Running parallel strings of Nimh D cells into Sevcon/Etek OK ? ( 100 amps )

15 posts / 0 new
Last post
lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
Running parallel strings of Nimh D cells into Sevcon/Etek OK ? ( 100 amps )

I figured I'd put this into the battery section as a start:

I am adding a rear wheel Etek drive into a Honda Insight , as an attempt at Plug in Hybrid assist type of
thing. The Sevcon will power a brushless Etek. For batteries I want to use cell strings from some salvaged
Honda battery backs. Strings of D cells.
To allow this to happen I will have to split the current amongst several strings of the cells
in parallel.
I understand that the charging will be somewhat difficult, perhaps needing a dc-dc converter per each of the parallel strings.
How about the discharge issues?
Will the strings be OK if they are simply put in parallel ? The draw will be 100 amps or so total. My concern is
that the strings share the load in a 'healthy' way.

Still looking for that easy solution for charging this, but that is another thread!

-Les

LinkOfHyrule
LinkOfHyrule's picture
Offline
Last seen: 14 years 9 months ago
Joined: Wednesday, October 17, 2007 - 14:54
Points: 730
Stupid subject being too long...*grumble grumble*

You should be fine. Good NiMH shouldn't be a problem with 5C (2 parallel). Even a single string wouldn't be a problem if you had batteries made for high-rate discharge.

The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai

andrew
andrew's picture
Offline
Last seen: 14 years 1 month ago
Joined: Tuesday, November 28, 2006 - 17:21
Points: 1361
Re: Running parallel strings of Nimh D cells into Sevcon/Etek

That's a really neat idea for a plug-in hybrid! What voltage will you be running the etek it? Will you be using two? Moving a large vehicle is probably too much for one etek, unless the motor is only required for low speeds (like less than 25 mph). I'd highly recommend using a perm PMG132 or etek RT for more power by running at 72v-96v, and using two of them.

Anyway, back to topic. The D cells from Honda Insights are high quality if I remember right. Much better than the chinese D cells, and probably have a much lower internal resistance. Can you find any information off the D cells? What's the manufacturer? Any spec sheets?

I think wiring strings in parallel will be the safest way to do it. That's how I would set it up. If on the other hand, you chose to wire parallel models of cells in series, this might pose a risk of damaging other cells if a cell is bad. A parallel module with a bad cell will require the good cell(s) to take all of the load, and then become fully discharged (because they don't have enough capacity with the "missing cell") and all may be reverse charged before the pack voltage drops low enough for the LV cutout.

On the other hand, if one of a long string of cells is bad, than that whole string voltage will drop some, and the other strings in parallel with it will take most of the load. Provided the LV cutoff is high enough, none of the other cells in the string with the bad cell should be damaged.

The hard part will probably be locating any bad cells. This is why it is advisable to separate a long string of cells into shorter 6-cell or so sections. That way, each section can have a connector, and be load tested independently of the others for diagnostics. Each 6-cell section would be in series with some other 6-cell sections to make a lager string that meets the pack votlage, and this larger string would then be in parallel with other larger strings of 6-cell sections in series.

It may even be possible to data log the voltage of each x-cell section that you decide upon with a pak trackr or some other device.

[url=/forum-topic/motorcycles-and-large-scooters/587-my-kz750-electric-motorcycle-project]KZ750 Motorcycle Conversion[/url]
[url=/forum-topic/motorcycles-and-large-scooters/588-fixing-my-chinese-scooter]900 watt scooter[/url]
Pic from http://www.electri

lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
Re: Stupid subject being too long...*grumble grumble*

Subject is long to be more informative. I like informative thread subjects.

I want to use 8 strings of 36 volts. The Etek is for adding power to the ICE from 0 to 75 MPH.
This should boost my MPG to near 100 mpg when at 60 mph .
I think it could accelerate the car in slow city traffic alone, but 6HP won't move it much past 25 mph, I'm pretty sure.
With 8 strings each string will only be supplying a little above 10 amps, nice and easy for the cells.

I'm reading with interest developments and talk on charging series parallel batteries. Doing a switching arrangement is very undesirable
in my opinion. There has to be a better way !
-Les

LinkOfHyrule
LinkOfHyrule's picture
Offline
Last seen: 14 years 9 months ago
Joined: Wednesday, October 17, 2007 - 14:54
Points: 730
Re: Stupid subject being too long...*grumble grumble*

LOL, I'm just kidding.

Hmm, why only 36V? Money issues?

The batteries will be fine in an 8p string. That's like 1C for most D cells.

Unfortunately, charging multiple strings of NiXX is a PITA. Other than switching around the strings, we haven't really found a way to charge them reliably.

Who's cells are you using anyway?

The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai

lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
brushless Etek rear drive for insight

The cells are from Honda packs. The $350 Sevcon controller I can get can only take 36 volts, so that would be the limit.
What controller would work with the brushless etek that can do more ?
I heard that the controllers that make trapazoid waveforms are not so good with the etek.
The Kelly controller is trapezoidal and accoring to some reports runs the etek loud and weak.
The only money issue for me is the controller.

Maybe a power mosfet switching setup can do the series string switching for charging? I need to monitor the strings for
faults as well. It's a hassle using the D cells but you can't beat them for cost and weight reasons.
I would try wet cell nicads, but the military surplus seems to have dried up on them, as far as I know.
-Les

andrew
andrew's picture
Offline
Last seen: 14 years 1 month ago
Joined: Tuesday, November 28, 2006 - 17:21
Points: 1361
Re: Stupid subject being too long...*grumble grumble*

lesd, I really hate to throw a wrench in your plans because it's a really neat idea. How did you calculate that you can reduce the fuel consumption to near 100 mpg at 60 mpg? Did you calculate the power required to move the Prius at 60 mph? The stored energy in 8 strings of D NiMH cells at 36v is small. The power output of an etek at 36v is not very significant when considering the power to move a car at 60 mph. It's not even very significant when considering the power to move a motorcycle at that speed. And the way an etek would be geared in a fixed gear ratio to provide torque up to 75 mph would mean that it would not be able to output much power at 80 motor amps at low speed. If the 80 amps is set as the battery current limit on the controller side, than the motor current will be much higher at most speeds, efficiency will be low at high current, and this poses a risk of frying an etek.

I'm sure it would be fun experimenting though anyway.

As far as charging a series parallel setup, this could be done with a battery management system (BMS). I don't know how to set this up, though I have some ideas. I recommend contacting some companies that do set up custom BMS systems and see if they'll sell any components, or at least give some consulting.
Here's one I know of: http://www.powerstream.com

[url=/forum-topic/motorcycles-and-large-scooters/587-my-kz750-electric-motorcycle-project]KZ750 Motorcycle Conversion[/url]
[url=/forum-topic/motorcycles-and-large-scooters/588-fixing-my-chinese-scooter]900 watt scooter[/url]
Pic from http://www.electri

lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
Re: Stupid subject being too long...*grumble grumble*

Well, first off, it's a Honda Insight. These cars routinely get 70 mpg stock, with conservative 'hybrid style' driving. No lead foot driving.
Some people even get 80 mpg.
There is a guy that runs a drop down scooter wheel ( 5th wheel) with a brushed etek, he gets the mileage I'm looking for.
My commute is 12 miles. I'm figuring that 240 D cells can push me from 70 mpg stock to 100 mpg with etek ,for that distance. 70% freeway at 60 mph.

I still am debating the etek idea, and maybe instead just having the extra packs feed the Honda hybrid motor. The Honda hybrid power
control system has been hacked to allow forcing the motor into assist mode as needed.
I'll check out the BMS link, thanks !
-Les

lesd, I really hate to throw a wrench in your plans because it's a really neat idea. How did you calculate that you can reduce the fuel consumption to near 100 mpg at 60 mpg?

andrew
andrew's picture
Offline
Last seen: 14 years 1 month ago
Joined: Tuesday, November 28, 2006 - 17:21
Points: 1361
Re: Stupid subject being too long...*grumble grumble*

How fast can the motor assist to when hacked?

Do you have a link to the guy's project that used a brushed etek setup? Any more information on it?

Anyway, I was running some calculations of energy. 240 D cells sounds like a lot, and it probably doesn't weigh much. There also probably not very bulky so that's nice. But at that speed, the weight of a small AGM pack wouldn't matter much because the losses will mostly be air drag on the highway. The same amount of usable energy at the high rate can be stored in 3 50 ah 12v batteries weighing about 110 lbs. The main issue would be the bulk. I'm not dictating your project, just throwing out some ideas. These would be much easier to charge, and it would be easy to throw on more capacity without too much expense, though maybe bulk and weight would become an issue.

[url=/forum-topic/motorcycles-and-large-scooters/587-my-kz750-electric-motorcycle-project]KZ750 Motorcycle Conversion[/url]
[url=/forum-topic/motorcycles-and-large-scooters/588-fixing-my-chinese-scooter]900 watt scooter[/url]
Pic from http://www.electri

lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
Re: Stupid subject being too long...*grumble grumble*

The stock Insight will add electric motor assist at pretty much any normal driving speed.
The site www.99mpg.com has info on the 'hack' and the 5th wheel.
His hack lets you divert more than stock levels of assist.
I am now leaning that way, as opposed to the etek, for helping out the gas mileage.
The charging would be a bit easier. Adding the battery power to the main pack would be slightly
tricky. Secondary packs of higher voltage would allow controlled current flow into the lower voltage
main pack. Lower voltage is 144 volts. ;)

The guy doing 99mpg.com ( Mike ) did the 5th wheel etek to allow EV only mode, and I don't need that.
( The insight can't go under electric power alone, as is the case on the Toyota systems )

So the lead acid batteries have about the same energy/ lb as Nimh ? I think the Honda packs are about 50 lb each
and I would use two of them for 240 cells.

andrew
andrew's picture
Offline
Last seen: 14 years 1 month ago
Joined: Tuesday, November 28, 2006 - 17:21
Points: 1361
Re: Stupid subject being too long...*grumble grumble*

I think I'm off on the calculation. I found this reference: insightcentral.net Battery Module. So the capacity is actually 6.5 ah (I had calculated 4.5 ah), and the weight is 48 lbs. Considering that you'll probably get about 90% of the nominal energy, than that's 1685 whrs for 240 cells, and 96 lbs.

Ok, that throws it off a bit. To store a comparable amount of energy with SLA's at 36v would require three of these: MPL80-12 and weigh about 172 lbs.

So the NiMH cells have about 80% more capacity for the weight. Unfortunately they don't give the size, so I don't know about the bulk comparison. Also, the NiMH would have better cold weather performance, and last probably at least three times as many cycles as the SLA.

[url=/forum-topic/motorcycles-and-large-scooters/587-my-kz750-electric-motorcycle-project]KZ750 Motorcycle Conversion[/url]
[url=/forum-topic/motorcycles-and-large-scooters/588-fixing-my-chinese-scooter]900 watt scooter[/url]
Pic from http://www.electri

reikiman
reikiman's picture
Offline
Last seen: 11 months 1 week ago
Joined: Sunday, November 19, 2006 - 17:52
Points: 8447
Re: Stupid subject being too long...*grumble grumble*

Hmm.. this looks interesting.

BTW a couple months ago I found some ecomodders who are using aerodynamics to get 90+ miles/gallon .. Improving aerodynamics seem to me an easier way to get higher efficiency than adding a 5th wheel.

lesd
Offline
Last seen: 12 years 11 months ago
Joined: Monday, August 6, 2007 - 23:25
Points: 14
Re: Stupid subject being too long...*grumble grumble*

*shrug* Did you see the picture of that Civic with the low CD ? ;) ;)

My Insight has a CD of 0.25 stock, and I kinda like the way it looks.
I am abandoning the extra power using an Etek for my Insight, instead going with just driving the
existing electric motor harder using extra batteries. For my commute of 12 to 15 miles, I should see over 100 mpg without too much effort. I have a Honda 600 that I may try the Etek in instead. That car weighs 1300 lbs but only gets 36 mpg with it's old tech 600cc air cooled engine.

You are totally right about CD being very important. It is the biggest factor when driving on the freeways at a speed of 60 or so. The CD of my old Honda 600 is probably pretty bad.

reikiman
reikiman's picture
Offline
Last seen: 11 months 1 week ago
Joined: Sunday, November 19, 2006 - 17:52
Points: 8447
Re: Stupid subject being too long...*grumble grumble*

Yeah, that one with the fish tail was a bit over the top. But it appeared there were some simple things that could be done which would improve efficiency. Such as cover the wheel wells and put a better nose on the front. Removing the side mirrors can be done but then how do you see behind you. I think it can be done with cameras facing rearward, with LCD screens so the driver can see. The Aptera motorcycle does this, and I recently found another guy who'd done this as well.

The problem is the car companies aren't designing for efficiency. They're designing for other things. If they put their mind to it they could design for efficiency and at the same time make it look pretty.

HCT
HCT's picture
Offline
Last seen: 9 months 2 weeks ago
Joined: Wednesday, February 7, 2007 - 08:16
Points: 137
Re: Running parallel strings of Nimh D cells into Sevcon/Etek

Andrew
Answering your question

Anyway, back to topic. The D cells from Honda Insights are high quality if I remember right. Much better than the chinese D cells, and probably have a much lower internal resistance. Can you find any information off the D cells? What's the manufacturer? Any spec sheets?

Believe that those cells are Panasonic HHR650D 6.5 A/H have been working with them for over 5 years assembled them into 36 volts 5,000 Watts power pack for airplane ( real size)
Those cells when charged up have impedance less than 2 milliohms , capable of deliver almost 300 Amps for quite awhile.Problem come from how to connect to them ,ended up welding 10-32 studs on both end .
Was told both Toyota and Honda have use those cells few years ago .
Now they have repackaged them for thermally efficient with air circulation .

Biggest problem with Nimh chemistry is their reluctance to accept a charge when hot .have seem charging efficiency dropped down to 25% at 45C .
Resolving the thermal situation is most important in using them in scooter , car or whatever ,

HEAT & LOSS & INEFFICIENCY RELATED DIRECTLY to IMPEDANCE

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • stuuno
  • marce002
  • Heiwarsot
  • headsupcorporation
  • OldTech

Support V is for Voltage