More Inverter Musings

[Maine Sail]

Just messin' around in the barn this afternoon and compared the factory ASUS 12V adapter vs. the factory ASUS 110V adapter run through a 150 watt inverter. This was powering my ASUS Net Book

Running the 12 volt adapter @ 62% of charge / 12.34 volts:

Running a 150 watt inverter and the ASUS 110V wall adapter @ 63% charge /12.33 volts (same programs running):

The Link 10 is not always agreeing with the clamp meter, could just be a speed response time issue, so I need to check my connections and make sure they are clean. Of course I did replace it because it was "buggy" to begin with.. The clamp meter is usually very accurate. Either way I have run both scenarios..

If we use the clamp meter it was 3.12 amps vs. 4.12 amps or a 33% increase in consumption using an inverter and 110V adapter vs. the 12V adapter.

If we use the Link 10 it was 3 amps vs. 3.8 amps or a 27% increase in consumption using an inverter and 110V adapter vs. the 12V adapter.

Not all 12V adapters are created equal. The factory ASUS adapter is designed to always feed the laptop 12 volts whether it sees 10 volts or 18 volts. This computer is quite finicky about it's 12V power input and it needs to be 12 volts not 14.2 or 10.5. The frustrating this is that I own about 5 12 volt adapters including a Targus and an iGo neither of which are compatible voltage wise. Most laptops can run on more flexible voltages but this one can't.:cussing: $349.99 for the computer and $46.99 for the ASUS 12V adapter..

 

[Roger Long]

This computer (Asus Eee PC) is quite finicky about it's 12V power input and it needs to be 12 volts not 14.2 or 10.5. The frustrating this is that I own about 5 12 volt adapters including a Targus and an iGo neither of which are compatible voltage wise.

I'm interested in your experience with this because I have the same computer. I bought an after market 12 v adapter on line that said it was for this computer and plugs into a 12 V lighter type outlet. Both in the car and on the boat, trying to use it results in the pop up window that tells you that you have either plugged in or disconnected external power flashing on and off constantly in the middle of the display. This tends to make navigation difficult because the boat stays centered in the display and is hidden by the flashing window.

The only work around I have found is to remove the battery. Risky in case of surges but it's a cheap computer and I've run it for many hours both in the car and on the boat without a problem. I've never had it glitch or freeze while running this way although I'm careful not to start or stop the engine with it running. It does make stopping for coffee a pain in the car though because, as you probably know, it takes this little machine about 20 minutes to get itself up and running (or, at least, it seems like it).

 

[Sumner]

Nice test, thanks. I hate running inverters since they waste so much electricity themselves. If I have to run one I'll get the smallest one I can do do the job.

I also don't understand why most laptop computers (we have 2 Dells) don't operate directly on 12 volts. Our Dells are 19 volts and I'll almost bet that the computer you have there doesn't work on 12 volts either. The 12 volt adapter is probably still using electricity to get the volts from 12 to whatever the computer needs.

I hope to setup a 'true' 12 volt computer..............

http://metanomy.org/node/9

http://www.mini-box.com/site/index.html

...........in the boat in the future. They are designed to be used in an auto or marine environment where you have low voltage during engine starting and also temperature extremes. They should be a lot better on power usage and you can configure them like any other computer and they could be mounted in an unused location to save space.

c ya,

Sum

Our Trips to Lake Powell, UT - Kootenay Lake, Canada - Priest Lake, ID

Our Mac Pages

Mac Links

 

[Maine Sail]

I'm interested in your experience with this because I have the same computer. I bought an after market 12 v adapter on line that said it was for this computer and plugs into a 12 V lighter type outlet. Both in the car and on the boat, trying to use it results in the pop up window that tells you that you have either plugged in or disconnected external power flashing on and off constantly in the middle of the display. This tends to make navigation difficult because the boat stays centered in the display and is hidden by the flashing window.

The only work around I have found is to remove the battery. Risky in case of surges but it's a cheap computer and I've run it for many hours both in the car and on the boat without a problem. I've never had it glitch or freeze while running this way although I'm careful not to start or stop the engine with it running. It does make stopping for coffee a pain in the car though because, as you probably know, it takes this little machine about 20 minutes to get itself up and running (or, at least, it seems like it).

Roger,

The adapter you are using is putting out the incorrect voltage and thus the computer accepts charge, denies charge, accepts, denies. This is an odd computer in that is has no tolerance for flexible voltages, as many do. People will try to sell you anything they can even if it may potentially cause harm to your computer.

I actually spoke with ASUS tech support about this and even iGo does not have an adapter that will work, only the ASUS 12V adapter. I'm sure someone makes one that actually works but I don't like to be a lab rat and do the testing myself, as I have already done once off eBay..

I have tested the ASUS adapter and it puts out the same voltage at 10.4 volts as it does at 14.4 volts a steady 12 volts.. If you bought one of those $12.95 12V adapters from eBay or one of the resellers of all things "knock off" I would consider discontinuing its use and going back to an inverter so you don't damage the computer.

Here is the actual email correspondence & response I got from ASUS:

Hi,

I am trying to find out what the acceptable input voltage range of an eeePc 1000HE would be?

1- Does it need to be 12 volts constant or can it work within a range of say 11 - 15 volts?

2- Does the DC power supply need to be regulated to a constant 12 volts? Will the 14.4 volts my car puts out when running damage the 1000HE ??

3- In short, is there an acceptable "range" of DC input voltage or must it be 12V fixed constant?

Here is the ASUS tech support response:

Dear Sir/Madam

It needs to be a constant 12v.

Thank you for your support of our products!

After this I called them and spent nearly an hour on hold and then was bounced to multiple individuals only to find out that this is in fact true, and the computer is very voltage intolerant. It was explained to me that this intolerance is partly how they make it so efficient and is but one of the reasons the battery can last 7 hours or more. I don't really understand the inner workings of computers so I have to take them at their word.

I too tried an knock off eBay POS ,"works with the ASUS Netbooks" charger, and had the same issues you had. "Plugged into external power source" on/off etc.. I took a screw driver, opened the adapter up, and the only thing inside was a little diode for the LED indicator light. There was NOTHING in the signal path what so ever to regulate the output power.. For the 10.95 I spent it went straight into the trash as a lesson learned...

 

[Roger Long]

The adapter you are using is putting out the incorrect voltage and thus the computer accepts charge, denies charge, accepts, denies. ... If you bought one of those $12.95 12V adapters from eBay .... It was explained to me that this intolerance is partly how they make it so efficient and is but one of the reasons the battery can last 7 hours or more.

I measured the voltage of the one I bought to be exactly 12 V with 13-14 V input from the boat and also from the car which I didn't check for input. Meter draws less than the computer though so it could be that it sags in real operation. Ten bucks is about what I paid before shipping. It's glued together but I can see what looks like a small power supply through vent holes.

Your battery lasts 7 hours? One of the disappointments of the EeePC for me was that the battery lasts about 30 minutes.

I couldn't locate the ASUS adapter on their web site. That's why I bought the cheap one. If you have a link or part number handy I'd appreciate it.

 

[Maine Sail]

I measured the voltage of the one I bought to be exactly 12 V with 13-14 V input from the boat and also from the car which I didn't check for input. Meter draws less than the computer though so it could be that it sags in real operation. Ten bucks is about what I paid before shipping. It's glued together but I can see what looks like a small power supply through vent holes.

Your battery lasts 7 hours? One of the disappointments of the EeePC for me was that the battery lasts about 30 minutes.

I couldn't locate the ASUS adapter on their web site. That's why I bought the cheap one. If you have a link or part number handy I'd appreciate it.

Sounds like you got a slightly better eBay charger than I did..


Google this part number: 90-OA00CA1100

You can buy it at New Egg: http://www.newegg.com/Product/Product.aspx?Item=N82E16834998923


I got mine off Amazon as it was the best deal at the time with free shipping...

 

[wonkodsane]

It makes me wonder if you couldn't build a cheap voltage regulator to incorporate into the circuit. See: http://www.rc-cam.com/dc-dc.htm

At least it isn't like the blackberry chargers, which won't charge off of standard USB ports without the BB drivers. Default USB port power is ~300mA and the BB "requires" 500mA to charge. The fix is to tell the computer to up the mA, and this is exactly what the BB drivers do.

The Admiral's Eee works about 10 hours on a charge, but hers is brand new.

 

[Don12364]

Clamp On Ammeter

Main,
Most clamp on ammeters only measure the AC component of current via the clamp. Those with a hall effect sensor do read the DC component but I have never been sure it is only the DC component and I worry that the reading may include some AC content.
Your 12v/110V adapter will likely be a switch mode unit so will be biting great chunks out of the 12V source battery so has a large AC element in its current.
I tried to read the type number of your meter but the photo was too pixellated. I recognise that the clamp meter read lower than the Link but I wonder what that reads too!

 

[Maine Sail]

Main,
Most clamp on ammeters only measure the AC component of current via the clamp. Those with a hall effect sensor do read the DC component but I have never been sure it is only the DC component and I worry that the reading may include some AC content.
Your 12v/110V adapter will likely be a switch mode unit so will be biting great chunks out of the 12V source battery so has a large AC element in its current.
I tried to read the type number of your meter but the photo was too pixellated. I recognise that the clamp meter read lower than the Link but I wonder what that reads too!

Actually the clamp meter read slightly higher than the Link 10, no that that matters much. When we are dealing with roughly 30% more energy consumption going from 12v to 110v back to 12v vs. just keeping it at 12V a few tenths here and there are most likely not a big deal..

I was reading up stream of the inverter. That black wire was coming directly off the back of the cigarette lighter. I also checked it on the hot lead and the reading was identical. Perhaps there is some dirty AC component in there but my meter has always been spot on when compared with DC specs on know quantities.

Either way we have inverter inefficiencies and converter inefficiencies going from 12V to 110V then back to 12V because this computer actually runs on 12V not 110V.

Meter is a Mastech MS-2108 (LINK).

If a Link 10 and a good quality DC clamp meter are not capable of measuring current draw, what is??

 

[walt]

Maine sail, interesting and useful experiment. I'm somewhat surprised about the 27% difference..

Last spring, I powered an about 150 watt "hardware store" inverter off a 12 volt battery and ran a light bulb. I put a shunt (very small value resistor) in the 12 volt DC line and took a O scope snap shot of the voltage accross the shunt which is equivalent to the current waveform. You can see that twice per AC cycle, the inverter "whacks" the 12 volt DC source to create the AC waveform. In this case, the light bulb power draw was a good percent of the inverter power rating so the duty cycle is high.

I've seen some expensive Lab instrument not agree with each other when measuring RMS current of a somewhat similar waveform (the AC input to a switcher power supply). However, I'm guessing that the battery monitors probably do a reasonable job with this since it would not be that hard to do.

I think Donalex could add to or correct this but I'm assuming that the battery monitor uses a differential voltage measurement (over the twisted pair wire) of the current shunt. This voltage is first filtered to at least 1/2 of the sample rate bandwidth and then a micro with an A/D converter samples the filtered voltage. The software integrates and scales the readings to get total charge (amp hours) and also displays a running average (amps). The sample rate is much faster than the display update. If the sample rate is fast enough (in the case of the waveform shown, there are 1 msec rise times), the battery monitor should do a pretty accurate job of determining how much charge is actually used by the battery. In the case of the waveform shown, I think inexpensive hardware would fairly easily be able to sample fast enough to include the 1msec details of the waveform. Since its inexpensive, good chance that battery monitor hardware is at least this good - and possibly a lot better.

Don't know about the current meter, but I would certainly believe what the battery monitor says - or at least its relevant to our problem of understanding battery draw from inverter. The 12 volt DC to DC converter could have much faster transients.. but I think Id still beleive what the battery monitor was measuring.

 

[Roger Long]

Google this part number: 90-OA00CA1100
You can buy it at New Egg: http://www.newegg.com/Product/Product.aspx?Item=N82E16834998923

It came and, sure enough, it solves the problem of the battery message blinking on and off. I tried the old adapter and, it even fixed that one! Just having the new adapter in the house evidently scared it into acting properly with the old one. Go figure.

I'm glad to have the new one though as it's better sealed against water and it's reassuring to have the proper ASUS part. Thanks for the link.

 

[Maine Sail]

It came and, sure enough, it solves the problem of the battery message blinking on and off. I tried the old adapter and, it even fixed that one! Just having the new adapter in the house evidently scared it into acting properly with the old one. Go figure.

I'm glad to have the new one though as it's better sealed against water and it's reassuring to have the proper ASUS part. Thanks for the link.

Glad I could help. I just thought of another option for your battery life.

When running on battery power right click on the battery icon in the lower left.

Select, ADJUST POWER PROPERTIES

Under POWER SCHEMES hit the drop down and select MAX BATTERY.

You can also manually adjust how your computer performs while on the battery..

 

[walt]

(techie stuff.. but a few might find this interesting)
Something I’ve wondered about for a while is how well do the battery monitors work with measuring “chopped DC” current. Why care about this? Because a lot of the devices used in boats have some level of “rapidly” changing DC current when they are operating and the battery montor must measure this. For example, the outboard alternator after the rectifier has current which increases and then collapses each time the a coil moves past a magnet. In my case with a magneto (small outboard), this current rises and collapses three times per revolution resulting in the current changing 250 time per second at 5000 RPM (current is always flowing into the battery because of the rectifier – but its rising and falling at 250 times per second). A pulse width modulated solar charge controller also results in chopping the DC current to the battery as a way to limit the charging. The image of the current waveform from an inverter in my previous post shows the input current getting “chopped” for each cycle of the AC waveform. The speaker in the stereo needs rising and falling current in order to produce the sound pressure modulation and even with some internal filtering, the stereo will to some extent modulate its current input with the music being played (watch your meter while music is playing). The devices Mainesail used in his test (inverter and DC to DC converter) both create chopped DC current waveforms. My lights are about the only thing I use which draw a steady current.

So I did a test on my linklite to see how well it handled a chopped DC current. I found out what most everyone here already knew, the battery monitor worked fairly well. (FYI, maybe your wondering why I did this - I design analog electronics for a living, this was something I needed to know). If you don’t read any further, I think the linklite does a pretty accurate job of determining how much charge flows in or out of the battery.

The test.. I made a simple circuit composed of a 10 ohm resistor and an electronic switch (N channel FET). The switch basically either connected the 10 ohm resistor to the 12 volt outlet or disconnected it. I had a lab signal generator which allowed me to change the frequency of the connect/disconnect and also change the duty cycle. I used three setting of duty cycle 20% (i.e., 20% on, 80% off), 50% and 80%. The picture of the scope shows the duty cycle.

My battery voltage was 12.8 volts at the 10 ohm resistor so when the resistor was connected to the voltage, I would have 1.28 amps flowing. When the load was switched so that it was on 80% of the time and off 20% of the time, the “average current” (and equivalent to RMS) would be 1.28*.8 = 1.024 amps. At 50% on/off, the average current would be 1.28*.5 = .64 amps and at 20% (i.e., off 80% of the time), the average current would be 1.28*.2 = .256 amps.

For the 10 ohm load not switched at all, the linklite showed 1.3 amps. Note that the resolution on the meter is only .1 amp.

Frequency / 20% / 50% / 80% (20, 50, 80 refer to the duty cycle of the load)

10hz / .3amp / .7 amp / 1.0 amp
100hz / .3 amp / .7 amp / 1.0 amp
1Khz / .3 amp / .6 amp / 1.0 amp
10Khz / .3 amp / .6 amp / 1.0 amp
100Khz / .2 amp / .6 amp / .9 amp
300Khz / .2 amp / .5 amp / .8 amp

I would consider this good results. 300Khz is much higher than anything the linklite actually needs to measure.. however, some DC to DC converters work up to near 500Khz so I threw this measurement in.

Next test, was to see how well the linklite determined accumulated charge (i.e., amp hours) for a chopped current. So I set the modulated load for 1Khz at 50% duty cycle. I.e., the load was on half the time and off half the time.. For the average voltage conditions during the test, the current dissipated was (12.15 volts / 10 ohms ) * .5 = .6075 amps.

I left this 1Khz switched current on the battery / monitor for 9.916 hours. The calculated amp hours would then be 9.916 hours * .6075 amps = 6.02 amp hours.

The linklite showed 6.6 amp hours. Not exactly the same… but given some minor inaccuracies in my test, I would say the linklite probably did a better job measuring than I did.