Current Server Time: 23:28 (Central Europe)

#apertus IRC Channel Logs

2020/07/30

Timezone: UTC


05:29
intracube
left the channel
05:31
intracube
joined the channel
06:07
BAndiT1983|away
changed nick to: BAndiT1983
08:26
mumptai
joined the channel
08:59
Bertl_zZ
changed nick to: Bertl
08:59
Bertl
morning folks!
09:13
kbeckmann
good morning!
09:14
kbeckmann
i think i will do a test today where i connect a logic analyzer to the control pins on the ft600 to visualize exactly how bad USB 3 + ftdi drivers are essentially.
09:15
Bertl
sounds like a plan!
09:18
se6ast1an
good day
10:55
Dest123
joined the channel
11:25
Bertl
off for now ... bbl
11:25
Bertl
changed nick to: Bertl_oO
11:48
Dest123
left the channel
14:28
intracube
left the channel
15:04
BAndiT1983
changed nick to: BAndiT1983|away
15:53
BAndiT1983|away
changed nick to: BAndiT1983
17:02
Bertl_oO
changed nick to: Bertl
17:08
Oscar
joined the channel
17:20
intracube
joined the channel
17:45
BAndiT1983
changed nick to: BAndiT1983|away
18:47
Oscar
left the channel
19:02
AndroUser
joined the channel
19:05
AndroUser
Test
19:05
AndroUser
left the channel
19:05
RexOrCine_OnTour
joined the channel
19:05
RexOrCine_OnTour
Tesr
19:06
RexOrCine_OnTour
left the channel
19:08
RexOrCine_OnTour
joined the channel
20:26
anuejn
kbeckmann: any results?
20:27
anuejn
(I am quite interested in seeing why the test results from bertl were so disappointing)
20:44
se6ast1an
wiki: codeformating has been updated
20:44
se6ast1an
https://wiki.apertus.org/index.php/Sandbox#Code_Formatting
20:44
kbeckmann
i did some testing today and honestly it looked pretty good
20:44
se6ast1an
use:
20:44
se6ast1an
<pre class="code">
20:44
se6ast1an
</pre>
20:44
se6ast1an
the space before the first <pre... is important
20:44
se6ast1an
kbeckmann: nice!
20:45
se6ast1an
can you share?
20:46
kbeckmann
sorry for the format, but i recorded a stream that i did a few hours ago. let me see if i can find the interesting timestamps. but MB/s-wise, i get 190 MB/s out of 200 which is the theoretical max.
20:47
kbeckmann
https://www.twitch.tv/videos/695349295 the interesting stuff starts at around 27 minutes
20:47
se6ast1an
that indeed sounds very promising!
20:47
kbeckmann
i attached a logic analyzer to the state machine and plotted the transitions
20:48
BAndiT1983|away
changed nick to: BAndiT1983
20:49
Bertl
kbeckmann: what logic analyzer is that?
20:50
Bertl
I recognize the GUI (Saleae) but the dvice looks like one of the chinese logic analyzers
20:50
Bertl
*device
20:51
kbeckmann
yep it's one of those cheap clones. sampling at 12MS/s
20:51
Bertl
okay, so you are running at less than 6MHz there?
20:52
kbeckmann
what i was looking for was significant delays, that we should see even with that
20:52
Bertl
well, we were 'good' up to about half the theoretical bandwidth limit
20:53
Bertl
so I wouldn't expect anything unusual at 6/10MHz :)
20:53
kbeckmann
mm.. but i was expecting to see a few us of delay actually
20:53
kbeckmann
but that didn't show up at all
20:57
Bertl
I suspect the issue we were seeing resulted from the driver not being able to handle the data properly
20:58
Bertl
i.e. not the FT60x side, but the receiver side
21:04
kbeckmann
i see. maybe you could test writing a loopback or constant source gateware and test it with the example code that just consumes data?
21:05
Bertl
yes, we have a generator HDL which does simply feed PRNG data all the time
21:06
Bertl
I'm currently swamped with unrelated stuff, but it should be better next week, if you like we can get a test setup you can access remotely or so?
21:06
kbeckmann
ah i see
21:06
kbeckmann
oh yeah sure, that sounds cool
21:07
Bertl
great! I'll keep you updated on the progress then
21:07
kbeckmann
sounds good :)
21:09
Bertl
btw, nice 'live stream'
21:09
Bertl
do you do that on a regular basis?
21:10
kbeckmann
thanks, it was my first so i probably just rambled :p
21:10
kbeckmann
i am thinking of doing that but not sure if people are interested
21:11
kbeckmann
i'll try to have a better focus next time and announce it earlier. this was basically just a test
21:11
Bertl
not bad for a first/test run
21:11
kbeckmann
thx
21:12
Bertl
I've seen a lot worse from established live streamers :)
21:12
kbeckmann
hehe
21:12
kbeckmann
i am not used to talk and think at the same time. always messed up when whiteboarding etc :p
21:12
Bertl
you'll get used to it ...
22:40
RexOrCine_OnTour
left the channel
22:52
BAndiT1983
changed nick to: BAndiT1983|away
23:00
Bertl
off to bed now ... have a good one everyone!
23:00
Bertl
changed nick to: Bertl_zZ