View previous topic :: View next topic |
Author |
Message |
asmallri
Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia
|
|
Posted: Thu Sep 29, 2005 5:05 pm |
|
|
True but generally this class of switch mode regulator has very limited input voltage range where the typical maximum input voltage for this class is 5.5volts. I typically designed for a maximum input voltage 18V. _________________ Regards, Andrew
http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!! |
|
|
ChickyMeal Guest
|
|
Posted: Fri Sep 30, 2005 1:47 am |
|
|
Guyss...
Im quite confuse here..
My circuit current flow = 0.03A
The power is needed for:
1. High Voltage - for GM Detector
2. 5V for other ICs - PIC and a few other chips
Which power supply solution is highly reccomended? |
|
|
Mark
Joined: 07 Sep 2003 Posts: 2838 Location: Atlanta, GA
|
|
Posted: Fri Sep 30, 2005 6:21 am |
|
|
First, I would try to get away from the 9V battery. I guessing that its just the typical 9V battery. Not much capacity and a big waste. The easiest solution would be to try and design the circuit to run on 3- 1.5V AA or AAA batteries. You circuit would probably need to work down to 3V in case the user wanted to use rechargeables which are dead at about 1V per cell. The next easiest solution would be to use 4- 1.5V cells with an LDO. You probably would not recommend using recharageables to you end user. If you want the best battery life, then go with a switcher. You can power your circuit with lower voltage cells. These can have larger capacity which means more battery life and the supply is more efficient. The switcher will cost you more however. |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1909
|
|
Posted: Fri Sep 30, 2005 8:53 am |
|
|
If you do end up going with a 4.5V or 3.0V (or something like that) supply, then there's a really neat low battery detection "trick" that someone posted here a long time ago. Sorry, can't remember who. This will only work if the PIC is powered directly from the battery - no regulator.
If you happen to have a spare A/D input channel (if your PIC has an A/D of course), you'll need to create a very simple zener diode voltage reference. That would be Vcc -----Resistor-----Zener diode----ground. The zener should be rated for a voltage less than your nominal supply voltage. Now you connect the resistor-zener junction to an A/D input. The A/D should be set up so that the high voltage reference is Vdd.
Now when the battery's voltage starts to fall, the A/D reading will start to increase. Very simple, and a very elegant solution. Like I said, this isn't my idea, and I'm sorry that I can't remember who posted it.
Hope someone finds this useful - I did. |
|
|
Neutone
Joined: 08 Sep 2003 Posts: 839 Location: Houston
|
|
Posted: Fri Sep 30, 2005 9:33 am |
|
|
newguy wrote: | If you do end up going with a 4.5V or 3.0V (or something like that) supply, then there's a really neat low battery detection "trick" that someone posted here a long time ago. Sorry, can't remember who. This will only work if the PIC is powered directly from the battery - no regulator.
If you happen to have a spare A/D input channel (if your PIC has an A/D of course), you'll need to create a very simple zener diode voltage reference. That would be Vcc -----Resistor-----Zener diode----ground. The zener should be rated for a voltage less than your nominal supply voltage. Now you connect the resistor-zener junction to an A/D input. The A/D should be set up so that the high voltage reference is Vdd.
Now when the battery's voltage starts to fall, the A/D reading will start to increase. Very simple, and a very elegant solution. Like I said, this isn't my idea, and I'm sorry that I can't remember who posted it.
Hope someone finds this useful - I did. |
That would be better if you powered the voltage divider from a pin that you turn off and on so that you dont waste any power. |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1909
|
|
Posted: Fri Sep 30, 2005 9:47 am |
|
|
Neutone - what a great idea!
I'm often in awe of the ideas that float around on this forum. |
|
|
MGP
Joined: 11 Sep 2003 Posts: 57
|
|
Posted: Fri Sep 30, 2005 3:17 pm |
|
|
asmallri wrote: | True but generally this class of switch mode regulator has very limited input voltage range where the typical maximum input voltage for this class is 5.5volts. I typically designed for a maximum input voltage 18V. |
Define "very limited input voltage". The last MSP430 design I did used a TI chip that had a 16V maximum input and quiescent current under 100uA and switched down to 3.3V. That's about a 5:1 voltage range, not something I'd call "very limited".
The original poster was talking about a 9V battery as a source. There's quite a few switching regulator IC's with low quiescent current that fits those requirements.
I do lots of battery powered commercial designs and there's never been a better selection of parts to squeeze every last electron out of your batteries. |
|
|
asmallri
Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia
|
|
Posted: Fri Sep 30, 2005 7:17 pm |
|
|
Quote: | Define "very limited input voltage". The last MSP430 design I did used a TI chip that had a 16V maximum input and quiescent current under 100uA and switched down to 3.3V. That's about a 5:1 voltage range, not something I'd call "very limited".
|
A matter of context. I mentioned that I usually assume a maximum input voltage of 18V and I was refering to high efficiency switching regulators (>=94%) at low output currents (10mA). There are several regulators that offer low quiecient current as you mentioned but generally when they are delivering a low current the efficiency if low (80% or less).
If you know of one that will work with an 18volt input and delivery >90% efficiency at 10mA could you let me know the device.
Code: | The original poster was talking about a 9V battery as a source.
|
Yes. My mistake as mentioned. _________________ Regards, Andrew
http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!! |
|
|
JimB
Joined: 25 Aug 2005 Posts: 65 Location: Huntington Beach, CA
|
|
Posted: Fri Sep 30, 2005 8:37 pm |
|
|
I have battery operated instrument project that I am going to use a dual battery supply, ±3V lithiums and plan on using a charge pump to develop the higher voltage for the LCD. Charge pumps accomplish this by charging and switching capacitors in and out to generate this voltage. Not much current capability, but simple, without any inductors required. I will then monitor the + 3 volts for battery status. The - 3 volts is less heavily loaded and should last longer than the +3.
Places to look are at Linear Technology and Maxum. You don't want to waste that 9Volt current in the drop through a linear regulator. 9 Volt batteries don't have a lot of milliamp hours of energy in them. |
|
|
MikeValencia
Joined: 04 Aug 2004 Posts: 238 Location: Chicago
|
|
Posted: Sat Oct 01, 2005 9:28 am |
|
|
This design last year uses a 9V battery and a MAX666 voltage regulator. It has very small quiescent current, has adjustable output, and has a pin to indicate low battery status, based on the resistor divider values you choose. "PCMPro" and "Kenny" led me to use these chips (i had a choice of the maxim or seiko chip), and i still haven't heard a complaint yet from an angry customer of a low battery chirp. By the way, the processor used is a PIC16LF627 (not the chip shown above).
|
|
|
ckielstra
Joined: 18 Mar 2004 Posts: 3680 Location: The Netherlands
|
|
Posted: Sat Oct 01, 2005 2:45 pm |
|
|
The MAX666 is a nice chip, but is a low drop linear regulator as several other types mentioned earlier in this thread. Any lineair regulator, low drop or not, will loose 44% of energy in converting from 9V to 5V.
Most switching regulators I've studied have a reduced efficiency when running at low output currents and higher input voltages. What we are looking for is a type number of a switching regulator with:
- A maximum input voltage of at least 10V, preferably up to 18V.
- Minimum 90% efficiency at 10mA output current
- Low quiescent current. |
|
|
Eugeneo
Joined: 30 Aug 2005 Posts: 155 Location: Calgary, AB
|
|
Posted: Sat Oct 01, 2005 5:17 pm |
|
|
Use a LM2575.. It works great for everything as long as you size the coil properly |
|
|
MGP
Joined: 11 Sep 2003 Posts: 57
|
|
Posted: Sat Oct 01, 2005 10:14 pm |
|
|
Here's a link to one IC I've used in the past. It's over 90% efficient down to around 1mA and will deliver up to 800 mA (although I only tested it to about 350mA in my design).
http://focus.ti.com/docs/prod/folders/print/tps62050.html
I can't really see the point of having much larger input voltage range than 10V for a battery application that's switched down to 5V or less. If it's an automotive or SLA battery application then having a very low quiescent current really doesn't mean much as the capacity of these cells is generally not small compared to the loads we're talking about.
At any rate, almost any modern switcher IC circuit will be better than a linear regulator for regulating a 9V battery to 5V or less, more so the higher the load current gets. |
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
|
Posted: Sat Oct 01, 2005 10:45 pm |
|
|
You can try MAX1921 http://pdfserv.maxim-ic.com/en/ds/MAX1920-MAX1921.pdf. It's in SOT-23.
Eugeneo wrote: | Use a LM2575.. It works great for everything as long as you size the coil properly |
LM2575 (just like my favorite LT1074) doesn't have synchronous rectification. Realistically, it makes the power supply no more then 85% efficient. And, yes, power supply chip manufacturers often stack the deck a little in their datasheets. |
|
|
Guest
|
|
Posted: Sun Oct 02, 2005 3:51 am |
|
|
Eugeneo wrote: | Use a LM2575.. It works great for everything as long as you size the coil properly |
Unfortunately the efficiency of this device is less that 75% at low values of Iout.
The maximum input voltage of this device is 5.5 volts.
Quote: |
http://focus.ti.com/docs/prod/folders/print/tps62050.html
I can't really see the point of having much larger input voltage range than 10V for a battery application that's switched down to 5V or less. If it's an automotive or SLA battery application then having a very low quiescent current really doesn't mean much as the capacity of these cells is generally not small compared to the loads we're talking about.
|
The applications I develop are intended to run for months of unattended operation from nominal 12 volt solar powered systems. The output voltage from these systems can exceed 16 volts which is why I design for a minimum of 18 volt operation. |
|
|
|