Server migration complete, Welcome to version 2.1.1



Welcome, Guest.
Please login or register.
Forgot your password?

+-Site Stats

Total Members: 15492
Latest: Robnation13
New This Month: 5
New This Week: 28
New Today: 5
Total Posts: 127035
Total Topics: 15585
Most Online Today: 67
Most Online Ever: 7634
(January 21, 2020, 02:14:03 AM)
Users Online
Members: 5
Guests: 47
Total: 52

uAmp -- powering from the pixel-side instead of controller-side

Started by Steve Gase, March 26, 2015, 12:14:31 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Steve Gase

I don't see any problem with this, but I thought some discussion might spark some ideas...

Jim @ SanDevice is about to introduce a tiny e1.31 controller that does not have power for pixels onboard.  it generates 4 ports like the e6804, but without the power to drive pixels.

as I thought more about this, i thought having the power injection close to the pixels might be more convenient in some cases. 

without a lot of details on the new controller, I suspect that a ground line must come out from the controller along with the data line...  this would allow the power injection to tie back to the ground on the controller.

i thought that having the uAmp applied to this setup might be nice... getting more distance between controller and first pixel.  BUT to do that we'd need a positive wire -- which would not come from the controller... it could only come from the power injection.

so, I have created this picture to represent how I think it would all work... a power bus would be created back from the injector to the uAmp.  Maybe multiple uAmps could be applied in sequence if the distance between first pixel and controller is REALLY great.  But the power only needs to reach the first uAmp.

Does this sound right?  |  110K channels, 50K lights  |  Nutcracker, Falcon, DLA, HolidayCoro


Technically, anything in the 'micro' family doesn't have a true 'power input' / 'power output'.  (This is generally true of SSC's so far as I know, too)  Obviously there are holes/pads on the board labeled as such, but this is more for clarity to the user than electronic function.

At the board level, Vin/Vout and GNDin/GNDout are directly connected by the thickest, widest and most direct traces I could fit.  So this 'backfeeding' power should be fine so far as the device is concerned.  You could even run the uAmp as a "T" off the main power line, etc.  All it needs is about 15mA.  All other power is just passing through the board.

Curious you show the e1.31 with "no power".  Since we transmit data as a 0-5V square wave, that 'power' has to come from somewhere.  Maybe you mean just no high current power, but I will assume the e1.31 at least has some voltage source to run it's 'brain' and send the data?

With that being the case, my main concern would arise as to how the node power supply and e1.31 power supply would interact if connected only by the ground line.  This would be especially true if the node / e1.31 power supplies happened to be on the two separate 110v legs of your house and/or the supplies happen to be at (even slightly) different voltages.

If you can clarify the source of the e1.31 power, it might help with a better prediction.


Burt Rumley
Winterville, NC


So there is power to the SanDevice controller, per the video.

I guess the easiest way to think about this...for me, anyway...would be to say, "Any power supply arrangement Jim is happy with for the controller, should be fine for the uAmp."  Though again, I'd worry about using two different power supplies and/or having them on different outlets of the house, which may mean different legs of 110 power. 

The uAmp will limit the output on the data high pulse to 5V max (relative to the local ground) and will tie the data low pulse to the local ground level.  So in that sense, it acts as a voltage leveler in addition to clarifying and regenerating the data.

Another option may be to put the single power supply at the nodes, then run power wires carrying a few dozen mA of power back to the uAmp and possibly the SanDevice controller.  Not exactly sure the current draw on the controller, but it's likely not terribly much.


I and others have asked Jim about the possibility of PoE but haven't heard back yet.  I think that would be the cleanest way to handle the nano and then all the deployed power supplies would be for pixel power only including the uAmps.  If not, I would just deploy power supplies in the field and use one common AC circuit to drive them with all the grounds tied together.
Burt Rumley
Winterville, NC


Some of it may depend on how he is generating the output.  The mosfet drive on the 'micro' family is pretty robust and tolerant of quite a lot...shorting the output to power, shorting to ground, over voltage, static discharge, etc.  The PIC output on the SSCs is probably the opposite end of the spectrum.  One tiny blip over 25mA and the PIC is generally toast.

At least on the surface, two separate power supplies tied only by ground seems a bit sketchy at best, though maybe it would run fine.  If I had to do it, I'd probably try to power the e1.31 box from the node power supply, too.  Given the uAmp would draw no more than 20mA (0.02A) power and taking a wild guess the e1.31 might be close to the same, you would then have something like this...which seems pretty reasonable.


Steve Gase

each port/line can be done the same way -- each with their own uAmp -- or each can take a different approach.
one may use 5v strings with a uAmp if the distances makes it useful... 
another may use 12v strings without the uAmp if the distances are short enough where it doesn't help.
strings may share the same power supply, or use different power supplies at different voltages.  |  110K channels, 50K lights  |  Nutcracker, Falcon, DLA, HolidayCoro


That is not what I was talking about.  He was referring to an issue with using separate power supplies and had suggested using the same supply for both the controller and the string and showed the wiring in his diagram.  My question is how is that going to work with the other three outputs.  Would you need to use the same supply for the other three strings as well?

Support FPP

+- Recent Topics

FS Kulp K-8B Controller, Complete by
Today at 06:28:08 AM

Fpp 6.2 ui password by Sawdust
Today at 06:19:14 AM

FPP RPi 4 GPIO Big Button question by tynmar
Today at 06:08:16 AM

FPP not releasing control to WLED when idle by dkulp
Today at 05:27:12 AM

Remote Relay Control by MGHaff007
Today at 05:03:25 AM

Unable to donate on PayPal by darylc
Today at 04:15:05 AM

Pi zero 2w by Sawdust
Today at 12:25:33 AM

I'm Stumped by saleencb
November 30, 2022, 11:49:40 PM

BRP Voting Plugin - Allow others to vote for your songs! by plaberge
November 30, 2022, 09:27:58 PM

Very unexpected error message by tirebytrlights
November 30, 2022, 09:23:37 PM

Powered by EzPortal
Powered by SMFPacks Menu Editor Mod