Gasoline Battery Charger using a PMG

[SPSInc]

Below is a link to a video showing a 48Vdc battery charging generator I have been tinkering with for a while. I used a 8 HP Honda engine and a 24 Pole PMG. I electronically control the engine speed to regulate the output of the PMG according to the load. For this setup I programmed the charging voltage to 56Vdc and the charging current is limited to 75Amps so peak power is 4200 Watts. Attached to the generator is a load bank and 48Vdc 200 A/hr battery string. At the beginning of the video I apply 60 Amps of load to the battery to start discharging it. I then start the engine. It has electric start and auto choke. The engine idles for a short period and then begins to increase the engine speed to raise the output voltage of the PMG. It does this slowly to walk into the charging load. The control system monitors the output current via a shunt. It will limit the charging current to 75 Amps so to not overload the engine. I then begin to turn off some load and the engine speed increases to raise the voltage but yet maintain only a 75 Amp charge. Once the voltage reaches 56Vdc the controller maintains 56Vdc regulation and the charging current begins to taper off. As I drop more load you can hear the engine speed slow to keep regulation.

http://www.youtube.com/watch?v=1CvYcCUZqOI

[Thob]

Nice setup - I like the way it controls the generator.  I'd be interested in any details you can share about the engine controller - what are you using?  Is it some kind of computer such as arduino, basic stamp, etc?  Also, what kind of actuator are you using to control the throttle?  Does it also sense engine RPM?

[SteveU.]

Hello EricA.
Very nice video you put up.
I really like the way you are acknowledging need to control both the charge voltage and the amperage for proper bank charging. Very smart for fuel consumption and engine longevity to use PM generators and engine speed to do this.
There are a couple of woodgas unit designer/manufactures  needing a variable RPM spark ignition engine charging system I will refer your way through your web site.

You diesel guys hating that aircooled gasoline sound he has a companion video with a watercooled Kohler diesel DC generating up on his youtube channel too.
Inline two cylinder 1800 RPM unit should be more pleasing for you to listen to.

Best Regards
Washington State Steve Unruh

[SPSInc]

Thob,

I used 2 different controls. One controller does the engine speed control. It uses a stepper motor to move the throttle and senses the engine speed. The other controller which is in the small box in front of the generator tells the engine speed controller what speed to run the engine at. This controller then measures the output voltage, output current and adjusts. I even incorporated the ability for remote battery sensing (to make up for cable voltage drop) and temperature compensation but I haven't integrated those features in yet. I used Atmel 8 bit AVR controllers.

EA

[SPSInc]

SteveU

Thank you for your kind words. I appreciate any referrals you wish to give.

I do have a diesel unit on my YouTube channel. It worked out to be a very nice package too. I also did a similar generator to this Kohler diesel using a Kubota DF972 on LP. I'll have to get it posted up there too.

EA

[Henry W]

EA,

That is a pretty neat PMG unit there. I enjoyed watching the clips.
I hope I can get my unit running as well as yours.
Thanks for posting.

Henry

[mobile_bob]

i like what i am seeing with your unit, nice compact unit
i do have a couple of questions

1. what are you using for the bridge rectifier?  with a 24 pole alternator turning fast enough to make
48volts nominal, the need for a fast recovery diode is something i have seen a need for in my testing.
curious what the manufacture of the units are.

2. how are you managing both amps and voltage regulation into a battery bank?  i can see voltage via engine speed and amperage determinate on the batteries acceptance, but am curious how you manage to control both volts and amps with just engine speed, unless...
you are using a controlled rectifier? igbt, mosfet or something else?

3. it is my understanding the alternators are no longer in production?  or is there an alternative?
i understand they are advertized at about 90% efficient, does that also factor in the rectifier losses?
rectifier losses can be significant and reduce the otherwise high advertized efficiency rating to something on the order of maybe 82-83% efficiency?  have you tested the alternator to see if the efficiency curve is fairly flat? or is it peaked at some particular output range?

pretty cool job!


thanks
bob g

[SPSInc]

Hi Bob,

1. I just use standard speed rectifiers. 24 pole at 3600 RPM produces 720 Hz. No need for high speed rectifiers for that. I use rectifiers from Vishay, Powerex, Crydom. Look at a VSKD91.

2. I manage to regulate both the voltage and the current by the speed of the engine. In a battery charging application you can lower the charging current by reducing the charging voltage. What I do is monitor both the voltage and amps. If the amps is greater than the limit then I regulate on the current or in a constant current mode. If the current is less then the limit I increase the speed until the limit is reached or the voltage becomes the regulated limit 56Vdc in this case. At that point I regulate the speed of the engine to maintain a constant voltage. I don't use any SCR, IGBT or MOSFETS.

3. At the moment they are not in production. That is subject to change. I have a number of them in house. I don't have the exact numbers but they are close to 90%. You would then need to add in the rectifier losses. Your 82-83% would be close. I have a source to get some of this information. I'll see what I can dig up on the eff. I haven't seen anything affordable to beat the efficiency of this design. Since you have a direct couple there isn't any loss in a belt drive, no bushes or bearings either. Use of Schottky diodes would reduce the rectifier loss, but you are very limited on the voltage rating of the Schottky.

BTW you can get @ 10KW out of this size generator. It measures about 11" OD and 4" thick. Weighs @ 40lbs.

Hope this helps

[Ronmar]

That is a pretty neat setup.  I wish those alternators were more readilly available...

[mobile_bob]

thanks for the reply

in my testing i had a horrible time with rectifiers, they would go into avalanche and basically the unit becomes a big electromagnetic brake... that at around 600 hz  i tried crydom and powerex with the same results...
i called the engineering dept of one of the manufactures and was told that there are only a couple manufactures that make rectifiers that can handle over 400hz, most of the power rectifiers are made to
operate at 60hz and at over 100hz they go into avalanche.

so i ended up going back to individual press in aftermarket diodes to build up my rectifier bridges,,, i will take a look at the one you listed. it would be very welcome to have something in that form factor to remotely mount.

as for managing both amps and voltage, i see now how our are doing it,, you are doing one or the other but not both at the same time?  i would guess that is effective when you have a microcontroller to monitor both and switch between both criteria making minute throttle changes as needed...
nice approach, simple enough and looks effective enough.

thanks for the clarification on the efficiency thing, it was something i wondered about late last night while going over a set of test reports i did about 3 years ago... when i tested and determined my alternator efficiency, i included the rectifier losses along with the excitation current which of course lowers the efficiency from gross to net which is necessary as we are needing DC anyway.

i would really like a copy of any documentation as to efficiency of those units, fascinating they are!

coupled to a watercooled engine, it would make a sweet small/compact cogen unit.

bob g

[rcavictim]

Very impressive charge and engine control system!

[SPSInc]

Hi Bob,

I can't say I've had any issues with the diodes at this frequency. I've built a lot of units using these modules and never have had a rectifier problem. What type of alternator were you using? What diodes were you having problems with? Have you looked at the AC wave form coming off the alternator? Does it have a lot of high frequency noise that could be the problem? Noise Spikes?

I believe this would make a great cogen application if one had a liquid cooled engine. I don't have much experience in the heat exchanger area but I know my way around a PMG and engine control system. I'd be willing to work with some body if they are looking to but something together.

[Tom Reed]

Yes very impressive! Is 56v the maximum for this or all these type alternators?

[SPSInc]

Tom,

56V isn't the max. This one could go a bit higher. I just set it there for a good charging voltage to play with. At no load 3600 RPM the output voltage is @ 70vdc.

The alternator stators can be wound to achieve any desired voltage. We used to use some that ran at 300Vdc+ that we would then attach to an inverter and build variable speed AC generators with.

[mobile_bob]

i remember taking a look at the waveform on the scope at the time, but i don't recall what i was
using for excitation current during the scope test, it might be the controller was impressing a high frequency
overlay on the waveform when i tested the rectifiers and if i used a dc source and a reostat with the scope it might have been much cleaner waveform.. i don't remember.

i do however remember talking to the engineering dept of one of the big oem's of those 3phase bridge rectifiers and they told me that their offerings were all made for 50/60 hz operation and  would exhibit
this problem over 100hz... i remember the gal asking around her office to see who might make a suitable unit and was advised at the time it was fugi if i recall correctly..

it never occurred to me to look further at the problem, because i could take care of the problem with off the shelf individual diodes at a very reasonable price.

when i saw your youtube clip, and the units you are using it prompted me to ask, so i will be looking at them again.

every time i take the word of some engineer as gospel and change direction it seems later i find out that i should have stayed the course to see if maybe there would be a different outcome.

the 555 alternator is a classic example, everyone in the world will tell you that automotive alternators are
around 50% efficient and as such are a waste of time for high efficiency charging... i listened to that for many years, until one day i decided to do some testing of my own...  to those that said "no" i say bullcrap!

bob g