00:12 <ldcd[m]> With 12 MSPs output rate I can put out a surprisingly not garbage 1MHz phase modulation on the tone but I start to hit FIFO stalls

00:14 <ldcd[m]> I can still put out 1MHz at 6MSPs but its not nearly as clean, can probably be solved by running the FIFO slower and using a CIC interpolator to smooth out the input into the PLL

00:28 <ldcd[m]> Not the cleanest 100kHz modulation in the world but not terrible

00:28 <ldcd[m]> Second order spur is down about 2.5 dB from the main spur

00:30 <ldcd[m]> This is straight from the Glasgow through a 10MHz HPF for a DC block and a 3dB attenuator into the spectrum analyzer

00:34 <ldcd[m]> With a 1 MHz the main lobe is at 4.7dBm, first order modulation product weirdly higher at 6.75, and then the second order at -3.5dBm

00:40 <ldcd[m]> The third harmonic at 144 MHz is at 1.4 dBm when unmodulated, and goes down to -13 when modulated with a full scale 1MHz phase sine wave with the two sidebands at -5 and -11

00:40 <ldcd[m]> not sure why they aren't symmetric

00:41 <ldcd[m]> Sorry realizing this may be a support only channel, is stuff like this welcome

00:59 <joshua_> it has been pretty quiet here recently but I have been watching this and enjoying watching it

01:00 <josHua[m]> oop we are still netsplit from irc/matrix (well, thi sgets mirrored to matrix / irc, but not the other way around); I responded there just now. but broadly I have been following this and enjoy seeing this 🙂

01:01 <threeflour[m]> I think it's very cool!

01:02 <josHua[m]> it has been on my list to read more carefully the architecture of what you have built

01:23 <esden[m]> Yeah we are still unidirectional netsplit... Catherine looked into it a while ago but I am not sure what the conclusion was.

01:24 <whitequark[m]> that it's a pain in the arse

01:24 <whitequark[m]> I know how to fix it, probably

01:24 <esden[m]> Yeah I figured...

01:24 <whitequark[m]> it would involve manual DB manipulation

01:24 <whitequark[m]> and that makes me want to drink vodka straight from the bottle

01:25 <whitequark[m]> so in the interest of my liver I'm going to do this at some point when I'm feeling better

01:25 <esden[m]> Yeah, definitely not rushing you to do anything. Just checking where things ended up at.

01:26 <esden[m]> Independently, I hope you feel better soon. 🙂

01:26 <whitequark[m]> I've been doing steadily worse in the last 10 months or so

01:26 <whitequark[m]> this will likely continue

01:33 <esden[m]> 😬

04:08 <josHua[m]> if you give me a webhook URL for this room I can stand up a unidirectional bridge at least from IRC to here that involves less vodka (I already have basically that thing handy), though it cannot disambiguate between matrix and discord, so it would not be able to proxy matrix to avoid an echo from discord users

04:11 <whitequark[m]> that would definitely make the current situation worse

04:14 <josHua[m]> well, I would have it ignore every user ending in [m], which does suck, but allows at least one bridge inbound without an echo

04:25 <whitequark[m]> that suffix isn't a certain indicator

04:25 <whitequark[m]> this solution is not welcome

04:28 <josHua[m]> it shall not be implemented, in that case

04:45 <ldcd[m]> Oh its fine I can just go back to monitoring catherine's irc logger like I would if I was on IRC anyways

05:08 <ldcd[m]> Ok this shape is actually expected, I just haven't done anything but QAM in a while

05:08 <ldcd[m]> This is what 4 MHz phase modulated with 100KHz looks like:

05:18 <ldcd[m]> Bumping up the modulation index really pushes the power out to higher order sidebands

05:18 <ldcd[m]> which makes sense now that I think about it

14:28 <cyborg_ar> that's a spikey boi

18:01 <ldcd[m]> Yeah, spikey Indeed

18:02 <ldcd[m]> for posterity here's what it looks like un-windowed with appropriately quantized frequenies

18:02 <ldcd[m]> which matches what the spectrum analyzer was showing

22:33 <thestumbler[m]> Did someone post a plug-on board with two DE-9 connectors and logic to RS232 transceivers for serial port sniffing? I forgot to bookmark it and now I can’t find the message.

22:36 <josHua[m]> this reminds me, if you wanted a capture with more analog bandwidth, I'm happy to apply the same setup as I applied above for the pulse generator to your thing

22:41 <ldcd[m]> That's pretty close to what I did at work (albeit without any load for the scope captures)

22:42 <ldcd[m]> Was going to do more testing today but I've got a nasty cold so I didn't end up going in

22:43 <ldcd[m]> I'll give the applet an update and send it over to you after work today, there are a couple potentially interesting things I wasn't able to resolve with my rigol that I didn't manage to get to on the nice scope yesterday

22:44 <josHua[m]> sounds good!