See attached diagram…
Project Goal: Use Spark Fun FIO 3v to drive 4 in. tall LED’s for Pinewood Derby Finish Line.
Challenge: The LED segments require at least 8v to forward bias.
Circuit Description: The MC sends serial commands to a MAX7219 7-SEG LED driver. The MAX7219 can only output 5v but my large LED segments require 8v minimum to forward bias. The outputs of the MAX7219 are then connected to the inputs of NEC PC2501L-4 optical isolators allowing the circuit to forward bias the LED’s at approximately 9-12v (TBD).
Problem: On the breadboard, it doesn’t work.
Troubleshooting:
PASS: 5v rail has 5v
PASS: 12v RAIL has 12v
PASS: Serial commands are received by the MAX7219 7-SEG LED driver
FAIL: MAX7219 7-SEG LED driver — no output
TROUBLESHOOTING RESULTS: Removing jumper from output of MAX7219 7-SEG LED driver to input of optical isolator allows test to pass.
PASS: MAX7219 7-SEG LED driver output
Conclusion: I have made a design mistake in how I connect the MAX7219 7-SEG LED driver to the optical isolator. The outputs of MAX7219 7-SEG LED driver are being driven low by the anode of the optical isolator IR diode.
Can anyone give me a suggestion?
Ok … I had a duh moment. When the opto-isolator anode is forward biased of course it sinks the input to ground, the IR-diode is forward biased! Reading voltage across the forward bias diode would only measure the voltage drop across it… not much.
So ok… now my problem is that the output of the opto-isolator emitter never seems to go to 12v…
Looking at the datasheet for the NEC, there is supposed to be a 30Ohm resistor to ground from each cathode. You probably burnt out the internal LEDs in the NEC.
Another thing, I would put U3 behind U2 as it wouldn’t have to loose 11V+ to heat. Put it behind U2 and it will only need to get rid of 7V. This would be more efficient.
I also don’t see any decoupling caps. These should be the first thing you add. .1uF and ~10uF between ground and VCC respectively. For overkill add another .1uF cap as close to the GND and VCC pins for the 7219. Add another .1uF cap on GND and VCC between U2 and U3.
wbaldwin@ksu.edu:
So ok… now my problem is that the output of the opto-isolator emitter never seems to go to 12v…
How you measuring it ? A DVM will be too slow to catch the scan pattern. You'd need an O-scope.
Thank you so much for your quick response.
Looking at the datasheet for the NEC, there is supposed to be a 30Ohm resistor to ground from each cathode. You probably burnt out the internal LEDs in the NEC.
The current limiter in the MAX7219 set by the ISET resister was suppose to protect the IR diodes but you are right. I need to look at that again and verify that the IR LEDS are still operational and that the ISET resister limits the current within limits for the IR diodes.
Another thing, I would put U3 behind U2 as it wouldn’t have to loose 11V+ to heat. Put it behind U2 and it will only need to get rid of 7V. This would be more efficient.
U2 and U3 are 95% efficient buck converters that aren’t suppose to get hot but I do like your logic here and will change the circuit design.
I also don’t see any decoupling caps. These should be the first thing you add. .1uF and ~10uF between ground and VCC respectively. For overkill add another .1uF cap as close to the GND and VCC pins for the 7219. Add another .1uF cap on GND and VCC between U2 and U3.
Great Point!
wbaldwin@ksu.edu:
Another thing, I would put U3 behind U2 as it wouldn’t have to loose 11V+ to heat. Put it behind U2 and it will only need to get rid of 7V. This would be more efficient.
>> U2 and U3 are 95% efficient buck converters that aren't suppose to get hot but I do like your logic here and will change the circuit design.
I'd look before I'd leap here. Revising the design means U2 now has to supply it's original load plus that of U3. Be sure you're not bumping into it's current or thermal limits. Also switchers have an efficiency curve that's dependent on input voltage and load. I'd run the math on both configurations using those curves and see which wins.
On the Optical Isolator re-reading the specifications… the max forward voltage should be 1.4v with a max forward current of 80mA (PEAK 1A) then looking back at the MAX7221 I can supply the optical isolator diode with a 1.5v forward bias at 10mA using a 66.7 ohm I-set resister on the MAX7221. If the MAX7221 limits the current to 10mA do I really need a current limiting resister on the cathode of the optical isolator diode?
wbaldwin@ksu.edu:
If the MAX7221 limits the current to 10mA do I really need a current limiting resister on the cathode of the optical isolator diode?
No but it's likely a thermal question. Resistors will take some of the power to be dissipated from the 7221, perhaps enough to not need a heatsink. Then again perhaps no heatsink is needed. Run the numbers.
How much current are you trying to pull into your LEDs? I would never recommend using an opto-coupler’s phototransistor to directly drive a load with any appreciable amount of current. Why not ditch the opto-couplers and just use some trusty N-FETs (in a low-side switch configuration)? Probably cheaper than opto-couplers, and definitely requires less hand-holding.