Notifications
Clear all

Pixel Power

35 Posts
10 Users
0 Reactions
9,723 Views
Posts: 1076
 Zeph
(@zeph)
Noble Member
Joined: 15 years ago

I think that with unbalanced lines driven and received by logic gates, determining or controlling the end impedances is the complicated part, not finding the cable impedance.

What's the impedance of a 74HCT541 output by itself? Or an ATMega328p output or a Propellor output? What does it become with a 100 ohm inline resistor? With a 200 ohm resistor to +5 or ground??

What's your pixel chip's DI impedance? How about with a 33 ohm inline resistor on the pixel PCB?

How about the transition from your long lead wire (eg: Catx) to the ribbon cable lead on a pixel string? Is there an impedance mismatch at the solder joint?

How about the impedance of the flexible PCB of a pixel strip?

Basically, knowing that your cable has a 100 ohm impedance vs 75 ohm isn't going to buy you a tremendous amount in terms of matching that impedance on the sending and receiving end. You may have to try experimenting with resistors to see which works best (in adjusting the unknown impendance of the chips to match the best impedance of the cable).

With something like RS-485 it's much less confusing, as you put the terminating resistor ACROSS the differential lines and use sender and receiver circuits designed for analog transmission purpose. Single ended connections going from logic output (of some technology on some PCB) to logic input (of some technology on some PCB) is a lot messier to try to calculate adjustments for based on the cable specs.

That's just my take, perhaps some of the more knowledgeable can shed more light on calculating the proper adjustments on both ends under these circumstances.

One other thing - if you want to preserve a 2ns edge, you do indeed need very high bandwidth. Sometimes it's OK to soften the edges of a square wave tho, and you don't need them to be absolutely crisp - I seem to recall seeing a scope picture of the pixel to pixel data waveform which wasn't super crisp, but the schmidt trigger on DI was OK with it. Losing the highest frequency components may not be all that bad. (Excessive ringing would not be a good thing tho).


Reply
Posts: 182
 slug
Topic starter
(@slug)
Reputable Member
Joined: 19 years ago

I ended up getting rid of the external power supply, and I'm just driving the strings with two null pixels on each run, and it's doing fine. Not sure if the external power supply would have brightened them up, but they seem fine. Except that my blues are red on the 3001 strips. I'll have to go into the controller tomorrow and try either GBR or RBG on the string. Not a huge deal. It just doesn't match the picture in my head. Thanks for all the help on this problem, and I'm going to renew my knowledge on transmission lines.


Reply
Posts: 352
(@pmscientist)
Reputable Member
Joined: 16 years ago
Zeph wrote:I think that with unbalanced lines driven and received by logic gates, determining or controlling the end impedances is the complicated part, not finding the cable impedance.

What's the impedance of a 74HCT541 output by itself? Or an ATMega328p output or a Propellor output? What does it become with a 100 ohm inline resistor? With a 200 ohm resistor to +5 or ground??

All good questions, but it's not necessary to get in to that amount of detail for this application. All you need to know to get a decent increase in the distance to your first pixel is the cable impedance. Shifting to 5V, if not done by the controller, is a good step as well.

Fact is, matching cable impedance does buy you a tremendous amount. It reduces the problem to one of voltage drop for at least 10's of feet. At that distance, degraded edges won't come in to play unless there's a damaged cable or something. Sure, in order to get 100m you may need to go to a balanced signal like rs485, but why bother if you only need 50ft. Plus, you're almost certainly not going to find a resistor that will give you an exact match after accounting for the other impedances.

Not sure that I'd agree that rs485 is less confusing than a resistor value. Is 'develop an RS485 interface' simpler than 'use an X ohm resistor'?

Sure, if you're doing some truly high speed stuff over long distances, a balaced line solves some problems, but it also adds several components. If you just have an edge problem, why not just use a schmit trigger? Doesn't require the supporting infrastructure of a balanced line, and is totally fine for the data rates of most pixels.

The whole point is to keep it simple. Matching the cable impedance with the proper resistor goes a long way to improving things. It's also dirt cheap and simple. This isn't analog stuff, it's not high speed or bandwidth stuff, why complicate things until it's really necessary?


Reply
Posts: 352
(@pmscientist)
Reputable Member
Joined: 16 years ago
slug wrote:I ended up getting rid of the external power supply, and I'm just driving the strings with two null pixels on each run, and it's doing fine. Not sure if the external power supply would have brightened them up, but they seem fine. Except that my blues are red on the 3001 strips. I'll have to go into the controller tomorrow and try either GBR or RBG on the string. Not a huge deal. It just doesn't match the picture in my head. Thanks for all the help on this problem, and I'm going to renew my knowledge on transmission lines.

Glad a couple null pixels got you running well.

I thought the external supply was the only thing powering the pixels. Did you have the pixels connected to the + side of both the controller and the external supply by chance?


Reply
Posts: 2229
(@plasmadrive)
Famed Member
Joined: 15 years ago
pmscientist wrote:All good questions, but it's not necessary to get in to that amount of detail for this application. All you need to know to get a decent increase in the distance to your first pixel is the cable impedance. Shifting to 5V, if not done by the controller, is a good step as well.

Fact is, matching cable impedance does buy you a tremendous amount. It reduces the problem to one of voltage drop for at least 10's of feet. At that distance, degraded edges won't come in to play unless there's a damaged cable or something. Sure, in order to get 100m you may need to go to a balanced signal like rs485, but why bother if you only need 50ft. Plus, you're almost certainly not going to find a resistor that will give you an exact match after accounting for the other impedances.

Not sure that I'd agree that rs485 is less confusing than a resistor value. Is 'develop an RS485 interface' simpler than 'use an X ohm resistor'?

Sure, if you're doing some truly high speed stuff over long distances, a balaced line solves some problems, but it also adds several components. If you just have an edge problem, why not just use a schmit trigger? Doesn't require the supporting infrastructure of a balanced line, and is totally fine for the data rates of most pixels.

The whole point is to keep it simple. Matching the cable impedance with the proper resistor goes a long way to improving things. It's also dirt cheap and simple. This isn't analog stuff, it's not high speed or bandwidth stuff, why complicate things until it's really necessary?

Actually using RS485 solves several issues for the most part. First off converters are cheap. $15 for a receiver and e6804 and e682 (I assume the J1S stuff also) will output differential already. I don't know about you but my time to mess with null pixels or adding resistor cleanly, is not worth the trouble. I tried it the first time around and never again.. what a pain!

Secondly not only distance is an issue. There are also noise implications for bundles of cables. (been there done that). Now granted not everyone has the same issues I did with the Plasma Icicles, but those conditions do exist... So I guess it could be a case by case thing... but I bought a bunch of Pixel Extenders and plan to use them whenever I have any distance or data cable bundles to deal with in regards to pixel stuff.


Reply
Page 5 / 7
Share: