As a final note, I have found that the Samsung tablet will draw an extra 0.7A if charged via the TPS2511 port, compared to being charged via the port which had D+ and D- shorted.
This is a surprising result, and probably means they are making use of this mode
compared to this much more standard mode for non-Apple chargers
The simple “D+ to D- shorted” mode does however seem to work 100% reliably for all the non-Apple stuff I have. It may not charge at the max possible speed but it will always charge.
For reference, Apple devices use one of these two and if you don’t do that, you get a very unsatisfactory result (usually zero charge unless the device is turned off)
It is difficult to get good data because all these devices vary their charging currents wildly according to state of charge, internal current draw (screen brightness, etc) and who knows what else.
Finished 
The two female USB sockets on the end of that cable have D+ and D- shorted together so they do the max charge rate on all non Apple devices. Apple ones have to use the other USB socket on the back, which uses that chip to detect there is an Apple device attached. I don’t use any Apple stuff in the plane but somebody else might.
Re lithiums, this is why I am using this power pack rather than some cheaper ones. I assume all consumer IT lithium batteries are LIPO and thus would burn pretty well…
Re the device grounding, I have just checked my Ipad 2 and its case is not grounded to USB ground. There is 100k of resistance. I wonder what the reasoning was… And my Samsung 8.4" tablet is open-circuit totally. In normal use the device being charged will not be touching the airframe anyway.
Peter wrote:
I forgot to draw in two 2.2uF 16V ceramic caps
I didn’t notice them in the photo. Now that you mention them I found them.
Peter wrote:
What do you think, Tom?
The whole thing should be quite good. Most likely the load device will connect USB ground and shield, so the USB power side will cease to float once the load device is plugged in.
You might do worse than adding one or two ferrite beads; also, it might make sense to use an USB cable length that is not Lambda/4 (or an integer multiple thereof) at 110MHz… And I might connect the two black wires at the negative terminal of the decaps…
Lithium polymer (LiPo) batteries certainly do end up with a reduced life as a result of fast charging, but the packs used by modellers are optimised for light-weigh (i.e. power density) at all costs. As they’ve been improving, acceptable charge rates have gone up, but I’d agree that 6C would be too much for most batteries, even if there were some that would cope with it.
LeSving’s car will probably use an LiFePo battery of some sort. These are safer, heavier, and have a much better life expectancy. Some of them can also be charged very fast – e.g. 6C. Also much less likely to do this:
(LiPo charging fire)
Which isn’t what you want at any height.
I forgot to draw in two 2.2uF 16V ceramic caps, at the input and on the output of the TPS2511.
Note that the whole output side of the DC-DC converter is still floating. The box is shielded to the airframe ground. The box also has all sides bonded, via the self tapping screws which cut into the removable lid and into the two end panels. The USB connector casing is connected to that, but not to the USB ground.
I think it’s damn hard to connect stuff to the airframe itself, unless one constructed a box which forms a part of the airframe. Then shielding could be much improved. But I think that in the absence of that, the best one can do is to shield the box and make sure “nothing” gets out of it.
What do you think, Tom? You know infinitely more about RF than I ever did.
You don’t particularily like decoupling, do you?
Here is the add-on USB output
The white block is an aluminium oxide thermally conductive pad.
I have been testing it with various USB devices and it really does work.
Via the TPS2511, the charging currents are:
Samsung Android 8.4" tablet at 1.8A
S6 phone at 1.5A
Lenovo T2 tablet at 1.75A
Ipad 2 (year 2011) at 1.3A
If these devices are connected to a USB port with nothing on the two data wires, you get:
Samsung Android 8.4" tablet at 0.5A
S6 phone at 0.5A
Lenovo T2 tablet at 0.5A
Ipad 2 (year 2011) at zero
If these devices are connected to a USB port with a short on the two data wires, you get:
Samsung Android 8.4" tablet at 2.3A
S6 phone at 1.4A
Lenovo T2 tablet at 2.4A
Ipad 2 (year 2011) at 0.33A
So, for the non-Apple devices, the benefit of doing anything except shorting the two data wires is nil (as one would expect) but for the Apple device (the Ipad) that “smart chip” is essential.
I remembered wrong. The guaranty of the battery is 8 years or 160000 km, the rest of the car is only 5 years or 100000 km The battery is “Lithium Ion” without any more specification. The car itself take care of optimizing the life of the battery. It is never truly fully charged and never truly fully empty. Maybe it goes between 20% and 80 %. It also monitors the temperature, charging faster at 0-20 degree C, slower above and below. Temperature is what most effects battery life. This may be different from car to car because they use slightly different batteries. The only thing the manual warns about is letting it stay uncharged for more than 4 months continuously. Exactly how the battery is taken care of may change also, through upgrades, when VW gets more real world data and experience (and who knows what tricks these VW software engineers are doing )
In principle though, the battery “lasts” longer than the car for VW electric cars, so there is nothing to worry about. It’s all about intelligent charging though.
I will put an extra USB socket on the back of my power unit, powered via that TI chip, and will see if it works better with any device than the other outlet which will have just D+ and D- shorted.
The thing about batteries getting trashed depends obviously on the chemistry. The model plane scene nowadays uses LIPO and there is a lot of feedback on the reduced battery life as a result of rapid charging which is obviously popular because at a hundred quid each people don’t want to buy half a dozen of them. What battery chemistry does your car use?
