ECOFLOW and Viofo

[Vortex Radar]

Yeah the timeout feature is awesome. I've got the Ecoflow on the bench right now doing some low bitrate parking mode recording with the A329 2CH. I've got most of my notes and everything ready. Just getting some updated datapoints for the time being.

The timeout feature doesn't work with a USB C port plugged in and the dashcam off. I haven't tested the USB A port yet, but I'll try that out after the parking mode test. If it works on the River 2 as well, I'll switch my custom USB C cable for a USB A one, especially since it would be nice to have the dashcam shut off after X hours without the battery continuing to self-drain.

Besides the larger size and DIY wiring, the only other catches I'm seeing are the longer startup times on the dashcams, especially if the battery fully drains, as well as the self-drain issue that I'm hoping to resolve with a different cable, though that won't be an issue for people who want their dashcams to record in parking mode indefinitely.

By the way, I'm definitely gonna give you a shoutout and kudos in the video, but I wanted to say thank you yet again. This stuff is awesome. 🙂

 

[GPak]

Thank you!
Yep, it takes 2 minutes to start powering the dash-cam from completely dead River3, which is noticeably long time.
But a completely dead battery with a useful capacity of about 190Wh is not such a common situation: in two months, my River3 never dropped below 45% charge.
It only takes 15 seconds to start powering the dash-cam from the battery that is turned off but not completely dead. (This is also due to the soft start, which is actually good)
15 second is kind of OK, by the time I buckle-up, checked the surroundings and ready to go the dash-cam is already recording.
(the dash cam itself (A229 Plus) takes about 20-25 seconds to boot and start recording).

It would be interesting to know how long it takes to start powering the dash-cam under similar conditions for dedicated batteries.

Most USB-A, 5V to 12V step-up converter-cables output around 12.0V.
This is too close to the HK4’s cut-off setting of 11.8V, so a special version of the HK4 cable without a cut-off may be needed.

I actually removed the voltage regulators from the HK4 cables and am supplying 5.0V directly to the dash-cams using a regular USB-A to pig tail power cables instead of a step-up converter-cables.
This way it's more efficient due to the lack of voltage conversion losses of the USB-A booster-cable as well as the HK4 regulator, and so far it is working well, but I still need to make sure it will work in extreme summer temperatures, with no DVR power starvation.

And yes, all power stations have significant self-discharge, I assume due to regulated circuits and more complex electronics, so it is important to use not the maximum declared capacity of the power station, but the useful capacity, tested under the power consumption of the DVR. (It is usually about 70-80% of the declared capacity)

 

[Vortex Radar]

Can you share a bit about your reference to it having ~190Wh of useful capacity? I just finished a parking mode test and got 64.5 hours on an A329 2CH in low bitrate mode. IIRC, I get ~24 hours on your typical 96Wh batteries, or thereabouts. Going from around 1 day to just over 2.5 days is in line with a jump from 96Wh to 256Wh.

 

[GPak]

Interesting!
Dedicated batteries are much more efficient due to their simpler design and unregulated output changing from 14.4V max to about 10.4V min compared to more complicated power stations providing regulated output of around 12.8V.
However, the difference between your result of 64.5 hours and my result of 46 hours is too big, so...
I actually decided to repeat my parking test to see what is going on.
My previous test with a USB-C to 12V cable lasted 46 hours, with the A229 Plus LBR draw averaging 3.85W (current measured with a clamp meter)
My River2 is now updated to a new firmware, and I'm using a USB-C to 15V cable (I can't find the old cable), everything else should be the same.
At the moment, my A229 Plus LBR draw is around 3.98W (0.265A current measured with the same clamp meter)
I don't think the A329, 2ch can be that much more efficient in LBR parking power draw compare to A229 Plus 2ch

 

[GPak]

I arrived at a average usable capacity of about 190 Wh based on:
- LBR parking time of about 46 h, depends on dashcam(s), wiring, settings, etc
- LBR parking power consumption of about 4.1 W on average, depends on dashcam(s), settings, etc
So 46 x 4.1 = 189 Wh
The rest of the 256 Wh capacity is eaten up by River2 self-discharge, 256 - 189 = 67Wh
The self discharge rate per hour would be then 67 / 46 = 1.46 W
These numbers are average, the real numbers depend on specific setup, settings, many parameters, some of which flactuate,
Will see what we get for my specific test setap at the end.

 

[GPak]

Here is a screenshot after exactly 24 hours of dash cam in LBT parking. (time is on upper left corner of the screenshot)
As you can see, there is 68% of capacity left, which is more than I expected!

The test is ongoing, I'll take the next screenshot when the time reaches 36 hours, and thanks to WiFi (love it), I can monitor and do it from work.

 

[Vortex Radar]

Ah great! Perhaps it's a new firmware thing? I just updated my firmware too.

Just finished around of testing on the 96Wh 70mai battery and got 23.5 hours with the A329 2CH in low bitrate. That's right in line with how Viofo's usually fare so it doesn't seem like the A329 is particularly more efficient than other models.

I'm also doing a slow discharge test with an HK4 in the USB C port, but no dashcam plugged in. The battery percentage thing I don't think is super accurate, but if we go by that, so far the River 2 is losing ~8%/day.

 

[GPak]

Of course, I missed the 36 hour mark, was busy at work, and somehow, I missed the exact time, but the River2 shut down between 43 and 44 hours. This is actually less than my previous test of 46 hours.
My power consumption, measured with a current clamp and a voltmeter, was about 4W. This does not include the USB-C cable and possibly the voltage conversion from 5V to 15V (not sure). Including all this, I think the total power consumption might be around 4.1-4.2W. (If I remember correctly, @rcg530 measured about 4.1W for the A229 Plus (2ch) without a USB-C cable in the system).

Between the two tests, the average runtime is about 44.5 hours, the average power consumption is about 4.15W, giving about 185Wh of usable capacity.

The most important and accurate comparison would be between the 44.5h runtime for the River2 and the 23.5h runtime for the 96Wh dedicated battery, the difference is 44.5/23.5=1.89 times or about 89% more for the River2.

The most surprising thing for me was that about half an hour before shutting down the battery was showing about 30% charge remaining, I was under impression that the charge indication was much more accurate, maybe because I never saw the battery go below 45% in my Jeep, usually cycling around 75%?

I hope you find time to retest River2 to verify the 64h runtime.
Your test of the 70mai battery pack showing 23.5h runtime matches to what @rcg530 reported for the A229 Plus (2ch) for a 96Wh battery if I recall correctly.

 

[Vortex Radar]

huh, I wonder why we're seeing such different results. Is there anyone else with one of these batteries who could test?

Even if we go with your numbers, we're looking at 2x the record time for a fraction of the price. Very impressive.

I also saw the battery remaining time being inaccurate. I saw it at 67%, then later to 61% or so, then shortly thereafter it was dead. For this reason, I'm not sure how accurate things would be when using that charge remaining percentage to gauge how much power it's losing per hour/day.

 

[Vortex Radar]

Alrighty, rerunning the A329 2CH test on the Ecoflow. It's giving me an estimated 84 hours right now with the dashcam pulling 6 watts, lol, so we'll have to see what it does over the weekend.

I also ran the A329 on the B-130A and got 24.5 hours which is in line with a 96 Wh battery.

 

[EricSan]

Is there anyone else with one of these batteries who could test?

I have an EcoFlow River 2 and a Vantrue E360 3-ch camera that I can run to see how long the battery holds up. I’ll charge up the R2 and see how long it lasts, then compare to my LTO battery. This will likely take 4-5 days for both batteries to run down.

 

[GPak]

I ran another test with exactly the same conditions, and this time I got an accurate River2 shutdown time of 45 hours and 54 minutes, or 45.9 hours. That's better than my second test, and very close to my original test of 46 hours.

The USB-C cable definitely draws some power, I measured it to be about 3-4°C hotter at the port connection than the surrounding area of the River3.

I ran this test because I updated the firmware right before the second test, and because some sophisticated battery management systems can learn from multiple cycles to optimize and show more reliable data.

And sure enough, this time I got 22% and about 10 hours remaining displayed, when River2 shutdown, which is considerably better than the previous test!
(Hopefully it will farther improve with more cycling)

Next I will test the River3 under the same conditions.

 

[EricSan]

Here is some power consumption and battery longevity data from my EcoFlow River2 and Vantrue E360 3-ch camera.

The River2 is a 256Wh battery that I made sure was fully charged before I started making any measurements.

I measured the E360's camera's parking mode power draw with two different values of inline precision power resistors (which showed 3.75w and 3.85w respectively) for an average power draw of 3.8w with 3 channels running in low bitrate 1920x1080 @15fps. The display on the River 2 showed 4w of power draw, which is pretty close to my measurements. Power was drawn directly from the USB-A port.

After a few minutes, the E360 switched to parking mode and the River 2 showed 54h of remaining capacity.
At the 12h mark, the River2 showed 38h and 76% charge remaining.
At the 24h mark, the River2 showed 27h and 54% charge remaining.
At the 36h mark, the River2 showed 15h and 30% charge remaining.
The River2 turned itself off at 38.5h.

So, with 3.8w of power draw, my EcoFlow 256Wh River2 lasted for 38.5h. This works out to 146Wh of usable battery capacity (57% of stated capacity). I don't know how old the actual batteries are because my River2 was purchased as a refurb (I think I only paid $120-130 for it), so I don't have any history info.

When I plugged the River2 into the wall outlet to start recharging it, the display indicated 25% remaining capacity (guess the SoC calibration was off) and it took the dashcam 10s to turn on.

Next, I'll do the same with my DIY LTO battery and see what kind of results I get. Interestingly, the BMS in my LTO pack shows 3.6w of power draw for the E360, so it’s pretty close to my measurement of 3.8w.

 

[GPak]

Thanks for the test @EricSan
I think a 38.5 hours of parking would be very good for a 3ch dash cam with LBR, but not for a power consumption of only 3.8W.
Can you please describe your setup and clarify the LBR power consumption of 3.8W for the E360 (3 ch).
- Was the dash cam connected using the HW kit, and if so, what was the cut-off setting? (Sorry, I'm not familiar with the E360, not sure if it comes with the HW kit).
- Was the dash cam and HW kit, if applicable, powered by:
12V USB-C cable, or
15V USB-C cable, or
12V USB-A converter cable, or
5V direct USB-A system

If the HW kit and/or any of the above cables were used, are they included in the measured 3.8W power consumption?

And I assume that River2 is set to 0%-100% charge and discharge limits by default.

 

[EricSan]

38h is good longevity overall, but I was surprised it wasn’t longer given the low power draw. The Vantrue camera has a USB-C port for power, so the camera was connected directly to the River2 USB-A port with the Vantrue provided USB-A to USB-C cord. I did not use the 12v to 5v regulator hardwire kit, this was just a straight 1-foot long 5v USB-A to USB-C connection. (Bummer, no regulator losses to account for)

The measured power draw includes the 1 foot USB power cable. The River 2 was fully charged and is set to charge to 100% and discharge to 0%. I was expecting to see many more hours for a 3.8w draw. This is why I said I am unaware of the history of my River2 because it was purchased as a refurb. I’m guessing the batteries are not running at 100% capacity anymore. I have another River2 to test (I originally purchased two), but I’m not sure I’ll get that opportunity - the Vantrue camera is going in my daughter’s car by the end of this week (spring break) for her drive back to college.

 

[GPak]

Completed the River3 parking duration test under exactly the same conditions as for the River2.
The Dashcam is A229 Plus (2ch, F/R), LBR parking mode, GPS is set to off for parking, a standard HK4 hardwire kit is used with a cut-off at 11.8V, powered via a USB-C-15V cable. (note that the HK4's 11.8V cutoff is irrelevant for this test, as the USB-C to 15V cable supplies around 15V at all times)

The average measured power consumption for the Dashcam + HK4 was about 3.9-4.0W without the USB-C-15V cable, with the USB-C-15V cable it is estimated to be 4.1-4.2W. (Assuming a 95% voltage conversion efficiency of the cable)

For this test I got the exact shutdown time of the River3 at 47.8 hour mark (47h and 48m).
This result is almost 2 hours better than 45.9h for the River2 !!
And this is despite the River3 being rated at 245Wh compared to 256Wh for the River2, meaning the River3 is more efficient !!

The average usable capacity is about 4.1W x 47.8h = 196Wh, or about 80% of the rated 245Wh.
The remaining 49Wh capacity is eaten up by River3 self-consumption or self-discharge.
The self-discharge rate per hour would be then 49Wh / 47.8h = 1.03W (compare to 1.46W for the R2)

The best part is that the App/displayed remaining charge percentage is pretty accurate !! (the remaining time isn't, but I don't use it anyway)

Also, once I started charging a completely dead River3, it took around 1.5 minute to start supplying power to the dash cam. (previous test was around 2 min)

Here is recorded data and screenshots: (as attachments are limited to 10 images - see next post for the rest of the images).

 

[GPak]

The rest of the images from the previous post are here:
The measurement time is indicated in the upper left corner of the image:
Note that the 45 hour mark is missing because I got distracted and missed it, I think it would have been 8%.

 

[Vortex Radar]

Just finished an A329 Pro LBR retest and this time around my results are more in line with yours. I got 40 hours and 13 min.

 

[GPak]

Yep, 40.2 hours is more in line with what @EricSan and I got.
Of course, we all used different Dascams, different test setups and wiring, which affects the results, but the ballpark numbers are close.
@Vortex Radar this is just a suggestion, but maybe you could contact Ecoflow and ask for a River3 test unit.
I think they would be interested, the Dashcam market is still relatively small, but for them it is completely new.
Based on the review, they might listen to recommendations on how to improve the River3 for dash cam use.
The reason is that:
The River3 is a new product, while the River2 is outgoing one.
The River3 has better functionality, is smaller in size and is more efficient with a longer parking time compared to River2.

 

[EricSan]

It bears noting that EcoFlow units have a max working temp of 45c/113F. They mean it, one of mine fried after a few consecutive days in a car with outside temps of 35c/95F - I don’t know what the temp inside of a dark blue car was. I needed to send it back for repair. This is why I built the LTO pack that @GPak designed, it has a much wider working temp range. So far, my LTO is proven from -22c/-8F (outside temp) to 55c/131F (in car temp). High temp that day was probably 92-95F/32-35c.