House Bank not charging from alternator

[sunchaser]

View attachment 137810

Steve
Then comes the very common two alternator setup. The CMS website drawings and discussions cover these scenarios in depth.

 

[DDW]

But why did the ACR not close at 13 causing the wakespeed to see a needed charge to house. Indeed if the start bat is already at 13 it should have combined with ACR.
What is missing here?
The ACR should close when one bank is 13 or above.
I think your problem is wiring like this
View attachment 137809

The start battery is at rest, about 12.8V. Start the engine and there is a 15 second ramp before the alternator puts out anything. Then it drives the start battery to 14.7 absorb. Wakespeed sense line looks at start battery voltage, says yeah good, 14.7V. Then it looks at the shunt current (house) and says, charge is down to <3A, in fact it is discharging, so we should be in float and switches to float. 30 seconds later, the ACR connects. Too late.

In your second diagram, engine starts up, 15 second delay before alternator puts out anything. Then the voltage on the house rises above 13.5 (the connect voltage for the ACR). In another 15 seconds, the Cummins, sitting on the start battery, says hey, I've not seen a charge for 30 seconds so I'm going to alarm, store this alarm in the record of alarms, and make you drill through 3 levels or menu to silence it. Then 15 seconds after that the ACR connects. Too late.

So while those pictures are nice, the reality doesn't match in this case.

 

[DDW]

Just to end my highjacking of the thread and return it, this is what I decided to do:

Now I'm at the boat and can see and try things, I measured the current requirements of the manual selection wire of the Blue Seas 7622. It has a red wire, pulled to +12 it forces the solenoid on, pulled to ground it forces it off, left to float it is in Automatic switching mode. This is normally tied to a SPDT On-Off-On rocker switch. The amount of current required to be supplied or sunk is tiny: 0.64 ma sourced from 12V to force on, 0.24 ma sunk to ground to force off.

I decided to Wire-Or the ignition circuit to the red control wire, through a diode and 1.2K 1/2W resistor in series. If the ACR mode switch is in the middle (off, or letting the red wire float), turning on the ignition will drive it high through the diode and resistor, causing the ACR to connect. It is then already connected when the engine is started, eliminating all the timing problems. If the mode switch is set to force on (12V) it cannot leak through the ignition as the diode blocks it. If the mode switch is set to force off (gnd), it will pull the red wire low. If the ignition is simultaneously on, there will be about 16 ma leaked through the diode and resistor, worst case about 0.2W so a miminum of 1/4W resistor is necessary.

This preserves the function of the mode switch (and its LED status) in all cases, while guaranteeing that in normal circumstances the solenoid is on when the engine starts, and off when it shuts down. That is what I have been doing manually for a year. If isolation is wanted (for example a very low house battery that you do not want to parallel during start) the mode switch can be set to Isolate and the ignition will have no effect.

It isn't ideal, as it will routinely parallel the house and start batteries before every start, but this isn't that horrible an idea in most cases, in the event of a really low house battery I can use the mode switch as described. It is much simpler to implement than the other alternatives, and seems to have as few or fewer drawbacks.

 

[JDCAVE]

The start battery is at rest, about 12.8V. Start the engine and there is a 15 second ramp before the alternator puts out anything...

I’m surprised by the lengthy lag for the alternator to be energized.

Jim

 

[sunchaser]

I’m surprised by the lengthy lag for the alternator to be energized.
Jim

But, even if a lag is there does your 2812 show your batteries are getting charged? It appears so.

 

[luna]

DDW.

Glad to read that you settled on the ignition circuit driven, "force on" scheme.

If the always paralleled at start up mode causes issues, (low voltage drop out of electronics or ?) perhaps the 7622's start isolation circuit could be enabled. It still should work, as I think all you have done is told the 7622 that the voltage is high enough that it must combine, even though the voltage is not.

The low current (0.24ma) to turn "off" explains why after a few years the SPDT switch fails due to oxidation. At that current and this use environment, it's a poor switch design choice.
You can look forward to pulling the switch and cleaning the silver plated contacts about every 2 years or so. I just did mine for the 3rd time last week.

 

[DDW]

I’m surprised by the lengthy lag for the alternator to be energized.

Jim

Yeah, a Balmar has a ramp delay you can set, but you can also set it down to a low value. On the Wakespeed, 15 seconds is the minimum time, it can be disabled completely. But the combine delay on the Blue Sea ACR is fixed at 30 seconds also. That is sufficient by itself to cause problems.

 

[CharlieJ]

@DDW #25:

If the sense line is connected to the house and the alternator is connected to the house, no charge to the Cummins QSB and it low voltage alarms.

Why does the Cummins black box report under voltage in <45 seconds?

 

[DDW]

Don't know - they won't give me the source code for the ECU. If they did, I could fix a few problems.

With the Balmar alternator, it would alarm anytime the engine idled because the Balmar could not break even at idle. The engine and alternator draw something around 20A just to stay running. I tried to have them turn up the idle but that was apparently a Board level decision at Cummins. Went back to a Delco 28Si, which will do about 90A at idle.

 

[JDCAVE]

But, even if a lag is there does your 2812 show your batteries are getting charged? It appears so.

Tom: Yes, in my case it did once I fixed my problem, but I was responding to DDW’s post.

Jim

 

[ronobrien]

The following Thread, House Bank not charging from alternator, was originally posted yesterday and was lost, together with a number of other threads. This was the approximate post, with minor revisions and I have also posted the reply by DDW.

We did a sea trial yesterday that went well. Back at the dock, plugged into shore power immediately and the Magnum 2812 went into bulk and then threshold charge. I thought, that’s strange and looked at the monitor. The bank was down 20 amp hours. It should have come back fully charged.

I came back to the boat the next day and checked all connections and fuses, on the alternator and at the battery bank. The voltage regulator is a Balmar 612. Again, all connections and fuses looked good. The alternator connects directly to a Mastervolt battery mate 1602 isolator. These systems have worked well in the past. Prior to the trial, the house bank was fully charged at 13.54 volts and float float charge.

I turned off the shore power and started up the boat and bring the revs up to 1200 in neutral. Testing with a multi meter, the alternator is putting out 18.34 volts and 71 amps. House bank indicated 12.54 volts and was not charging. It was covering the house loads of ~20 amps. The voltage at the Batterymate studd to the generator battery was 17.36 volts, from the Batterymate stud to the starter battery, 17.9 volts and the Batterymate stud to the house bank, 12.54 volts. End of test.

Right now I’m concluding there is an issue with the 612 regulator, either a faulty connection or wire harness issue. For some reason it is not regulating the alternator. I will test this tomorrow and if I am unable to resolve it, I will take the alternator to one of my trusted shops, Brighouse Electric for bench testing and ask their opinions on the subject. Brighouse does alternators, starters and winch motors. That’s all they do. I’ve used them in the past and trust them.

Jim

I'm not familiar with the setup or the specific alternator but 18+ VDC is enough to cook batteries, unless that is measured before a regulator. Voltage drop testing (which it seems as though you have done) will point to where the issue lives.