Re: [EVDL] EV Digest, Vol 118, Issue 2

[Justin Kenny via EV]

The Tesla "FSD" computer is actually "dual redundant" as mentioned in this
Tesla Autonomy Day clip from 3 years ago:
https://youtu.be/Ucp0TTmvqOE?t=4648 However to your point, 2 is not enough;
it's only enough to tell you *someone* is wrong, but not who. Hence why in
my job with hardware testing we need a minimum of 3 devices tested to
indicate which one is the outlier, if any. Given that "FSD" even now is
still a level 2 system where the driver is fully responsible for the
actions of the car, I suppose they don't care, if one computer doesn't
agree with the other, it kicks back to you, the responsible driver. Very
annoying and dangerous if this happens around a sharp turn or construction
zone, as (non-FSD) autopilot tends to do, however.

- Justin


> Message: 4
> Date: Wed, 10 Aug 2022 21:44:09 -0400
> From: "mark hanson" 
> To: "'Electric Vehicle Discussion List'" 
> Subject: [EVDL] Self Driving Not Ready for Prime Time, EVLN: Tesla
> can't see kids
> Message-ID: <000b01d8ad23$d9e84c50$8db8e4f0$@gmail.com>
> Content-Type: text/plain; charset="us-ascii"
>
> Hi Folks,
>
>
>
> I was heading up rt81 freeway from Roanoke VA to Harrisonburg yesterday in
> my Tesla Y to get flushed away at the Massanutten H20 Park and an hour into
> the trip the self driving or "auto pilot" jerked the steering wheel into
> the
> left lane & I jerked it back (luckily between cars).  I thought using self
> driving was safe on straight highways, "what could possibly go wrong?".
> Then about a 1/2 hour later (just using the cruise control), it hit the
> brakes and I hit the accelerator (almost got rear ended).  My son said
> "dad,
> why don't you just drive the car and forget about the self driving
> nonsense".  Words of wisdom.  Elon should just keep making a nice EV and
> forget the fluffy stuff that causes problems.
>
>
>
> The fundemantal "fatal flaw" is self driving needs at least 3 computers to
> "vote" and kick out the one that gets a glitch (cosmic rays, brain farts
> etc).  All the cars attempting this foolishness are using only *one*
> microcomputer.
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
HELP: http://www.evdl.org/help/

Re: [EVDL] EV Digest, Vol 106, Issue 27

[Justin Kenny via EV]

I would be curious to see some real world comparison/results doing a drive
like that coasting as much as possible vs. using full regen. I used to
drive a 2011 Leaf, and I struggled to get more than 4mi/kWH driving as
efficiently as I could (minimizing regen by coasting) with no A/C, etc.,
yet now I drive a Model 3 and I find it almost impossible to get less than
4mi/kWH with all the comforts available and driving like a maniac ;)
Clearly the vehicle's efficiency has a way bigger impact, but of course it
is probably still better for the environment to be reusing a 10 year old EV.

- Justin


> --
>
> Message: 3
> Date: Tue, 24 Aug 2021 19:06:15 + (UTC)
> From: Lawrence Rhodes 
> To: "ev@lists.evdl.org" 
> Subject: Re: [EVDL] Regen Efficiency vs Coasting
> Message-ID: <608585047.1164471.1629831975...@mail.yahoo.com>
> Content-Type: text/plain; charset=UTF-8
>
> When you are at the top of a hill after the climb the best way to descend
> is to select no regen until you reach the speed limit and then use regen to
> keep your vehicle at the speed limit. My Leaf has 6 regen selections. N for
> no regen, D for some, Deco for more, B and Beco for the most. Using these
> settings I can in a way coast and not use my mechanical brakes for long
> distances down hill. EVs are never in neutral unless you have a conversion
> with a clutch. If the vehicle moves the motor turns. That is why you can't
> flat tow an EV. So the most efficient way down a hill is No regen until you
> are in fear of a speeding ticket. Lawrence Rhodes
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
ARCHIVE: http://www.evdl.org/archive/
LIST INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org

Re: [EVDL] EV Digest, Vol 105, Issue 7

[Justin Kenny via EV]

Autopilot will only safely bring you to a stop if autopilot can remain
engaged according to its rules that entire time. If a driver engages
autopilot, and the Tesla determines that it is a situation it can't handle
(e.g. a sharp curve in the road), it will kick you out with very little
warning. I think this skews the autopilot safety results.. for truly
dangerous situations it can't handle, well surely autopilot wouldn't be
engaged in an accident, since it would have kicked the user out of it
moments before the problem. I use autopilot quite a lot and am happy with
its ability to reduce driving fatigue for long boring drives, but I would
never trust it to be totally failsafe if I didn't touch the wheel, I've had
to make a couple quick wheel grabs for the situations I described.

- Justin


> Message: 4
> Date: Thu, 8 Jul 2021 11:05:30 -0700
> From: "(-pEEf-)" 
> To: Peri Hartman , Electric Vehicle Discussion List
> 
> Subject: Re: [EVDL] Fwd: New York Times article about teen death by
> Tesla pic up truck on auto pilot is click bait.
> Message-ID:
>  bfkvaxzb1n8pkxbdz+k0qbxgk7...@mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> It does this already (somewhat).  If you don't acknowledge the attention
> "nag" it requests, which is to place some slight force on the wheel from
> time to time, it will set off an alarm and ask you to take over.  If you
> don't, it will turn on the hazard flashers and slow to a stop.   Of course
> it's possible to keep force on the wheel enough to satisfy the nag, and
> still not pay attention to the road.
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
Address messages to ev@lists.evdl.org
No other addresses in TO and CC fields
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
ARCHIVE: http://www.evdl.org/archive/
LIST INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org

Re: [EVDL] Question about charging small LiIon batteries

[Justin Kenny via EV]

I've had pretty bad luck with MP2307 boards, a lot of them that I use are
defective. The LM2596 boards I've used have all worked however, they use
the much better 10-turn potentiometers. You could try and replace the
variable resistor with your own potentiometer or fixed one, it is a bit of
a challenging rework since it's such a small SMD part. It won't have any
current limiting either so it might not work very well, but I guess it's
better than diodes.

- Justin

On Tue, Jun 2, 2020 at 9:05 PM  wrote:

> Justin Kenny's suggestion earlier reminded me of something, so I went
> looking through my spare parts and found some regulators boards that have
> variable MP2307 regulators. I originally got them to use with 3.3V
> microcontrollers, but never used them because they have flaky outputs. I'm
> pretty sure the problem with them is the cheap variable resistors, I think
> if I replace those with a fixed voltage divider they might be ok.
>
>
> However, I think the problem with them is the cheap variable resistors
> they use
> June 2, 2020 6:30 PM, justin.sout...@xtra.co.nz wrote:
>
> Peter,
> Can i suggest a lm723 voltage regulator configured as a shunt regulator. I
> have used the datasheet design with solar panels. Lee may be able to
> recommend a better design than the barebones design in the datasheet. As
> the solar panel is a current source of limited output you could probably
> forgo the load resistor and simply dump all the excess energy into a large
> enough power transistor. The standby current of the device is typically
> 1.7ma with 30v source according to the datasheet so id expect it to be
> lower in your application. The addition of an output diode would disconnect
> the reg once the solar input was removed.
> Regards,
> Justin
> On Jun 3, 2020 10:58, Peter VanDerWal via EV  wrote:
>
> Yes, Lee has already pointed this out, so I am looking into putting the
> zener in parallel to the PV panel and then a schottky, etc. in series to
> the battery.
>
> The problem I have with the cheap CV/CC circuits from china is that in my
> experience they have a high failure rate and, just like buttered toast, the
> failures tend to happen in the worst possible mode.
>
> June 2, 2020 3:17 PM, "Justin Kenny"  jkenn...@gmail.com?to=%22Justin%20Kenny%22%20)> wrote:
> Not sure if you understood the previous 2 messages, the Zener diode will
> drain your battery, and a diode will not stop it from overcharging. I
> really suggest to use the right tool for the job, $3.25 for a
> constant-current constant-voltage converter isn't much cost to save your
> battery pack from being killed prematurely, or at worst preventing a
> fire/destruction by overcharging: https://www.amazon.com/dp/B00LPK0IKQ (
> https://www.amazon.com/dp/B00LPK0IKQ)
> - Justin
> On Tue, Jun 2, 2020 at 2:57 PM  e...@vanderwal.us)> wrote: I'm thinking of using two panels, one tilted
> 45 degs to the east, and the other tilted 45 deg to
> the west. That should provide a fairly flat power output through the day
> with something like .375
> ma max. If I put a 4.3V zener across the panel, then that should bypass
> some of the current, so I should only see a swing from perhaps 0.2V to 0.3V
> and the 0.2V will likely occur when the panel voltage is lower.
> At any rate a 0.1V variation is close enough for my purposes.
> I've got a selection of diodes on order, so I'll see what happens when I
> test it on the bench.
>
> Pete.
>
> June 2, 2020 1:51 PM, "jkenny23 via EV"  ev@lists.evdl.org)> wrote:
>
> > Another very important note; diodes don't drop voltage the same at all
> > current levels. Meaning; they will NOT stop your 4.5V panel from
> > overcharging your battery to 4.5V if you don't have a dedicated charging
> IC.
> >
> > If you want to charge your battery to an unusual voltage, I highly
> recommend
> > using an adjustable CC/CV (constant-current/constant-voltage) buck
> > converter. They're widely available from China on eBay and Aliexpress,
> and
> > probably available from the US through Amazon. This way you can set a
> safe
> > current limit (say 400mA), and a voltage limit that is lower than 4.2V
> (say
> > 3.9 or 4.0V). Then you could even use multiple panels in series and not
> > worry about charging, the input limit for the common LM2596 modules is
> 35V.
> >
> > Here's some data for a common 1N5819 Schottky diode illustrating the
> issue
> > of voltage drop as current tapers down at the end of charging:
> > 1.0A - 0.41V
> > 0.5A - 0.35V
> > 0.1A - 0.28V
> > 20mA - 0.23V
> >
> > --
> > Sent from: http://electric-vehicle-discussion-list.413529.n4.nabble.com
> (http://electric-vehicle-discussion-list.413529.n4.nabble.com)
> > ___
> > UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub (
> http://www.evdl.org/help/index.html#usub)
> > ARCHIVE: http://www.evdl.org/archive/index.html (
> http://www.evdl.org/archive/index.html)
> > INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org (
> 

Re: [EVDL] Question about charging small LiIon batteries

[Justin Kenny via EV]

Suit yourself, I hope you are using protected 18650 cells. If not please
use one of these to protect the batteries from overcharging/undervoltage:
https://www.amazon.com/dp/B07KSPYMJ2

- Justin

On Tue, Jun 2, 2020 at 3:58 PM  wrote:

> Yes, Lee has already pointed this out, so I am looking into putting the
> zener in parallel to the PV panel and then a schottky, etc. in series to
> the battery.
>
> The problem I have with the cheap CV/CC circuits from china is that in my
> experience they have a high failure rate and, just like buttered toast, the
> failures tend to happen in the worst possible mode.
>
> June 2, 2020 3:17 PM, "Justin Kenny"  >
> wrote:
>
> Not sure if you understood the previous 2 messages, the Zener diode will
> drain your battery, and a diode will not stop it from overcharging. I
> really suggest to use the right tool for the job, $3.25 for a
> constant-current constant-voltage converter isn't much cost to save your
> battery pack from being killed prematurely, or at worst preventing a
> fire/destruction by overcharging: https://www.amazon.com/dp/B00LPK0IKQ
> - Justin
> On Tue, Jun 2, 2020 at 2:57 PM  wrote:
>
> I'm thinking of using two panels, one tilted 45 degs to the east, and the
> other tilted 45 deg to
> the west. That should provide a fairly flat power output through the day
> with something like .375
> ma max. If I put a 4.3V zener across the panel, then that should bypass
> some of the current, so I should only see a swing from perhaps 0.2V to 0.3V
> and the 0.2V will likely occur when the panel voltage is lower.
> At any rate a 0.1V variation is close enough for my purposes.
> I've got a selection of diodes on order, so I'll see what happens when I
> test it on the bench.
>
> Pete.
>
> June 2, 2020 1:51 PM, "jkenny23 via EV"  wrote:
>
> > Another very important note; diodes don't drop voltage the same at all
> > current levels. Meaning; they will NOT stop your 4.5V panel from
> > overcharging your battery to 4.5V if you don't have a dedicated charging
> IC.
> >
> > If you want to charge your battery to an unusual voltage, I highly
> recommend
> > using an adjustable CC/CV (constant-current/constant-voltage) buck
> > converter. They're widely available from China on eBay and Aliexpress,
> and
> > probably available from the US through Amazon. This way you can set a
> safe
> > current limit (say 400mA), and a voltage limit that is lower than 4.2V
> (say
> > 3.9 or 4.0V). Then you could even use multiple panels in series and not
> > worry about charging, the input limit for the common LM2596 modules is
> 35V.
> >
> > Here's some data for a common 1N5819 Schottky diode illustrating the
> issue
> > of voltage drop as current tapers down at the end of charging:
> > 1.0A - 0.41V
> > 0.5A - 0.35V
> > 0.1A - 0.28V
> > 20mA - 0.23V
> >
> > --
> > Sent from: http://electric-vehicle-discussion-list.413529.n4.nabble.com
> > ___
> > UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
> > ARCHIVE: http://www.evdl.org/archive/index.html
> > INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
> > Please discuss EV drag racing at NEDRA (
> http://groups.yahoo.com/group/NEDRA)
>
>
>
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
ARCHIVE: http://www.evdl.org/archive/index.html
INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)

Re: [EVDL] Question about charging small LiIon batteries

[Justin Kenny via EV]

Not sure if you understood the previous 2 messages, the Zener diode will
drain your battery, and a diode will not stop it from overcharging. I
really suggest to use the right tool for the job, $3.25 for a
constant-current constant-voltage converter isn't much cost to save your
battery pack from being killed prematurely, or at worst preventing a
fire/destruction by overcharging:  https://www.amazon.com/dp/B00LPK0IKQ

- Justin

On Tue, Jun 2, 2020 at 2:57 PM  wrote:

> I'm thinking of using two panels, one tilted 45 degs to the east, and the
> other tilted 45 deg to
> the west. That should provide a fairly flat power output through the day
> with something like .375
> ma max. If I put a 4.3V zener across the panel, then that should bypass
> some of the current, so I should only see a swing from perhaps 0.2V to 0.3V
> and the 0.2V will likely occur when the panel voltage is lower.
> At any rate a 0.1V variation is close enough for my purposes.
> I've got a selection of diodes on order, so I'll see what happens when I
> test it on the bench.
>
> Pete.
>
> June 2, 2020 1:51 PM, "jkenny23 via EV"  wrote:
>
> > Another very important note; diodes don't drop voltage the same at all
> > current levels. Meaning; they will NOT stop your 4.5V panel from
> > overcharging your battery to 4.5V if you don't have a dedicated charging
> IC.
> >
> > If you want to charge your battery to an unusual voltage, I highly
> recommend
> > using an adjustable CC/CV (constant-current/constant-voltage) buck
> > converter. They're widely available from China on eBay and Aliexpress,
> and
> > probably available from the US through Amazon. This way you can set a
> safe
> > current limit (say 400mA), and a voltage limit that is lower than 4.2V
> (say
> > 3.9 or 4.0V). Then you could even use multiple panels in series and not
> > worry about charging, the input limit for the common LM2596 modules is
> 35V.
> >
> > Here's some data for a common 1N5819 Schottky diode illustrating the
> issue
> > of voltage drop as current tapers down at the end of charging:
> > 1.0A - 0.41V
> > 0.5A - 0.35V
> > 0.1A - 0.28V
> > 20mA - 0.23V
> >
> > --
> > Sent from: http://electric-vehicle-discussion-list.413529.n4.nabble.com
> > ___
> > UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
> > ARCHIVE: http://www.evdl.org/archive/index.html
> > INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
> > Please discuss EV drag racing at NEDRA (
> http://groups.yahoo.com/group/NEDRA)
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
ARCHIVE: http://www.evdl.org/archive/index.html
INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)

Re: [EVDL] custom battery

[Justin Kenny via EV]

Well I assumed by disconnect you meant a manual one (e.g. big on/off
switch), which would still have to take all the power required through it.
That'd be the service disconnect I linked from EV West. The battery
management system would control the contactor, which also switches all of
the required power for your application, and that would handle over temp,
over voltage, under voltage, etc. fault conditions. None of those *should*
happen in a well designed system, so it shouldn't need to hard-switch
hundreds of amps, but it should be designed to be able to without welding
contacts shut or arcing etc., and that's why you need a proper high
voltage/high current DC contactor, and not just a big relay that
technically can handle hundreds of amps.

- Justin

On Sun, Apr 19, 2020 at 7:55 PM Peri Hartman  wrote:

> Why shouldn't I use a disconnect. Or am I using the wrong term:
> something like a contactor that breaks the power connection if the temp
> rises too much. I'm figuring I'll need two: one for the load and one for
> the charger power. The latter should be easy: 120VAC, 15A.
>
> With 6kW or more, if something goes wrong, there'll be a lot of heat in
> the wrong place.
>
> Peri
>
> -- Original Message --
> From: "jkenny23 via EV" 
> To: ev@lists.evdl.org
> Cc: "jkenny23" 
> Sent: 19-Apr-20 6:16:39 PM
> Subject: Re: [EVDL] custom battery
>
> >Well you shouldn't need to use the disconnects; but yes, those amperage
> >levels are used commonly in EVs of course. EV West has pretty good prices
> >for stuff like this (service disconnect, rated for max current your system
> >would see):
> >
> https://www.evwest.com/catalog/product_info.php?cPath=2_42_id=267=na4hdbkd5u5i6u7ne1lltda641
> >
> >And here's an example of the contactor you'd need to control with the BMS
> >(quite cheap for this model, good deal actually):
> >https://www.ebay.com/itm/293546196201
> >
> >--
> >Sent from: http://electric-vehicle-discussion-list.413529.n4.nabble.com/
> >___
> >UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
> >ARCHIVE: http://www.evdl.org/archive/index.html
> >INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
> >Please discuss EV drag racing at NEDRA (
> http://groups.yahoo.com/group/NEDRA)
> >
>
>
-- next part --
An HTML attachment was scrubbed...
URL: 

___
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
ARCHIVE: http://www.evdl.org/archive/index.html
INFO: http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)

Re: [EVDL] 18650 cell level fuse wire

[Justin Kenny via EV]

I'm the owner of the Leaf Rex channel Cor mentioned earlier. I highly
caution against using PCB fuses in general, I've done testing on those
also: https://youtu.be/CMlpCX0bug8 If you have a relatively large open
circuit voltage in the system you are trying to protect, it will
create a conductive plasma and char the PCB, leaving behind a
"resistor" which can continue to burn, though hopefully
self-extinguish in keeping with the "flame retardant" part of FR-4.

- Justin

On Mon, Jan 21, 2019 at 1:09 PM  wrote:
> Message: 13
> Date: Mon, 21 Jan 2019 15:25:38 -0500
> From: Martin Klingensmith 
> To: Lee Hart via EV 
> Subject: Re: [EVDL] 18650 cell level fuse wire
> Message-ID: <623a4924-a24a-52f3-b720-2ee3c2f69...@nnytech.net>
> Content-Type: text/plain; charset=utf-8; format=flowed
>
>
> IIRC I've seen "fusible links" on PCBs. They're a short section of
> copper trace that is much narrower than the rest of the trace. My
> understanding is that they're used for an extra level of protection or
> where the designer thinks they're really clever. For all modern PCB
> materials it's not *too* unsafe because they aren't (supposed to be)
> flammable.
>
> -
>
> Martin K
___
UNSUBSCRIBE: http://www.evdl.org/help/index.html#usub
http://lists.evdl.org/listinfo.cgi/ev-evdl.org
Please discuss EV drag racing at NEDRA (http://groups.yahoo.com/group/NEDRA)