DIY Electric Car Forums banner

1 - 20 of 22 Posts

·
Registered
Joined
·
70 Posts
Discussion Starter #1
I=92d like to modify my Link-10 E-meter to put out data through a serial
port. Apparently the version I own (which is all that can be had now) wh=
ich
does not have a serial port can be modified to have a serial port by adding
back the components that were not populated when it was built.

To do this, I need to know what the missing components were. Lee Hart
helpfully explained to me that he did it and it was fairly easy to do when
one had a Link-10 with a serial port for comparison. It was a matter of
adding one IC and some resistors.

My problem is that I don=92t have a Link-10 with a serial port for comparis=
on
and I have no means of getting or borrowing one. (Actually, I might. =
I=92ll
look into borrowing.)

In the mean time, is there any chance that any of you happen to have
detailed pictures of the inside of a Link-10 E-meter that has a serial
port? Ones that are detailed enough that I could read the part number on
the IC and the values of the resistors? (I almost asked Lee specifically,
but I=92d rather he spent the time on his Sunrise EV2 project.)

If somebody has them, I promise that I=92ll put details of my efforts on a =
web
site for future people going this direction.

Thanks,
Jake Oshins


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #2
Yes, I do have pictures of E-meter guts. Both with
and without the serial port option. I don't have
access to the pictures right now. If you email me a
picture of what you have (it will be interesting to me
to compare against what I have) I can tell you the
differences between yours and mine. I've performed
the upgrade several times and it has always worked. =

Some emeters come with the SN75155 chip and two caps
(no missing resistors involved) in which case, all you
have to do is add a 9 pin connector, others don't come
with those parts.

All my comments assume you have a surface mount
e-meter. I've only seen one e-meter with leaded
components, and have not studied it to determine the
difference. For what it's worth, this version had 3
PCBs total (vs the more common surface mount one with
only two PCBs).

As I said above, the missing chip is an SN75155. It's
an 8 pin surface mount chip, located on the same side
as the EPROM (the "BOTTOM"). It's the chip closest to
the 9-pin connector, and pin 1 is located farthest
away from the 9-pin (closest to the display). There
are no resistors involved; just two caps. One is
simply a 0.1uF decoupling cap, located to the left of
the chip. You can see a trace between the bottom pad
and pin 4 of the IC. This is C29. The other
potentially missing cap is C8, wich is 470 pF. It is
horizontally mounted (right angle to C29), located to
the right of C14, and to the left of R4, just below
the HC573 address demultiplexer chip. C14 is the 22uF
50V surface mount cap just to the right of the chip.

Pin 2 of the 9 pin goes to pin 5 of the 75155, and is
the RS-232 data input to the CPU. The TTL version of
this signal is on pin 3 of the 75155, and goes to pin
24 of the CPU (the RxD pin). Pin 3 of the 9 pin goes
to pin 7 of the 75155, which is the RS-232 data from
the CPU, to your e-meter opto-isolator. The TTL
version of this signal comes from pin 25 (RxD) of the
CPU, and goes to pin 2 of the 75155. It is
interesting that the hardware is set up to allow
bidirectional communication. I have not disassembled
the EPROM to see what commands it might accept.

I don't have all my notes in front of me, but several
other pins of the 9 pin connector do other,
non-standard functions. In other words, might not be
a good idea to connect all 9 pins directly to a serial
port; I would just connect pins 2, 3, and 5.

If for some reason, your firmware doesn't have the
RS-232 option enabled (I have not seen this yet), I
might be able to program you an eprom with this
option. Actually, I know I can, but I'm not too sure
how legal it might be.

This is from memory, but I seem to remember one
version of Lee's e-meter opto-isolator having a typeo
in the pinouts. If you study the data sheet it should
be obvious. I made up a few PCBs a while back
implementing this, along with options to connect it
directly to an RS-232 serial port (like Lee's
schematic), or to a TTL port of a microprocessor
(logic inverted). I think I still have one unused
one somewhere, if you are interested.

Do you have these parts on hand? If not, I'm sure I
can send you some. Otherwise it might be hard to make
the minimum order for most places. Or I can just do
the upgrade for you.

I have also made what I call an "E-Meter sucker". It
records all the data from an e-meter to a secure
digital memory card. My one prototype is pretty ugly,
and bigger than it needs to be, but since it's doing
it's job for me, I have not finished a more compact
version. The reason why I'm mentioning this is
two-fold; one is to brag a little :), the other is,
if there is much interest, maybe it might encourage me
to finish up designing the smaller version.

- Steven Ciciora

--- Jake Oshins <[email protected]> wrote:

> I=92d like to modify my Link-10 E-meter to put out
> data through a serial
> port. Apparently the version I own (which is all
> that can be had now) which
> does not have a serial port can be modified to have
> a serial port by adding
> back the components that were not populated when it
> was built.
> =

> To do this, I need to know what the missing
> components were. Lee Hart
> helpfully explained to me that he did it and it was
> fairly easy to do when
> one had a Link-10 with a serial port for
> comparison. It was a matter of
> adding one IC and some resistors.
> =

> My problem is that I don=92t have a Link-10 with a
> serial port for comparison
> and I have no means of getting or borrowing one.
> (Actually, I might. I=92ll
> look into borrowing.)
> =

> In the mean time, is there any chance that any of
> you happen to have
> detailed pictures of the inside of a Link-10 E-meter
> that has a serial
> port? Ones that are detailed enough that I could
> read the part number on
> the IC and the values of the resistors? (I almost
> asked Lee specifically,
> but I=92d rather he spent the time on his Sunrise EV2
> project.)
> =

> If somebody has them, I promise that I=92ll put
> details of my efforts on a web
> site for future people going this direction.
> =

> Thanks,
> Jake Oshins
> =

> =

> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
> =



__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around =

http://mail.yahoo.com =


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #3
Thanks for this detail. It's tremendously helpful.

>From this, I can tell that both of my E-meters happen to be fully populated,
except for the nine-pin D connector. So I don't need to take you up on your
generous offers.

It was interesting to look at the two of them, though. They both came in
packaging labeled "Xantrex." But one of them has "Cruising Equipment Co.
Seattle Washington" on the case. This one is conformal coated on the
inside, making the chips really hard to decipher. On the plus side, all the
resistors and capacitors are silk-screened on the board. The other one has
no silk screen, but also no coating, so the chips are easier to make out.

Thanks again,
Jake Oshins


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf
Of Steven Ciciora
Sent: Tuesday, November 06, 2007 2:53 AM
To: Electric Vehicle Discussion List
Subject: Re: [EVDL] Picture of an E-Meter (Link-10) with a serial port?

Yes, I do have pictures of E-meter guts. Both with
and without the serial port option. I don't have
access to the pictures right now. If you email me a
picture of what you have (it will be interesting to me
to compare against what I have) I can tell you the
differences between yours and mine. I've performed
the upgrade several times and it has always worked. =

Some emeters come with the SN75155 chip and two caps
(no missing resistors involved) in which case, all you
have to do is add a 9 pin connector, others don't come
with those parts.

All my comments assume you have a surface mount
e-meter. I've only seen one e-meter with leaded
components, and have not studied it to determine the
difference. For what it's worth, this version had 3
PCBs total (vs the more common surface mount one with
only two PCBs).

As I said above, the missing chip is an SN75155. It's
an 8 pin surface mount chip, located on the same side
as the EPROM (the "BOTTOM"). It's the chip closest to
the 9-pin connector, and pin 1 is located farthest
away from the 9-pin (closest to the display). There
are no resistors involved; just two caps. One is
simply a 0.1uF decoupling cap, located to the left of
the chip. You can see a trace between the bottom pad
and pin 4 of the IC. This is C29. The other
potentially missing cap is C8, wich is 470 pF. It is
horizontally mounted (right angle to C29), located to
the right of C14, and to the left of R4, just below
the HC573 address demultiplexer chip. C14 is the 22uF
50V surface mount cap just to the right of the chip.

Pin 2 of the 9 pin goes to pin 5 of the 75155, and is
the RS-232 data input to the CPU. The TTL version of
this signal is on pin 3 of the 75155, and goes to pin
24 of the CPU (the RxD pin). Pin 3 of the 9 pin goes
to pin 7 of the 75155, which is the RS-232 data from
the CPU, to your e-meter opto-isolator. The TTL
version of this signal comes from pin 25 (RxD) of the
CPU, and goes to pin 2 of the 75155. It is
interesting that the hardware is set up to allow
bidirectional communication. I have not disassembled
the EPROM to see what commands it might accept.

I don't have all my notes in front of me, but several
other pins of the 9 pin connector do other,
non-standard functions. In other words, might not be
a good idea to connect all 9 pins directly to a serial
port; I would just connect pins 2, 3, and 5.

If for some reason, your firmware doesn't have the
RS-232 option enabled (I have not seen this yet), I
might be able to program you an eprom with this
option. Actually, I know I can, but I'm not too sure
how legal it might be.

This is from memory, but I seem to remember one
version of Lee's e-meter opto-isolator having a typeo
in the pinouts. If you study the data sheet it should
be obvious. I made up a few PCBs a while back
implementing this, along with options to connect it
directly to an RS-232 serial port (like Lee's
schematic), or to a TTL port of a microprocessor
(logic inverted). I think I still have one unused
one somewhere, if you are interested.

Do you have these parts on hand? If not, I'm sure I
can send you some. Otherwise it might be hard to make
the minimum order for most places. Or I can just do
the upgrade for you.

I have also made what I call an "E-Meter sucker". It
records all the data from an e-meter to a secure
digital memory card. My one prototype is pretty ugly,
and bigger than it needs to be, but since it's doing
it's job for me, I have not finished a more compact
version. The reason why I'm mentioning this is
two-fold; one is to brag a little :), the other is,
if there is much interest, maybe it might encourage me
to finish up designing the smaller version.

- Steven Ciciora

--- Jake Oshins <[email protected]> wrote:

> I=92d like to modify my Link-10 E-meter to put out
> data through a serial
> port. Apparently the version I own (which is all
> that can be had now) which
> does not have a serial port can be modified to have
> a serial port by adding
> back the components that were not populated when it
> was built.
> =

> To do this, I need to know what the missing
> components were. Lee Hart
> helpfully explained to me that he did it and it was
> fairly easy to do when
> one had a Link-10 with a serial port for
> comparison. It was a matter of
> adding one IC and some resistors.
> =

> My problem is that I don=92t have a Link-10 with a
> serial port for comparison
> and I have no means of getting or borrowing one.
> (Actually, I might. I=92ll
> look into borrowing.)
> =

> In the mean time, is there any chance that any of
> you happen to have
> detailed pictures of the inside of a Link-10 E-meter
> that has a serial
> port? Ones that are detailed enough that I could
> read the part number on
> the IC and the values of the resistors? (I almost
> asked Lee specifically,
> but I=92d rather he spent the time on his Sunrise EV2
> project.)
> =

> If somebody has them, I promise that I=92ll put
> details of my efforts on a web
> site for future people going this direction.
> =

> Thanks,
> Jake Oshins
> =

> =

> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
> =



__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around =

http://mail.yahoo.com =


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #4
Hey all,

Here are pictures of the guts of 2 E-Meters with serial ports. One is an
early production version, SN 3246, the other is later SN 26450. Very
different boards.

http://ironandwood.org/emeter.html


Rush
Tucson, AZ
www.ironandwood.org
www.Airphibian.com
www.TEVA2.com


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #5
Jake Oshins wrote:
> Thanks for this detail. It's tremendously helpful.
>
> From this, I can tell that both of my E-meters happen to be fully
> populated, except for the nine-pin D connector. So I don't need to
> take you up on your generous offers.
>
> It was interesting to look at the two of them, though. They both
> came in packaging labeled "Xantrex." But one of them has "Cruising
> Equipment Co. Seattle Washington" on the case. This one is conformal
> coated on the inside, making the chips really hard to decipher. On
> the plus side, all the resistors and capacitors are silk-screened on
> the board. The other one has no silk screen, but also no coating,
> so the chips are easier to make out.

I took a look inside two of my E-meters; one with and one without serial
port. Both are "E-meters" made by Cruising Equipment. The one without
the serial port (rev.4) was made in 1997; the one with it (rev.6) in
2001. The older board is not conformally coated; the later one is.
Neither main board has silkscreening on either side to identify part
locations.

Steven Ciciora Sent:
> Yes, I do have pictures of E-meter guts. Both with and without the
> serial port option. I don't have access to the pictures right now.
> If you email me a picture of what you have (it will be interesting to
> me to compare against what I have) I can tell you the differences
> between yours and mine. I've performed the upgrade several times and
> it has always worked. Some emeters come with the SN75155 chip and
> two caps (no missing resistors involved) in which case, all you have
> to do is add a 9 pin connector, others don't come with those parts.

On the ones I have, the one without the serial port left out:

- 9-pin male D-connector
- SN75155 8-pin surface mount chip, mounted on bottom of board,
straight toward the front from pin 1 of the D connector, pin 1
end toward the front of the meter.
- C29: 0.1uf ceramic capacitor, surface mount, on the pads near
pins 3-4 of the SN75155.
- C8: 470pf ceramic capacitor, surface mount, on bottom of board
near pins 7-8 of the 20-pin SN74HC573 chip.
- 100 ohm resistor, on top side of board at the extreme right rear
corner between two capacitors.

There might be more; but these are what I remember having to add. I
don't know the capacitor values; they aren't marked and I'd have to take
them off the board to measure them.

> All my comments assume you have a surface mount e-meter. I've only
> seen one e-meter with leaded components, and have not studied it to
> determine the difference. For what it's worth, this version had 3
> PCBs total (vs the more common surface mount one with only two PCBs).
>
>
> As I said above, the missing chip is an SN75155. It's an 8 pin
> surface mount chip, located on the same side as the EPROM (the
> "BOTTOM"). It's the chip closest to the 9-pin connector, and pin 1
> is located farthest away from the 9-pin (closest to the display).
> There are no resistors involved; just two caps.

See above. The resistor completes the ground path of the D-connector,
pin 5. It and the 75155 are what usually blow up if you connect the
serial port to something that is not isolated.

> One is simply a 0.1uF decoupling cap, located to the left of the
> chip. You can see a trace between the bottom pad and pin 4 of the
> IC. This is C29. The other potentially missing cap is C8, which is
> 470 pF. It is horizontally mounted (right angle to C29), located to
> the right of C14, and to the left of R4, just below the HC573 address
> demultiplexer chip. C14 is the 22uF 50V surface mount cap just to
> the right of the chip.

Thanks! I didn't have the component designators for these parts. Your
data also corroborates my findings.

> Pin 2 of the 9 pin goes to pin 5 of the 75155, and is the RS-232 data
> input to the CPU. The TTL version of this signal is on pin 3 of the
> 75155, and goes to pin 24 of the CPU (the RxD pin).

Yes, though the stock E-meter software ignores serial input data.

> Pin 3 of the 9 pin goes to pin 7 of the 75155, which is the RS-232
> data from the CPU, to your e-meter opto-isolator. The TTL version of
> this signal comes from pin 25 (RxD) of the CPU, and goes to pin 2 of
> the 75155. It is interesting that the hardware is set up to allow
> bidirectional communication. I have not disassembled the EPROM to
> see what commands it might accept.
>
> I don't have all my notes in front of me, but several other pins of
> the 9 pin connector do other, non-standard functions. In other
> words, might not be a good idea to connect all 9 pins directly to a
> serial port; I would just connect pins 2, 3, and 5.

Agreed. Though the two boards I just looked at left off the parts that
connected to these other pins, other E-meters may have some of them
installed.

> If for some reason, your firmware doesn't have the RS-232 option
> enabled (I have not seen this yet), I might be able to program you an
> eprom with this option. Actually, I know I can, but I'm not too
> sure how legal it might be.
>
> This is from memory, but I seem to remember one version of Lee's
> e-meter opto-isolator having a typeo in the pinouts. If you study
> the data sheet it should be obvious.

I remember that, too. I swapped the optos, which inverted the data as I
recall.

> I made up a few PCBs a while back implementing this, along with options to connect it directly to
> an RS-232 serial port (like Lee's schematic), or to a TTL port of a
> microprocessor (logic inverted). I think I still have one unused
> one somewhere, if you are interested.

I should have some of my Companion boards later this week, if all goes
well. This is the third revision; I hope it's finally worth shipping!
The Companion is a 2" round board that mounts on the back of the
E-meter, and includes the prescaler, DC/DC, and optoisolator for the
serial port (things that should have been in the E-meter all along).

> I have also made what I call an "E-Meter sucker". It records all the
> data from an e-meter to a secure digital memory card. My one
> prototype is pretty ugly, and bigger than it needs to be, but since
> it's doing it's job for me, I have not finished a more compact
> version. The reason why I'm mentioning this is two-fold; one is to
> brag a little :), the other is, if there is much interest, maybe it
> might encourage me to finish up designing the smaller version.

Cruising Equipment made a really nice version of this. It had 2 megs of
RAM to hold data, a GPS receiver input to add position to the stored
data, a real-time clock to time-stamp the data, and a wireless output so
you could download the data without a cable. Now *that's* the gadget I'd
like to have! But they wanted $1000 for it, so it didn't sell.

How about a little board with an RS-232 port to receive the E-meter
data, another port for a $100 GPS receiver, and a USB port to either
send it to your laptop, or plug in a USB-to-WiFi modem to send it
wirelessly? There's probably something already on the market to do this;
we'd just need the software.
--
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

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #6
Thanks Lee. Whenever I state something absolutely (no
resistors, only caps), I usually end up wrong :) The
funny thing is, I sort-of remembered having to add a
resistor, but at 2 in the morning, I couldn't find all
my notes, and by looking at two e-meters, I only
noticed the two caps. You can see the 100 Ohm
resistor in the pictures that Rod Hower took.

As for the 470 pF, that's the value I measured after
removing two of them (actually, I think one was closer
to 485 pF, and one was closer to 460 pF, but 470 pF is
the closest common value).

- Steven Ciciora

> On the ones I have, the one without the serial port
> left out:
>
> - 9-pin male D-connector
> - SN75155 8-pin surface mount chip, mounted on
> bottom of board,
> straight toward the front from pin 1 of the D
> connector, pin 1
> end toward the front of the meter.
> - C29: 0.1uf ceramic capacitor, surface mount, on
> the pads near
> pins 3-4 of the SN75155.
> - C8: 470pf ceramic capacitor, surface mount, on
> bottom of board
> near pins 7-8 of the 20-pin SN74HC573 chip.
> - 100 ohm resistor, on top side of board at the
> extreme right rear
> corner between two capacitors.
>
> There might be more; but these are what I remember
> having to add. I
> don't know the capacitor values; they aren't marked
> and I'd have to take
> them off the board to measure them.


__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #7
Steven Ciciora wrote:
> Thanks Lee. Whenever I state something absolutely (no
> resistors, only caps), I usually end up wrong :) The
> funny thing is, I sort-of remembered having to add a
> resistor, but at 2 in the morning, I couldn't find all
> my notes, and by looking at two e-meters, I only
> noticed the two caps. You can see the 100 Ohm
> resistor in the pictures that Rod Hower took.
>
> As for the 470 pF, that's the value I measured after
> removing two of them (actually, I think one was closer
> to 485 pF, and one was closer to 460 pF, but 470 pF is
> the closest common value).

I never saw the pictures that Rod Hower took. They must have gotten
stripped when he posted them to the EV list.

Thanks for posting the capacitor values. They aren't marked in the meter
I looked at now. When I upgraded one the last time, the E-meter I looked
at had marked capacitors, but I wasn't smart enough to write down their
values for future reference.

--
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

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #8
That's actually what I'm trying to build. I'm starting with the Atmel
AT90USBKey and adding two serial ports, one for the E-meter and one for
talking to Bruce Sherry's version of EVilbus. Ultimately, I'm trying to
build a system that will stop a Manzanita Micro PFC charger cold once a
certain number of positive amp-hours have been put into a battery pack. I'm
working with some surplus Saft NiMH batteries and my first attempt at a BMS
(testing on a single module) let the cells charge to the point of damage, as
evidenced by significant swelling. I was trying to just go by voltage and
temperature, but that wasn't enough. So now I want to track state of
charge, which (for me) is most easily done by connecting to a device that I
already own which is good at tracking state of charge.

The AT90USBKey contains some flash memory that I plan to use for data
recording. It contains a USB interface. It also contains an SPI interface
that I'm planning on using for connecting a couple more UARTs. Maxim sent
me some sample chips that provide both the drivers and the logic.

If somebody can explain to me exactly why I'd want a GPS for input, too, I'm
all ears. I doubt that I'll bother to buy a GPS, though, so that might
wait.

I plan to make my USB device provide one HID interface for configuration and
two or more communications class interfaces for passing through the serial
data from the E-meter, the EVilbus and perhaps GPS.

It's also worth noting that I'm doing this partly for the experience gained
in building a USB device. I understand that what I'm doing may be overkill
for my needs. But I'm a kernel-mode programmer and building hardware and
firmware essentially adds to a body of knowledge that I want anyhow.

- Jake Oshins


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf
Of Lee Hart
Sent: Wednesday, November 07, 2007 9:46 AM
To: Electric Vehicle Discussion List
Subject: Re: [EVDL] Picture of an E-Meter (Link-10) with a serial port?

(snip)

> I have also made what I call an "E-Meter sucker". It records all the
> data from an e-meter to a secure digital memory card. My one
> prototype is pretty ugly, and bigger than it needs to be, but since
> it's doing it's job for me, I have not finished a more compact
> version. The reason why I'm mentioning this is two-fold; one is to
> brag a little :), the other is, if there is much interest, maybe it
> might encourage me to finish up designing the smaller version.

Cruising Equipment made a really nice version of this. It had 2 megs of
RAM to hold data, a GPS receiver input to add position to the stored
data, a real-time clock to time-stamp the data, and a wireless output so
you could download the data without a cable. Now *that's* the gadget I'd
like to have! But they wanted $1000 for it, so it didn't sell.

How about a little board with an RS-232 port to receive the E-meter
data, another port for a $100 GPS receiver, and a USB port to either
send it to your laptop, or plug in a USB-to-WiFi modem to send it
wirelessly? There's probably something already on the market to do this;
we'd just need the software.
--
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

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #9
Jake Oshins <[email protected]> wrote:
> Ultimately, I'm trying to
> build a system that will stop a Manzanita Micro PFC charger cold once a
> certain number of positive amp-hours have been put into a battery pack.

I'm defintely interested in this too. I'm hoping to use the eMeter to
tell the charger when all the amp hours are put back. But I need to
put back more than 100%. After the eMeter has told the charger it's
put enough back, I'd like the charger to go into to taper down mode
for a specific period of time. It does this now, but that is based on
reaching a voltage limit and doens't guarentee I've returned enough
amp hours back. Instead of stopping the charger cold why not trigger
it to go into taper down mode like when it reaches the voltage limit?

Dave Cover

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #10
dave cover wrote:

>
Jake Oshins <[email protected]> wrote:
> > Ultimately, I'm trying to
> > build a system that will stop a Manzanita Micro PFC charger
> > cold once a certain number of positive amp-hours have been
> > put into a battery pack.
>
> I'm defintely interested in this too. I'm hoping to use the
> eMeter to tell the charger when all the amp hours are put
> back. But I need to put back more than 100%. After the eMeter
> has told the charger it's put enough back, I'd like the
> charger to go into to taper down mode for a specific period
> of time. It does this now, but that is based on reaching a
> voltage limit and doens't guarentee I've returned enough amp
> hours back. Instead of stopping the charger cold why not
> trigger it to go into taper down mode like when it reaches
> the voltage limit?

I believe that you can use an "off-the-shelf" E-Meter/Link10 to do this. You either need a unit with the low battery alarm output option, or you need to open your unit and install the FET that is needed to implement this feature (pulls pin 7 to ground when active).

I believe (but have not confirmed) that if you manually set the charge efficiency factor appropriately, for instance to 90%, then on recharge, the charge Ah are weighted by 0.9 such that when the Ah count reaches 0Ah, 110% of the Ah removed have actually been returned.

The low battery alarm feature is controlled by F10-12. You could then set the alarm to come on anytime the battery is discharged below 10% and to go off at 0%, and use the alarm output to either switch off the charger when it goes inactive, or to signal the charger to reduce its charge rate, etc.

Of course, one should experiment with this themselves first to determine if it will or won't work. The operation can be verified before modifying one's meter (should it be found not to have the alarm option installed), since all meters provide a visual alarm indication.

Cheers,

Roger.

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #11
I wrote:

> I believe (but have not confirmed) that if you manually set
> the charge efficiency factor appropriately, for instance to
> 90%, then on recharge, the charge Ah are weighted by 0.9 such
> that when the Ah count reaches 0Ah, 110% of the Ah removed
> have actually been returned.

Just to be clear, I know the preceding to be true, and have confirmed it.

What I have not confirmed is that the low battery alarm setpoints are affected in the same way by the CEF value. The manual states only that they are based on rate/peukert adjusted Ah.

Cheers,

Roger.

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #12
PFC chargers don't really expose this ability externally. I suppose you
could tear apart the PFC and replace the current limit pot with a digital
pot. But I don't want to do that. I just want to assert the regbus line
that the regulators themselves would assert. The PFC is built to
immediately stop current when that line is pulled. You can try to PWM it,
but if the charger responds at all, you'll just end up putting the charger
into a stop-start-stop-start cycle. (Believe me, I know. I wrote the code
that Rich is currently shipping in his MK3 regs and I tried really hard to
get the charger to taper down. It just wasn't worth it.)

So my plan is to let the regs work, limiting the charger based on both
voltage and temperature, until there are some number of positive amp-hours
on the E-meter. I haven't yet chosen that number, and I probably won't
until I run some more tests. At the point where the battery pack hits this
to-be-determined number of positive amp-hours, the charger stops.

- Jake Oshins

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf
Of dave cover
Sent: Wednesday, November 07, 2007 3:47 PM
To: Electric Vehicle Discussion List
Subject: Re: [EVDL] Picture of an E-Meter (Link-10) with a serial port?

Jake Oshins <[email protected]> wrote:
> Ultimately, I'm trying to
> build a system that will stop a Manzanita Micro PFC charger cold once a
> certain number of positive amp-hours have been put into a battery pack.

I'm defintely interested in this too. I'm hoping to use the eMeter to
tell the charger when all the amp hours are put back. But I need to
put back more than 100%. After the eMeter has told the charger it's
put enough back, I'd like the charger to go into to taper down mode
for a specific period of time. It does this now, but that is based on
reaching a voltage limit and doens't guarentee I've returned enough
amp hours back. Instead of stopping the charger cold why not trigger
it to go into taper down mode like when it reaches the voltage limit?

Dave Cover

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #13
If the E-meter uses Peukert's in that figure, it is blatantly wrong to
use it to determine the amount of charging needed. E-meter uses their
own interpretation of Peukert's and it won't give the right figure for
AH needed.

Danny

>I'm defintely interested in this too. I'm hoping to use the eMeter to
>tell the charger when all the amp hours are put back. But I need to
>put back more than 100%. After the eMeter has told the charger it's
>put enough back, I'd like the charger to go into to taper down mode
>for a specific period of time. It does this now, but that is based on
>reaching a voltage limit and doens't guarentee I've returned enough
>amp hours back. Instead of stopping the charger cold why not trigger
>it to go into taper down mode like when it reaches the voltage limit?
>
>Dave Cover
>
>
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #14
Jake,

I have a board that Chris Brune put together that allows me to control
the PFC's current from the REGBUS. My software runs a 2khz PWM signal
onto REGBUS pin #2 (through a low-pass filter). I didn't let it run
for a long time, but long enough to prove the design (several minutes
at each setting of 10%, 20%, 30% ... on up to full current).

What version of PFC charger were you working with? I thought most of
them now had the "analog cutback" feature that does allow it to taper
the current in response to activity on REGBUS pin #2. Check with Rich
to see what your charger has.

Just for grins, did you try using a pot between REGBUS pin #1 (+5)
and pin #4 (ground), with the wiper on pin #2? With this setup, you
should be able to control the current with the pot, ranging from no
current up to the current set by the 'current control pot' on the PFC.

Ralph


Jake Oshins writes:
>
> PFC chargers don't really expose this ability externally. I suppose you
> could tear apart the PFC and replace the current limit pot with a digital
> pot. But I don't want to do that. I just want to assert the regbus line
> that the regulators themselves would assert. The PFC is built to
> immediately stop current when that line is pulled. You can try to PWM it,
> but if the charger responds at all, you'll just end up putting the charger
> into a stop-start-stop-start cycle. (Believe me, I know. I wrote the code
> that Rich is currently shipping in his MK3 regs and I tried really hard to
> get the charger to taper down. It just wasn't worth it.)
>
> So my plan is to let the regs work, limiting the charger based on both
> voltage and temperature, until there are some number of positive amp-hours
> on the E-meter. I haven't yet chosen that number, and I probably won't
> until I run some more tests. At the point where the battery pack hits this
> to-be-determined number of positive amp-hours, the charger stops.
>
> - Jake Oshins
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf
> Of dave cover
> Sent: Wednesday, November 07, 2007 3:47 PM
> To: Electric Vehicle Discussion List
> Subject: Re: [EVDL] Picture of an E-Meter (Link-10) with a serial port?
>
>
Jake Oshins <[email protected]> wrote:
> > Ultimately, I'm trying to
> > build a system that will stop a Manzanita Micro PFC charger cold once a
> > certain number of positive amp-hours have been put into a battery pack.
>
> I'm defintely interested in this too. I'm hoping to use the eMeter to
> tell the charger when all the amp hours are put back. But I need to
> put back more than 100%. After the eMeter has told the charger it's
> put enough back, I'd like the charger to go into to taper down mode
> for a specific period of time. It does this now, but that is based on
> reaching a voltage limit and doens't guarentee I've returned enough
> amp hours back. Instead of stopping the charger cold why not trigger
> it to go into taper down mode like when it reaches the voltage limit?
>
> Dave Cover
>
> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
>
> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #15
From: Danny Miller
> If the E-meter uses Peukert's in that figure, it is blatantly wrong
> to use it to determine the amount of charging needed. E-meter uses
> their own interpretation of Peukert's and it won't give the right
> figure for AH needed.

No, the E-meter does not use Peukert for charging. Peukert is only used for the empty-full "fuel gauge" bar graph at the top. The amphours and KWH displays count true amphours and watthours.

However, the E-meter does apply the CEF (Charge Efficiency Factor) during charging. Basically, if you set the CEF to 90%, then the amphours during charging count up at 90% of the actual current x time rate.


--
"Excellence does not require perfection." -- Henry James
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart-at-earthlink.net

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #16
Lee Hart asked:
> How about a little board with an RS-232 port to receive the E-meter
> data, another port for a $100 GPS receiver, and a USB port to either
> send it to your laptop, or plug in a USB-to-WiFi modem to send it
> wirelessly? There's probably something already on the market to do
> this; we'd just need the software.

Jake Oshins replied:
> That's actually what I'm trying to build. I'm starting with the
> Atmel AT90USBKey and adding two serial ports, one for the E-meter and
> one for talking to Bruce Sherry's version of EVilbus. Ultimately,
> I'm trying to build a system that will stop a Manzanita Micro PFC
> charger cold once a certain number of positive amp-hours have been
> put into a battery pack. I'm working with some surplus Saft NiMH
> batteries and my first attempt at a BMS (testing on a single module)
> let the cells charge to the point of damage, as evidenced by
> significant swelling.

Nicads need to be restrained so they don't swell. They usually put them
in a rigid metal box, or stack them up and put a plate on each end with
steel bands to keep them under pressure.

But I agree; we do need a good way to control a PFC charger to get a
proper profile for the various batteries that people use with them.

> I was trying to just go by voltage and temperature, but that wasn't
> enough. So now I want to track state of charge, which (for me) is
> most easily done by connecting to a device that I already own which
> is good at tracking state of charge.

Good plan. Nicads work well with amphour-counting charging algorithms.

> If somebody can explain to me exactly why I'd want a GPS for input,
> too, I'm all ears. I doubt that I'll bother to buy a GPS, though, so
> that might wait.

Once you have speed, position, and altitude data, you can make
meaningful vehicle efficiency measurements (KWH/mile etc.)

The GPS is an easy way to get speed, altitude, and acceleration data.
You don't need to tap into the vehicle's speedometer or odometer, or add
a 5th wheel.
--
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

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #17
Lee Hart wrote:

> > If somebody can explain to me exactly why I'd want a GPS for input,
> > too, I'm all ears. I doubt that I'll bother to buy a GPS, though, so
> > that might wait.
>
> Once you have speed, position, and altitude data, you can make
> meaningful vehicle efficiency measurements (KWH/mile etc.)

Indeed, I have calculated the power required to push my car at various
speeds using a GPS. I hope to confirm my analysis when I finish my
conversion.

http://carrott.org/cgi-bin/twiki/view/Main/RollingResistance

I wrote the program to generate the graphs, if you want a copy (I could
even be convinced to make it easy to use) then please ask.

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #18
I don't believe the GPS speed includes the vertical component. It's of
VERY minimal significance anyways in calculating wheel speed.

Did you attempt to address the uphill/downhill or acceleration factor?
I mean if you know you weigh X lbs, then you can actually deduct how
much energy was used to increase your elevation or increase your speed.
Or in coasting or going downhill you should add that kinetic/potential
energy to the energy figure provided to the wheels for the purposes of
determining drag losses.

I believe with a proper algorithm and enough acceleration/coasting
and/or alt changes it's possible to approximate the vehicle's mass
without having to guess or weigh it.

Danny

----- Original Message -----
From: Tom Parker <[email protected]>
Date: Friday, November 9, 2007 5:37 am
Subject: Re: [EVDL] Picture of an E-Meter (Link-10) with a serial port?
To: Electric Vehicle Discussion List <[email protected]>

>
Lee Hart wrote:
>
> > > If somebody can explain to me exactly why I'd want a GPS for
> input,
> > > too, I'm all ears. I doubt that I'll bother to buy a GPS,
> though, so
> > > that might wait.
> >
> > Once you have speed, position, and altitude data, you can make
> > meaningful vehicle efficiency measurements (KWH/mile etc.)
>
> Indeed, I have calculated the power required to push my car at various
> speeds using a GPS. I hope to confirm my analysis when I finish my
> conversion.
>
> http://carrott.org/cgi-bin/twiki/view/Main/RollingResistance
>
> I wrote the program to generate the graphs, if you want a copy (I
> couldeven be convinced to make it easy to use) then please ask.
>
> _______________________________________________
> For subscription options, see
> http://lists.sjsu.edu/mailman/listinfo/ev
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #19
Lee Hart said:

> Nicads need to be restrained so they don't swell. They
> usually put them in a rigid metal box, or stack them up
> and put a plate on each end with steel bands to keep
> them under pressure.

Do you expect NiMH to need the same restraint as NiCd?

Thanks,
Jake Oshins



_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #20
From: Jake Oshins
Lee Hart said:
>> Nicads need to be restrained so they don't swell.

> Do you expect NiMH to need the same restraint as NiCd?

Yes. They also pressurize their cases during charging, and will swell if given the chance. The little cylindrical nicads and nimh cells are pressure vessels; but the big flat cells like the Ovonic ones or the Prius used need strong restraints.

--
"Excellence does not require perfection." -- Henry James
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart-at-earthlink.net

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 
1 - 20 of 22 Posts
Top