I think I've found my generator head, a 7.5KW Stamford Newage

[TimSR2]

http://www.frontierpower.com/stamford/bc-162.pdf

Actually a 12 kw generator in prime service.  Major bargain if they can be adapted.

[Simtech]

A good parts list from Cummings Generator

https://www.cumminsgeneratortechnologies.com/www/en/support/aftermarket/PartsLists/BC_PartsList_01.pdf

40 D.E. ADAPTOR BRACKET SAE 5 . . . . . . . . . . . . . . . . . . .130-21602 . . . .1
41 D.E. SCREEN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .130-21325 . . . .2
44 D.E. BEARING KIT . . . . . . . . . . . . . . . . . . .  . . . . . . . . . .45-0363 . . . .1
45 D.E. ENDBRACKET . . . . . . . . . . . . . . . . . . . . . . . . . . . .130-20773 . . . .1
46 ADAPTOR KIT SAE 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0719 . . . .1
. . ADAPTOR KIT SAE 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0619 . . . .1
. . ADAPTOR KIT SAE 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0620 . . . .1
. . ADAPTOR KIT SAE 5 . . . . . . . . . . . . . . . . . .  . . . . . . . . . .45-0621 . . . .1

[BruceM]

Using caps alone for filtering rectified single phase power has some issues; your power factor is crap, since most of the load is on the leading top quarter of the waveform, only above the capacitor voltage. If your generator head is way larger than the total load, and your other loads aren't fussy about waveform, it will work fine, though fuel efficiency will suffer some.  Your DC voltage output is about 1.4x the AC input RMS voltage. This is the cheapest solution, usually.

If an inductor is used for primary filtering on rectified single phase, with capacitors as secondary, you can have an excellent power factor, no waveform distortion to affect other loads, and  least loading of your generator.  Your DC output voltage =  AC input RMS voltage.

Recently was messing with a design for a 120V battery bank charger that used 3 small caps, dynamically (solid state) switched in combination before the primary inductive filter to regulate the charge current. I had tested this on grid power and it worked like a champ.  On my ST-3 this same design was a total bust at 1200 watts of charge and above. The AVR didn't regulate the voltage properly as the waveform was trashed. Between the low voltage and trashed waveform, I got control relay chatter.  I redesigned the AVR using a true RMS chip.  Worked, but I had no capacity left for other loads due to poor PF, and the RMS chip was so slow in responding that there was no more flicker compensation, so my shop heat/light floods were driving me buggy.

I finally abandoned my not so "brilliant, simple, patentable" switched capacitor design with custom printed circuit boards    and went to a homebrew motorized Variac and straight inductive filtering.   It's (embarrassingly) low tech and simple and works like a champ.

[Henry W]

A good parts list from Cummings Generator

https://www.cumminsgeneratortechnologies.com/www/en/support/aftermarket/PartsLists/BC_PartsList_01.pdf

40 D.E. ADAPTOR BRACKET SAE 5 . . . . . . . . . . . . . . . . . . .130-21602 . . . .1
41 D.E. SCREEN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .130-21325 . . . .2
44 D.E. BEARING KIT . . . . . . . . . . . . . . . . . . .  . . . . . . . . . .45-0363 . . . .1
45 D.E. ENDBRACKET . . . . . . . . . . . . . . . . . . . . . . . . . . . .130-20773 . . . .1
46 ADAPTOR KIT SAE 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0719 . . . .1
. . ADAPTOR KIT SAE 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0619 . . . .1
. . ADAPTOR KIT SAE 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . .45-0620 . . . .1
. . ADAPTOR KIT SAE 5 . . . . . . . . . . . . . . . . . .  . . . . . . . . . .45-0621 . . . .1

Been through this. I spent lots of time over 8 months ago doing exactly what you are trying to do.

If you look, the stubshaft is not listed. That is the part that is no longer available and is not listed in the parts list. They have been all bought out by a company in Mexico. They are no longer manufactured. And Cummins is not willing to have them manufactured and would rather see these gen-heads rust away.

Henry

[Simtech]

Thanks hwew, I'll quit barking up that tree and just focus on fabbing an end bearing housing.  It doesnt need to be pretty, just work well enough to hold the side load of a pulley and belt.
I tried to call Cummings today and they're closed anyway.

Bruce, could you maybe expound on the specifics of your rectifier/filter?  Perhaps a circuit diagram?

[dieselgman]

It may not be too difficult to turn out a stub shaft for that application... just a thought.

dieselgman

[Henry W]

Well I been holding off on writing this. I have a contact number from the engineer that has helped find the stub shaft info that is no longer available. What I am planning to do is see if it is possible to release a drawing of the stub shaft. Since it is obsolete and they have no intrest in manufacturing them I feel why not ask. They would not be loosing anything. If they would release a drawing I have the means to get them machined on a C&C.

Henry

[XYZER]

IMO the shaft is the easy part....the end plate to carry the bearing will take a fair size lathe and material to fabricate. Not a stopper just some fab and setup.

[Henry W]

This is true.
Lets get the price on all the other parts listed.
I'll see if I can come up with a drawing for the stub shaft than there would be no guessing or trial fitting. Just get it machined.
Once the cost is figured than members can decide if it is worth it.

Henry

[BruceM]

Simtech,
My AC charger is part of a larger custom system which I designed for a 120VDC (nominal) battery bank.  Each 12V battery in series has an individual temperature controlled, 3 voltage shunt regulator (float, bulk/absorption, equalize), with shunt current feedback.  The PV controller is analog; it will charge at 3 preset net charge currents. (Float 0.1A, Bulk- Max PV, Absorption- 2.5A)

The Battery Bank Charge Controller is a Picaxe 40X2 based board which I designed. It manages the PV charge controller, the 10 battery shunt regulators, and the Variac control for the AC charger- adjusting it once per second as needed to keep the battery shunt regulators from overheating during the bulk charging.  It presently stops charging when the batteries are each at bulk voltage and net charge current is down to 3 amps.  That's not full, but it's close enough that even on a dark day the PV system can finish the charge.

The two stage inductive filtering of my AC charger (see schematic below) is not necessary, but I was trying to reduce the 10Hz "Lister flicker", which I find bothersome on lower wattage incandescent lights.  I don't find it bothersome at the present level.  60Hz ripple is down to about 30mv on 145VDC at 10 amps. 10Hz variation is about 1.4V.

If I had been feeling more ambitious I could have used active linear regulation to get rid of the 10Hz Lister Flicker, by just regulating to a steady voltage a few volts below the peak. That could eliminate the second inductor, and reduce the size of the primary inductor as well. But the big toroidal transformers I used were surplus ebay buys, and this is just a prototype.

Since you're planning on using a PWM charge regulator, all you need is bulk DC within it's input specs. The Midnight 200 can handle rectified 120VAC in terms of peak voltage. No filter needed for 3 phase. For single phase, you just have decide on inductive vs capacitive filtering.

[Simtech]

Bruce:  Oy!  Thats a bit more complicated that I want to go, I ran the numbers for a 4000W combined inductor/capacitor filter and I came up with 19 Henrys for the inductor.  Is that right??! Thats a pretty big inductor.

I havent mucked about with component level electronics in a long time so I could easily be wrong.

[BruceM]

Hi Simtech,
Yes, doing a 120V battery bank charge controller properly is a bit involved. You can't just charge the bank as one big series battery and hope that frequent equalization will keep all the batteries in sync. There is just too much variation in individual battery charge acceptance and internal resistance.  you must apply the individual battery shunt regulator approach used by the electric car folks. There are lots of advantages of this method, and higher voltage banks in general, but I won't go into that.

Choke size isn't anywhere near your multi-Henry figure. It will be a small fraction of a Henry, depending on your allowable ripple.  A free program,  PSUD is good tool for doing a thorough design of an LC single or multistage filter.

Also:
http://www.toroid.com/custom_transformers/dc_filter_chokes.htm
has a table for selecting values. 

The rule of thumb by audio folks is that the critical choke size (minimum for primary filter) is L=E/I (where I is in milliamps).

Here's the two critical equations:

The minimum current is aproximately equal to:

Imin (mA) = [Vin(RMS)] / [L(H)]

You can calculate the ripple of the choke-cap system as follows:

[Vrip(RMS)] / [VDC] = 1 / [ 6 * sqrt(2) * 2 * PI * f * L * C]

L = inductance
f = AC line frequency
C = capacitance
VDC = DC voltage at cap
PI = 3.14...

For a filter in front of the Midnight 200 classic, for a single phase generator, you would want a simple LC filter, unless you have excess capacity and are going to just do capacitive filtering.  In that case if you know your allowable ripple voltage:

Delta V = I / [2 * f * C]

Toroids are much less prone to DC saturation, so I always use surplus toroid transformers for chokes. The utter BS/cheap bastard method I often use is to just make sure the winding you use as a choke is rated for more than the DC current.  You don't need a toroid as large as the primary transformer, but it won't hurt.  I've not had trouble with DC saturation using the BS/cheap bastard method and toroid transformers. A conventional E-I core transformer is another matter altogether; they can't normally be used unless they have an air gap, due to DC saturation.

If your head will do 3 phase, 120V, with a common ground to the battery bank negative, you don't need any filtering at all, using the Midnight 200 classic charge controller, according to their literature.  If a single phase generator head is used, ask Midnight how much ripple is allowable. It will probably be quite large, feeding a 36 or 48V battery bank from 120VDC (nominal). You might be able to use a single 1000 watt toroid transformer core, using a low voltage secondary winding for the choke.

[mbryner]

19 Henrys for the inductor

The inductor I'm using w/ my Classic200 is no where near that large and it works.

Delta V = I / [2 * f * C]

Bruce:   So for 1 V ripple voltage @ 60 Hz and 40,000 mA (i.e. 40 Amps) --> C = 333 (?)   microfarads?

Wish I had an o-scope.  I just ordered 2 x 3000 microF caps from Digikey. Probably way over did it.

[mbryner]

Here is the pdf that Halfcrazy (Midnight Solar employee) gave me describing his setup for charging 48 V batt bank w/ a Classic:

DC_charging_w_listeroid.pdf

[Tom Reed]

Very interesting post Doc. No inductor just rectify the AC and some caps. Interesting his 6/1 roid has what look like 1000 rpm flywheels.