EV Digest 5276
Topics covered in this issue include:
1) Re: Bizarre Amphour Meter Idea
by Lee Hart <[EMAIL PROTECTED]>
2) Re: quartz-halogen lamps to make a current source Re: Looking to build a
badder-boy charger
by Lee Hart <[EMAIL PROTECTED]>
3) Re: Values for transformer and other bad boy parts
by Neon John <[EMAIL PROTECTED]>
4) Re: Congratulations to Matt Graham and Lowell Simmons!
by Lowell Simmons <[EMAIL PROTECTED]>
5) Re: Good Boy charger (Re: Looking to build a badder-boy charger)
by Lee Hart <[EMAIL PROTECTED]>
6) RE: Good Boy charger (Re: Looking to build a badder-boy charger)
by "Roger Stockton" <[EMAIL PROTECTED]>
7) Re: The batts went flat...
by Lee Hart <[EMAIL PROTECTED]>
8) RE: The batts went flat...
by "Roger Stockton" <[EMAIL PROTECTED]>
9) RE: Good Boy charger (Re: Looking to build a badder-boy charger)
by Mike Phillips <[EMAIL PROTECTED]>
10) Re: Help with Specs
by "Jonathan Smith" <[EMAIL PROTECTED]>
11) Re: Good Boy charger (Re: Looking to build a badder-boy charger)
by Christopher Zach <[EMAIL PROTECTED]>
12) E-Meter phantom currents WAS Re: The batts went flat...
by "Michaela Merz" <[EMAIL PROTECTED]>
13) Re: 59 dollar Huffy 20" Chopper bike.
by Lee Hart <[EMAIL PROTECTED]>
14) Re: LED Headlights and the DOT
by Nick Viera <[EMAIL PROTECTED]>
15) Re: Good Boy charger (Re: Looking to build a badder-boy charger)
by Lee Hart <[EMAIL PROTECTED]>
16) Re: LED Headlights and the DOT
by Danny Miller <[EMAIL PROTECTED]>
17) Re: LED Headlights and the DOT
by "Andre' Blanchard" <[EMAIL PROTECTED]>
18) Re: Good Boy charger (Re: Looking to build a badder-boy charger)
by Lee Hart <[EMAIL PROTECTED]>
19) Re: LED Headlights and the DOT
by "Andre' Blanchard" <[EMAIL PROTECTED]>
20) RE: Good Boy charger (Re: Looking to build a badder-boy charger)
by "Roger Stockton" <[EMAIL PROTECTED]>
21) Re: Good Boy charger (Re: Looking to build a badder-boy charger)
by Mike Phillips <[EMAIL PROTECTED]>
--- Begin Message ---
Dale Curren wrote:
>
> ** Reply to message from Eric Poulsen <[EMAIL PROTECTED]> on Thu, 23 Mar 2006
> 09:14:03 -0800
>
> > Hmmm, looking at this more, it looks like this would work well.
>
> would you use the bicycle odometer? I was thinking the odometer might not
> accept the pulses as fast as the chip spits them out.
I originally suggested using a clock, which has a chip that divides a
32.768 KHz crystal down to 1 pulse per second. You would set up the
V-to-F converter chip to output a frequency near this as your maximum
counting rate.
A bicycle speedometer's maximum pulse-counting rate is considerably
lower; perhaps 10 per second. You could use a correspondingly bigger
capacitor on the V-to-F chip so its maximum frequency was 10 Hz, but
that requires a large-value precision capacitor; not cheap!
So, the better way to do it is to add a frequency divider chip, like a
CMOS 4040. Setup the V-to-F chip for a high frequency so the capacitor
is reasonable, and divide it down to 10 Hz or so for the bicycle
speedometer.
--
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 ---
Steven Ciciora wrote:
>
> Do quartz-halogen lamps have a positive temperature
> coefficient similar to filiment bulbs (making them
> emulate constant current sources)? I guess I could
> measure someday, but it sounds like you have done that
> already.
It's the tungsten filament that's responsible for the constant current
behavior. Quartz-halogen lamps still have a tungsten filament, so it
works for them, too.
The constant current effect is useful, but not very precise. You will
have something like a 2:1 change in current over a 10-1 change in
voltage.
--
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 Fri, 24 Mar 2006 07:24:56 -0800, "Tom Shjarback"
<[EMAIL PROTECTED]> wrote:
>Lee, forgive me. Your comment reminds me of elementary
>school. I am seeking information, not what you sent.
>Please refrain from reading and replying to any of my post
>in the future.
But Lee was correct. No one goes out to buy parts to make a bad boy
charger. A correct charger would be not much more expensive. You do
a bad boy charger using things laying around in your boneyard. Some
cords here, a variac there, an unknown and unlabeled bridge over
there. It helps to have a large boneyard (Min's a 7000 sq ft
warehouse :-)
The army would call it an improvised charger. Just like the POWs held
by the germans who made radios out of bits of wire and anything else
they could find.
If your question is, "how do I know what to look for at the boneyard?"
Then that's easy. Define the highest voltage your pack requires.
Determine what available AC voltage is available. Decide what current
you want to charge at. From there it's simple ohm's law stuff. If a
big ole' boat anchor transformer happens to be laying around like
there was in my shop and it's the proper voltage, then by all means
use it.
Whenever I go to a ham radio festival I go with a large shopping list
of stuff that I need for pending projects. My "round tuit" list. Very
little has anything to do with radios.
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 ---
We finally got the 944 down the track with a respectful time. I have been
building 96 and 156 volt cars in school for several years, but going 240 has
definitely been a new mindset. The 944 has a .975 solid driveshaft with 26
splines. We spun the driveshaft splines at Battery Beach Burnout. Since then
we have torn up one transaxle, split the driveshaft to tranxaxle spline coupler
(944's have a stationary driveshaft with one coupler in the front to the motor
and a second factory designed coupler in the rear) and exploded the chain
coupler between our two 8" XP motors. (These were done just around our school
and at my home.. My students can pull all the front batteries and motors out in
1 1/2 hours now. We machined a beefier front coupler, welded gussets into the
rear coupler and off to Moroso we went. I had only driven the car home from
school and back (30 mile roundtrip) one time before going to the track. You
might guess I was a little apprehensive about running a h!
ard pass.
We will be leaving April 5th for the EV Challenge in Raleigh NC with this
car. It wouldn't leave much time if we broke. I decided to attempt a soft
launch to minimize coupler distress. My main goal was to test for vibrations
and such. No burnout, Kuhmo 225 50 R16 radials, soft launch, (2.12- 60 ft )
Zilla shifts from series to parallel then it's time to keep the pedal to the
metal. The only problem came when the sunroof decided to come unlatched at
around 90 MPH and the night stars started to shine overhead. (got it back with
only minor scratches) I was very pleased with the 14.55 at 93 MPH. I made a
second pass but controller shut down because sli battery voltage too low.
Still was a 15.5 at 89. This would actually be my new record since time had to
be backed up since ont a sanctioned nedra event. I'm not too worried. There
will be more times to come. Thanks for all those who have helped me get to
this point. Not to bad for a 3200 lb car. Oh yeah Rich I will !
be
getting with you as soon as I can raise some more money. I'm getting tired of
breaking my pack down into two strings and charging one with my 156 Zivan and
the other with an old Lester.
Charles Whalen <[EMAIL PROTECTED]> wrote:
Congratulations to Matt Graham and Lowell Simmons for their outstanding
performances at Moroso Racetrack in West Palm Beach last night! Lowell set
a new world record of 14.55 seconds in Miramar High School's Porsche 944,
beating the previous record by almost 2 seconds! Matt Graham also made
history in his Nissan 240SX by being only the third (full-bodied) car ever
to break the 100mph barrier in the quarter mile! It was a great night for
EVs at the track! Racers and fans alike were (literally) blown away by
Matt's and Lowell's cars and their performances! We had lots of fun,
excitement, high fives and jubilation!
Way to go Lowell and Matt!
Charles Whalen
---------------------------------
Blab-away for as little as 1ยข/min. Make PC-to-Phone Calls using Yahoo!
Messenger with Voice.
--- End Message ---
--- Begin Message ---
Mike Phillips wrote:
> Hmmm. I charge my truck on 240vac when the pack is down to 290v
> loaded, 325v unloaded. Both way under 373v. In fact the pack peaks at
> 379/380 volts at end of charge. The charger's current output on my
> Emeter is less than zero as the charger ramps up.
Mike, these numbers don't add up. If your charger is nothing but a
bridge rectifier, then it can't produce more than 1.4 times the AC input
voltage. 240vac x 1.4 = 336vdc. And that is at zero charging current. If
the pack voltage is less, then you get charging current. How *much*
current is unpredictable; you are depending on the random resistance of
your wiring, connectors, etc.
Check your meter to see if it is honest. See what your actual AC line
voltage is, and the actual pack voltage.
>> This means that at 240V AC input (which can easily vary 10%) you
>> will need to make sure that your lowest battery voltage *ever* when
>> starting the charger is 240 x 110% x 1.41 = 373V DC
Only if you want zero charging current. As soon as you draw any current
at all, the DC voltage sags to something less than 1.4 times the AC
voltage.
>> If you discharge your battery pack below that voltage and plug the
>> charger in, you will get an uncontrolled large current through the
>> rectifier-inductor-output diode and blow something....
Normally, your house wiring, fuses, circuit breakers, connectors, AC
line cords, series inductor, etc. all total up to more than an ohm of
resistance. If it was one ohm, for example, then your charging current
is:
240vac x 1.4 = 336vdc at 0 amps
- 1a x 1ohm = 335vdc at 1 amp
-10a x 1ohm = 326vdc at 10 amps
-100a x 1ohm = 236vdc at 100 amps etc.
These are actually the peak charging current; the average will be lower
because it all flows at the peak of each AC line cycle.
This is one reason Bad Boys are such a bad idea. To get 10 amps of
charging current requires 50-100 amps of peak current thru your wiring,
breakers, connectors, etc. This is hard on them! Overheated wiring,
burned up connectors, and tripped breakers result.
--
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 ---
Mike Phillips [mailto:[EMAIL PROTECTED] wrote:
> So that means every USE truck has that problem. So by design
> every USE truck main board will die an ugly death, guaranteed.
I'm not disputing Chris' observations regarding Dolphin failures, but
question the purported mechanism. It seems to me that Chris was
responding to an explanation of why a PFC pre-regulator circuit on its
own makes a poor charger and has the inherent limitation of not being
able to charge in a controlled manner into too low of a pack voltage, or
indeed of not being able to survive the attempt.
However, the reason for this limitation is that PFC pre-regulator
circuits are *boost* converters, and so can only regulate to an output
voltage that is greater than the peak of the input AC voltage. It would
surprise me greatly to hear that the Dolphin charger is a pure boost
convertor topology, for a number of reasons, including that the output
of a PFC pre-regulator boost convertor has "lots" of 120Hz ripple,
something sealed batteries do not want.
If the Dolphin chargers tend to fail when attempting to charge into
unusually low battery pack voltages, the more likely cause is that the
designers did not anticipate it being connected to a seriously
over-discharged pack. A better solution to this problem might be
putting a relay/contactor in that will prevent the charger from starting
unless the pack voltage is above some minimum level. A 325V
open-circuit voltage is 12.5V/battery, which is well within the normal
operating range for lead-acid. I find it very had to believe that the
Dolphin designers would not have anticipated charging battery packs with
an open-circuit voltage of 12.0V/module (312V) or somewhat lower, but
they certainly might not have considered having to charge a battery that
has somehow been allowed to discharge below, say, 200V.
Finally, if someone wanted to use a PFC pre-regulator as a charger, one
possibility for using it with arbitrary pack voltages would be to feed
it from a variac so that the peak of the AC input voltage can be varied
to be less than the pack voltage.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
damon henry wrote:
> sitting in the garage with the main circuit open... The only energy
> flow is the few milliamps from the 12v battery to the emeter, and
> that battery is dedicated to the emeter.
It might be that you have wired it so the E-meter's supply current is
flowing thru the shunt. This causes a slow but steady discharge rate.
There should be NOTHING connected to the battery negative except the
wire to the shunt. This way, ANY current in/out of the battery is
tracked by the E-meter.
If the E-meter is powered from the propulsion battery, either directly
or with a DC/DC, then you really are discharging the battery, and the
E-meter is correct in showing it.
If the E-meter is powered by a separate 12v battery, then this battery
should connect to the E-meter's + and - supply terminals. The negative
does NOT go to the propulsion battery's negative.
Also, remember that the E-meter shorts the grounds from the propulsion
pack, its own power supply, and the RS-232 port all together. Besides
the safety hazard, this can cause ground fault paths, and this current
will show up on the E-meter.
For example, your propulsion pack is "floating" (no connection to
ground). You have a resistive path from pack+ to ground from a dirty
battery top; but no current flows because this is the only path to
ground (the rest are clean).
Now you add an E-meter. You power it from the 12v accessory battery,
whose ground is tied to the chassis. Now you have added a SECOND path,
from pack- to ground via the E-meter. Now a leakage current flows (and
the E-meter will dutifully accumulate it). This current flows from pack+
to ground, to 12v battery-, to E-meter-, to shunt, and thus to pack-.
--
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 ---
damon henry [mailto:[EMAIL PROTECTED] wrote:
> I'll give this a shot, since the wire I used is simply self
> twisted and I have lots of ethernet lying around, but where
> is the noise coming from? I can understand when the
> controller or charger are active, but this happens with no
> charger plugged in, the DC 2 DC powered down, and the bike
> just sitting in the garage with the main circuit open.
Give it a try if you like, but it isn't RF noise that is causing your
problem, and I wouldn't use ethernet (CAT5, etc.) cable in an EV because
it is [usually] solid rather than stranded wire and will fail from
vibration. A pair of wires twisted using a hand drill until they have a
few turns per inch is all you need to reject noise.
Assuming you are using the standard 500A/50mV shunt with your E-Meter,
then 0.4A corresponds to a voltage of 40 micro volts being sensed by the
E-Meter from the shunt. 0.4A is kind of high in my experience for
E-Meter error reading, but is perhaps not unbelievable. I have seen
current errors result from loose connections: check the
tightness/integrity of the connections at the shunt and at the screw
terminals on the back of the E-Meter. I have also seen current sense
errors result from the shunt sense connections making even slight
contact with the high current connections to the shunt, though this
usually results in error while current is flowing rather than error
while idle.
Another possibility is a wiring error, although I believe you stated you
have reviewed the wiring to eliminate that possibility. Basically, if
the wiring results in the E-Meter drawing any current through either of
the shunt sense wires, you will have a current offset error. This sort
of wiring error usually involves having used one of the current sense
leads (e.g. from the load side of the shunt) to also provide the
negative side of the pack voltage sense connection, or even worse, for
the negative supply connection to the E-Meter. The shunt sense wires
must only connect to the sense terminals on the shunt, and the current
sense terminals on the E-Meter. The only connections to the battery
side of the shunt should be the high current cable to the battery and
the sense lead to the E-Meter. All other pack -ve connections (other
than the other current sense lead) should be made at the high current
load terminal of the shunt.
Good luck,
Roger.
--- End Message ---
--- Begin Message ---
Lee,
I've cross checked all instrument readings. The charger as far as I
know is a boost topology only. So it can produce up to at least 395vdc
in my experience.
Roger,
The Dolphin software has a minimum on AC voltage variable in software.
It's threshold is 90vdc minimum at the pack for the charger to startup.
So it's clear that the designers thought of these things. It's just
funny that the charger is basically boost only. Maybe controlling the
fets makes via the uP is what makes it work when the pack is lower than
the charger output. The schematic that one member generated of the
charger shows it to be a boost only with filtering on the input.
I think that a pfc supply is a good foundation for the power side of a
charger. I'm starting there for my own education.
Rich,
I like the idea of opto isolation. I'd be really surprised if it didn't
exist already. Optos can't drive the fets directly in this case, but
could be used to control a driver circuit. I will be getting my Dolphin
repair fixture rolling and will get my head wrapped around this board
better. I complain to OEM's all the time about their cheezy driver
circuitry, because I repair their hardware.
Mike
--- Roger Stockton <[EMAIL PROTECTED]> wrote:
> Mike Phillips [mailto:[EMAIL PROTECTED] wrote:
>
> > So that means every USE truck has that problem. So by design
> > every USE truck main board will die an ugly death, guaranteed.
>
> I'm not disputing Chris' observations regarding Dolphin failures, but
> question the purported mechanism. It seems to me that Chris was
> responding to an explanation of why a PFC pre-regulator circuit on
> its
> own makes a poor charger and has the inherent limitation of not being
> able to charge in a controlled manner into too low of a pack voltage,
> or
> indeed of not being able to survive the attempt.
>
> However, the reason for this limitation is that PFC pre-regulator
> circuits are *boost* converters, and so can only regulate to an
> output
> voltage that is greater than the peak of the input AC voltage. It
> would
> surprise me greatly to hear that the Dolphin charger is a pure boost
> convertor topology, for a number of reasons, including that the
> output
> of a PFC pre-regulator boost convertor has "lots" of 120Hz ripple,
> something sealed batteries do not want.
>
> If the Dolphin chargers tend to fail when attempting to charge into
> unusually low battery pack voltages, the more likely cause is that
> the
> designers did not anticipate it being connected to a seriously
> over-discharged pack. A better solution to this problem might be
> putting a relay/contactor in that will prevent the charger from
> starting
> unless the pack voltage is above some minimum level. A 325V
> open-circuit voltage is 12.5V/battery, which is well within the
> normal
> operating range for lead-acid. I find it very had to believe that
> the
> Dolphin designers would not have anticipated charging battery packs
> with
> an open-circuit voltage of 12.0V/module (312V) or somewhat lower, but
> they certainly might not have considered having to charge a battery
> that
> has somehow been allowed to discharge below, say, 200V.
>
> Finally, if someone wanted to use a PFC pre-regulator as a charger,
> one
> possibility for using it with arbitrary pack voltages would be to
> feed
> it from a variac so that the peak of the AC input voltage can be
> varied
> to be less than the pack voltage.
>
> Cheers,
>
> Roger.
>
>
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 ---
>
> Victor is using supper caps in a way you are talking , but his batteries
> have a wide voltage swing and can't put out the BIG accelerating amps he
> sometime wants , I could see our idea working on batteries like these ,
> kind of , but not lead .
Thanks for the favor of a reply. No, I wish to avoid spending an outrageous
amount of money. The cost of a new set of Trojans was plenty.
What prompted the inquiry was observing the announcement and discussion on
this list of the e3 Super Cell -- a battery + super cap combo.
Acceleration is sufficient, I was hoping to extend range and life of
batteries.
--- End Message ---
--- Begin Message ---
So that means every USE truck has that problem. So by design
every USE truck main board will die an ugly death, guaranteed.
Only if you plug it into 240 volts.
The problem with the Dolphin is kinda related to their solution. The
engineers saw this and attempted to fix it by having a pair of big FETs
on the rectified side of the DC bridge which were controlled by the
computer. If the computer saw that the line voltage was above the pack
voltage, it would turn off the FET till the sine wave on the line
voltage dropped below pack voltage. In that way, the pack would
slooooowly charge up at first (only catching power from the bottom of
the sine wave), then ramp up as the pack voltage came up (and more and
more of the wave was converted to energy).
The problem however is that these on-off's are not synced to the charge
pump rate of the Dolphin's IGBT/motor step-up stunt. So the FETs wind up
carrying a lot of load on those switch transition times. Eventually they
fail, and then the fun begins as they short closed, high voltage DC is
now on the gate, and it merrily rips through the telemetry circuit
looking for frame ground. Telemetry has the +15 tied to frame ground
because it's a ref point for the ground fault detection circuits. It's
isolated from the pack due to the +300-15v isolation xfrmr, but not from
the AC line.
USE figured this out after a year, and started outfitting trucks and
cars with a 240-208 transformer. Lower peaks meant the circuit would
never engage, life is better. Downside is this cuts performance on 120
volt charges (now down to 90 or so) and adds weight. The other solution
is to just use a Magnecharger which bypasses the whole mess.
I'm not disputing Chris' observations regarding Dolphin failures, but
question the purported mechanism. It seems to me that Chris was
responding to an explanation of why a PFC pre-regulator circuit on its
own makes a poor charger and has the inherent limitation of not being
able to charge in a controlled manner into too low of a pack voltage, or
indeed of not being able to survive the attempt.
However, the reason for this limitation is that PFC pre-regulator
circuits are *boost* converters, and so can only regulate to an output
voltage that is greater than the peak of the input AC voltage. It would
surprise me greatly to hear that the Dolphin charger is a pure boost
convertor topology, for a number of reasons, including that the output
of a PFC pre-regulator boost convertor has "lots" of 120Hz ripple,
something sealed batteries do not want.
I believe that it is. They have a ton of filters on the input side to
prevent this IIRC. They also have pretty good power factor if I recall
correctly; I'll check again tonight with my meter.
If the Dolphin chargers tend to fail when attempting to charge into
unusually low battery pack voltages, the more likely cause is that the
designers did not anticipate it being connected to a seriously
over-discharged pack. A better solution to this problem might be
putting a relay/contactor in that will prevent the charger from starting
unless the pack voltage is above some minimum level. A 325V
open-circuit voltage is 12.5V/battery, which is well within the normal
operating range for lead-acid. I find it very had to believe that the
Dolphin designers would not have anticipated charging battery packs with
an open-circuit voltage of 12.0V/module (312V) or somewhat lower, but
they certainly might not have considered having to charge a battery that
has somehow been allowed to discharge below, say, 200V.
Right. They did anticipate this, but with a FET instead of a relay. But
you need a really big FET to take these sorts of pulses. A standard 30a
one will not cut the mustard.
Chris
--- End Message ---
--- Begin Message ---
> Mark wrote:
>
> The most common cause is noise, make sure you are using good quality
> twisted pair wire from the shunt to the emeter, and route it away from
> other lines if possible. I like to use the pairs from a stripped piece
> of Ethernet patch cord (stranded) since its fairly affordable and has a
> high twist per inch ratio.
Well, my emeter is the other way around. It shows 0.5 discharge with
everything out and disconnected. Even with the shunt shorted out, it still
shows the same phantom discharge. There shouldn't be any noise as the
system
is down and disconnected.
Michaela
--- End Message ---
--- Begin Message ---
Lee Hart wrote:
>> It's no doubt heavy, with high rolling resistance, and harder to ride
>> than a normal bike. Probably made in China, too. I would think it
>> would be a *bad* choice to make into an Ebike.
Mark Grasser wrote:
> You make a lot of assumption here. Rolling resistance is "tires".
Agreed. These bikes have huge balloon tires, not high-pressure skinny
tires. That usually means high rolling resistance.
> Hard to ride, so are those ricer bikes and they sell more of those then
> Harleys by far.
What is a ricer bike?
Bikes come in all sorts of configurations. They range from very good
(easy to ride, handles well, great brakes, etc.) to very bad (all style,
hard to ride, uncomfortable, poor brakes, etc.)
But people (especially kids) are incredible adaptable. They can learn to
ride almost anything. So, it has become common to build really terrible
bikes, and sell them on the basis of looks rather than performance. I
think the assumption is that the customer can't tell the difference
between good and bad.
> Made in China, now that is one thing I am very familiar with...
> we are about to be edged out. They get much better at what they do
> and they will be better then us
It's a complex problem. Many Americans seem to think that nothing except
price matters. It doesn't matter how well it works; it doesn't matter
how long it lasts; it doesn't matter who or what was harmed making it;
it just has to be CHEAP!
The Chinese are ruthless pragmatists. You want cheap? We give you cheap!
They are certainly capable of building higher quality products; but all
the stupid rich Americans want is cheap!
With regards to Chinese bikes. My BEST projects (see
www.bestoutreach.org) use a lot of old bikes. The kids take them apart,
fix them, and use the parts to build their electric vehicles. We find
that the older the bike, the easier it is to use. The bolts are
standard, the parts are generic, everything comes apart, and it's easy
to fix. But most of what we get are newer Chinese-made bikes. They only
lasted a short time before something broke, and then they were
discarded. They are nearly impossible to fix, and it's difficult to use
their parts for anything else.
Such a bike is a poor foundation for building a quality E-bike. Instead
of a cheap, low-quality, poor handling bike; you'll have an EXPENSIVE
low quality, poor handling bike that happens to have electric assist.
Does the styling make it worth it?
--
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 ---
Hi,
Peter VanDerWal wrote:
I was, however, refering to conversions. For a conversion, LEDs offer
little, if any, advantages.
The actual savings in power is miniscule compared to the power required to
move a full size vehicle so you won't see any increase in range.
Although I disagree; I do understand that this argument is pretty
subjective. However, I just want to point out two facts that I think
have been previously missed in this discussion.
First, the rapid response time of LEDs compared to incandescent lamps
hasn't been mentioned. IMHO, this alone is a good pro for LEDs, because
it makes them more noticeable and attention-grabbing (especially in
Brake light applications).
Second, while the power-savings of using LEDs is insignificant compared
to the power running through the HV system, and thus range, it can be
significant with respect to the 12-Volt system.
Especially if your EV is like mine where you've got an MR2 P/S pump and
lots of other 12-Volt loads (averaging around 1000-Watts/84-Amps peak);
but due to the unavailability of affordable, high-power DC/DC converters
for EV use, have to use a wimpy ~300-Watt/30-Amp DC/DC converter to
attempt to keep the 12-Volt system from heavily sagging or dying. In
this case, LEDs can make a big difference towards keeping the DC/DC
converter from constantly being overloaded, and thus keeping the 12-Volt
system happy.
My 0.333 kWh,
--
-Nick
1988 Jeep Cherokee 4x4 EV
http://go.DriveEV.com/
http://www.ACEAA.org/
--------------------------
--- End Message ---
--- Begin Message ---
Rich Rudman wrote:
> Folks the solution is a Opto coupled gate drive circuit. You kill
> the gate driver and not the rest of the powerstage, and logic.
It depends, Rich. Many of the cheap optocouplers just use clear plastic
between the LED and phototransistor. If one side's device overheats, it
melts a hole in the barrier and SHORTS. I've seen this many times.
If your circuit is such that it is possible to burn up the
phototransistor, then you need a "glass isolated" coupler; they use a
little glass panel between LED and phototranstor. Or, look for "high
isolation" types that have the LED and phototransistor on opposite ends
of the leaframe, and use a light pipe to channel the light between them.
> Fuses won't be fast enough. Fuses at this speed put out the fire...
> and that's about it.
They can be fast enough. We tried them in a motor controller that had
the shorted opto problem. They worked; but it was cheaper to replace the
opto with a more expensive part with better fail-safe isolation.
Later, we found that we could put resistors in both legs of the opto, so
if/when it did short, the resistors would act as fuses. This was the
cheapest solution.
Another good idea is to use gate transformers. It's not hard to design a
transformer so it is virtually impossible for a primary/secondary short.
Transformers are also great because they can transfer power, not just
signal, so you don't need isolated power supplies.
> The better idea is to find out what is deficient in the gate drive
> and the physics of the Fet/IGBT that cause them to fail. Fix it.
> Band-aids only work so well.
Agreed. Though accidents happen. Band-aids are useful when they do.
I like belts-and-suspenders circuits. This part shouldn't fail -- but
maybe I missed something; something I didn't think of might happen. So I
include a backup system to limit the damage. This means a circuit might
have 10 parts to actually do the job, and 10 more there "just in case".
It costs more, but is a lot more reliable!
--
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 ---
As a fairly off topic observation, is anybody else as annoyed as I am
when some sort of traffic/vehicle application PWMs their lights?
While the eye doesn't respond directly to the pulsing at the freq they
use, when the object is moving quickly in relation to your field of view
it appears as a sequence of flashes. I don't know if this is done for
some ignorant electrical reason or if it was supposed to be more
noticible. I find it annoying and visually confusing when trying to
drive at night.
It's not like it's going to make me run into something but it's just
annoying IMHO and does not improve safety. Anybody else ever peeve over it?
Danny
Nick Viera wrote:
Hi,
Peter VanDerWal wrote:
I was, however, refering to conversions. For a conversion, LEDs offer
little, if any, advantages.
The actual savings in power is miniscule compared to the power
required to
move a full size vehicle so you won't see any increase in range.
Although I disagree; I do understand that this argument is pretty
subjective. However, I just want to point out two facts that I think
have been previously missed in this discussion.
First, the rapid response time of LEDs compared to incandescent lamps
hasn't been mentioned. IMHO, this alone is a good pro for LEDs,
because it makes them more noticeable and attention-grabbing
(especially in Brake light applications).
Second, while the power-savings of using LEDs is insignificant
compared to the power running through the HV system, and thus range,
it can be significant with respect to the 12-Volt system.
Especially if your EV is like mine where you've got an MR2 P/S pump
and lots of other 12-Volt loads (averaging around 1000-Watts/84-Amps
peak); but due to the unavailability of affordable, high-power DC/DC
converters for EV use, have to use a wimpy ~300-Watt/30-Amp DC/DC
converter to attempt to keep the 12-Volt system from heavily sagging
or dying. In this case, LEDs can make a big difference towards keeping
the DC/DC converter from constantly being overloaded, and thus keeping
the 12-Volt system happy.
My 0.333 kWh,
--- End Message ---
--- Begin Message ---
At 01:46 PM 3/24/2006, you wrote:
<< sinp >>
First, the rapid response time of LEDs compared to incandescent lamps
hasn't been mentioned. IMHO, this alone is a good pro for LEDs, because it
makes them more noticeable and attention-grabbing (especially in Brake
light applications).
<< snip >>
--
-Nick
1988 Jeep Cherokee 4x4 EV
http://go.DriveEV.com/
http://www.ACEAA.org/
--------------------------
Personally I think this kind of stuff is just an arms race. Like the third
brake light after a while they are on every vehicle and drivers just
compensate by paying even less attention to driving.
__________
Andre' B. Clear Lake, Wi.
--- End Message ---
--- Begin Message ---
Mike Phillips wrote:
> The [Dolphin charger] schematic that one member generated of the
> charger shows it to be a boost only with filtering on the input.
Interesting; they sold a NON-isolated charger to the general public?
> I think that a pfc supply is a good foundation for the power side of a
> charger. I'm starting there for my own education.
There are some problem, as these failures indicate. One is that the pack
DC voltage has to be more than 1.4 times the AC input voltage, or the
charging current is uncontrolled. Another is that the output current is
inherently pulsating, so you need large expensive filter capacitors to
manage ripple and noise. And, due to the large capacitors, the inrush
current when you first connect it to the AC supply can be very large.
I designed a 12kw PFC charger. It dealt with the inrush problem by
placing MOSFETs in series with the large filter capacitors. These
MOSFETs were used as current-limiting resistors during startup to manage
inrush, then turned fully-on for normal operation. This was
straightforward and required no high-side driver. But this charger had
an output transformer (a full bridge 50 KHz switcher), so I didn't have
to manage the inrush caused by connecting batteries directly to the PFC
output.
I designed another 10kw charger that was a pure boost converter, with
the batteries directly connected to the PFC output. In this case, the
input bridge rectifier was actually made up with two diodes and two
SCRs. The SCRs were phase controlled to limit the inrush and charging
current when the battery pack was too low. It could charge even a
totally dead (0v) pack at low current. The power factor during precharge
and this low-current "recovery" mode was not as high, but it doesn't
matter because it only lasts a short time and is at low current.
> I complain to OEM's all the time about their cheezy driver circuitry,
> because I repair their hardware.
Good! One only hopes that the comments actually reach the ears of the
engineers that designed the things. Sadly, my experience is that most
companies isolate the engineers as thoroughly as possible from any real
customers.
--
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 ---
At 01:59 PM 3/24/2006, you wrote:
As a fairly off topic observation, is anybody else as annoyed as I am when
some sort of traffic/vehicle application PWMs their lights?
While the eye doesn't respond directly to the pulsing at the freq they
use, when the object is moving quickly in relation to your field of view
it appears as a sequence of flashes. I don't know if this is done for
some ignorant electrical reason or if it was supposed to be more
noticible. I find it annoying and visually confusing when trying to drive
at night.
It's not like it's going to make me run into something but it's just
annoying IMHO and does not improve safety. Anybody else ever peeve over it?
Danny
Nick Viera wrote:
You could put some of these on your car and send a message back.:)
http://eagle-systems.net/lm1.htm
__________
Andre' B. Clear Lake, Wi.
--- End Message ---
--- Begin Message ---
Mike,
> The Dolphin software has a minimum on AC voltage variable in
> software. It's threshold is 90vdc minimum at the pack for the
> charger to startup. So it's clear that the designers thought
> of these things.
This sounds a bit confused. You mention a minimum on "AC" voltage
parameter, but then state that its threshold is 90vdc minimum at the
pack, which sounds like a minimum battery voltage value.
I can understand the charger having both parameters, but not using one
for both purposes.
> It's just funny that the charger is basically boost only.
Yes, this is odd, considering that the designers apparently knew it
would have to handle charging battery packs whose voltage is less than
the peak line voltage.
> I think that a pfc supply is a good foundation for the power
> side of a charger. I'm starting there for my own education.
I agree. A straightforward approach is a PFC pre-regulator to provide a
non-isolated ~400VDC bus and then an isolated buck convertor to provide
current and voltage control for whatever the pack voltage happens to be.
Chris,
> The problem with the Dolphin is kinda related to their solution.
An interesting band-aid; I understand the problem better now.
> > including that the output of a PFC pre-regulator boost
> > convertor has "lots" of 120Hz ripple, something sealed
> > batteries do not want.
>
> I believe that it is. They have a ton of filters on the
> input side to prevent this IIRC. They also have pretty
> good power factor if I recall correctly; I'll check again
> tonight with my meter.
Filters on the input side will do nothing for this. The problem is due
to the fact that the PFC *is* good; good PFC means that the input
current is sinusoidal at 60Hz (and is in phase with the input line
voltage). This is achieved by the boost convertor *forcing* its output
current to be a rectified 60Hz sine wave, which results in 120Hz ripple
on the 400V bus. If the output of the boost convertor feeds the
batteries directly then there will be 120Hz ripple seen by the
batteries. Placing another conversion stage, such as a buck convertor,
between the 400V bus and the batteries allows the buck convertor to
reject a large amount of the 120Hz ripple and provide the batteries with
near pure DC.
> > A better solution to this problem might be putting a
> > relay/contactor in that will prevent the charger from
> > starting unless the pack voltage is above some minimum
> > level.
> Right. They did anticipate this, but with a FET instead
> of a relay. But you need a really big FET to take these
> sorts of pulses. A standard 30a one will not cut the
> mustard.
There is a significant difference between what they did and what I am
suggesting. My suggestion is to have a relay/contactor powered by the
battery such that if the battery voltage is below some level, the
Dolphin is not allowed to charge at all. Their solution was to try to
allow the Dolphin to operate in a reduced manner if the battery was
below some level, and their approach results in the charger experiencing
stresses it cannot survive.
Given your description of their eventual "solution" of a 240<->208
transformer, I would revise my suggestion to using a relay whose coil is
powered from the battery pack to switch a 240<->208 transformer into the
input side of the Dolphin when pack voltage is below some level and line
voltage is above some level. This way the input voltage is not reduced
except when necessary to protect the charger from damage. The 240<->208
transformers are autotransformers and are relatively compact (a 7.5kVA
unit comes in a metal box less than 6"x6"x10" and weighs just a few
pounds) and cheap, so this might not be too ugly a kludge.
Cheers,
Roger.
--- End Message ---
--- Begin Message ---
I don't know if the charger is isolated or not. I think Chris or Andy
know the answer. Mine is because the truck was done so late in the
design cycle that it got a 240 to 208 step down transformer to
supposedly save the fets from smoking. Consequently my charger may
never blow up. The charge rate is lower however. But my truck is rare
having this transformer. No other truck I've seen has it, and I know
the cars don't have it either.
Can a buck-boost system also be power factor corrected?
I love a learning curve, but if you have a 10kw charger that works that
well, have you ever considered selling it or the docs/software?
As for the OEM's, I know them personally, so I hope they listen. They
are motivated because instead of a customer buying a new unit when
theirs smokes, the customer sends it to me for repair. So the OEM makes
no money. In fact, it's gotten to the point where the customers ask if
the unit is rebuildable before they will buy it! Yet another place to
cost the OEM money! Some units are not rebuildable because the parts
are $5 each in quantities of 1000 only.
My goal is to give my truck a better charger. I'd love to have a
Magnecharger, but it's alot of added hardware and money, that my truck
didn't come with. So making an onboard charger that can use the liquid
cooling system and sit inside the nice aluminum case is a major plus to
me. I think there is enough volume to install a new design. It would be
even nicer if the pwm from the control hardware could be used to drive
the new design. Basically a drop in replacement charger.
Mike
--- Lee Hart <[EMAIL PROTECTED]> wrote:
> Mike Phillips wrote:
> > The [Dolphin charger] schematic that one member generated of the
> > charger shows it to be a boost only with filtering on the input.
>
> Interesting; they sold a NON-isolated charger to the general public?
>
> > I think that a pfc supply is a good foundation for the power side
> of a
> > charger. I'm starting there for my own education.
>
> There are some problem, as these failures indicate. One is that the
> pack
> DC voltage has to be more than 1.4 times the AC input voltage, or the
> charging current is uncontrolled. Another is that the output current
> is
> inherently pulsating, so you need large expensive filter capacitors
> to
> manage ripple and noise. And, due to the large capacitors, the inrush
> current when you first connect it to the AC supply can be very large.
>
> I designed a 12kw PFC charger. It dealt with the inrush problem by
> placing MOSFETs in series with the large filter capacitors. These
> MOSFETs were used as current-limiting resistors during startup to
> manage
> inrush, then turned fully-on for normal operation. This was
> straightforward and required no high-side driver. But this charger
> had
> an output transformer (a full bridge 50 KHz switcher), so I didn't
> have
> to manage the inrush caused by connecting batteries directly to the
> PFC
> output.
>
> I designed another 10kw charger that was a pure boost converter, with
> the batteries directly connected to the PFC output. In this case, the
> input bridge rectifier was actually made up with two diodes and two
> SCRs. The SCRs were phase controlled to limit the inrush and charging
> current when the battery pack was too low. It could charge even a
> totally dead (0v) pack at low current. The power factor during
> precharge
> and this low-current "recovery" mode was not as high, but it doesn't
> matter because it only lasts a short time and is at low current.
>
> > I complain to OEM's all the time about their cheezy driver
> circuitry,
> > because I repair their hardware.
>
> Good! One only hopes that the comments actually reach the ears of the
> engineers that designed the things. Sadly, my experience is that most
> companies isolate the engineers as thoroughly as possible from any
> real
> customers.
> --
> 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
>
>
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 ---
