Another pixel contr...
 
Notifications
Clear all

Another pixel controller

292 Posts
18 Users
0 Reactions
81.5 K Views
Posts: 217
(@bob-moody)
Reputable Member
Joined: 18 years ago

Just a quick report on the new ENC28J60 code ...

With a .10 timing grid (same testing I have been using) its right on the money. It doesnt appear to drop any pixels and the chase is smooth.
However at .05 there seems to be a recurring annomoly in that with each pass down the string there will be a dropped pixel that decrements with each subsequent pass.
Better explination... lets say it misses pixel 30, on the next pass it would miss pixel 29, then on the next it might hit 28 but on the next pass after that it would miss 27 (hope that makes sense?)

My current loop is about 30 seconds long so the last chase pass comes up short about 11 pixels (thats good in a way in that it lets me see when the sequence restarts)
When it does restart, it may make a coupel of complete passes before the annomoly shows back up...

I havent run it yet any faster than .05 on the timing grid.

Bob


Reply
jchuchla
Posts: 8698
(@jchuchla)
Illustrious Member
Joined: 15 years ago

That would indicate it's dropping one date slot every x number of frames. I'd guess it's not every frame. If it were you'd see that dropped pixel move 10 bulbs in one second.

--Jon Chuchla--

Sent from my iPhone using Tapatalk


Reply
Posts: 313
Topic starter
(@ee351)
Reputable Member
Joined: 13 years ago
Bob Moody wrote:Just a quick report on the new ENC28J60 code ...

With a .10 timing grid (same testing I have been using) its right on the money. It doesnt appear to drop any pixels and the chase is smooth.
However at .05 there seems to be a recurring annomoly in that with each pass down the string there will be a dropped pixel that decrements with each subsequent pass.
Better explination... lets say it misses pixel 30, on the next pass it would miss pixel 29, then on the next it might hit 28 but on the next pass after that it would miss 27 (hope that makes sense?)

My current loop is about 30 seconds long so the last chase pass comes up short about 11 pixels (thats good in a way in that it lets me see when the sequence restarts)
When it does restart, it may make a coupel of complete passes before the annomoly shows back up...

I havent run it yet any faster than .05 on the timing grid.

Bob

Comparing this to your post #201, would you say v3 is faster, slower, or the same as the v2 enc driver?


Reply
Posts: 313
Topic starter
(@ee351)
Reputable Member
Joined: 13 years ago

One more thing, does anyone know the rate at which LOR sends udp packets? For example, if the timing grid is set for 0.05, does it send the packets once every 1/20th of a second (50ms) or does it send the same packet twice (1/40th of a second, every 25ms)? Maybe using wireshark while LOR is running could shed some light on this.


Reply
Posts: 217
(@bob-moody)
Reputable Member
Joined: 18 years ago
EE351 wrote:Comparing this to your post #201, would you say v3 is faster, slower, or the same as the v2 enc driver?

Not sure how to answer that one since LOR is still sending data the same rate it was I'm not sure how to quantify the question.
I dont belive the annomoly exists in V2..
If my ENC's from China would EVER get here I could actually run V2 and V3 side by side .. but for now .. thats a moot point..

I'll see if I can find the LOR sending rate either from LOR or through Wireshark.

Bob


Reply
Posts: 217
(@bob-moody)
Reputable Member
Joined: 18 years ago

Here is the answer I got the LOR data rate...
1. "The timing grid does not affect comm speed in any way."
2. "It will send no faster than once every 23 milliseconds, which corresponds to something like 43 or 44 frames per second"

Is that what you were looking for ?

Bob


Reply
jchuchla
Posts: 8698
(@jchuchla)
Illustrious Member
Joined: 15 years ago

Lol. No. That's a fancy way of dancing around the answer.
1) well if the timing grid doesn't affect it, then what does?
2) 44fps is the maximum allowed by the dmx spec. (And e1.31 sACN by extension). So that answer just says "we're within spec".
Typical rates from other software are generally a lot lower.

--Jon Chuchla--

Sent from my iPhone using Tapatalk


Reply
Posts: 217
(@bob-moody)
Reputable Member
Joined: 18 years ago

Well since the 1st reply I got contained a question back to me...which was UDP to what....I figured the timing grid answer referring to the comm rate may have referring to the LOR dongle and the rs4xx serial specs....


Reply
jchuchla
Posts: 8698
(@jchuchla)
Illustrious Member
Joined: 15 years ago

Yeah. That's the sort of response if expect when you ask a technical question from a commercial software company. Automatically defensive.
As EE mentioned. Our best bet for a real answer is probably wireshark.

--Jon Chuchla--

Sent from my iPhone using Tapatalk


Reply
Posts: 217
(@bob-moody)
Reputable Member
Joined: 18 years ago

On another note.... My other 3 enc28j60's showed up today... Now I can side by side v2 and v3.


Reply
Page 23 / 30
Share: