On point missing from this discussion is the input voltage to the regulator, It's the input-output differential, times the current, that determines the heat dissipation, not just the current alone. If the input voltage to the LM2940 is only 5.5V, say, the heat dissipated at 1A would only be 1/2W. OTOH, with an input voltage of 9V, the same part would have to dissipate 4W -- unlikely without additional heatsinking.
-Phil
In a pure linear regulator such as the 7805 you would be right, but these are LDO.
Anything below the 6.5V (maybe higher) or so shifts to a higher current consumpton... dramatically so. I just stick with the 7.5 volt supply to avoid even odder problems.
In a pure linear regulator such as the 7805 you would be right, but these are LDO. Anything below the 6.5V (maybe higher) or so shifts to a higher current consumpton... dramatically so.
Although the quescent current is higher at low input-output voltage differentials, it's still only a fraction of the output current. So my statement remains accurate with LDOs.
I got my Prop Mini today. It certainly is cute! I'll try playing with it over the Christmas holidays. Thanks for everyone's suggestions on how to handle the power supply. I'll start by using the Prop Mini 5V supply driving only a couple of RGB LEDs and probably move to an external 5V supply when I try to drive the 16 LED ring.
No, I would say that they think you can get 200ma out of the 1a regulator because it is in a tiny SMD on a board without a heat sink.
Parallax buys these regulators in reels. There is no reason to buy more inventory when a 500ma regulator and a 1 amp regulator are about the same price. But it is the thermal design context that decides the actual maximum.
In the old days, the 7805 would just melt the solder and slide off a BasicStamp when you demanded too much power for the heat budget. But the newer votage regulators just have a thermal shutdown to prevent that. Over 600ma, you may risk having mysterious shut downs.
Speaking of the 7805, I remembered that I have a few of those and am wondering if I can just use one to create a 1A 5V supply to power my LED ring. Is there any reason that would be a bad idea? I assume I can power the 7805 with the same supply that I use for the Prop Mini.
Another way to approach this is to use a 5V regulated power supply, which are cheap as dirt because everything is mini-USB nowadays, and power the Prop Mini through its 5V pin instead of Vin, ignoring the 5V regulator. (A similar trick has worked forever on Protoboards.) I have a prototype commercial product which has been running for two months like this without interruption and flawlessly.
Another way to approach this is to use a 5V regulated power supply, which are cheap as dirt because everything is mini-USB nowadays, and power the Prop Mini through its 5V pin instead of Vin, ignoring the 5V regulator. (A similar trick has worked forever on Protoboards.) I have a prototype commercial product which has been running for two months like this without interruption and flawlessly.
If I do that won't there be a problem because I'm feeding power into the output pin of the 5V regulator? Should I just remove the 5V regulator from the Prop Mini?
If I do that won't there be a problem because I'm feeding power into the output pin of the 5V regulator? Should I just remove the 5V regulator from the Prop Mini?
Sorry for all of the questions but now I'm wondering how long the wires can be between the LED ring and the power supply and Propeller pin? My plan had been to mount the ring on the back of a clear plastic snowflake ornament and run wires from the ring to the trunk of the tree where I would place the Prop Mini and power supply. A regulated wall wart will power the 7805 regulator powering the LED ring and also the Prop Mini through its 5V output as suggested by localranger. I won't be using the 5V regulator on the Prop Mini itself. So my question is, how long can the wires be between the LED ring and the Prop Mini and 7805 power supply?
In a pure linear regulator such as the 7805 you would be right, but these are LDO.
Anything below the 6.5V (maybe higher) or so shifts to a higher current consumpton... dramatically so. I just stick with the 7.5 volt supply to avoid even odder problems.
I took another look at the LM2940-5 pdf. The Low Dropout feature causes a huge climb in quiesent currnet. At 1 amp load, anything below 9V supply will increase the current consuption by the regulator itself. It starts out at 40ma, and climbs to 120ma or so.
IOW, 9 volt seems to generate less heat that 7.5 volts with a 1 amp load. But if you are just having 500ma or less, the 7.5 volt supply might be more reasonalbe. (There are several charts that have to be juggled to figure this all out. You may be running the Propeller at low capacity, say 100ma with pins demanding another 100ma for a total of 200ma.)
Good to hear you will use an independent supply. The real question is how long do you want the wires to be? Or need to be? Voltage drop is pretty much rated to wire size as well as distance. Under 10 feet, I'd just use solid core telepone wire. I think that is about 22 guage.
David, the schematic is calling out the LM2937, but the part on the board is an LM2940 which can push 1A. The part marking is L53B and on the LM2940 datasheet, that equates to an MPN of "LM2940IMP-5.0/NOPB". The LM2937 would have a marking of L73B.
I confirmed that the schematic was incorrect, and the Propeller Mini has always used the LM2940. It has been corrected.
I confirmed that the schematic was incorrect, and the Propeller Mini has always used the LM2940. It has been corrected.
Thanks Steph! I set aside my problem of how to power my LED ring from the Prop Mini and instead am using it with a ActivityBoard which seems to have plenty of 5V capacity. Once I get the software working the way I want I'll go back to the Prop Mini for the actual project because it's smaller and cheaper.
IOW, 9 volt seems to generate less heat that 7.5 volts with a 1 amp load.
It may be that you didn't understand Phil's post. The power used to operate the regulator is significantly less than the total power being dissipated in the regulator. In the example here, supplying 9V to a 3V3 regulator driving a 1A load is dissipating 5.7W on top of whatever the regulator circuitry draws for its own internal use.
This, of course, is the whole motivation for switching regulators vs linear regulators.
@UserName
I understood Phil's post. LDO regulators have a big low end jump in current consumption, even when quiescent. All along, I have been very wary that trying to coax more than 100ma above what the Propeller might use is likely to shut down the thermal protection in the LM2940.
At least the device will run cooler in the off state at 9VDC, and it may also run cooler in the on state if the load is 100ma or so above the Propeller's demands of roughly 300ma tops.
1A 5VDC output at any input voltage seems impossible to me. .... just not adequate heat sinks.
Loopy thanks for bringing out attention to that Iq spike - I wasn't previously aware of it.
It means the regulator itself can be adding up a watt or so, just for "operating", on top of what its dissipating to regulate what goes to the load, and that's significant.
I finally got around to wiring up a Prop Mini to an AdaFruit 16 RGB LED ring. I ended up following the suggestion of using an old USB phone charger that puts out regulated 5V connected directly to the 5V pin on the Prop Mini. The charger drives the LED ring directly and the 5V regulator on the Prop Mini is not used. I connected P15 on the Prop Mini to the control input on the LED ring and that seems to work fine in this setup. I now need to do a bit of packaging and I'll be ready to put the ornament on the tree and make a video for posting. Thanks everyone for your suggestions!
I finally got around to wiring up a Prop Mini to an AdaFruit 16 RGB LED ring. I ended up following the suggestion of using an old USB phone charger that puts out regulated 5V connected directly to the 5V pin on the Prop Mini. The charger drives the LED ring directly and the 5V regulator on the Prop Mini is not used. I connected P15 on the Prop Mini to the control input on the LED ring and that seems to work fine in this setup. I now need to do a bit of packaging and I'll be ready to put the ornament on the tree and make a video for posting. Thanks everyone for your suggestions!
Here's a link to a YouTube video of my ornament. Not as beautiful as Duane's but at least it's done before Christmas!
I have another question about powering the Prop Mini. Is it save to feed regulated 3.3v into the 3.3v pin on the Prop Mini as long as VIN remains unconnected?
I have another question about powering the Prop Mini. Is it save to feed regulated 3.3v into the 3.3v pin on the Prop Mini as long as VIN remains unconnected?
Some here would probably so no but I have done it many times before and gotten away with no problems.
Well, this isn't really a Propeller or even a Parallax project but it is the next step in the Christmas ornament project I did last year. I decided I wanted to try using a controller that wasn't quite as expensive as the Prop Mini so I first switched to the AdaFruit Trinket with an ATtiny85 chip on it. That worked but was still $7 for the board. I then ordered a few ATtiny85 chips from Digikey and tried wiring one up on its own. The result is an ornament with the same AdaFruit 16 RGB LED ring, half of one of their Perma-Proto Small Mint Tin Size protoboards and an ATtiny85 chip. It works exactly the same as the ornament I created last year and runs the same code but the total cost is fairly low. The half protoboard is about $1.33 and the ATtiny85 is around $1.50. The only other parts are a 10uf cap and a 2.1mm barrel jack to connect to a power cable. The ATtiny85 has to be programmed separately because the board I made for the ornament doesn't have a programming interface but I created a board for programming on another protoboard.
The 6 pin header on the brown board connects to an AVR programmer which also powers the chip.
Comments
In a pure linear regulator such as the 7805 you would be right, but these are LDO.
Anything below the 6.5V (maybe higher) or so shifts to a higher current consumpton... dramatically so. I just stick with the 7.5 volt supply to avoid even odder problems.
-Phil
edit: link broken from last forum upgrade: https://forums.parallax.com/discussion/152388/propeller-based-xbee-remote-for-christmas-play-time-machine-prop
No need whatsoever for removing the 5V regulator.
I took another look at the LM2940-5 pdf. The Low Dropout feature causes a huge climb in quiesent currnet. At 1 amp load, anything below 9V supply will increase the current consuption by the regulator itself. It starts out at 40ma, and climbs to 120ma or so.
IOW, 9 volt seems to generate less heat that 7.5 volts with a 1 amp load. But if you are just having 500ma or less, the 7.5 volt supply might be more reasonalbe. (There are several charts that have to be juggled to figure this all out. You may be running the Propeller at low capacity, say 100ma with pins demanding another 100ma for a total of 200ma.)
Good to hear you will use an independent supply. The real question is how long do you want the wires to be? Or need to be? Voltage drop is pretty much rated to wire size as well as distance. Under 10 feet, I'd just use solid core telepone wire. I think that is about 22 guage.
I confirmed that the schematic was incorrect, and the Propeller Mini has always used the LM2940. It has been corrected.
It may be that you didn't understand Phil's post. The power used to operate the regulator is significantly less than the total power being dissipated in the regulator. In the example here, supplying 9V to a 3V3 regulator driving a 1A load is dissipating 5.7W on top of whatever the regulator circuitry draws for its own internal use.
This, of course, is the whole motivation for switching regulators vs linear regulators.
I understood Phil's post. LDO regulators have a big low end jump in current consumption, even when quiescent. All along, I have been very wary that trying to coax more than 100ma above what the Propeller might use is likely to shut down the thermal protection in the LM2940.
At least the device will run cooler in the off state at 9VDC, and it may also run cooler in the on state if the load is 100ma or so above the Propeller's demands of roughly 300ma tops.
1A 5VDC output at any input voltage seems impossible to me. .... just not adequate heat sinks.
It means the regulator itself can be adding up a watt or so, just for "operating", on top of what its dissipating to regulate what goes to the load, and that's significant.
http://www.youtube.com/watch?v=vS_nWI7Z_n8
Here is the source code I ended up using.
ws2812.zip
Some here would probably so no but I have done it many times before and gotten away with no problems.
The 6 pin header on the brown board connects to an AVR programmer which also powers the chip.