Author |
Share Topic Topic Search Topic Options
|
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Topic: Texman 182g Mods Posted: 18 Feb 2019 at 6:11pm |
|
|
 |
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Posted: 18 Feb 2019 at 6:07pm |
|
|
 |
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Posted: 18 Feb 2019 at 6:03pm |
I have the recirc system up and going. I have tested some and it works. The placement of the thermostat sensor for a the recirc side is still ongoing. My first attempt was in a place too exposed and basically caused the water heater and pump to run continuously. I had my whole fresh water tank at 60 deg when freezing outside. No surprise then I am low on propane and it will be cold for several days and I don’t have time to test right now. So I wintrrezed the pod for now to focus on exposed pipes. The 182g has three water pipe exposures. Two are low point drains and I think I have ideas to insulate those as I can’t remove them. The third is the supply line that runs from the tank to inside the pod. It is about three fry long and totally exposed and is a definite problem. Scratching my head on that right now. Reroute maybe, insulate it triple? Not sure yet. That is definitely the place for the thermostat sensor.
The system runs from one switch. It is really a on-off-on switch called a dpdt with led indicators. When switched to red, the valve controlling whether cold water flows normally is closed. This also powers the thermostat that controls the two valves that allow hot water to recirc and return via the overflow line and ultimately the fresh water tank. I had some water spew from overflow so I had to reduce flow accordingly. I will post some pics.
|
|
 |
offgrid
Senior Member
Joined: 23 Jul 2018
Online Status: Offline
Posts: 5290
|
Posted: 15 Feb 2019 at 11:55am |
And I seem to be constantly relearning stuff I used to know. Comes with age I guess.
|
1994 Chinook Concourse
1995 RV6A Experimental Aircraft
2015 Rpod 179 - sold
|
 |
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Posted: 15 Feb 2019 at 11:06am |
off grid-that actually make sense to me now.  i have been learning....
|
|
 |
offgrid
Senior Member
Joined: 23 Jul 2018
Online Status: Offline
Posts: 5290
|
Posted: 15 Feb 2019 at 11:04am |
The traditional way to measure dc current is using a current shunt. Those are rock solid and accurate, I'd recommend them over hall effect sensors for permanent installations. For temporary clamp ons, the hall effect sensors are great, just zero them immediately before each measurement and keep them away from magnetic fields.
I last got bit on this trying to use my hall effect clamp on to measure brake magnet current and wondered why it was different every time I took a reading. Duh. 
|
1994 Chinook Concourse
1995 RV6A Experimental Aircraft
2015 Rpod 179 - sold
|
 |
GlueGuy
Senior Member
Joined: 15 May 2017
Location: N. California
Online Status: Offline
Posts: 2702
|
Posted: 15 Feb 2019 at 10:32am |
Congratulations on getting it together and outlining for the rest of us some of the ins and outs. Hall effect sensors are great, in fact there's one in that Klein meter of yours (and mine). It's currently the only way I know to get a non-invasive measure of DC current. A couple of things you did not mention about them is that they require an accurate bias-voltage in order to operate, and that both the bias voltage mechanism as well as the hall-effect circuit need to be temperature compensated. That's something you may not find in less expensive iterations.
None-the-less, I think it will work very well for you. We have a few dozen neighbors who have off-grid solar systems, and I know of a few that have a similar power-in/power-out meters separate from their charge controllers and inverters.
|
bp
2017 R-Pod 179 Hood River
2015 Ford F150 SuperCrew 4WD 3.5L Ecoboost
|
 |
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Posted: 15 Feb 2019 at 9:11am |
Here is my take on the meter at this point:
wiring: Red for meter power, yellow wire measures watts and mine is connected to + battery cable, white is ground. only problem is the lead to the donut (Hall sensor) is only about 12" long. It must be extended unless you are mounting in the battery box. Other reviewers had already done that and said no change in readings were noted. I added about 60" to the sensor with 20 ga solid copper. it has three wires that have to be matched. My sensor has the lead from the battery to the board passing through it behind the board. The meter itself does not come with a bezel for mounting, so plan accordingly. Getting the meter cleared to zero for current readings should be done prior to mounting i think. Since mine was mounted, i had to jump wires from the meter straight to the battery to get a zero measurement to start with. you have to have power to the meter to clear the current measurement, so if you are running on battery, the sensor must be not in place. and, the meter will not clear unless the sensor is present. The instructions are ok, but lack proper or even normal English so a bit had to follow on some items.
THe meter readings for current are right on with my Klein at currents above 1 amp. The meter only measures current to 1/10th of an amp. So it may just be rounding when looking at milliamps. not sure yet. I really like that the meter keeps a cumulative amp in/amp out measurement and it is settable to the total amp hours of your particular battery setup. As long as accuracy is good, it should give me close to a fuel tank like setting for the batteries. Relying on voltage reading is good as well, but when batteries are under load the voltage does not read the same as a resting battery. The meter allows you to clear the cumulative amps in/out when needed to establish a baseline. For instance when you just started a boondock session and want to see if you are over or under what you are getting from solar.
edit: and after doing some reading hall sensors are "notorious" for being inaccurate.
quote :Since the Hall sensor is not directly connected to the current-carrying
wire, outside forces can cause significant error in the magnetic field
measurements [1]. The earth’s magnetic field alone can cause a 0.4 A
error, not to mention the fields generated by other coils, conductors,
and electric motors / generators internal to the vehicle [1]. Being
in-circuit means an IBS allows far fewer errors from outside
interference compared to a Hall Effect sensor. The maximum current
sensing error under any in-car condition for an IBS unit should be 0.5 %
plus offset (30 mA), which is the same error that can be observed due
to the earth’s magnetic field in the Hall Effect sensor just by changing
directions with 80 A of current flowing [1].
so i think at this point i will conclude with this: it is good enough for who it is for. At least for now anyway. A shunt system is more accurate but has some disadvantages as well.
|
|
 |
texman
Senior Member
Joined: 24 Jul 2014
Location: TeXas
Online Status: Offline
Posts: 446
|
Posted: 14 Feb 2019 at 2:19pm |
the only thing we use it for is our air mattress which is 120vac, 0.9amp and 105 watts on the rating. I am getting a 30 amp inline fuse for the inverter and using 10 ga wire. the wire is only 2 feet long from inverter to the board connection. i will put the fuse in that 2 ft span. no hair dryers allowed.
|
|
 |
offgrid
Senior Member
Joined: 23 Jul 2018
Online Status: Offline
Posts: 5290
|
Posted: 14 Feb 2019 at 2:01pm |
Many higher quality inverters do have a standby mode where they draw a few tens on mA to run the control board only, then turn onthe power section every say 10 seconds long enough to see if there is a load on the ac line. If there is they stay on, otherwise they shut down the power section off.
Yes you should have a fuse or circuit breaker on the inverter positive line if that is coming off the battery. It is there for the same reason you have cb’s in your panel, to protect the conductors from burning up from shorts and faults. So you size the wires and fuse or cb the same way the electric code would require you to do it for your ac household circuits. That is, you would size the conductors for the max rated inverter input current and then select the fuse or cb for that wire size. That max current should be in your inverter manual, if not it is going to around the inverter max output power rating divided by 12 divided by the inverter efficiency, which is usually around 80-90 percent, so around 60A for your inverter. That will be 4 awg wire. I know that sounds high but you’re dealing with a lot of current at only 12V. Many bigger inverters are run with 2/0 or 4/0. By doing it that way you won’t have nuisance cb trips or fuses blowing if you use the inverter at full load, and most importantly you won’t have circuits burning up. The cb or fuse should go as close to the battery positive terminal as you can so it protects the whole circuit from faults and shorts.
|
1994 Chinook Concourse
1995 RV6A Experimental Aircraft
2015 Rpod 179 - sold
|
 |