There are a lot of other advantages as well. The old e682 WizNet/Propeller thing has a lot of restrictions on number of universes (12 in unicast I think?) which can really limit the amount of data it can receive. Also, it cannot support a bunch of modern things like the Virtual Strings and per-virtual-string settings like begin/end nulls, gamma, brightness, etc.... that the BBB string code in FPP can provide. The other "extras" like the OLED and ability to use Wifi to run in remote mode and such are also great additions.
In my opinion it is the game changer known as "Play from a local fseq file" would drive this decision for me. That removes the heavy load on the network and allows your shows to grow without incurring network congestion.
DrNeutron wrote:An E682 on its own is just a pixel controller which requires E1.31 data from a show computer/Pi/BBB.
A PB controller IS the show computer with FPP loaded onto it and a copy of the sequences you want to play.
If you had all FPP controllers you could sync them all via wifi and have them run the sequences locally on each.
There would be no lag as each FPP controller would only send sync packets over wifi and not E1.31 data to keep the sequences in sync.
This would eliminate cat5 cable to each controller....
Ok, I can see those as benefits. Although, I've personally had lots of problems with the wifi syncing. I have 6 P10 panels run by BBB and octoscroller. The P10 panels will every so often, glitch, slow down then run fast to catch up. I know I have a good wifi signal and have tried a few different settings except changing the BBB to multisync. I haven't done much more with it cause right now it's mostly used for my "Tune to" sign.
dkulp wrote:There are a lot of other advantages as well. The old e682 WizNet/Propeller thing has a lot of restrictions on number of universes (12 in unicast I think?) which can really limit the amount of data it can receive. Also, it cannot support a bunch of modern things like the Virtual Strings and per-virtual-string settings like begin/end nulls, gamma, brightness, etc.... that the BBB string code in FPP can provide. The other "extras" like the OLED and ability to use Wifi to run in remote mode and such are also great additions.
Dan, Yes the E682 only supports 12 universes in Unicast and only 7 in Multicast. I'm neurotic about control w/ my electronics and only run Unicast so I have complete control over the whole system. Right now I still have 4 ports available on my show E682, but as my display grows, I won't have to buy a new controller so I'll keep this setup and integrate it into my show as needed. Although, running FPP adds a lot of complexity to the setup vs. the basic firmware of the E684. I don't completely understand all the nuances that FPP provides such as virtual strings.
Does this PocketBeagle adaptation allow the E6804 to be upgraded as well (albeit with a 40 wire ribbon cable extension)?
Does this allow the faster 40fps?
Thank you guys for all your knowledge.
kpken wrote:Thanks Scott, that is very helpful, 🙂
I set it up identical to your AM26C31 with pin 4 and 12 (Enable pins) to Gnd, I will take a look at the datasheet again and see if there is something I can test.
I think you are right that it is sourcing current some where, just need to find it, and why it is doing it on all 6 of them.
I will also recheck the one that is connected to the 4 single ports w/ power is also getting warm (little harder, its under the OLED).
Attached is SCH/PCB snip of one of the ports.
Please let me know if you have any other ideas, thanks.Update: I checked that there is no pins to Gnd that shouldn't be, dif outputs Ohm good with term resistor, no shorts to Gnd.
I finally got a chance to do some more testing on these IC's, I removed the IC's from the board, and soldered to adapter board and connected them to the same power with some term resistors, and they didn't get hot.
I did find that some of them didn't have the correct input/output characteristics as noted in the datasheet.
I found that some of the IC's would only swap polarity on the outputs when tied to gnd instead of +5 as noted in datasheet.
There were also a couple of them that one of the outputs would hold last state when wire disconnected, and would only change when connecting to Gnd or +5.
I found that I had a total of 3 out of 17 IC's (mostly new ones) that acted as they should of changing state when connecting to +5, and then back when removed.
I put these 3 IC's on the board and they all worked, but still got warm/hot, so I was hunting around on Maxim's site to see about a replacement (and of course the package sizes are not the same) but found a diagram in the datasheet for a transmit only vrs a multidrop network with responding slave units, and it only had terminating resistors on the receiver side, and some notes that they are not needed on the transmit side, since there is no reflection back to the sender.
So, I removed all the resistors, and hooked up a long cable (50 ft or so), and it now works great without getting warm.
What is strange is the test setup didn't get hot, and it had all the same connections as the main board, only difference was it had 1/4 watt wire resistors vrs 1/8 watt SMT on the board.
Hopefully this will help someone else troubleshoot their issues.
Scott, maybe do some testing with your boards with diff ports and see if the transmit side terminating resistors are even needed for the max wire length, at max data rate, to eliminate a few components and reduce the load on these IC's.
Does the USB port support multiple devices attached? I'm considering using it for both an Ethernet adapter and for an audio adapter to connect to my FM transmitter.
Sent from my SM-N986U using Tapatalk
