Karel details plans for Power Mangement

[BKeeper]

I just received my DR!

I don't know if this has already been posted.

Here is an official statement from karel indicating plans for DR power management implementation:

During my absence there have been a few topics in regard to the battery life. In order to maintain an overview, I have locked all these topics and redirected the discussion to this new thread, where I would like to start with giving a bit more insight in the power consumption and battery life of the iRex Digital Reader 1000.

In the current software version 1.0, we have only implemented very limited power management. While the iRex Digital Reader has a hardware architecture that is capable of utilizing a number of advanced power management functionalities including a suspend-to-RAM (standby mode) and suspend-to-Flash (hibernate mode). Once these are implemented the battery life will effectively be several days.

Although the following data is not 100% correct, it should give you a good view on what to expect of the battery life, so please do not pin me down on this data.

The iRex Digital Reader has a battery of 1300 mA (milliampere) and with software version 1.0 the reader consume about 150 mA per hour, resulting in about 8 hours of battery life. With suspend-to-RAM or standby mode, which we will introduce in a next software update, the reader will go into standby mode after x seconds. While in standby mode the power consumption will scale back to less than 50 mA. Good for about 24 hours of standby time.

In addition to the standby mode we will also be looking to introduce a sleep mode or suspend-to-Flash. This means that if the reader is not being used for x minutes it will go into sleep mode. While in sleep mode the power consumption is reduced to about 1 mA. To 'wake up' the reader from hibernate takes about 10 seconds.

In the coming days we will come back to you with more information about our software road map and what software updates you can expect to see in the future.

IMO this indicates something is seriously wrong (again) with their hardware.

Now they are talking about 24h of standby. Not that this figure is not for continuous use. I would expect 12 hours. I don't like that.

Then in March 09 iRex will obviously launch a revision with a larger battery.

Ah don't you love iRex...

[Nate the great]

In the current software version 1.0, we have only implemented very limited power management.

Any portable device that doesn't have power management cannot be version 1.0; it isn't done yet. At most it can be version 0.9.

[ericshliao]

Well, that means we may have a DR1000 2nd edition just like iLiad.

[bob_ninja]

I find it curious that in standby mode (suspend to memory) it still uses 50 mA compared to 150 mA while operational. That sounds high. It implies using only 66% less power in standby!!!! I wonde how much power is needed for memory alone?

Looking back at Cybook measurements:
https://www.mobileread.com/forums/sho...t=25555&page=2

Power input of the processor, sleeping mode: 0.04 W 87 hours for fully loaded battery
Power input of the processor, running mode: 0.53 W 7 hours for fully loaded battery

The "sleeping mode" is same as standby (as all the state is in memory) so should be equivalent to DR1000 standby. Note power consumption is more than 10x lower!!!!

Assuming 3.7V, Cybook draws 0.53 / 3.7 = 143 mA while active (comparable to DR), but only 0.04 / 3.7 = 11 mA while sleeping!!!!!

They will either have to fix standby mode power use or add another 1300 mAh battery. Otherwise DR users will have to recharge anywhere from every day to at least once per week, even AFTER the power mgmt modes are implemented!

[BKeeper]

Originally Posted by bob_ninja

I find it curious that in standby mode (suspend to memory) it still uses 50 mA compared to 150 mA while operational. That sounds high. It implies using only 66% less power in standby!!!! I wonde how much power is needed for memory alone?

Exactly!. The CPU consumes waay too much power even in standby.
That's why I posted this, because I can't belive iRex is presenting this numbers as a solution. It's unacceptable.

This was the one thing they had to get right after the humiliation of having to lauch a battery replacemt program for the original iLiad.

Funny, when I placed my order I convinced myself by thinking that no matter how bad battery life was, it would always be better than the iLiad, but it seems I underestimeted iRex.

[bob_ninja]

BKeeper,
I posted in the other thread:

Now according to these numbers, "light reading" used up internal battery after 5 hours. Of course, since battery is never depleted fully, I will guess/estimate about 1000 mAh (to be conswervative meaning 300 mAh still left).
1000 mAh / 5 hrs = 200 mA

However, Karel in the other thread stated:
"with software version 1.0 the reader consume about 150 mA per hour"

So once again they seem to be underestimating true power use. Seems like a pattern of overestimating device capabilities.

I suppose battery could be defective or not fully broken in, but highly unlikely. A more likely explanation is that the current software version is still not optimized for power use and subsequent versions might clean up these rough edges.

However, I wouldn't be too quick to give up on DR!!! As the other thread points out, you can have some AAs recharge DR and extend runtime a great deal when away from a PC. I think the entire battery/runtime issue needs to be measured against the benefits (like PDF handling which is far superior to my Cybook). Perhaps its features are more important than the short runtime.

USB power packs using AAs should increase DR runtime many times, depending on how many AAs you are willing to carry. So there is a reasonably painless option (no hardware mods needed) to fixing it.

On the other hand Cybook now has zoom and scrolling for PDFs to address A4 PDF issue. However, it is still very clumsy (hard to scroll) and it still cannot handle large/complex docs that DR seems to be able to handle.

Anyway, I guess iRex (mis)information doesn't help their image, so I understand your frustration.

[wallcraft]

All the reports we have seen indicate that the following is a wildly off the mark:

The iRex Digital Reader has a battery of 1300 mA (milliampere) and with software version 1.0 the reader consume about 150 mA per hour, resulting in about 8 hours of battery life.

The DR doing nothing at all lasts about 5.5 hours or about 235 mA per hour. We know this isn't due to a faulty battery, because allovertheglobe got entirely consistent power drain from an external battery pack, see Power on the go - one practical solution. Perhaps karel is talking about the, unreleased, DR1000. Does anyone know how much power turning off Wacom saves?

[Morffius]

I can't wait to see what adding wifi and Bluetooth to the mix will do. What should we expect then, a normal runtime of 2hrs!?

[bob_ninja]

LOL
Forgot WiFi/Bluetooth
Ouch, not sure how they'll explain very short runtime when wireless is enabled?!??!

[pilotbob]

From reading iRex's "official" statement from marketing (we know how reliable those are) I feel they made a few basic mistakes.

1. The battery they use for the device is much to small. They should have used something twice as big. (Or at least make it user swappable so you could put in a second battery rather than waiting for a charge.) They should have planned for battery size for at least 12hrs of continous non-standby use. Would this have made the device heavier.. sure but a 1/2 pound or so... but ask most if they would like .5lb less or more battery life?

2. They are running the CPU at full speed. I doubt the reader needs the CPU to be at full speed most of the time. Step down the CPU voltage unless the reader is doing something really important like initial open/index of a book. They can run at full speed if the thing has external power. Perhaps the CPU they use doesn't support this?

3. Shipping the device before the battery management features were done. I know shipping is a feature... but it is much harder to fix a bad reputation than loose a few months of sales. The few months will end, the bad rep will go on for a long time.

BOb

[Shaggy]

Originally Posted by bob_ninja

LOL
Forgot WiFi/Bluetooth
Ouch, not sure how they'll explain very short runtime when wireless is enabled?!??!

The way the iLiad works is that Wifi is only enabled when you're connecting to the network to download something. It is disabled the majority of the time.

Edit: I assume the DR will be similar.

[Shaggy]

Originally Posted by pilotbob

2. They are running the CPU at full speed. I doubt the reader needs the CPU to be at full speed most of the time. Step down the CPU voltage unless the reader is doing something really important like initial open/index of a book. They can run at full speed if the thing has external power. Perhaps the CPU they use doesn't support this?

They mentioned on their forum that the CPU they are using is already very efficient, and they did not see a significant difference when they throttled it down. I think they did mention that they were still planning on throttling in a future update since any little bit helps, but that it wasn't a dramatic difference like it was on the iLiad.

[axel77]

IMHO this battery issues is a sign Irex is going down the drain. I really liked the open platform devices they provided, but with managing decission like this, to bring out devices unfinished like this.... If they don't get straight fast, I personally would not see a bright future...

[Morffius]

I thought there was meant to be a browser on the Reader as well, which would probably mean continuous WiFi.

Originally Posted by Shaggy

The way the iLiad works is that Wifi is only enabled when you're connecting to the network to download something. It is disabled the majority of the time.

Edit: I assume the DR will be similar.

[BKeeper]

Originally Posted by bob_ninja

However, I wouldn't be too quick to give up on DR!!! As the other thread points out, you can have some AAs recharge DR and extend runtime a great deal when away from a PC. I think the entire battery/runtime issue needs to be measured against the benefits (like PDF handling which is far superior to my Cybook). Perhaps its features are more important than the short runtime.

I'm certainly not giving on the DR just yet. I bought it to read A4 and it's great for that.

But iRex cannot make the same mistakes all over again, and they have.
Wanna hear something really scary, I have it on good authority that iRex is explicitly using the term beta firmware with their retail chain.

Do you realize the significance of this?
This makes us beta testers, which I'm not, beta testers don't pay for product testing, they get the product for free (maybe with a deposit) and sign NDAs.

They simply cannot do that. (again)
The first time they were protected because most of the customers signed a release. But this time I haven't and this is false advertisement.
No need to lie. They could have said look here is our reader we expect 12h of battery life. and with that info it was up to me to decide if it was worth €600. But by inflating perceived value they are effectively stealing from me.

Again, I got my DR because I thought that given iRex's history they simply could not afford to get power management wrong, again, to lie to their customers (again),

I thought they'd rather ship the DR with a car battery if they had to.

Now it looks like we are the ones getting external batteries anyway.

I'm not giving them a free pass this time, either they fix it. or give partial refund, or I will notify EU's consumer protection authorities.

I'm not saying I don't like my DR, I'm just saying that iRex simply cannot do this. They should have launched DR in December.They have rushed the product and they shouldn't have.