, 61, 63, 55, 67, 51, 59, 65, 57,
       69, 53, 61, 63, 55, 67, 51, 65, 69, 53, 64, 56, 68, 52, 66, 58, 54,
       62, 64, 56, 68, 52, 66, 58, 54, 62, 64, 56, 68, 52, 66, 58, 52])

So it looks like these are all within the desired subband.