Power On but no Auto Record

[Isoprop]

My Mobius has a battery in it, and it should be perfectly capable of providing the necessary voltage to the board with it. As to the charging circuit, if this isn't a cheap design as you seem to imply, it should be designed well enough to withstand voltage spikes and/or drops that a car can have, otherwise the device shouldn't be marketed as a dash cam, should it? And what comes to the I/O pin sniffing for the voltage, low voltages are no issue and high voltage spikes, if any, can be handled to an extent, if the board is properly designed.

As you know, the camera is not specifically designed as a DashCam with interference suppression circuitry, voltage stabilizers, large smoothing capacitors and all the other electronics necessary for a traditional DashCam. It's just not possible to build all this extra circuitry into such a small case. The Mobius also doesn't have a G-sensor or GPS receiver or monitor, so I think it should be obvious that the camera isn't a typical DashCam but rather a camera designed for many usages.
The Mobius was designed to accept a stabilized and clean power supply of 5V following USB standards. It was NOT designed to accept a 'dirty' power supply. The manual states very clearly that you need to use a quality charger.
If there is a battery in your Mobius then you must ensure a clean and stable 5V supply. If this is not the case the charger IC chip will constantly switch it's charging circuitry between on and off. These fast on/off charging cycles will not only heat the charging IC but will also heat and destroy the internal battery and possibly destroy the Mobius at the same time. In the past, bad 5V chargers have been the cause of destroyed and melted cameras.

 

[mobnam]

The Mobius was designed to accept a stabilized and clean power supply of 5V following USB standards. It was NOT designed to accept a 'dirty' power supply. The manual states very clearly that you need to use a quality charger.
If there is a battery in your Mobius then you must ensure a clean and stable 5V supply. If this is not the case the charger IC chip will constantly switch it's charging circuitry between on and off. These fast on/off charging cycles will not only heat the charging IC but will also heat and destroy the internal battery and possibly destroy the Mobius at the same time. In the past, bad 5V chargers have been the cause of destroyed and melted cameras.

Given that I don't seem to have any issues with auto-start nor the 8 inch extension cable, when I have the mobius hooked up to my PC, this could indeed all be related to the psu. I have tried twisting the sensor cable similarly as in the car installation but it seems to work fine when hooked with the pc.

I need to dig out my pocket scope some day and see what is going on in the external power line when the device is hooked up to the car. Or maybe I will try first with a laptop in the car as an alternative psu.. If that takes care of the image quality issue, then I have a root cause. If the IQ issue remains with the laptop as the psu, then I suppose there must be some other interference hitting the 8 inch cable.

Regarding the psu which I purchased from JooVuu, it is tiny. It can't have big capasitors in it, but I can add some to the output. I saw one 2200uF and one 1000uF cap in my component stash, so I can try those. If I'm lucky, that 3200uF of capacitance will sort out both issues.

I will try these when I'm back with the car.. In a week or so.

 

[mobnam]

I took the cable apart and opened every connection, turned the adapter around and even turned the longer cable around. And I did not stop there, I gave the inside of the Mobius casing a bit of aluminium tape, continued on over the whole cable length to the sensor back-plate, and on top of that I wrapped the cable then with black insulating tape. Tin-foil stuff, probably only had a placebo-effect..

Then I went for a test drive and to my surprise, there were no issues with the IQ, except my lens was unfocused, and there was the usual rather bad dash-reflection which I suppose is a feature of the wide-angle lens and the car, more than Mobius. Now I focused the lens again and am left with the reflection and startup issues.

For the latter, I recorded one engine startup from the ecu fuse, with my "proudly-made-in-china" pocket oscilloscope, check out the "paushed" picture.

Looks like there is only a very short drop of voltage, and it should be fairly easy to filter that out. I suppose I need to take a measurement also from the mobius power line input, basically from that hardwire kit output, to see if it actually triggers the scope by dropping as well.

 

[mobnam]

Now I measured the actual Mobius power input during starting of the engine. Drops all the way to 1 volts, so that's it then. That drop lasts for around half a second during these fast summer startups.

Poor Mobius fw doesn't know how to check the voltage another time after a few seconds, combined with a hardwire kit which doesn't provide much of capacitive buffering.. and you have a non-working mobius. I'll see if those two caps I have, make a difference.

 

[mobnam]

Ok, I experimented with the caps now. I soldered two caps (1000uF and 2200uF) to a small circuit board, totaling 3200uF, and of course two leads out of the board for connecting. Using test cables with clamp type of ends to connect it all, I started the engine and recorded with the scope what happens.

The voltage seems to drop like a rock in a few milliseconds down to 3.5 volts or so, if my memory is not failing me, which was a bit of a surprise to me. It was as if something really sucked a lot of juice during startup - it can't be the Mobius, so it had to be the hardwire kit itself when it shuts down, or then my old capasitors were somewhat bad. From there on, the voltage dropped like I expected it to drop - with a curve.

In hindsight, I should've connected these 10v caps in series, making it safe up to 20v, which works for the car 12v, and then putting them to the hardwire kit input with a diode preventing the car from sucking the juice from them. That would have probably worked a bit better.

But anyhow, these two caps worked! It really was enough to just put these to the hardwire kit output, basically to the USB power lead. At least during the summer they do the trick. For the winter startups which are eating a lot longer from a -20C car battery, the story might be different. If the problem resurfaces, I know how to fix it.

Thus, I no longer have the auto-start issue, after:
1) the 8" extension cable is no-longer giving transmission errors to the Mobius, after I swapped the adaptor and the cable and did some tin-foil placebo shielding.
2) the voltage of the Mobius input doesn't drop too much during engine startup

I'm left with the dashboard reflections, but that discussion certainly belongs to some other topic.

 

[BobDiaz]

mobnam,

Remember TC = R x C

In this case 3200uF = 0.0032F. The "R" depends on the current being drawn, which should be in the order of around 30 to 35 ohms. Let's see, 0.0032F x 35 ohms = 0.112 seconds. However at that time the voltage on the capacitor is at about 37% of the starting voltage. So based on the value of capacitors you have, I would expect the voltage to drop like a rock.

 

[mobnam]

So based on the value of capacitors you have, I would expect the voltage to drop like a rock.

You are right. But this isn't a simple RC circuit and the oscilloscope certainly confirms that, the power drain from the usb connector is not constant.

I'm speculating here, but I believe that what happens is that the voltage quickly drops to a level where the Mobius starts to use its internal battery. This voltage where the external power is dropped off, is probably a bit higher than what is the internal battery voltage (my guess: about one diode or around 0.6-0.7V higher). The circuitry which is responsible for this selection of power, is probably not much in contact with the cpu in the device. I believe the device is using some sort of simple I/O pin for sniffing the external power for recording stop purpose, and these capacitors that I added are holding that voltage long enough and just high enough for the I/O to read as 1, after Mobius has switched to using its internal battery. The power drain from the capacitors is not that big at that point of time, since most likely the juice for recording is coming from the internal battery.