EV Digest 5080
Topics covered in this issue include:
1) RE: Mk3 Regs
by [EMAIL PROTECTED]
2) Re: Tire Ignition Sequence Photos of White Zombie???
by Victor Tikhonov <[EMAIL PROTECTED]>
3) RE: Tire Ignition Sequence Photos of White Zombie???
by "Roger Stockton" <[EMAIL PROTECTED]>
4) RE: I want to build a PWM DC motor controller
by "Roger Stockton" <[EMAIL PROTECTED]>
5) "Posicharge" item on eBay
by Marvin Campbell <[EMAIL PROTECTED]>
6) FW: See photos of world's best electric scooter
by "Lawrence Rhodes" <[EMAIL PROTECTED]>
7) Re: I want to build a PWM DC motor controller
by "Stefan T. Peters" <[EMAIL PROTECTED]>
8) Re: EV digest 5078
by Jimmy <[EMAIL PROTECTED]>
9) RE: Neg supply for E-meter
by Cor van de Water <[EMAIL PROTECTED]>
10) RE: Neg supply for E-meter
by "Roger Stockton" <[EMAIL PROTECTED]>
11) RE: Neg supply for E-meter
by Cor van de Water <[EMAIL PROTECTED]>
12) Re: I want to build a PWM DC motor controller
by Rod Hower <[EMAIL PROTECTED]>
--- Begin Message ---
Hey Rich.. How about making the 232/evil bus simply a daughter board that
snaps onto the Reg board using a single header and four nylon standoffs much
like the add on daughter boards common to video graphics boards for higher
end video in a computer?? That way your Reg boards are all the same and the
added bus support can be an in-field upgrade??
Just a thought
Pedroman
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Rich Rudman
Sent: Tuesday, January 10, 2006 12:38 PM
To: [email protected]
Subject: Re: Mk3 Regs
Hey Chris and listers..
I have been busy all weekend building up my Fiero and shoveling Weeks of
charger repairs and orders, so.. I have not been sitting here playing "EV
lister".
What I have been hounding all my Evil buss supports for is simple and
robust. We don't need a Evil Buss that can play MP3 sounds..
Bruce finally got the time to get code cut that works, but... we need some
work on it still.
The bridge is a special loaded MK3 board that has the 232 drivers and Evil
buss terminators on it. We could cut a speical pcb just for this function,
but I insisted that all the Regs be the same. This idea may be taken any
way.... Dropping the 232 stuff could cut the normal Reg's Realestate by
quite a bit. Having any Reg be just the one you need is also a good idea in
the real world of BMS field support.
I gotta go.... more appointments to make..
Rich Rudman
Manzanita Micro
----- Original Message -----
From: "Chris Brune" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Sunday, January 08, 2006 8:44 PM
Subject: Re: Mk3 Regs
> Rich,
> I can't believe a day and half and not one response to this... I'll
> bite, bring on the details!!! I'm very excited to see what you guys
> have put together. There's been a
lot
> of talk about doing this by a number of people nice to see some action
> by somebody. I know that Ralph and I have been working on and off
> (mostly
off)
> for more than a year on this. I suspect that if what Bruce and Sheer
> have put together works it will make all the recent EVILbus protocol
discussions
> mute.
>
> Is the "bridge" reg a bus master?
>
> Details???
>
> Regards,
> Chris Brune Tigard, OR
> 93 Honda Del Sol
>
--- End Message ---
--- Begin Message ---
Perhaps misunderstanding Roger, I said that one must at least
use dedicated film scanner which usually does 2400 lpi at least
(consumer grade), not flat bed one with film attachment.
Using drum scanner is 'course better, but for the photo
on the mag you may not need all the resolution actual
film possess, this requirements in general depend on the editor
of the publication (been there).
Victor
Roger Stockton wrote:
Victor Tikhonov [mailto:[EMAIL PROTECTED] wrote:
'Course I meant film scanner, usually 2400 lpi resolution
in average (not a generic flat bed scanner attempting to scan
negatives as any other paper).
I'm talking about 14,000dpi drum scanners used professionally to convert
analog images to digital, not consumer grade junk. 2400dpi is nowhere
near sufficient to read all the information available on an analog
negative/positive and is representative of a flatbed scanner attampting
to scan film with an attachment; even dedicated consumer film scanners
are higher resolution than this.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
Victor Tikhonov [mailto:[EMAIL PROTECTED] wrote:
> Using drum scanner is 'course better, but for the photo
> on the mag you may not need all the resolution actual
> film possess, this requirements in general depend on the
> editor of the publication (been there).
Complete agreement Victor!
The amount of information on the film is almost always going to be much
more than is required for print or web publication. Consider that a
drum scan of a 35mm frame at 14000dpi works out to over 100Mb per image,
yet as Jeff observed, 3 megapixels is overkill for web publication, and
5 megapixels is OK for upto about 8x10 prints, and 8 megapixels is fine
for many professional purposes. Prints are only about 300dpi, which
works out to just under 8 megapixels for an 8x10 print.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
Stefan T. Peters [mailto:[EMAIL PROTECTED] wrote:
> Just the ability to control each battery separately. This allows for
> automatic mix-and-match, battery fail over/replacement, and gradual
> system upgrades to a very high limit.
Right, but we have also been able to identify that the penalties
associated with this approach include higher cost and lower efficiency
than a normal PWM controller. We've been able to quantify (at least
ballpark) the efficiency loss; I think that if we could quantify the
cost difference we'd be in a much better position to appreciate whether
or not the potential benefits (cost savings) of the BMS functionality is
sufficient to outweigh the efficiency and cost hits.
> Sorry about the snappiness.
No worries; I did include a smiley in my response to indicate that I
wasn't bothered by it at all.
> I think you can make one set of circuits that can be reused
> in various ways to perform all operation aspects of a EV.
Maybe, but I suspect it is unlikely that such a set of circuits will
then be optimal for any single purpose. Consider that a circuit
intended to be used in a contactor-type controller can use components
that switch much more slowly than those in a circuit intended for use in
a PWM-type controller operating at any reasonable frequency. The
slow-switching circuit can take advantage of simpler, perhaps cheaper,
devices such as contactors and SCRs while the faster one is pushed into
using devices such as MOSFETs.
> > I understand this, however, to evaluate what merit this
> > approach has we must consider a collection/combination
> > that approximates the features, etc. available from
> > conventional alternatives and see if that collection
> > actually offers a benefit that would encourage a user
> > to choose it over the alternative.
>
> Ah-ha! Now we have it. To evaluate the system structured in
> such a way as to give the same features and benefits as a
> retail PWM controller, you would configure it as such:
>
> Data Bus:
>
> System Controller -> PWM-based Motor Controller
>
> HV Bus:
>
> Batteries (series) -> PWM-based Motor Controller -> Motor
I don't buy it. ;^> This configuration certainly is possible, and would
duplicate the features and benefits of a traditional PWM controller,
except that I don't really see this being an attractive option for any
end user. The feature it adds over a traditional controller is merely
that the brain/micro is a separate blob than the controller proper, so
one can replace one without replacing the other. I don't see this as a
selling feature by itself.
I think that to really attract people away from traditional one-box PWM
controllers the system has to offer some additional features or some
sort of cost advantages, etc. It seems that this is only possible when
one considers the configurations that include individual battery
switching (providing the additional BMS-type features) and the
series/parallel switching (to mimic the current multiplication features
of a standard PWM controller). It also happens that this is the only
arrangement that anyone is presently working on...
> > I have not yet been able to identify a single advantage to the end
> > user of your system over the status quo.
> >
>
> Modularity. Component reuse. Less waste with many parallel
> open-source EV projects. I guess EE and NE/SE are much further
> apart in their practices then I originally thought.
Nope, you just missed a key part of what I wrote: "to the end user".
The benefits that you cite are of course why modularity and reuse are
important to *us* (i.e. those developing things), but the end user could
care less about what makes our lives easier or more productive unless he
derives some direct benefit such as a cheaper product.
The idea of being able to plug blocks together as required to assemble
one's particular EV is very attractive, but if it results in a more
costly, less efficient (and potentially less reliable due to the
increased wiring & connections) system than one could otherwise build it
becomes a tough sell.
> Data Bus:
>
> System Controller -> multi-stage series/parallel Motor Controller
>
> HV Bus:
>
> Batteries (many strings) -> multi-stage series/parallel Motor
> Controller -> Motor
This is, unfortunately, looking more and more like modern LEGO, that is,
instead of basic generic building blocks that one rearranges into
whatever their imagination can envision (a la traditional LEGO) there
are custom blocks designed for specific tasks.
My description of a 120V system using 4 subnet switches stacked to allow
combining 5 24V strings is based on the traditional LEGO system view. I
assumed you were proposing building a general-purpose subnet switch
which at its most basic needs only to allow 2 input strings to be
combined in series or parallel to its output. An end user would then
take as many of these basic blocks and stack or otherwise arrange them
as required to build his particular system; he would not look for a
special purpose 4-input switch because he has 4 strings, or a 7-input
switch because he has 7 strings at the present time. Once you get to
this mentality you are really saying the system consists of a generic
brain blob (with application-specific code), and a special purpose
controller blob (either a PWM controller or a contactor controller,
etc.)
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
Aerovironment manufactured the quickchargers for the luggage trucks at LAX.
Twenty minutes, baby, and you're done.
Too bad quick charging won't work for passenger vehicles...
Haw!
Marv
> From: Andrew Letton <[EMAIL PROTECTED]>
> Date: Tue, 10 Jan 2006 10:21:09 -0800
> To: [email protected]
> Subject: Re: "Posicharge" item on eBay
>
> Google Aerovironment.
> They have done extensive work on fast charging.
> Evidently they've spun off a separate charger company:
> http://www.posicharge.com/index.html
> I'd guess that this thing was some prototype test device, though I
> wouldn't put anything built by Aerovironment and "homemade" in the same
> category.
> I do have to agree on the price though...
> cheers,
> Andrew
>
> Neon John wrote:
>
>> On Tue, 10 Jan 2006 08:25:47 -0800, [EMAIL PROTECTED] wrote:
>>
>>
>>
>>> Any idea what this was used on:
>>> http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&item=4603646892
>>>
>>>
>>>
>>
>> Looks like a homemade charge/discharge datalogger to me. Given that
>> it's 24 volts and has "aero" in the name, I imagine it either has
>> something do do with charging aircraft batteries or perhaps
>> self-propelled luggage carts.
>>
>> Nifty setup but the guy's smokin' dope on the price. IMO :-)
>>
>> John
>> ---
>> John De Armond
>> See my website for my current email address
>> http://www.johngsbbq.com
>> Cleveland, Occupied TN
>> A foolish consistency is the hobgoblin of little minds.-Ralph Waldo Emerson
--- End Message ---
--- Begin Message ---
From: "vanrescuers" <[EMAIL PROTECTED]>
Subject: FW: See photos of world's best electric scooter
http://www.geocities.com/robert04mat/BEST_ELEC_SCOOTER.html
Ride comfort is top-drawer! Almost weather-proof, at low speeds. Has
more inherent stability than the Segway, imho.
Just a few problems with it .. produces smelly fumes, and some gas.
Lawrence Rhodes
Bassoon/Contrabassoon
Reedmaker
Book 4/5 doubler
Electric Vehicle & Solar Power Advocate
415-821-3519
[EMAIL PROTECTED]
--- End Message ---
--- Begin Message ---
Roger Stockton wrote:
Stefan T. Peters [mailto:[EMAIL PROTECTED] wrote:
Just the ability to control each battery separately. This allows for
automatic mix-and-match, battery fail over/replacement, and gradual
system upgrades to a very high limit.
Right, but we have also been able to identify that the penalties
associated with this approach include higher cost and lower efficiency
than a normal PWM controller. We've been able to quantify (at least
ballpark) the efficiency loss; I think that if we could quantify the
cost difference we'd be in a much better position to appreciate whether
or not the potential benefits (cost savings) of the BMS functionality is
sufficient to outweigh the efficiency and cost hits.
Well, so far I have this for a BatPack style setup:
System Controller = $38
Battery Controller (with 18V/500A switch board) = $61 each
Motor Controller (with 2 stage series parallel board) = $8 + $4 + TBD
Yeah, I know that last one is the big one. I have a feeling it will be
around $150-200 for the two string 1000A series-parallel version. A
straight PWM version for use with Battery Sensors instead of controllers:
System Controller = $38
Battery Sensor = $8 + $4 each
Motor Controller = $8 + $4 + depends on voltage, for a 24V/600A one it's
~$80. Maybe someone could give a price for a 10Khz+ 96V/600A MOSFET
power stage with smoothing caps and freewheel diode that can take a
logic-level PWM input.
Which puts this setup as the likely cheapest given new parts. Now if you
happen to have a couple of 500A/100V+ SPDT relays lying around, this
would be the cheapest.
System Controller = $38
Motor Controller (with 4 stage parallel/series board) = $8 + $4 + ~$40
Sorry about the snappiness.
No worries; I did include a smiley in my response to indicate that I
wasn't bothered by it at all.
I think you can make one set of circuits that can be reused
in various ways to perform all operation aspects of a EV.
Maybe, but I suspect it is unlikely that such a set of circuits will
then be optimal for any single purpose. Consider that a circuit
intended to be used in a contactor-type controller can use components
that switch much more slowly than those in a circuit intended for use in
a PWM-type controller operating at any reasonable frequency. The
slow-switching circuit can take advantage of simpler, perhaps cheaper,
devices such as contactors and SCRs while the faster one is pushed into
using devices such as MOSFETs.
But they can use the exact same micro circuit. And a contactor style
series/parallel switch is just what a BatPack setup needs. This is why
*where* you split it up is of such importance. When someone presents
such a difficulty, you adjust the dividing line until you have a setup
that doesn't compromise how optimal the controller can be. Of course you
wouldn't use the contactor-type board if you are making a PWM setup, you
use one of the PWM boards with the micro board they both share.
Family makeup so far:
System Controller = One board, micro circuit with power supply and bus
terminator.
Motor Controller = Two (or more) boards:
a) micro circuit with cpu/power supervisor & bus interface (three chips)
b) power switching board, varies according to purpose, likely one
for series/parallel rearrangement of multiple strings, one for PWM
manipulate of motor supply (with one input). Combinations possible since
micro board supports up to 4 analog/digital inputs and 4 analog/digital
outputs.
c) sensor board for isolated electrical measurements (high voltage,
optional)
Battery Sensor/Controller = Two (or more) boards:
a) same micro board as above
b) sensor board for isolated electrical measurements (low voltage)
c) power switching board, not the same as above (optional, for full
battery control)
- or -
d) power balancer board (optional, for battery balancing when not
using board c)
I understand this, however, to evaluate what merit this
approach has we must consider a collection/combination
that approximates the features, etc. available from
conventional alternatives and see if that collection
actually offers a benefit that would encourage a user
to choose it over the alternative.
Ah-ha! Now we have it. To evaluate the system structured in
such a way as to give the same features and benefits as a
retail PWM controller, you would configure it as such:
Data Bus:
System Controller -> PWM-based Motor Controller
HV Bus:
Batteries (series) -> PWM-based Motor Controller -> Motor
I don't buy it. ;^> This configuration certainly is possible, and would
duplicate the features and benefits of a traditional PWM controller,
except that I don't really see this being an attractive option for any
end user. The feature it adds over a traditional controller is merely
that the brain/micro is a separate blob than the controller proper, so
one can replace one without replacing the other. I don't see this as a
selling feature by itself.
I guess is depends on how often you've roasted a controller playing
around. But it's not a significant point for all potential users, as you
have pointed out. But then, someone like that could come along, build
the PWM-based Motor Controller, flash the "stand-alone" firmware (which
would be the standard firmware with a throttle monitoring section that
replaces the bus com section of the code), and have a nice PWM
controller. No extra wiring required.
I have been thinking about the benefits seen from removing the System
Controller and having the Motor Controller do it's responsibilities. I
guess it depends on how flexible you want this system to be. Putting the
System Controller code in the Motor Controller would work fine, but
reduces what it can do since cycles and pins are now taken up doing the
Motor Controllers job (modulate/switch power to the motor while
monitoring it). This would mean someone like me would need to toss in
another micro board for my LCD/LED dash instead of just using the System
Controller.
So it comes down to saving $26 and one length of shielded wire versus
plenty of room to modify and expand the system... but then you don't
really save that much, because before the Motor Controller shares the
same micro board as the Battery Controllers/Sensors (4Mhz, 14 pin
micro), while the System Controller had a big beefy micro (20Mhz, 40pin)
that could do just about anything you could ever need it to, no matter
how many devices you attach to it. So I think that if you want to make
a "monolithic" PWM controller, you would just build a System Controller
board (which has 8 analog input/outputs and 2 analog/digital outputs and
8 digital only outputs), and wire it directly in the same box to the PWM
power switching board mentioned above. Then you would have pretty much a
higher voltage Alltrax controller, since it would use the same CPU even.
I don't want those who wish for flexibility have to suffer for those who
don't have a need for it. You can build two boards in one box much
easier then splitting one board into two. So I kept them separate, for
now. Plus it's much easy to debug in such an early phase of development.
I still think those wanting a "set it and forget it" setup and who would
be willing to go through the trouble of building their own
controller/BMS is a minority. But they are allowed for easy enough as
shown above. ;)
I think that to really attract people away from traditional one-box PWM
controllers the system has to offer some additional features or some
sort of cost advantages, etc. It seems that this is only possible when
one considers the configurations that include individual battery
switching (providing the additional BMS-type features) and the
series/parallel switching (to mimic the current multiplication features
of a standard PWM controller). It also happens that this is the only
arrangement that anyone is presently working on...
BMS *can* effectively be done with BatPack style Battery Controllers.
They are just an extreme. You can use boards a, b, and d listed above to
make a handy full-featured BMS/balancer setup on the cheap that
interfaces with the existing system controller. I notice that you
haven't mentioned the "can upgrade it bit-by-bit without tossing much
along the way" benefit. Which is a severely important one to me at
least. If you the type of person performing a DIY motor controller
build, you are not going to make it once and leave it alone, are you? If
you are looking to make and forget, you are going to use standard retail
parts.
Nope, you just missed a key part of what I wrote: "to the end user".
The benefits that you cite are of course why modularity and reuse are
important to *us* (i.e. those developing things), but the end user could
care less about what makes our lives easier or more productive unless he
derives some direct benefit such as a cheaper product.
I just didn't think you meant that kind of end user. I see the primary
potential user of such a project as someone who will want to tinker
with, constantly upgrade, and regularly modify their EV. In other words,
a DIYer. A hobbyist. An experimenter. To them ease of upgrade-ability
and customize-ability (and fix-ability) is typically something they care
about quite a bit. This wouldn't be a good idea for a typical soccer mom
who wants to pick up the groceries everyday. I would shudder to think of
that person building and operating their own EV motor controller.
The idea of being able to plug blocks together as required to assemble
one's particular EV is very attractive, but if it results in a more
costly, less efficient
How would a PWM setup as described above be less efficient? Or more
expensive?
(and potentially less reliable due to the
increased wiring & connections) system than one could otherwise build it
becomes a tough sell,
Yes, there is that technicality. A wire with a plug on each end is
potentially less reliable then a PCB trace with a solder pool on each
end. Having all the (spaghetti-style) wiring and connectors in my 1979
Honda Accord still working fine to this day does point to the fact that
is can be reliable enough though. But then a shielded wire rated to much
higher environmental extremes then any of the silicon, carrying little
current with similarly shielded and rugged "snap-lock" connectors is not
the wiring that would fail in any EV, is it? Wouldn't that likely be the
HV wiring, which is stressed so much more?
This is, unfortunately, looking more and more like modern LEGO, that is,
instead of basic generic building blocks that one rearranges into
whatever their imagination can envision (a la traditional LEGO) there
are custom blocks designed for specific tasks.
It only looks that way on the surface ;) The different variations of
controllers are built from a few mostly shared PCB boards. The only time
you diverge at the board level (creating variations) is when, as you
pointed out above, you would have sub par performance reusing an
existing part.
My description of a 120V system using 4 subnet switches stacked to allow
combining 5 24V strings is based on the traditional LEGO system view. I
assumed you were proposing building a general-purpose subnet switch
which at its most basic needs only to allow 2 input strings to be
combined in series or parallel to its output. An end user would then
take as many of these basic blocks and stack or otherwise arrange them
as required to build his particular system; he would not look for a
special purpose 4-input switch because he has 4 strings, or a 7-input
switch because he has 7 strings at the present time. Once you get to
this mentality you are really saying the system consists of a generic
brain blob (with application-specific code), and a special purpose
controller blob (either a PWM controller or a contactor controller,
etc.)
I would indeed say that (but not use the blob word). Actually the system
consists of a number of PCB boards designed to be reused wherever
possible. A "brain blob" is a $38 powerful micro board with bus
management circuitry and plenty of future expansion potential enclosed
in a box. This would be intended to be placed in the the passenger
compartment of the vehicle, but could be placed in the box with the
"controller blob" parts for a minimalist setup. The "controller blob" is
at least two boards: a $8 special purpose micro board with power
management features (PWM generator, dual Op Amps, voltage reference,
etc) and external power & watchdog monitoring for safety, and one of a
couple different power switching board designed for high voltages which
can be cascaded. You can add a small voltage\amperage sensor board if
you wish. And the "battery blobs" are the same $8 micro board as the
"controller blob" connected to the a voltage\amperage sensor board with
optionally either a power switching/bypass board designed for low
voltages or a power balancing board for active battery balancing
(haven't got that far yet).
So far, that sounds like a good starting point, eh?
(would've been a lot simpler to just wait for the detail diagram, but
not as much fun I suppose) :P
--
Stefan T. Peters
--- End Message ---
--- Begin Message ---
Steve and all,
Check out Ebay # 7579213053. It is the best dc to dc
supply you can get for $10, you can even trickle
charge your aux. battery if you like. It weighs a
couple pounds and no hassle, just hook up 4 wires and
maybe a fan.
Jimmy
From: "STEVE CLUNN" <[EMAIL PROTECTED]>
To: [email protected]
Subject: Re: Neg supply for E-meter
Date: Tue, 10 Jan 2006 08:00:32 -0500
I kind of have the same problem you do , with my e
mower , pack voltage
is
96v , and I have no 12v alx battery . I was thinking ,
e meter takes 80
ma
max , min pack voltage will be 70 , max voltage to
meter 35v , so to
power
the meter off traction pack , I could have a 400 ohm
resistor ( 10
watts
min) for the traction pack + going to the meter b+ ,
at the meter b+
have a
regulator to ground to keep the meter b+ for going
over 35v . The
regulator
would have to pass at least 250 ma to deal with 120v
battery charge
voltage
and e meter in sleep mode using just 20 ma.
steve clunn 3
----- Original Message -----
From: "damon henry" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Monday, January 09, 2006 1:50 PM
Subject: RE: Neg supply for E-meter
> Actually, there is very little wiring on the
motorcycle and most of
it I
> have replaced already. If it weren't for the way
the dumb lights are
> made, I could easily isolate my whole 12volt system
from the frame.
> Unfortunately all the lights mounting hardware is
built expecting the
> negative side of the power equation from the frame,
so even if I ran
a
> seperate ground wire to each, I would also have to
figure out a way
to
> isolate the individual hardware piece from the
frame. That is just
not
> worth the effort...
>
> damon
>
>
>>From: Mike & Paula Willmon <[EMAIL PROTECTED]>
>>Reply-To: [email protected]
>>To: [email protected]
>>Subject: RE: Neg supply for E-meter
>>Date: Mon, 09 Jan 2006 00:45:34 -0900
>>
>>Not to change the subject from the E-meter problem
but 10 of those
babies
>>would give the requisite 30A at 12V (nominal) to run
an accessory
system.
>>Keep 'em cool enough, the input wires protected and
you wouldn't have
to
>>worry about isolating their chassis' from your frame
ground. Throw a
Trace
>>(now Xantrex)C-40 charge controller inline and
there's an inexpensive
>>alternatve to DC-DC converters. More versatile too.
>>
>>Mike
>>
>>-----Original Message-----
>>From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]
>>Behalf Of Cor van de Water
>>Sent: Sunday, January 08, 2006 1:09 AM
>>To: [email protected]
>>Subject: RE: Neg supply for E-meter
>>
>>
>>I suggest to look at (used) laptop power supplies.
>>They are always switchers (you can use them from
>>100 - 240V AC, so they should be fine up to 360V
>>DC, which limits the voltage (during eq charge)
>>to about 22 x 12V batteries or 264V system.
>>They have plenty of output current (3+ Amp) and
>>often deliver around 15 - 18V.
>>
>>One currently on Ebay for $5 +$8 shipping:
>>Item nr 6837357331
>>
>>You can still look in your junk-box for old
>>power supplies - make sure they have a wide
>>input voltage range.
>>
>>Success,
>>
>>Cor van de Water
>>Systems Architect
>>Email: [EMAIL PROTECTED] Private:
http://www.cvandewater.com
>>Skype: cor_van_de_water IM:
[EMAIL PROTECTED]
>>Tel: +1 408 542 5225 VoIP: +31 20 3987567 FWD#
25925
>>Fax: +1 408 731 3675 eFAX: +31-87-784-1130
>>Proxim Wireless Networks eFAX: +1-501-641-8576
>>Take your network further http://www.proxim.com
>>
>>
>>-----Original Message-----
>>From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]
>>Behalf Of Rush
>>Sent: Saturday, January 07, 2006 2:40 PM
>>To: [email protected]
>>Subject: Re: Neg supply for E-meter
>>
>>
>>Lee,
>>
>>Would this one work -
http://datasheet.astrodyne.com/PWB.pdf
>>
>>Model AFC-15S, output V- 15vdc, output A - 0.66,
output W - 10W,
input V -
>>85-265vac?
>>
>>sorta pricy, $39 from them.
>>
>>Thanks
>>
>>Rush
>>Tucson AZ
>>www.ironandwood.org
>>
>>
>>----- Original Message -----
>>From: "Lee Hart" <[EMAIL PROTECTED]>
>>To: <[email protected]>
>>Sent: Saturday, January 07, 2006 12:29 PM
>>Subject: Re: Neg supply for E-meter
>>
>>
>> > From: damon henry
>> >> I still have a couple more questions specific to
my application.
I
>>don't
>> >> currently have an accesory battery, just a DC-DC
converter. The
>>negative
>> >> side of my DC-DC is connected to my frame. The
E-meter manual
states
>> >> that I must connect pin one to both the negative
side of my
traction
>>pack
>> >> as well as the negative supply for the E-meter.
This of course
>> >> connects
>> >> the negative side of my traction pack to the
negative side of my
DC-DC
>> >> and my frame. This sounds bad...
>> >
>> > Correct! You can't power the E-meter directly
from your 12v
accessory
>>power.
>> >
>> >>The manual offers two options...
>> >
>> > There is another option, which I prefer. Get a
small switching
power
>>supply
>> > that can run directly off your traction pack, and
which has an
isolated
>>12v
>> > output to power the E-meter (and nothing else).
The E-meter takes
so
>>little
>> > power that this can be a "wall wart" or old
laptop power brick.
Just
>>make
>> > sure it is a SWITCHING power supply, not one with
a 60hz
transformer!
>> >
>> > Almost all low-power switchers have "universal"
inputs and work on
AC
>> > or
>> > DC, at anything from 90-300v. For example, I use
an Astrodyne
15vdc
>> > 0.5amp switcher. It is rated for 90-264vac input,
but actually
works
>>from
>> > 60-350vdc. It's a little potted "brick", and
would even work
>> > underwater.
>> >
>> > The advantage of this approach is that the little
switcher is more
>>efficient
>> > than having to run the main DC/DC just to power a
*second* DC/DC
to get
>> > the 25ma or so the E-meter needs. The Astrodyne I
use draws 4-9ma
from
>> > my 132v propulsion pack, depending on whether the
E-meter is
asleep or
>>in
>> > full sunlight. A normal "full-size" DC/DC would
draw over 20ma
even
>> > with
>>NO
>> > load on its output.
>> > --
>> > Lee Hart
>> >
>> >
>> >
>>
>
>
--- End Message ---
--- Begin Message ---
Sounds like either:
- voltage divider, feeding a follower transistor which can withstand 120V.
The current through the resistors should be low, under 5mA. The transistor
carries the same current as the E-meter and dissipates between 50 and 90V
at that current.
If the max current of the E-meter is 100mA, the transistor will take
up to 9W, much more than Lee's solution but likely cheaper.
- big zener diode + resistor.
The resistor will need to carry MORE current than the E-meter ever draws,
so this is even less efficient, likely you will dissipate 20W continuous
and you will need a zener that can handle up to 200mA at 15V or more,
so a 3+ Watt zener. This is a rather crude solution....
BTW if your pack is 96V (8x12?) how can your min voltage be 70?
I would think you need to keep it over 1.75 - 1.67V per cell, so
about 80 - 84V?
Regards,
Cor van de Water
Systems Architect
Email: [EMAIL PROTECTED] Private: http://www.cvandewater.com
Skype: cor_van_de_water IM: [EMAIL PROTECTED]
Tel: +1 408 542 5225 VoIP: +31 20 3987567 FWD# 25925
Fax: +1 408 731 3675 eFAX: +31-87-784-1130
Proxim Wireless Networks eFAX: +1-501-641-8576
Take your network further http://www.proxim.com
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of STEVE CLUNN
Sent: Tuesday, January 10, 2006 5:01 AM
To: [email protected]
Subject: Re: Neg supply for E-meter
I kind of have the same problem you do , with my e mower , pack voltage is
96v , and I have no 12v alx battery . I was thinking , e meter takes 80 ma
max , min pack voltage will be 70 , max voltage to meter 35v , so to power
the meter off traction pack , I could have a 400 ohm resistor ( 10 watts
min) for the traction pack + going to the meter b+ , at the meter b+ have a
regulator to ground to keep the meter b+ for going over 35v . The regulator
would have to pass at least 250 ma to deal with 120v battery charge voltage
and e meter in sleep mode using just 20 ma.
steve clunn 3
----- Original Message -----
From: "damon henry" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Monday, January 09, 2006 1:50 PM
Subject: RE: Neg supply for E-meter
> Actually, there is very little wiring on the motorcycle and most of it I
> have replaced already. If it weren't for the way the dumb lights are
> made, I could easily isolate my whole 12volt system from the frame.
> Unfortunately all the lights mounting hardware is built expecting the
> negative side of the power equation from the frame, so even if I ran a
> seperate ground wire to each, I would also have to figure out a way to
> isolate the individual hardware piece from the frame. That is just not
> worth the effort...
>
> damon
>
>
>>From: Mike & Paula Willmon <[EMAIL PROTECTED]>
>>Reply-To: [email protected]
>>To: [email protected]
>>Subject: RE: Neg supply for E-meter
>>Date: Mon, 09 Jan 2006 00:45:34 -0900
>>
>>Not to change the subject from the E-meter problem but 10 of those babies
>>would give the requisite 30A at 12V (nominal) to run an accessory system.
>>Keep 'em cool enough, the input wires protected and you wouldn't have to
>>worry about isolating their chassis' from your frame ground. Throw a Trace
>>(now Xantrex)C-40 charge controller inline and there's an inexpensive
>>alternatve to DC-DC converters. More versatile too.
>>
>>Mike
>>
>>-----Original Message-----
>>From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
>>Behalf Of Cor van de Water
>>Sent: Sunday, January 08, 2006 1:09 AM
>>To: [email protected]
>>Subject: RE: Neg supply for E-meter
>>
>>
>>I suggest to look at (used) laptop power supplies.
>>They are always switchers (you can use them from
>>100 - 240V AC, so they should be fine up to 360V
>>DC, which limits the voltage (during eq charge)
>>to about 22 x 12V batteries or 264V system.
>>They have plenty of output current (3+ Amp) and
>>often deliver around 15 - 18V.
>>
>>One currently on Ebay for $5 +$8 shipping:
>>Item nr 6837357331
>>
>>You can still look in your junk-box for old
>>power supplies - make sure they have a wide
>>input voltage range.
>>
>>Success,
>>
>>Cor van de Water
>>Systems Architect
>>Email: [EMAIL PROTECTED] Private: http://www.cvandewater.com
>>Skype: cor_van_de_water IM: [EMAIL PROTECTED]
>>Tel: +1 408 542 5225 VoIP: +31 20 3987567 FWD# 25925
>>Fax: +1 408 731 3675 eFAX: +31-87-784-1130
>>Proxim Wireless Networks eFAX: +1-501-641-8576
>>Take your network further http://www.proxim.com
>>
>>
>>-----Original Message-----
>>From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
>>Behalf Of Rush
>>Sent: Saturday, January 07, 2006 2:40 PM
>>To: [email protected]
>>Subject: Re: Neg supply for E-meter
>>
>>
>>Lee,
>>
>>Would this one work - http://datasheet.astrodyne.com/PWB.pdf
>>
>>Model AFC-15S, output V- 15vdc, output A - 0.66, output W - 10W, input V -
>>85-265vac?
>>
>>sorta pricy, $39 from them.
>>
>>Thanks
>>
>>Rush
>>Tucson AZ
>>www.ironandwood.org
>>
>>
>>----- Original Message -----
>>From: "Lee Hart" <[EMAIL PROTECTED]>
>>To: <[email protected]>
>>Sent: Saturday, January 07, 2006 12:29 PM
>>Subject: Re: Neg supply for E-meter
>>
>>
>> > From: damon henry
>> >> I still have a couple more questions specific to my application. I
>>don't
>> >> currently have an accesory battery, just a DC-DC converter. The
>>negative
>> >> side of my DC-DC is connected to my frame. The E-meter manual states
>> >> that I must connect pin one to both the negative side of my traction
>>pack
>> >> as well as the negative supply for the E-meter. This of course
>> >> connects
>> >> the negative side of my traction pack to the negative side of my DC-DC
>> >> and my frame. This sounds bad...
>> >
>> > Correct! You can't power the E-meter directly from your 12v accessory
>>power.
>> >
>> >>The manual offers two options...
>> >
>> > There is another option, which I prefer. Get a small switching power
>>supply
>> > that can run directly off your traction pack, and which has an isolated
>>12v
>> > output to power the E-meter (and nothing else). The E-meter takes so
>>little
>> > power that this can be a "wall wart" or old laptop power brick. Just
>>make
>> > sure it is a SWITCHING power supply, not one with a 60hz transformer!
>> >
>> > Almost all low-power switchers have "universal" inputs and work on AC
>> > or
>> > DC, at anything from 90-300v. For example, I use an Astrodyne 15vdc
>> > 0.5amp switcher. It is rated for 90-264vac input, but actually works
>>from
>> > 60-350vdc. It's a little potted "brick", and would even work
>> > underwater.
>> >
>> > The advantage of this approach is that the little switcher is more
>>efficient
>> > than having to run the main DC/DC just to power a *second* DC/DC to get
>> > the 25ma or so the E-meter needs. The Astrodyne I use draws 4-9ma from
>> > my 132v propulsion pack, depending on whether the E-meter is asleep or
>>in
>> > full sunlight. A normal "full-size" DC/DC would draw over 20ma even
>> > with
>>NO
>> > load on its output.
>> > --
>> > Lee Hart
>> >
>> >
>> >
>>
>
>
--- End Message ---
--- Begin Message ---
Cor van de Water [mailto:[EMAIL PROTECTED] wrote:
> Sounds like either:
> - voltage divider
> - big zener diode + resistor.
Naa. Steve's pack is made up of either 12V or (more likely) 6V blocks,
and the E-meter is spec'd to accept something like 9V-35V for its
supply. There is rumour of some people having observed smoke at the
high end of this range, but Steve could certainly still power his meter
by tapping the pack at 12, 18, or 24V. The 18V or 24V taps are most
recommended since the 12V tap might sag below the E-Meter's minimum
req'd under heavy loads. The 18V tap is likely the best compromise
since it keeps furhter away from the high end of the input range. Note
also that the E-Meter itself includes a switching supply and draws
*less* current when you feed it a higher supply voltage.
Just remember that the E-Meter supply ground is internally connected to
the traction pack negative, so you *must* tap off the negative end of
the pack, not somewhere else along the string!
Now, was Steve really wondering about powering his meter, or was he also
wondering about the prescaler required on the voltage sense lines when
using it with a pack that exceeds 50V? His 96V pack will exceed 100V on
charge, so he'd need to use a 500V prescaler, which as I recall is a
divide by 10 (vs the 100V prescaler which is a divide by 2). I believe
the commercial prescalers use a resistor in series with the positive
voltage sense line so that it forms a voltage divider with the E-Meter's
input resistance. You want to ensure the voltage at the E-Meter sense
input never exceeds 50V. The commercial prescalers include a zener to
voltage sense negative to ensure this.
The DIYer could get put a 1M pot in series with the voltage sense input
to their meter, connect it to 24V of batteries (so that the voltage at
the meter is guaranteed to remain below the 50V smoke-release point),
configure the meter for a 100V or 500V prescaler as appropriate, and
then adjust the pot until the voltage displayed by the E-Meter agrees
with the known voltage at the batteries (e.g. measured with a
trustworthy DMM). Unhook everything and measure the resistance of the
pot and then substitute an appropriate value fixed resistor. Add a <50V
zener from the meter side of this resistor to the pack negative sense
lead to ensure that even when the line is disconnected from the meter
its voltage will not rise to unsafe levels. The zener leakage current
may affect the accuracy of the voltage readings so don't choose one too
low in voltage compared to the voltages you expect to see in normal use.
E.g. a 24V zener on the output of a 500V prescaler would prevent you
from reading voltages over ~240V, but shouldn't significantly affect
normal readings with a peak pack voltage under 200V.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
Drawback of "tapping" is the unbalance...
Especially if the E-meter is permanent connected....
BTW - Vicor makes DC/DC supplies for all kinds of weird
voltage ranges. For example the VI-2T1 supplies 12V from
an input of 66 - 160V DC.
http://www.vicorpower.com/documents/datasheets/ds_vi-200.pdf
Cor van de Water
Systems Architect
Email: [EMAIL PROTECTED] Private: http://www.cvandewater.com
Skype: cor_van_de_water IM: [EMAIL PROTECTED]
Tel: +1 408 542 5225 VoIP: +31 20 3987567 FWD# 25925
Fax: +1 408 731 3675 eFAX: +31-87-784-1130
Proxim Wireless Networks eFAX: +1-501-641-8576
Take your network further http://www.proxim.com
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of Roger Stockton
Sent: Tuesday, January 10, 2006 6:23 PM
To: [email protected]
Subject: RE: Neg supply for E-meter
Cor van de Water [mailto:[EMAIL PROTECTED] wrote:
> Sounds like either:
> - voltage divider
> - big zener diode + resistor.
Naa. Steve's pack is made up of either 12V or (more likely) 6V blocks,
and the E-meter is spec'd to accept something like 9V-35V for its
supply. There is rumour of some people having observed smoke at the
high end of this range, but Steve could certainly still power his meter
by tapping the pack at 12, 18, or 24V. The 18V or 24V taps are most
recommended since the 12V tap might sag below the E-Meter's minimum
req'd under heavy loads. The 18V tap is likely the best compromise
since it keeps furhter away from the high end of the input range. Note
also that the E-Meter itself includes a switching supply and draws
*less* current when you feed it a higher supply voltage.
Just remember that the E-Meter supply ground is internally connected to
the traction pack negative, so you *must* tap off the negative end of
the pack, not somewhere else along the string!
Now, was Steve really wondering about powering his meter, or was he also
wondering about the prescaler required on the voltage sense lines when
using it with a pack that exceeds 50V? His 96V pack will exceed 100V on
charge, so he'd need to use a 500V prescaler, which as I recall is a
divide by 10 (vs the 100V prescaler which is a divide by 2). I believe
the commercial prescalers use a resistor in series with the positive
voltage sense line so that it forms a voltage divider with the E-Meter's
input resistance. You want to ensure the voltage at the E-Meter sense
input never exceeds 50V. The commercial prescalers include a zener to
voltage sense negative to ensure this.
The DIYer could get put a 1M pot in series with the voltage sense input
to their meter, connect it to 24V of batteries (so that the voltage at
the meter is guaranteed to remain below the 50V smoke-release point),
configure the meter for a 100V or 500V prescaler as appropriate, and
then adjust the pot until the voltage displayed by the E-Meter agrees
with the known voltage at the batteries (e.g. measured with a
trustworthy DMM). Unhook everything and measure the resistance of the
pot and then substitute an appropriate value fixed resistor. Add a <50V
zener from the meter side of this resistor to the pack negative sense
lead to ensure that even when the line is disconnected from the meter
its voltage will not rise to unsafe levels. The zener leakage current
may affect the accuracy of the voltage readings so don't choose one too
low in voltage compared to the voltages you expect to see in normal use.
E.g. a 24V zener on the output of a 500V prescaler would prevent you
from reading voltages over ~240V, but shouldn't significantly affect
normal readings with a peak pack voltage under 200V.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
I think the term that best applies here is 'Kluge'.
I can throw together parts that I have available for a
fraction
of the cost most people would pay. In mass production
however it would not be economically feasable for the
masses.
Otmar has produced a control that is quite expensive,
but realizing the economies of scale is quite cheap
(low volume production, lots of engineering and
resources to create a control with minimal resources
and low volume buying power). If Otmars control was
produced in high volume with engineering cost spread
out it would be hard to compete in cost with your
proposed scheme. I would guess that in parts alone
Otmars control would be at least double in cost to a
Curtis control since he uses top of the line high cost
components. However double the cost in components
produces triple the performance of a Curtis, and
reliability proven by 1/4 mile racers like John
Wayland.
Even though I design controllers for a living (much
smaller scale than Otmars, usually less than 5Hp
presently), I would buy Otmars control for my next
conversion considering the time and effort it would
take to create something similiar. The time I spend
creating a similar control is far more than the cost
of an existing control that has been tested
extensively.
I'm not trying to discourage your efforts, just trying
to offer a different perspective in creating a control
for on road EV's. This is no simple task, and for
those that are not diehard engineers designing
controls should be left to the experimenters willing
to spend more money and time than simply buying an
existing control. Good luck on your project,
hopefully you'll come up with a good solution.
Rod
--- "Stefan T. Peters" <[EMAIL PROTECTED]>
wrote:
> Roger Stockton wrote:
> > Stefan T. Peters
> [mailto:[EMAIL PROTECTED] wrote:
> >
> >
> >> Just the ability to control each battery
> separately. This allows for
> >> automatic mix-and-match, battery fail
> over/replacement, and gradual
> >> system upgrades to a very high limit.
> >>
> >
> > Right, but we have also been able to identify that
> the penalties
> > associated with this approach include higher cost
> and lower efficiency
> > than a normal PWM controller. We've been able to
> quantify (at least
> > ballpark) the efficiency loss; I think that if we
> could quantify the
> > cost difference we'd be in a much better position
> to appreciate whether
> > or not the potential benefits (cost savings) of
> the BMS functionality is
> > sufficient to outweigh the efficiency and cost
> hits.
> >
>
> Well, so far I have this for a BatPack style setup:
>
> System Controller = $38
> Battery Controller (with 18V/500A switch board) =
> $61 each
> Motor Controller (with 2 stage series parallel
> board) = $8 + $4 + TBD
>
> Yeah, I know that last one is the big one. I have a
> feeling it will be
> around $150-200 for the two string 1000A
> series-parallel version. A
> straight PWM version for use with Battery Sensors
> instead of controllers:
>
> System Controller = $38
> Battery Sensor = $8 + $4 each
> Motor Controller = $8 + $4 + depends on voltage, for
> a 24V/600A one it's
> ~$80. Maybe someone could give a price for a 10Khz+
> 96V/600A MOSFET
> power stage with smoothing caps and freewheel diode
> that can take a
> logic-level PWM input.
>
> Which puts this setup as the likely cheapest given
> new parts. Now if you
> happen to have a couple of 500A/100V+ SPDT relays
> lying around, this
> would be the cheapest.
>
> System Controller = $38
> Motor Controller (with 4 stage parallel/series
> board) = $8 + $4 + ~$40
>
>
> >
> >> Sorry about the snappiness.
> >>
> >
> > No worries; I did include a smiley in my response
> to indicate that I
> > wasn't bothered by it at all.
> >
> >
> >> I think you can make one set of circuits that can
> be reused
> >> in various ways to perform all operation aspects
> of a EV.
> >>
> >
> > Maybe, but I suspect it is unlikely that such a
> set of circuits will
> > then be optimal for any single purpose. Consider
> that a circuit
> > intended to be used in a contactor-type controller
> can use components
> > that switch much more slowly than those in a
> circuit intended for use in
> > a PWM-type controller operating at any reasonable
> frequency. The
> > slow-switching circuit can take advantage of
> simpler, perhaps cheaper,
> > devices such as contactors and SCRs while the
> faster one is pushed into
> > using devices such as MOSFETs.
> >
> >
>
> But they can use the exact same micro circuit. And a
> contactor style
> series/parallel switch is just what a BatPack setup
> needs. This is why
> *where* you split it up is of such importance. When
> someone presents
> such a difficulty, you adjust the dividing line
> until you have a setup
> that doesn't compromise how optimal the controller
> can be. Of course you
> wouldn't use the contactor-type board if you are
> making a PWM setup, you
> use one of the PWM boards with the micro board they
> both share.
>
> Family makeup so far:
>
> System Controller = One board, micro circuit with
> power supply and bus
> terminator.
>
> Motor Controller = Two (or more) boards:
> a) micro circuit with cpu/power supervisor & bus
> interface (three chips)
> b) power switching board, varies according to
> purpose, likely one
> for series/parallel rearrangement of multiple
> strings, one for PWM
> manipulate of motor supply (with one input).
> Combinations possible since
> micro board supports up to 4 analog/digital inputs
> and 4 analog/digital
> outputs.
> c) sensor board for isolated electrical
> measurements (high voltage,
> optional)
>
> Battery Sensor/Controller = Two (or more) boards:
> a) same micro board as above
> b) sensor board for isolated electrical
> measurements (low voltage)
> c) power switching board, not the same as above
> (optional, for full
> battery control)
> - or -
> d) power balancer board (optional, for battery
> balancing when not
> using board c)
>
>
> >>> I understand this, however, to evaluate what
> merit this
> >>> approach has we must consider a
> collection/combination
> >>> that approximates the features, etc. available
> from
> >>> conventional alternatives and see if that
> collection
> >>> actually offers a benefit that would encourage a
> user
> >>> to choose it over the alternative.
> >>>
> >> Ah-ha! Now we have it. To evaluate the system
> structured in
> >> such a way as to give the same features and
> benefits as a
> >> retail PWM controller, you would configure it as
> such:
> >>
> >> Data Bus:
> >>
> >> System Controller -> PWM-based Motor Controller
> >>
> >> HV Bus:
> >>
> >> Batteries (series) -> PWM-based Motor Controller
> -> Motor
> >>
> >
> > I don't buy it. ;^> This configuration certainly
> is possible, and would
> > duplicate the features and benefits of a
> traditional PWM controller,
> > except that I don't really see this being an
> attractive option for any
> > end user. The feature it adds over a traditional
> controller is merely
> > that the brain/micro is a separate blob than the
> controller proper, so
> > one can replace one without replacing the other.
> I don't see this as a
> > selling feature by itself.
> >
> >
>
> I guess is depends on how often you've roasted a
> controller playing
> around. But it's not a significant point for all
> potential users, as you
> have pointed out. But then, someone like that could
> come along, build
> the PWM-based Motor Controller, flash the
> "stand-alone" firmware (which
> would be the standard firmware with a throttle
> monitoring
=== message truncated ===
--- End Message ---
