PDA

View Full Version : Olsen 64 Channel Power Question



ghall426
08-17-2012, 11:43 PM
A 64 Channel Olsen 595 controller is powering 64 optos at 20mA each. That comes up to 1.3A when all channels are on. On top of that, most have 64 on board LEDs that use 10-20mA each as well.

Even at the 1.3A, shouldn't the trace widths be wider, like .025" or larger? And isn't 1.3A more current than can safely go through a CAT5 cable or CAT5 connector?

P. Short
08-18-2012, 06:05 AM
The usual DIY SSRs use MOC3023 optos, which get by with 6ma or so of current each, and the power LEDs also use a lot less than 20 ma each.

ghall426
08-29-2012, 08:22 PM
OK one more question. My opto/SSR chips pull 20mA and I designed my 595 board without the Darlington buffer chip since I didn't notice the 70mA total current limit of the 595. I surprisingly made it through season 1 completely without a hitch with 184 channels. Am I setting myself up for future failure if I don't fix this design flaw? I am hesitant because I'll also need to alter my SSR boxes to sink instead of source the current.

P. Short
08-30-2012, 01:36 PM
I don't know if anyone is going to be able to give you a definitive answer.

First, you only run into the spec limits if more than three of the channels are on continuously at the same time, although to be conservative you should assume that all eight are on at the same time. Also, that spec might only apply at one extreme of p/s voltage or temperature.

Second, it depends on the failure mechanism. If the mechanism is that the chip self-destructs in some way, then you're in trouble. If, however, the mechanism is that some operating parameter fails to meet spec (e.g. Voh varies out of bounds or some timing spec is not met) then you're probably OK.

I just don't think that anyone out here can give you an accurate answer, all that they can do is provide the most conservative response.

LightUp
08-30-2012, 06:49 PM
How do you know that each opto draws 20mA? Traces will heat up and then open (like a fuse) when its rating are severly exceeded. Before that they can show a change colour.
I prefer to see actual measurements to base decisions on. How sure can you be to say that next season's sequences will not cause the currents to exceed manufacture's specifications?

Side Note:
When I was designing electronic stuff for work in the '80s, we had an internal document that layed out what each electronic part was not to exceed. This ensured a "reliable" product. Our designs were for internal use and not sold on the commercial market. I don't recall the exact numbers off hand but let's say a resistor must not exceed 80% of its rated power at the highest expected ambient tempurature. Or an IC must not be operated at beyond 75% of its maximum parameters, be it voltage, current, power, etc. If you don't follow good design rules then you set yourself up for future troubles.
As an example, someone designed an automatic volume control amplifier using a FET as the variable resistor. It turned out that the units failure rate exceeded the requirements. I was asked to find out why. The FET met the manufacturer's specification but the designer overlooked the gate voltage spec with regards to the actual control voltage range available. To prevent a redesign of the board I chose a new FET with a narrower range of gate voltage so that it was always within the available control voltage. That fixed the problem.
If you can't measure something then you end up assuming, or guessing. Guessing may be democratic, and we can all vote, but physics doesn't care, nor change, based on votes. :)