EV Digest 5549
Topics covered in this issue include:
1) Re: Another way to get isolated DC-DC....
by Lee Hart <[EMAIL PROTECTED]>
2) Getting BACK into the EV Building BiZ
by Steven Lough <[EMAIL PROTECTED]>
3) Re: Another way to get isolated DC-DC....
by "Mike Phillips" <[EMAIL PROTECTED]>
4) Re: Nilars, etc...
by Doug Weathers <[EMAIL PROTECTED]>
5) Re: LED Matrix (was) the Mark Brueggemann Meter
by "Joe Vitek" <[EMAIL PROTECTED]>
6) Re: the Mark Brueggemann Meter
by Lee Hart <[EMAIL PROTECTED]>
7) Nilar batts
by Reverend Gadget <[EMAIL PROTECTED]>
8) Re: LED Matrix (was) the Mark Brueggemann Meter
by Lee Hart <[EMAIL PROTECTED]>
9) Re: Nilar batts
by "Mike Phillips" <[EMAIL PROTECTED]>
10) Re: Nilar batts
by "Michael Mohlere" <[EMAIL PROTECTED]>
11) 2006 Power of DC Sponsors
by Chip Gribben <[EMAIL PROTECTED]>
12) Nimh meltdown
by Mike Phillips <[EMAIL PROTECTED]>
13) Re: the Mark Brueggemann Meter
by "Peter Shabino" <[EMAIL PROTECTED]>
14) Re: the Mark Brueggemann Meter
by "Peter Shabino" <[EMAIL PROTECTED]>
15) Treehugger TV segment on plugin hybrids, shows some of conversion process.
by "Mike Ellis" <[EMAIL PROTECTED]>
16) Re: the Mark Brueggemann Meter
by Lee Hart <[EMAIL PROTECTED]>
17) Need Advice -- Gel pack rejuvenation plan
by Meta Bus <[EMAIL PROTECTED]>
18) Re: the Mark Brueggemann Meter
by "Peter Shabino" <[EMAIL PROTECTED]>
19) Re: How Hot Does a Terminal Get?
by "Chuck Hursch" <[EMAIL PROTECTED]>
20) Re: Question about IGFETs
by "John Luck Home" <[EMAIL PROTECTED]>
21) RE: LED Matrix (was) the Mark Brueggemann Meter
by "Roger Stockton" <[EMAIL PROTECTED]>
22) Re: Reverse Contactor
by Wayne <[EMAIL PROTECTED]>
23) Re: Lee's regs in action
by Tim Wong <[EMAIL PROTECTED]>
--- Begin Message ---
Mike Phillips wrote:
>
> Lee,
>
> Any idea of a fet gate isolation circuit? These main boards for our
> USE vehilces are no longer available and the traces & substrate die
> with great ease.
There are lots of ways this can be done. But without knowing the
physical situation, I'm not sure which ones could be retrofitted the
easiest. Examples include:
- small individual DC/DCs to power each gate driver
- special optocouplers for gate driving
- gate drive transformers
- SCR crowbars to lock down the gates
--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
--- End Message ---
--- Begin Message ---
Seems like( ICE ) Engine Failure always has something to do with me
getting my EV-Building Juices going... Back in 1980, it was the failure
of our friends VW Bug, that was the Message-from-above, that finally got
me going converting "ION-1" into a True-Electric. ( I sold the ICE
engine to them, and that left a big gaping Hole in the VW Trans-axle,
where a nice Prestolite fit JUST PERFECT !
This time it is the failure of my little Yamaha 125cc Scooter. So as we
speak... it is being ELECTROFIED ! with help from Dave Cloud of Cloud
Racing... ( Ya see... I no longer have a shop of my own like I did back
in 1980-81 ) ( Lough Motors GMC )
So the HUNT is on for a few componants that Dave doesn't have at his
"Skunk-Works" in Woodinville...
1: DC/DC converter from 48 to 14, or perhaps 60 to 14vdc
2: 48v or possibly 60v Charger... Yes, I know a PFC-20 would be a
wonderful unit... But Dave says we may already be "Over Budget" having
to re-manufacturere a whole rear end of the Scooter, to replace the
integral motor/trans/CVdrive that most all scooters come with...
3. A nice Matched SET of gauges (small-ish) for voltage ( 0 to 75-ish)
and Amperage... 0 to 200 - with Shunt.
Dave has worked out the gear ratio. And we feel the Scooter will have as
much, or more performance than the origional ICE unit. ( if we can keep
the weight down
--
Steven S. Lough, Pres.
Seattle EV Association
6021 32nd Ave. N.E.
Seattle, WA 98115-7230
Day: 206 850-8535
Eve: 206 524-1351
e-mail: [EMAIL PROTECTED]
web: http://www.seattleeva.org
--- End Message ---
--- Begin Message ---
Here is what our resident expert wrote some time ago about our IGBT's
in our chargers. See if this helps.
Quote\
"I measured across the power supply of the UC2708 that drives the gate
of the IGBTs. Chris made the point maybe it drops under load (lots of
switching). Its not a charge pump its a transformer. Each IGBT driver
circuit has a transformer driven by a 12 volt square wave (P12V in the
telemetry list), the output of the xformer powers the IGBT
driver. There's also 2 optoisolators, one drives the driver the other
is feedback to tell the computer that the voltage across the xistor
did in fact go down. Thats why there's 3 wires to each IGBT driver
(emitter, gate, collector) not just 2 (gate emitter). BTW an IGBT is
just a bipolar transistor with a small fet connecting collector to
base (integrated gate bypolar transistor), its got low gate
capacitance and lots of voltage capability but otherwise acts like a
fet except that the minimum on voltage is around .9 volts. Turns out
for power fets thats pretty typical anyway."
End quote\
Mike
--- In [EMAIL PROTECTED], Lee Hart <[EMAIL PROTECTED]> wrote:
>
> Mike Phillips wrote:
> >
> > Lee,
> >
> > Any idea of a fet gate isolation circuit? These main boards for our
> > USE vehilces are no longer available and the traces & substrate die
> > with great ease.
>
> There are lots of ways this can be done. But without knowing the
> physical situation, I'm not sure which ones could be retrofitted the
> easiest. Examples include:
>
> - small individual DC/DCs to power each gate driver
> - special optocouplers for gate driving
> - gate drive transformers
> - SCR crowbars to lock down the gates
> --
> Ring the bells that still can ring
> Forget the perfect offering
> There is a crack in everything
> That's how the light gets in -- Leonard Cohen
> --
> Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
>
--- End Message ---
--- Begin Message ---
On Jun 6, 2006, at 5:43 AM, Mike Chancey wrote:
In case you haven't seen it, the revised Album has a search by
location feature.
Wow, that's slick! Nice job!
Thanks,
Mike Chancey,
'88 Civic EV
Kansas City, Missouri
EV Photo Album at: http://evalbum.com
My Electric Car at: http://www.geocities.com/electric_honda
Mid-America EAA chapter at: http://maeaa.org
Join the EV List at: http://www.madkatz.com/ev/evlist.html
In medio stat virtus - Virtue is in the moderate, not the extreme
position. (Horace)
--
Doug Weathers
Las Cruces, NM, USA
<http://learn-something.blogsite.org/>
--- End Message ---
--- Begin Message ---
I am a lurker who rears his ugly head from time to time. I love the EE banter
(as I am one
myself) due to my tinkering love since I was... before I can remember. I was a
tech first
and then went and got my BSEE 11 years after high school.
<ugly head rearing>
> > Except you're not counting the thousands of semiconductor devices
> > (diodes, transistors, etc.) inside the PIC....
Yes, but who cares to count all that since it is already designed. The PIC is
still only
one part as far as the design is concerned. There are also quite a few
transistors,
diodes, etc. in the '3914. The PIC just has more of them.
> > "smarter" that it doesn't NEED a micro to do its job.
You're right, it doesn't NEED a micro, but a micro would be cheaper with fewer
parts as
Zik pointed out. I am one who loves "old school" electronics and loves the
'3914 solution.
I have played with them since 1980 (high school). However, I also like the PIC
solution.
Speaking of elegant, my criteria on elegance in design is; simplicity, parts
count, and
reliability (and coolness factor too). I think both are "elegant". They are
both simple
(from a design point of view), simple to use, parts count is low (relatively),
and both
can be built to be extremely reliable (automotive temp range parts).
If you aren't familiar with PICs, the '3914 is the way to go, if you are, the
PIC is the
way to go. Or, take your "pic" on which one is "cooler". I like them both and
maybe a
version of each should be designed and built.
--
joe
--- End Message ---
--- Begin Message ---
Doug Weathers wrote:
Peter Shabino wrote:
>> LM3914 data sheet... will get you the basic hookup info for the chip.
>> Now to get a X/Y display you will need some extra parts. Ten 10k
>> resistors (5% 1/4W ok) and 2 octal inverters...
> OK, that seems pretty clear... Would they need a separate power supply?
Yes. The LM3914 doesn't need a regulated 5v supply; it can run straight
off a 12v battery. But if you use a TTL family inverter, you also need a
5v power supply to run it. If you use transistors instead, you don't
need an extra power supply.
>> If you would like a set of schematics or a board file let me know. I
>> have the layout tools and can work on it this weekend.
There is so little to this circuit that a PC board layout may not be
justified unless you plan to build a lot of them. The PC board will cost
more than everything else combined, and take longer to lay out then just
building it.
Build your prototype on one of those white plastic breadboards. Every
hole is a socket; this makes it easy to play with the color LEDs in the
array to get an appropriate good/fair/bad graph. It's also a fast way to
figure out the resistor values to get suitable scales for the voltage
and current.
When it works, you have your circuit. Get a piece of perfboard with
holes on 0.1" centers. Solder single-row IC sockets to it to make your
10x10 LED matrix. Hand wire the rest of the parts. The highly regular
nature of this circuit means the traces on the perfboard are already
close to your circuilt.
--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
--- End Message ---
--- Begin Message ---
The thing about the Nilar battery is that even though
the assembly is different, the chemistry is still
NiMH. They are still limited to 10 amp hr, and will
NOT like to be paralleled. These are not issues to be
delt with by a beginner. That is a lot of money to go
up in smoke.
Gadget
visit my websites at www.reverendgadget.com, gadgetsworld.org,
leftcoastconversions.com
--- End Message ---
--- Begin Message ---
Zik Saleeba wrote:
> Take it easy man. I'm not looking for a pissing competition here.
No, I wasn't trying to argue at all. I'm seriously encouraging you to go
ahead, and build one with a micro. Lots of people prefer a micro-based
solution. See how it turns out.
> these days microcontrollers are intended for exactly these kinds of
> applications. It's easier to design, less external components to go
> wrong, cheaper and usually more reliable.
To a man with a hammer, every problem looks like a nail. :-)
If you know how to program PICs, and don't know how to do analog design,
then a PIC makes sense for you. Conversely, if you know analog design,
then the LM3914 makes more sense.
> The cheapest microcontrollers cost 50 cents in quantity
Except that a 50 cent micro can't do the job.
> and they're commonly used for things as simple as making an LED flash.
Except that you can buy an LED that already has the flasher inside for
less than that.
> There's basically no concept of overkill where microcontrollers are
> concerned - they're cheap and small enough for the most menial of tasks.
> For instance modern cars use a system called CAN which sends signals
> around the car digitally and each cluster of instruments or lights
> etc. has a tiny microcontroller which talks CAN on one side and turns
> lights etc. on and off on the other side. This is much simpler and
> cheaper than using an old style wiring loom since the one wire goes to
> the whole car rather than having tens of different cable trails to get
> wrong.
Making things more complex adds enormously to the design cost, but can
reduce the cost per unit slightly. Even if the extra complexity doesn't
reduce cost, it has benefits to the manufacturer (keeps things secret,
monopolizes repairs and replacement parts, etc.) So, large manufacturers
favor complex solutions. They spread the development cost over a million
units, and make it back by saving a few cents per unit, and in increased
repair and replacement costs.
If you were going to build a thousand Brueggemann meters, a micro makes
sense. If you wanted a million meters, a custom designed IC would make
even more sense. But here, most people want exactly one. Thus, I think
the analog design is a better choice.
> But personally I'd go with the reduced part count I'd get with a PIC
> (and the lower costs etc).
>
> 1 x 16F913 microcontroller ($2.49 from microchip direct)
> 1 x 7805 regulator ($1)
> 100 x 3mm red LEDs ($2)
> 3 x decoupling capacitors (<$1)
> 2 x resistors for a voltage divider for the battery level (<10c)
>
> That's it.
Is it? A few things I notice:
- Where's your voltage reference? The 7805 is pretty poor at this.
(The LM3914 has an internal reference.)
- How do you measure the voltage across the shunt? The micro needs an
opamp to boost the 50mv shunt voltage to 5v for the internal A/D
converter to work. (The LM3914 can measure this directly.)
- What sets LED current? You'll need ten resistors to set LED current.
(The LM3914 has current drivers built in.)
- Can the 16F913 source enough current for reasonable LED brightness?
(The LM3914 does up to 30ma. But I think both approaches will need
high side drivers, i.e. ten transistors or equivalent)
- LM3914 is $1.65 from www.jameco.com, so two of them are cheaper than
the 16F913 and 7805.
So, I still think you have more parts and a higher cost with the micro.
On the other hand, you could program the micro to do lots more things.
The 10x10 LEDs could display battery voltage, or a smiley face, etc.
Here's a reference design for a micro-based approach. The Velleman Mini
Message Board kit (MK124) is a 3.5" x 2.5" PCB with a 5x7 array of LEDs.
It uses a PIC 16C588, 7805, 5 diodes, 10 resistors, 3 pushbuttons, and 3
capacitors. Its software implements a tiny "scrolling message" display.
This kit costs $24.95 from www.jameco.com. I think it's a pretty fair
estimate of what the final parts and cost of a micro-based Brueggemann
meter would be like.
In contrast, the Vellemann Audio VU meter kit (K4304) uses the LM3914
and costs $16.99. But it includes an opamp, rectifier, and logarithmic
scale circuits. A better comparison would be the DC Battery Voltage
Monitor kit (VM-1) for $7.95, but it is poorly designed and uses two
LM324s.
--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
--- End Message ---
--- Begin Message ---
Have you tried paralleling Nimh?
Mike
--- In [EMAIL PROTECTED], Reverend Gadget <[EMAIL PROTECTED]> wrote:
>
> The thing about the Nilar battery is that even though
> the assembly is different, the chemistry is still
> NiMH. They are still limited to 10 amp hr, and will
> NOT like to be paralleled. These are not issues to be
> delt with by a beginner. That is a lot of money to go
> up in smoke.
>
> Gadget
>
>
>
>
> visit my websites at www.reverendgadget.com, gadgetsworld.org,
leftcoastconversions.com
>
--- End Message ---
--- Begin Message ---
More good information - I had not heard anything about these batteries not
liking to be in parallel, etc.....guess I'm back to "good ol' lead acid!!!".
Thanks again for all the input - you folks have already saved me a ton of
money and heart ache!!!
Mike
From: "Mike Phillips" <[EMAIL PROTECTED]>
Reply-To: ev@listproc.sjsu.edu
To: Reverend Gadget <ev@listproc.sjsu.edu>
Subject: Re: Nilar batts
Date: Tue, 06 Jun 2006 19:55:17 -0000
Have you tried paralleling Nimh?
Mike
--- In [EMAIL PROTECTED], Reverend Gadget <[EMAIL PROTECTED]> wrote:
>
> The thing about the Nilar battery is that even though
> the assembly is different, the chemistry is still
> NiMH. They are still limited to 10 amp hr, and will
> NOT like to be paralleled. These are not issues to be
> delt with by a beginner. That is a lot of money to go
> up in smoke.
>
> Gadget
>
>
>
>
> visit my websites at www.reverendgadget.com, gadgetsworld.org,
leftcoastconversions.com
>
--- End Message ---
--- Begin Message ---
Thanks to all of this year's sponsors who made the Power of DC a
success. We really appreciate it!
1) QuickCable provided tools to give away to the winners
http://www.quickcable.com
2) Jim Husted from Hi-Torque Electric made us a really nice 6 inch
motor to give away as a prize.
http://www.hitorqueelectric.com
3) Shawn Wagonner and Matt Graham from SEVO (Suncoast EV Outfitters)
donated an electric scooter motor and 36-volt controller
http://www.suncoast.net
4) Otmar Ebenhoech from Cafe Electric sponsored the event and
provided two t-shirts
http://www.cafeelectric.com
5) Rich Rudman from Manzanita Micro
http://www.manzanitamicro.com
6) Dave Stensland from Megawatt Motorworks
http://www.megawattmotorworks.com
7) Roderick Wilde from EV Parts provided some T-shirts to give away
http://www.evparts.com
8) Ken Koch from KTA Services
http://www.kta-ev.com
9) SkooterCommuter
http://www.skootercommuter.com
10) Battery Warehouse
11) Maryland Energy Administration
http://www.energy.state.md.us
12) Baltimore Clean Cities Campaign
http://www.energy.state.md.us/programs/transportation/baltimore.htm
13) Virginia Solar Council
Chip Gribben
NEDRA Power of DC
http://www.powerofdc.com
--- End Message ---
--- Begin Message ---
Sunday morning I had 3 of the 5 nimh packs totally go up in smoke, two
with unknown damage.
It's a very complicated process using nimh AND paralleling them. But I
need facts so I'm giving it a try.
For some reason after many successful charge/discharge cycles the nimh
array decided not to reach it's normal final charge voltage overnight.
So the charger kept putting in a couple amps and the array just got
hotter and hotter until 114 of the 190 modules cracked open and spewed
gasious KoH all over the place. The built in vent system failed 100%.
Instead the temperature measuring well split open.
So I am adding a protection circuit to pull a pack out of the loop if
it gets too hot. That means each pack will have a thermal fan circuit,
a thermal relay circuit and a current monitoring circuit.
I didn't charge the array to it's max voltage. I didn't want to mess
with shooting for absolute peak voltage. So I settled on 323v. The
peaking of 5 nimh packs has the potential disadvantage of not having
them all peak at the same time. From my testing 323vdc is just at the
very long, actual peak, not past it. Peaking the pack before 323 just
resulted in far less charging and an easy peak to shut off the charger
with, far too soon.
After every few charge discharge cycles I would separate the packs via
the relays and charge each one individually. This insured that they had
the same final voltage in them. After doing this once, they stayed very
close on consequent charge/discharge cycles.
I was up to 90+ miles worth of testing. Range was 19.7 miles max. I
hope the 5 replacements are in as good of condition.
All in the name of research.
Mike
Here's to the crazy ones.
The misfits.
The rebels.
The troublemakers.
The round pegs in the square holes.
The ones who see things differently
The ones that change the world!!
www.RotorDesign.com
--- End Message ---
--- Begin Message ---
PDF of the schematic are at http://www.wire2wire.org/Brueggermann_meter.pdf.
Note this assumes a 0 to 1.2V input referanced to ground. With some more
tweeking this can be changed and offset (ie full scale is 4 to 7V input ref
to ground). Also if you replace the 620 ohm resistor with CDS cell circuit
you can get a auto dimming display.
Have not tried any of this but should work. Also see more comments below.
From: Doug Weathers <[EMAIL PROTECTED]>
Reply-To: ev@listproc.sjsu.edu
To: ev@listproc.sjsu.edu
Subject: Re: the Mark Brueggemann Meter
Date: Tue, 6 Jun 2006 01:11:57 -0600
Peter, thanks for taking the time to reply.
On Jun 2, 2006, at 8:14 AM, Peter Shabino wrote:
First grab the data sheet at http://cache.national.com/ds/LM/LM3914.pdf
Got it. It's mostly understandable.
That will get you the basic hookup info for the chip. Now to get a X/Y
display you will need some extra parts. 10 10k resistors (5% 1/4W ok) and
2 octal inverters (pick inverters that have a high state that can drive
you leds (push pull output not OC)).
to hook this up wire both LM3914 as normal except for the LEDs. Now on the
X LM3914 connect the banded (-) side of all leds to this chip (10 leds in
each column x 10 columns). On the Y LM3914 connect the 10 10K resistors
between each output and +V. Also connect each output to one input on a
inverter. The output from the inverter to the (+) side of one row of 10
leds.
How this works is each LM3914 connect its output to ground when its "on".
So for the X row we can just use it as the (-) side of the led. If the
output is "off" the pin will be floating and the LED will not turn on. For
the Y row we need a logic high at the led to turn it on. Since the LM3914
can only drive a low we need to invert the signal to turn on the led. The
pull up resistor and inverter takes care of this.
OK, that seems pretty clear. Lee Hart and Joe Smalley (off list) suggested
transistors instead of the inverters.
I understand how the transistors work in this circuit, but I'm hazy on the
inverters. Would they need a separate power supply? I think the
transistors can be powered by the LM3914, which in turn is powered by the
voltage source that it's measuring.
What's your opinion?
The transistors and inverters do the same job. Take a active logic low and
convert it to a active logic high. The MC14049UBDR2 is also a CMOS part and
will run off of 3 to 18V just like the LM3914. There for the entire circuit
can run of a single 12V supply. Now if you signals are not referanced to
ground then life will get tricky as you will need to do some issolation work
to get this thing to run.
The best situation would be to have the current shunt in the ground leg of
the pack and also have a small DC-DC (or zener setup) to provide the 12V
supply to just this circuit. In that case build up a R network to divide
pack voltage down to 1.2V (connect it from the + side of the pack to the
motor side of the shunt) full scale and hook that to the column input +.
Connect the battery side of the shunt to the row input +. Connect the DC-DC
(or Zener) input to + pack and the motor side of the shunt. Connect the -
output to the motor side of the shunt and to ground on the circuit. connect
the + output of the DC-DC to the +12V in the circuit.
Clear as mud.....
Questions / comments?
Later,
Wire
If you would like a set of schematics or a board file let me know. I have
the layout tools and can work on it this weekend.
That seems a lot to ask, but I'd really like a schematic. I don't know
what I'd do with a board file, but others might. You can't post files to
the list, so how about posting them to www.evforge.net? Perhaps make it an
EV project there? Or send the files to me and I can post them on my
website for the rest of the list.
I have Eagle running, and Xcircuit shouldn't be far behind. Plus a really
slick-looking package for the Mac called MI-SUGAR. And of course I can
read PDFs. And ASCII :)
Thanks!
Later,
Wire
--
Doug Weathers
Las Cruces, NM, USA
<http://learn-something.blogsite.org/>
--- End Message ---
--- Begin Message ---
missed the questions at the end.. more below.
From: Doug Weathers <[EMAIL PROTECTED]>
Reply-To: ev@listproc.sjsu.edu
To: ev@listproc.sjsu.edu
Subject: Re: the Mark Brueggemann Meter
Date: Tue, 6 Jun 2006 01:11:57 -0600
Peter, thanks for taking the time to reply.
On Jun 2, 2006, at 8:14 AM, Peter Shabino wrote:
First grab the data sheet at http://cache.national.com/ds/LM/LM3914.pdf
Got it. It's mostly understandable.
That will get you the basic hookup info for the chip. Now to get a X/Y
display you will need some extra parts. 10 10k resistors (5% 1/4W ok) and
2 octal inverters (pick inverters that have a high state that can drive
you leds (push pull output not OC)).
to hook this up wire both LM3914 as normal except for the LEDs. Now on the
X LM3914 connect the banded (-) side of all leds to this chip (10 leds in
each column x 10 columns). On the Y LM3914 connect the 10 10K resistors
between each output and +V. Also connect each output to one input on a
inverter. The output from the inverter to the (+) side of one row of 10
leds.
How this works is each LM3914 connect its output to ground when its "on".
So for the X row we can just use it as the (-) side of the led. If the
output is "off" the pin will be floating and the LED will not turn on. For
the Y row we need a logic high at the led to turn it on. Since the LM3914
can only drive a low we need to invert the signal to turn on the led. The
pull up resistor and inverter takes care of this.
OK, that seems pretty clear. Lee Hart and Joe Smalley (off list) suggested
transistors instead of the inverters.
I understand how the transistors work in this circuit, but I'm hazy on the
inverters. Would they need a separate power supply? I think the
transistors can be powered by the LM3914, which in turn is powered by the
voltage source that it's measuring.
What's your opinion?
If you would like a set of schematics or a board file let me know. I have
the layout tools and can work on it this weekend.
That seems a lot to ask, but I'd really like a schematic. I don't know
what I'd do with a board file, but others might. You can't post files to
the list, so how about posting them to www.evforge.net? Perhaps make it an
EV project there? Or send the files to me and I can post them on my
website for the rest of the list.
board files is IBM speak for gerber / technical data to build a PCB (we also
call harddrives DASD just to confuse folks). If its needed I can lay this
out in a afternoon (hard part will be finding the free afternoon). Also
before you order any raw cards someone should protoboard it once and verify
it works as expected.
Later,
Peter
I have Eagle running, and Xcircuit shouldn't be far behind. Plus a really
slick-looking package for the Mac called MI-SUGAR. And of course I can
read PDFs. And ASCII :)
Thanks!
Later,
Wire
--
Doug Weathers
Las Cruces, NM, USA
<http://learn-something.blogsite.org/>
--- End Message ---
--- Begin Message ---
http://www.treehugger.com/files/2006/06/treehuggertv_plug_in_prius.php
--- End Message ---
--- Begin Message ---
Peter Shabino wrote:
>
> PDF of the schematic are at http://www.wire2wire.org/Brueggermann_meter.pdf.
It's a good start! The bias resistors to create the expanded scale
voltmeter, and 0-50mv shunt ranges are missing.
The 4049 has very low source current capability (a few ma). That's
probably too low to get sunlight-readable brightness. I still think a
transistor is the way to go.
If you must use an IC, there are octal high-side drivers. But they are
not very common, and so more expensive.
--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
--- End Message ---
--- Begin Message ---
I have a string of buddy pairs, 8G31DT (size 31 gel 12v Deka
Dominators), which are currently dead-flat, and have been lying in that
state for months. I am going to attempt to bring them back from the
dead, and get some useful percentage of a SOC-- I'll be happy if I can
recover 25% of their original 200ah (paired) nominal 20-hour capacity.
The string of 24 paralleled pairs (288v) have a manufacturer-recommended
bulk charge rate of 14.1v (at 68f), which equates to 338.4v. I am
attempting this recovery in Florida, intending an overnight charge for
several nights (low amperage), where I expect an average temp of around
76-78f-- which lowers my target bulk voltage to around 14v max (per
Deka's temp/charge range chart).
The 10kw charger I inherited is dead too, so...
I am resorting to a nice PFC desktop supply, which gives me a steady
351.9vdc, and is rated for a maximum of around 1300 watts (with 110vac
input, which is what I'll be employing). [I feel more comfortable with
this lower-power supply than trying to pour 10kw into the dead pack, anyway]
Assuming the PFC supply is 80% efficient (it proclaims a power factor of
0.995 at full load), that means I will be offering around 3 amps to the
string-of-resistors-formerly-known-as-a-battery-pack. I will, however,
also be a good half-volt over-voltage, which concerns me.
If necessary, I could throw another battery pair onto the string, which
would make my constant-voltage PFC match up almost perfectly, but I am
curious if over-voltage might actually help recover from the sulphation
which no doubt threatens to stratify my string.
Before I hook all this up and throw the switch (open my pack relays and
then turn my little PFC loose), can someone help me predict what to expect?
Thanks.
--- End Message ---
--- Begin Message ---
Should source 10mA with no issues most of the leds I am working with these
days will burn your eyes out at that level (bright red idiot lights so
operators don't pull cards out of the tester with power up). Transistors
will work just involves more breadboarding to test the circuit. (and a extra
page to the logics to get them all to fit in;).
Later,
Wire
From: Lee Hart <[EMAIL PROTECTED]>
Reply-To: ev@listproc.sjsu.edu
To: ev@listproc.sjsu.edu
Subject: Re: the Mark Brueggemann Meter
Date: Tue, 06 Jun 2006 16:33:55 -0500
Peter Shabino wrote:
>
> PDF of the schematic are at
http://www.wire2wire.org/Brueggermann_meter.pdf.
It's a good start! The bias resistors to create the expanded scale
voltmeter, and 0-50mv shunt ranges are missing.
The 4049 has very low source current capability (a few ma). That's
probably too low to get sunlight-readable brightness. I still think a
transistor is the way to go.
If you must use an IC, there are octal high-side drivers. But they are
not very common, and so more expensive.
--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart_at_earthlink.net
--- End Message ---
--- Begin Message ---
Lee Hart wrote Friday, June 02, 2006 10:32 AM:
> Chuck Hursch wrote:
> > I will be re-making on Saturday some of the 2/0 cables for my
> > EV's pack. I have one post that has turned a 250-degF
Celsi-Dot
> > black, and one that has turned a 200-degF dot.
>
> How much current are you drawing? 2/0 should handle a *lot* of
current!
> For it to reach 200-250 deg.F something is seriously wrong.
On the hills around here with my 96V pack, I am pulling 300-350A
through the battery string. There's also a freeway merge up a
hill that will get me for that amperage too. To go from one town
to another, each in there own little valley, you have to go over
a 300-500 ft ridge. They say Marin County EV driving is more up
and down than across. If I don't need to be on the freeway to go
somewhere else when I come home from work in the late evening, I
take the long way around out towards San Quentin with less of a
climb, and if nobody is following me up my hill at night, I'll
keep the foot up and keep the amp draw down to ~250A and crawl
up at 13-15 mph.
>
> > Yesterday I took a stab at trying to figure out where the
heat is
> > coming from... A few months ago, I had taken a shot at 15A
with
> > the car's heater, but only pulled zeros to three places to
the
> > right of the decimal point (millivolts) when measuring
between
> > the post and the cable lug.
>
> A 1 mv drop at 15amps is 0.00066 ohms, which is good. At 200
amps, that
> only produces 2.6 watts.
2.6W should not heat up a battery post, I would think, if the
draw is only for 30s to a minute. The heat will radiate out from
the post into the air, into the cable, and into the battery.
>
> What do you get if you measure from battery post to battery
post? This
> will include *all* the connection resistances;
post-to-terminal,
> terminal-to-wire, the wire itself, wire-to-terminal, and
terminal to
> post at the other end.
I was able to get a reading at 75A across the whole cable, and I
think it was up into the second digit. Unfortunately, I did not
write the number down.
The number at this point is irrelevant except for learning
purposes - the 10" cable was turned into a 7" cable on Sat,
chopping off the problem end (the copper did not look
discolored). I will put that cable in place of one of the 7"
cables in the rear that's having problems. I'm also going to run
the 75A test again, and measure across all the cables. I did put
my 4' emergency cable in place of where the 10" cable was, and my
post is now cool (at least for a quick full-out (300-350A for
about 35-45s) up my hill - gets me about 17-18mph, depending on
peppy the batteries are). Been there and done that already.
We ran out of time to make more cables down on the Peninsula, so
I need to get my own crimp tool. The one down there is a
Quickcable tool (Hex crimp?), which I think goes for about $100.
If I have an EV and occasionally need to make up a new cable, and
running down to the Peninsula implies a 100-mi round-trip drive,
or trying to shoehorn it into a continuing trip, as happened on
Saturday, I need to have my own tool.
>
> > Anyways, 75A and 000s on the voltmeter display made me think
of
> > Bill Dube' posting where he described the use of a Wheatstone
> > bridge to figure out the resistance (I have a vague
recollection
> > of this posting).
>
> A Wheatstone bridge is a circuit for *comparing* two
voltages -- some
> unknown against a reference.
>
> What you need here is a more sensitive voltmeter. Even many
inexpensive
> ones have a 200mv or 400mv scale, which resolves to 0.1
millivolt. Use
> it to measure the voltage drops with as high a load current as
you can
> conveniently provide.
I can go to 1mV on the voltmeter I carry in the car. However, it
just occurred to me my Radio Shack voltmeter may have those
2/400mV scales, so I will have to check that out.
>
> Another useful circuit is a 4-wire micro-ohmmeter. Basically,
it's a
> power source and a resistor to provide a known current, and a
sensitive
> voltmeter that measures the voltage drop caused by that
current. The
> 4-wire part comes because you use it with special alligator
clips that
> have the metal rivet removed and replaced with a nonconducting
pin, and
> insulating washers so the two halves of the clip don't
electrically
> connect to each other. The power source goes to one half of
each clip,
> and the measuring meter to the other half of each clip. The
current in
> the measuring halves of the clips are now zero, so they don't
produce
> any voltage drop. Thus, the 4-wire technique eliminates the
resistance
> of the connecting leads as a source of error.
I think that is what Cor was talking about. Will have to find a
power supply, but one can probably be had on eBay for a
reasonable price.
Thanks for the tips, folks! I suspect this most-problematic-post
heating problem was due to high resistance in the cable crimp,
but why it has waited all these years to start showing is beyond
me. That 10" cable was made in 2001. Full of noalox, so
corrosion should not be a problem, with heat-shrink tubing with
melt-able liner. ?? A bit of thermal cycling that got worse and
worse?
Chuck
--- End Message ---
--- Begin Message ---
Thankyou Peter - just what I needed.
John
----- Original Message -----
From: "Peter Shabino" <[EMAIL PROTECTED]>
To: <ev@listproc.sjsu.edu>
Sent: Tuesday, June 06, 2006 4:46 AM
Subject: RE: Question about IGFETs
Quck and dirty instructions are at
http://www.galco.com/circuit/igbt_testing.htm
The only way to fully understand if there is damage is to curve trace the
part or test it under load. The above test will tell you if its functional
or DOA.
later,
Wire
From: "John Luck Home" <[EMAIL PROTECTED]>
Reply-To: ev@listproc.sjsu.edu
To: <ev@listproc.sjsu.edu>
Subject: Question about IGFETs
Date: Mon, 5 Jun 2006 19:44:06 +0100
I have an IRG4BC40U IGBT device in my field controller on my SEPEX motor
that has been subject to a large voltage/current spike when the armature
current collapsed due to a broken brush and connection.
Does anyone know how I can measure one of these devices with a multimeter
on Ohms to determine if it is serviceable - or should I just replace it.
It has an accompanying centre tapped freewheel diode there for protection
which seems to read "diode like" with an ohmmeter, but I cannot test the
IGBT apart from being able to say it is not short circuit.
I also have a couple of PCB tracks to mend where the collapsing armature
field must have induce an enormous voltage/current surge through my double
poled reversing relays.
Advice would be welcome.
John - with a dead EV, and dead broke !!
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.1.394 / Virus Database: 268.8.2/356 - Release Date: 05/06/2006
--- End Message ---
--- Begin Message ---
Joe Vitek wrote:
> You're right, it doesn't NEED a micro, but a micro would be
> cheaper with fewer parts as Zik pointed out. I am one who
> loves "old school" electronics and loves the '3914 solution.
> I have played with them since 1980 (high school). However, I
> also like the PIC solution.
The PIC solution may be cheaper if you already have the PIC development
tools (~$200 for PICSTART+), otherwise, the simple analog solution is
vastly cheaper.
Reliability of the PIC approach hardware may be similar to that of the
LM3914 approach, but the overall reliability of the PIC approach will
depend largely on the software reliability... and this depends greatly
upon the skill of the individual programmer.
The main disadvantage of the PIC approach is development time; even
though the software required is pretty simple, it will take longer to
develop, test and debug the software than to simply wire up a pair of
LM3914s.
I'm also an EE and do embedded software professionally, so it is fairly
unusual for me to argue against a microcontroller solution ;^>
In this case, I think even the LM3914 approach is going the wrong way...
The beauty of the original dual movement meter is its extreme
simplicity. Going to LEDs vastly complicates the project (you need to
monitor both voltage and current; voltage is easy, just a voltage
divider across the pack, but for current you need an amplifier stage to
boost the 50mV peak signal from a shunt to something usable by either
the PIC or LM3914). The LED approach also dramatically increases the
power consumption of the meter.
The advantage of using a micro for this project is, IMHO, that one could
program the relationship between the current and voltage signals into
the micro so that it can drive a much simplified display; perhaps
providing only a 3-colour output, or a simple bargraph with multiple
steps of each colour for better resolution of the pack SOC/SOH.
While it seems intuitively obvious to those of us who work with micros
that they are a great solution for this project, especially if one has
aspirations of mass producing the device for sale, one thing that has
been glossed over is that the relationship between voltage and current
and the colours on the display matrix/dual movement meter/etc. must be
determined by trial and error by each user for his particular battery
pack. This means that the curve must "programmed" by each user
populating the LED matrix with the appropriate colour LED at the
appropriate location for *his* particular battery, unless one used an
array of tri-colour LEDs (more complex hardware and software to drive)
and provided a means for the user to program the appropriate display
characteristic into the micro in his display (significantly increasing
the complexity of the micro's software).
Have fun with whichever approach you prefer to take with this display;
I'd be using a dual movement meter similar to the original guage if I
wanted a graphical display, and a micro if I wanted the ability to
transform the graphical display into something simpler for the
non-technical user to interpret.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
Large thanks for the info, everyone...
Now, where can I go to see an example of the wiring?
I'm intrigued by this as well as Jack Knopf's
"Negative 48's" direct-drive automatic tranny.
Borrowing from the old Confucius tale, I guess I
WAS conceived in the back seat of a car with an
automatic transmission...I grew up to be a shiftless bastard...
Many thanks for the leads...
Wayne White
At 12:58 AM 6/4/2006, you wrote:
Simply, it's two SPDT contactors wired together
so that the current flow through the motor field
can be reversed, thus reversing the direction.
Wayne wrote:
Hello,
While crusing around the EV Album, I ran across
Andreas Sagvåg's 1986 Peugeot 205 XL (144v
DC). What intrigued me was the comment under
"Drivetrain" wherein Andreas notes: "Standard
5-speed transmission without clutch. Locked at
5th gear. Reverse contactor installed."
I honestly haven't read or heard anything about
a reverse contactor (if I did its long forgotten).
This seems to me to be a pretty nice way to avoid some/all shifting.
Could someone direct me to some references with
regards to a reverse contactor, its
installation and proper uses, advantages or disadvantages?
Thanks in advance...
Wayne White
If those of us who "can", "do" then those of us
who "can't" won't suffer as much from the high prices of excess.
If those of us who "can", "do" then those of us
who "can't" won't suffer as much from the high prices of excess.
--- End Message ---
--- Begin Message ---
Are you crimping the zeners into the Waytek eyelet, or soldering it in?
Also, the pr2 lamp spec says "Ave Life : 15". If that is 15 hours, it seems
like they will burn out pretty quickly - any idea why they don't?
Chris/Lee - would either of you be willing to post a picture of one of your
regulators (or email to me directly...)?
Thanks,
Tim Wong
--- End Message ---
