The page states that the backlight driver should use 243mA at 3.7V. Mine uses about 130mA initially then jumps to 310mA after a minute and the inductor gets hot enough to burn my finger. When started at room temperature the driver uses 130mA at 6V, but if I run it at 3.7v until it gets hot and then connect it to 6v it will use 440mA. Obviously there is some loss of efficiency when the inductor gets hot.
I have experienced exactly the same thing you saw – not sure you found more details? I have found that, especially if i run it off of 5V instead of 3.3V, it can get extremely hot. I ran it at 5V for a long time before I touched it, so it did not appear to self-destruct, but I think the 3.3v is much safer. Not sure if this represents a critical design flaw or not.
Funny, for me it seems to be just the opposite! At 3.7 volts it always gets hot. At 5v it seems that if it is initially cold it will stay cold. If it does get hot though, it will get extremely hot at 5v, even more so than at 3.7v.
I think that the inductor is not meant to handle those kinds of currents. When it gets hot the resistance increases, leading to less efficiency and even more heat. When I get some free time I think I will remove the inductor and replace it with something really chunky that can handle more current.
There is note that says, Pins 8 and 9 on the Nokia LCD connector will be connected with a solder bridge. Have you done this? Can you tell me how to do this?
I was having the same problems, so for future reference let me describe how I fixed it. This is the Nokia breakout SKU: LCD-08600.
Powering the backlight at 3.3V the boost IC on the breakout got really hot. Powering it at 5V it was even worse; I turned it off pretty quick for fear of smoking the board.
I tried putting a 33ohm resistor in series with the backlight pin. At 3.3V this made the display a little dimmer, but the heat was completely gone. Unfortunately it also caused sporadic flickering in the brightness.
At 5V with a 33ohm resistor the display is quite bright, the boost IC is staying cool and I haven’t seen any flicker. So there you go. Worth a shot for the desperate.
(Does anybody who knows how the boost circuit works care to comment on why this solution might have worked?)
cpolley: Interesting. I have no idea why that works.
Last week I finally had some time to work on my project again. I figured that I would use the circuit as-is and just ignore the heat. Sure enough, once I mounted the display in a plastic box the inductor fried (I was running it at 3.7v). The funny thing is that once the inductor fused itself into a short the backlight would only work when my circuit had no power supply bypass capacitor. I think the reason was that the wires going to the board were acting as an inductor. OK, I’m getting off topic so I’ll just tell how I fixed it. I installed my own inductor that was actually capable of handling the currents involved. This was a 220uH inductor (same as what the board comes with) that was rated at 500mA (which is what the board was drawing). Now the backlight works perfectly and only draws about 100mA. Judging by the sizes of the various inductors availably from Digikey I can say that there is no way the one used by SparkFun is rated for the amount of current being drawn. This inductor is huge (probably 1/2 to 3/4 inch long) so it is off to the side of the board, connected with wires. Maybe a slightly smaller inductor would have worked. I wonder what Nokia uses.
Sounds like the inductor was poorly chosen & should be replaced with a higher-current rated unit. There are some pretty small units that are rated for a decent current - spend a while with the parametric search on Digikey.
I’ve made this mistake before, and inductor saturation can do weird things. It’s not just the resistive losses in the coil - once the inductor starts to saturate, its inductance drops, and the rate of current increases in the switchmode circuit (at a certain point the switchmode control chip won’t be able to regulate properly).
This thing is huge (compared to the typical size of electronics components) and I ended up having to run wires from the inductor to the board so that I would be able to still mount the LCD on the front panel of my box. You might try and get by with a smaller one, but I just went for the one that had a current rating as high as what the board was drawing.
dstahlke:
I ordered it from Digikey. The part number is:
M8294-ND (CHOKE RF HI CURRENT 220UH .508A)
This thing is huge…
You're not kidding - that's big! The issue with the existing circuit is that resistor R4 sets the current limit on the inductor to more than 1amp, which is much higher than needed for this circuit, and easily enough that the inductor will saturate. I'd suggest changing R4 to something higher (say, 1 ohm), which should ensure the inductor doesn't saturate.
If you really want to change the inductor, I’d suggest a shielded type, as it will reduce EMI. Something like DR73-221-R or DR125-221-R might be a good choice, depending on what saturation current you need.
I really “don’t want to change the inductor”. What I want is this circuit to work and not overheat! I guess I will have to go look up the MC34063 datasheet and see what makes the most sense for the amount of current the backlight will draw. Anyone take a physical reading on the current used at the likely 7 volt supply?
fll-freak:
I really “don’t want to change the inductor”. What I want is this circuit to work and not overheat! I guess I will have to go look up the MC34063 datasheet and see what makes the most sense for the amount of current the backlight will draw. Anyone take a physical reading on the current used at the likely 7 volt supply?
Based on the figures given on the sparkfun page, I'd say it would be a bit under 100mA.
From the MC34063 datasheet, the overcurrent comparator has a nominal 300mV level. Changing R4 to (say) 1.2 ohms you could expect 250mA short-circuit current, so this should be OK for this application. Note that R4 seems to be maybe a 0603 size resistor, so you might want to change to a slightly bigger package, since it could dissipate a bit more than its rated power during normal operation.
I too noticed that inductor getting stinky-hot. I just changed one of the feedback resistor values on the 34063 (R2) from 3K3 to 3K92. This drops the backlight voltage from ~7V to ~6V. I saw a marked drop in current consumption, the inductor was much happier and the only penalty is a slightly dimmer backlight.
It worked for me. YMMV, depending on amibent conditions etc…
I took the suggestion of changing R4. I first removed R4 by simply “wipping” it off. I picked a 1.2 Ohm 1/8 watt resistor I had and carefull soldered it into place. The inductor runs much cooler. The wet look in the photo is IPA from washing the flux off.
So I have had this board sat in a box I ordered for all these years…
Finally had a use for it and dug it out.
I now have a massive burn on my thumb. Thanks Sparkfun! It’s pretty damn painful and I thought I had fried the board connecting it up wrong, did some searching on Sparkfun for this board and found this thread.
Next time, how about you email a warning to customers about known burn hazards on boards?
@Pyrofer - I’m very sorry you burned yourself. We have gotten better recently about mass emails regarding potential issues, but that doesn’t help much for older things. Let me know if there’s anything further I can do.
