Post by thread:[OT]: GPS current consumption

I'm looking around for the GPS chipset that uses the lowest current. So far, if this survey is any guide, http://yona_n.tripod.com/gps/gps-survey.html the Philips ones appear to have the lowest consumption. If anyone has an alternative I'd be glad to hear about it But there's an added complication - the client wants the GPS part to be asleep most of the time (to save batteries) and update the GPS data to the PIC every 10 - 20 seconds. This seems to be not possible, as the way I read it, if you power-down the GPS, it takes at least 45 seconds after power-up for "time to first fix". Is that correct, or is there a work-around ? Meaning that in the client's application the GPS would have to be on all the time. His original intention was for a unit to run for 15 - 20 days on 3 or 4 AA batteries (and it has other power- hungry things to do as well, not just GPS) but it looks to me as if you could quite easily stuff a set of batteries in half a day. He's short of space too, otherwise I'd suggest a larger battery. Even 6 x AA is going to be pushing it -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> I'm looking around for the GPS chipset that uses the lowest > current. So far, if this survey is any guide, > http://yona_n.tripod.com/gps/gps-survey.html > the Philips ones appear to have the lowest consumption. If > anyone has an alternative I'd be glad to hear about it > Meaning that in the client's application the GPS would have to > be on all the time. His original intention was for a unit to run for > 15 - 20 days on 3 or 4 AA batteries (and it has other power- > hungry things to do as well, not just GPS) That's say 3000 mAH/20 days/24 hours =~ 6 mA average. That's pretty low for a GPS. There is a Casio watch with GPS in - presumably it has a better battery life than that. Any chance of using eg solar charging to add energy to the system? Might some sort of inertial navigation allow you to use a lower average current with GPS update less frequently? I undertand some Japanese in-city systems do this to allow navigation in heavily built up areas. RM but it looks to me as {Quote hidden}> if you could quite easily stuff a set of batteries in half a day. He's > short of space too, otherwise I'd suggest a larger battery. Even > 6 x AA is going to be pushing it > > -- > http://www.piclist.com hint: The PICList is archived three different > ways. See http://www.piclist.com/#archives for details. > > > -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> That's say 3000 mAH/20 days/24 hours =~ 6 mA average. I was being slightly pessimistic about the battery capacity. 1600mAh looked more likely, giving around 10 days. The best GPS consumption I could see was > 35mA, and there's 50mA (not negotiable) on top of that for the other functions. Which is pretty much battery murder > There is a Casio watch with GPS in - presumably it has a better > battery life than that Wonder what they're doing that's so different. Must be conditional > Any chance of using eg solar charging to add energy to the system? He hoped so, but there's no way to predict how much light the unit will get or what orientation it will be in. My experience with solar cells is that you need to get them lined up very well to get any sort of useful charging voltage out of them > Might some sort of inertial navigation allow you to use a lower > average current with GPS update less frequently? What he wants is a big ask for the current technology. 10m accuracy, 15 second updates, no bigger than a cellphone. 20 day lifetime with AAs is the cruncher - IMHO it's simply not possible -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

The GPS from liapac IIRC had a 'sleep' mode which allowed it to wake up and get its first fix every 20 seconds or so. It might be enough for you. You might want to tell your customer to wait a year though. Since cell phones are getting GPS there are a lot more chip fabs that have single chip low power gps designs in the works, some of them even including much of the analog front end on the chip. These should be on the market in a year, and should be cheap in five. -Adam Jinx wrote: {Quote hidden}>I'm looking around for the GPS chipset that uses the lowest >current. So far, if this survey is any guide, > >http://yona_n.tripod.com/gps/gps-survey.html > >the Philips ones appear to have the lowest consumption. If >anyone has an alternative I'd be glad to hear about it > >But there's an added complication - the client wants the GPS >part to be asleep most of the time (to save batteries) and update >the GPS data to the PIC every 10 - 20 seconds. This seems to >be not possible, as the way I read it, if you power-down the GPS, >it takes at least 45 seconds after power-up for "time to first fix". >Is that correct, or is there a work-around ? > >Meaning that in the client's application the GPS would have to >be on all the time. His original intention was for a unit to run for >15 - 20 days on 3 or 4 AA batteries (and it has other power- >hungry things to do as well, not just GPS) but it looks to me as >if you could quite easily stuff a set of batteries in half a day. He's >short of space too, otherwise I'd suggest a larger battery. Even >6 x AA is going to be pushing it > >-- >http://www.piclist.com hint: The PICList is archived three different >ways. See http://www.piclist.com/#archives for details. > > > > > > > -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> > There is a Casio watch with GPS in - presumably it has a better > > battery life than that > > Wonder what they're doing that's so different. Must be conditional It's terrible! :-) Specs http://www.casio.com/gps/spec.html General http://www.casio.com/gps/top.html CR2 Lithium 5 hours battery life with 1 second update rate. 4 hours non stop auto. 12 hours with auto 20 second update rate. Presumably the GPS and watch proper have separate batteries :-) (When the time fails to return the time would have been .... ). -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> The GPS from liapac IIRC had a 'sleep' mode which allowed it > to wake up and get its first fix every 20 seconds or so. It might > be enough for you Thanks, handy to know > You might want to tell your customer to wait a year though Haha - he's been in my face since he asked me on Wednesday. Waiting a year would kill him. Hmmm, maybe.......... > Since cell phones are getting GPS there are a lot more chip fabs > that have single chip low power gps designs in the works That's the way I was starting to think. He wants a product that's not viable just yet -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> > > There is a Casio watch with GPS in - presumably it has a better > > > battery life than that > > > > Wonder what they're doing that's so different. Must be conditional > > It's terrible! :-) > > Specs http://www.casio.com/gps/spec.html > General http://www.casio.com/gps/top.html I sent that to my client as an example of what to expect from current technology. Now he's asked me if I can make one !! Yeah, sure, that's not so hard. Dragging a 40Ah battery around is going to get a little tiresome though ;-) I'm afraid he's going to have to let this one go -- http://www.piclist.com hint: To leave the PICList spam_OUTpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

How about doing periodic GPS nav, and using inertial navigation in the interim. If you can tolerate the accumlated errors between GPS fixes, might be a way to go. Accelerometers are getting small and cheap. > tiresome though ;-) I'm afraid he's going to have to let this one go > > -- > http://www.piclist.com hint: To leave the PICList > .....piclist-unsubscribe-requestKILLspam@spam@mitvma.mit.edu > > -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

> How about doing periodic GPS nav, and using inertial navigation > in the interim. If you can tolerate the accumlated errors between > GPS fixes, might be a way to go. Accelerometers are getting > small and cheap How large would you expect the errors to be ? He's aiming for 10 - 15m positional accuracy updated every 15 seconds He's asked me if it's possible to use the unit's movement to charge it. All I know is that it's being shaken (possibly worn on a person). If that's the case, ie no large or fast movement, then 80mAh charging seems a little hopeful. Especially if one objective is to keep weight and size down -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Hi, > > How about doing periodic GPS nav, and using inertial navigation > > in the interim. If you can tolerate the accumlated errors between > > GPS fixes, might be a way to go. Accelerometers are getting > > small and cheap > > How large would you expect the errors to be ? He's aiming for > 10 - 15m positional accuracy updated every 15 seconds The GPS errors alone are bigger than that ! In my tests my apartment windows moves more than 30 meters ! The cumulative errors from the acelerometer can get pretty big but if the movement is reasonably constant it might work. The error tends to be quite random if you make frequent readings and integrate them well, so part of it will cancel out. Best regards, Alexandre Guimaraes -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

> > How large would you expect the errors to be ? He's aiming for > > 10 - 15m positional accuracy updated every 15 seconds > > The GPS errors alone are bigger than that ! In my tests my > apartment windows moves more than 30 meters ! Time for a word with the builder ! > The cumulative errors from the acelerometer can get pretty big > but if the movement is reasonably constant it might work. The error > tends to be quite random if you make frequent readings and > integrate them well, so part of it will cancel out Unfortunately there's nothing constant. The unit is going to be subject to random movement in orientation, speed and direction. IOW, at any given time it could be anywhere in any position. The biggest problem is still power. Even with the best I can find that's readily available, it requires a really hefty battery to run at 80mAh continuously for 20 days. I estimated 130-150 x 1200mAh NiCd AA, which is "slightly" larger than the cellphone size he wanted -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Jinx wrote: {Quote hidden}>>> How large would you expect the errors to be ? He's aiming for >>> 10 - 15m positional accuracy updated every 15 seconds >> >> The GPS errors alone are bigger than that ! In my tests my >> apartment windows moves more than 30 meters ! > > Time for a word with the builder ! > >> The cumulative errors from the acelerometer can get pretty big >> but if the movement is reasonably constant it might work. The error >> tends to be quite random if you make frequent readings and >> integrate them well, so part of it will cancel out > > Unfortunately there's nothing constant. The unit is going to > be subject to random movement in orientation, speed and > direction. IOW, at any given time it could be anywhere in any > position. The biggest problem is still power. Even with the > best I can find that's readily available, it requires a really > hefty battery to run at 80mAh continuously for 20 days. > > I estimated 130-150 x 1200mAh NiCd AA, which is "slightly" > larger than the cellphone size he wanted Jinx, NiMH AA = 1800mA, forget NiCd man! their days are oooover! Try to make a deal with US Government, their error reading is around centimeters, since they have the error correction algoritm (random error sent down on purpose - of course, wouldn't you do it?). They 'probably' may have some nice gps gadgets that can run for more than 30 days with a 3.6V battery pack, under water! What are you laughing at? don't you believe on it? If they can build a sub that can stay down deep for 6 months without refueling, water, air, food, why can't you produce a gps unit that can work 20 days with a AA pack? I know you can. :) /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ Wagner Lipnharski - UST Research Inc Orlando FLorida - USA - http://www.ustr.net /_/_/_/ Atmel AVR Consultant /_/_/_/ -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Hi, > > The cumulative errors from the acelerometer can get pretty big > > but if the movement is reasonably constant it might work. The error > > tends to be quite random if you make frequent readings and > > integrate them well, so part of it will cancel out > > Unfortunately there's nothing constant. The unit is going to > be subject to random movement in orientation, speed and > direction. IOW, at any given time it could be anywhere in any > position. The biggest problem is still power. Even with the > best I can find that's readily available, it requires a really > hefty battery to run at 80mAh continuously for 20 days. It might be worth a try with the acelerometer.. If it is truly random it will cancel itself ! It is almost impossible to do it for long term measurements but to update the position for some seconds may work quite well. If is with a person you also have the phisical limits to cancel noise and random fast movements out. It may work, really.. Power can be very low for acelerometers. Best regards, Alexandre Guimaraes -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> Jinx, NiMH AA = 1800mA, forget NiCd man! their days are oooover! Yes, there are plenty of other types around now, and I was being conservative about the NiCd because of their self-discharge (which may or may not be important but it's usually prudent to under-estimate) > Try to make a deal with US Government My soul already has a "Sold" tag on it. What kind of deal are you suggesting or are you toying with me ;-) > their error reading is around centimeters, since they have the > error correction algoritm (random error sent down on purpose - > of course, wouldn't you do it?). They 'probably' may have some > nice gps gadgets that can run for more than 30 days with a 3.6V > battery pack, under water! What are you laughing at? don't you > believe on it? I don't doubt it for a minute, but presume that GIs get first use of them and the civilian version is some time away. The pdf that Alan posted yesterday has a lot of commentary about what is and what will be regarding GPS chipsets and applications. It seems to me that I could spend a lot of time/money/energy hacking something together from exotic technologies and finish just in time to see squillions of cheap-as-muck bells-and-whistles chipsets hit the shelves -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> and random fast movements out. It may work, really.. Power can > be very low for acelerometers Well, he hasn't told me what the application is, and to be honest I haven't asked - just been answering his general questions. Maybe it's time to get some specific details. It could be that there's an alternative to GPS, which he, as a non-technical person, hasn't even considered because he's latched on to GPS as what he thinks he wants/needs -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> their error reading is around centimeters, since they have the > error correction algoritm (random error sent down on purpose - I think you will find that this is obtained using the L2 frequency and data stream which I believe has a 10x data rate over the L1 data stream. The L1 frequency and data stream is just a "rough get you in the ball park quickly" fix and is the only data stream available to the open market. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

This may be a completely "off the wall" suggestion, but I had watched a show where ships use GPS and LORAN to accurately get their position within a few inches. Depending on the app it might be worth a look. > Unfortunately there's nothing constant. The unit is going to > be subject to random movement in orientation, speed and > direction. IOW, at any given time it could be anywhere in any > position. -- -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Hi all, It seems to me as though some of the information given in this thread is a little outdated (not necessarily the L1 vs L2 freq, I just replied to this email since it was convenient). First of all, selective availability, AFAIK, is now off most of the time (except for an occasional day or so when there is a hightened terror alert or similar). This means that they no longer add intentional jitter to the signal so that now the primary error you have to contend with is atmospheric induced phase jitter. Also, as I understand it, the typical error achieved with GPS is highly dependent on your environment. Alexandre mentioned 30 m, but it sounded to me like he was talking about indoors or very close to a building. From what I've been told, errors of 5 meters or so are now typical in open areas. Now, to get rid of the atmospheric jitter, you can use dual frequency capability (I thought that L2 was usable by the general public, but it was just that the receivers to do this are very expensive and large, but I could be wrong) or you can do differential GPS, which is where you have two receivers and you only want to measure the vector from one to the other very accurately (you don't care as much about their combined position on the earth or if you do, you make one of them stationary and placed at a known position). DGPS, when using carrier phase corrections and done in a wide open area, can get you to the 1 to 2 cm accuracy range, AFAIK. As for using inertial navigation, that has the potential to give you very accurate results even with a fairly poor GPS, but it would end up being expensive and take lots of work. It would also depend on the application (i.e., it may not work well if the motion can be truly arbitrary as you said, Jinx). Sean At 09:14 AM 9/30/2002 +0100, you wrote: {Quote hidden}> > their error reading is around centimeters, since they have the > > error correction algoritm (random error sent down on purpose - > >I think you will find that this is obtained using the L2 frequency and data >stream which I believe has a 10x data rate over the L1 data stream. > >The L1 frequency and data stream is just a "rough get you in the ball park >quickly" fix and is the only data stream available to the open market. > >-- >http://www.piclist.com hint: The list server can filter out subtopics >(like ads or off topics) for you. See http://www.piclist.com/#topics ------------------------------------------- Introducing NetZero Long Distance Unlimited Long Distance only $29.95/ month! Sign Up Today! http://www.netzerolongdistance.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> > How large would you expect the errors to be ? He's aiming for > > 10 - 15m positional accuracy updated every 15 seconds Another option to look into is DGPS (differential GPS). It requires a stationary transmitter transmitting error correction signals and a DGPS receiver. About half of the portable GPS manufacturers offer DGPS units. The FAA has been considering a nation wide DGPS system for precision GPS approaches for several years as soon as they get done negotiating with the military. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> How large would you expect the errors to be ? He's aiming for > 10 - 15m positional accuracy updated every 15 seconds I found out at last under an NDO what the application is. As expected, there's a much better alternative to what he originally proposed. My suggestion to him now is to have the GPS unit fitted with a small AM receiver that can be contacted to request position. It also transpires that even an hourly request is sufficient, not every 15 seconds. As the GPS module itself can be powered-down for long periods this means a substantial reduction in battery power and makes the product viable -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Jinx wrote: > ... The best GPS consumption I could see was > 35mA, > and there's 50mA (not negotiable) on top of that for > the other functions. Which is pretty much battery murder... From http://rd10.net/r.asp?ZXU=5843&ZXD=9948 GPS receiver IC extends battery life The PointCharger SE4100 SiGe GPS receiver consumes one-third the power of competing bipolar solutions-enabling longer battery-powered operation. The part integrates an IF filter, a VCO, tank circuitry, and an LNA onto a single chip with a typical current draw of 10 mA from a 2.7-V supply. . . Housed in a 24-pin LPCC package, the device supports additional power-saving modes offered by its companion baseband chips. ($3.50 ea/10,000-samples available now.) SiGe Semiconductor Ottawa, Ontario, Canada Information 613-820-9244 Fax 613-820-4933 infoKILLspamsige.com http://www.sige.com/ -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.