ARB compressor failures and fixes: Relays, leaks, and quick couplers

[Canyon Trekker]

Yep, I've been noticing hard cranks in the morning even with the Odyssey AGM batteries, so I've been tracking down the source. This relay failure may be one of them. Otherwise it will be the Midland radio or Samlex power Inverter.

Hope it's not the Samlex, I have the same one.

 

[BroncOh]

Hope it's not the Samlex, I have the same one.

I know this may not be the right thread for this for posterity reasons. I just completed my 90 minute parasitic draw test with the Fluke 117. Factory spec is 25mA to 50mA. This was with the ARB completely removed from truck.

Average draw: 31mA
Min draw: 26mA
Max draw: 196mA
Present draw at time of test completion: 39mA

 

[Canyon Trekker]

I know this may not be the right thread for this for posterity reasons. I just completed my 90 minute parasitic draw test with the Fluke 117. Factory spec is 25mA to 50mA. This was with the ARB completely removed from truck.

Average draw: 31mA
Min draw: 26mA
Max draw: 196mA
Present draw at time of test completion: 39mA

Do you think the .196 A could just be a glitch? Sometimes a lead or probe movement can cause such things. Otherwise looks normal? I'll have to check mine out. I don't have the 117 but I have other current meters. Does your 117 have the current clamp option?

 

[BroncOh]

Do you think the .196 A could just be a glitch? Sometimes a lead or probe movement can cause such things. Otherwise looks normal? I'll have to check mine out. I don't have the 117 but I have other current meters. Does your 117 have the current clamp option?

I don't think it was a glitch.

I have a high current Uni-Trend UT216C clamp meter, but it doesn't work good for these low currents. I used the 117 in series with auto power-off disabled. With dual batteries on the 6.7L, the procedure is to disconnect one of the batteries.

What I'd really like is to monitor the total battery consumption that occurs overnight and not just the mA reading.

 

[Canyon Trekker]

I don't think it was a glitch.

I have a high current Uni-Trend UT216C clamp meter, but it doesn't work good for these low currents. I used the 117 in series with auto power-off disabled. With dual batteries on the 6.7L, the procedure is to disconnect one of the batteries.

What I'd really like is to monitor the total battery consumption that occurs overnight and not just the mA reading.

I know they make data loggers and watt hour meters, mostly for AC unless you get into high end test equipment. But there are possibly low-cost alternatives such as this:

Amazon.com

If I was still working, I would have access to a Keysight 34970 or 34972 and Benchlink software which can do all of what you are looking for. But it's not cheap. (The Benchlink software is free) The above unit sounds like it might do what you're looking for though.

Just FYI

keysight data logger - Search Videos

www.bing.com

 

[BroncOh]

@Canyon Trekker I just tested the Midland radio by itself with the meter, and it only drew a max of 7mA (tested for 1 hour).

I'm thinking the capacitors in the Samlex are what is eating up some of my battery, as it's not truly ever 100% off.

 

[Canyon Trekker]

@Canyon Trekker I just tested the Midland radio by itself with the meter, and it only drew a max of 7mA (tested for 1 hour).

I'm thinking the capacitors in the Samlex are what is eating up some of my battery, as it's not truly ever 100% off.

I would think that the 31-ma average draw would be insignificant to the battery charge but if it is in fact causing the issue, it may be drawing more than you have seen so far. A good test obviously would be to disconnect it and see if it changes anything.

Capacitors rarely go intermittent but for sure, they can. But I've only seen that a few times in my career and they were generally signal capacitors, not electrolytic, though electrolytic failures can cause other bizarre symptoms to signals. Electrolytics tend to short or open. When they open, they do not regulate well, if at all. In all devices with a "warm start" where a signal energizes the device, the primary power supply usually has a live section for the logic. Many times, this is just a 5-volt circuit just for the logic. So, if there was "activity" in the processor of the Samlex and you had an open electrolytic, that may cause the indication of intermittent jump in current draw. However, in inverters, there are usually several electrolytics in parallel on the primary side of the power supply and the likeliness that they are all open is near impossible. Especially if it works ok otherwise.

That's not to say the Samlex could not cause the issue, other components can break down as well. And Switching circuits are unfortunately full of intermittent issues. (Sometimes caused by insufficient capacitors...) Cold weather may affect them as well.

I'm going to try and measure my current draw tomorrow, I also sent a message to Samlex for more information on quiescent draw in the standby position, both with and without the remote meter plugged in.

 

[BroncOh]

I would think that the 31-ma average draw would be insignificant to the battery charge but if it is in fact causing the issue, it may be drawing more than you have seen so far. A good test obviously would be to disconnect it and see if it changes anything.

Capacitors rarely go intermittent but for sure, they can. But I've only seen that a few times in my career and they were generally signal capacitors, not electrolytic, though electrolytic failures can cause other bizarre symptoms to signals. Electrolytics tend to short or open. When they open, they do not regulate well, if at all. In all devices with a "warm start" where a signal energizes the device, the primary power supply usually has a live section for the logic. Many times, this is just a 5-volt circuit just for the logic. So, if there was "activity" in the processor of the Samlex and you had an open electrolytic, that may cause the indication of intermittent jump in current draw. However, in inverters, there are usually several electrolytics in parallel on the primary side of the power supply and the likeliness that they are all open is near impossible. Especially if it works ok otherwise.

That's not to say the Samlex could not cause the issue, other components can break down as well. And Switching circuits are unfortunately full of intermittent issues. (Sometimes caused by insufficient capacitors...) Cold weather may affect them as well.

I'm going to try and measure my current draw tomorrow, I also sent a message to Samlex for more information on quiescent draw in the standby position, both with and without the remote meter plugged in.

Thank you! Hopefully they reply soon. Worth noting I do use the remote panel which requires the red switch on the unit itself be left in the off position, so this soft-switch functionality probably draws some current too.

I'll probably test this today for parasitic draw like I did with the Midland radio, to see what it draws for about an hour.

 

[BroncOh]

@Canyon Trekker

Finished a 1 hour test of just the inverter and it drew:

53mA max
2mA average
negative 38mA min (this is odd). Wonder if there's some reverse leakage somewhere in a FET intrinsic body diode or regular diode.

 

[adrummingdude]

Well It's good to see how they perform over the long haul. I have yet to install mine and I may opt for a heavier relay than the factory. I always thought ARB was at the top of the game, but all this makes me think it's kind of average.

Meanwhile, I have some auto parts store special compressor that is housed in a steel ammo can that just keeps truckin right along. Granted, it takes a while to pressure up one of these big tires, gets real hot and makes a WHOLE lot of noise but hey, 100 bucks all in.

 

[RedZilla]

Meanwhile, I have some auto parts store special compressor that is housed in a steel ammo can that just keeps truckin right along. Granted, it takes a while to pressure up one of these big tires, gets real hot and makes a WHOLE lot of noise but hey, 100 bucks all in.

Yup, simple and dirty probably wins in this situation. If my ARB kicks the bucket ever once I get it installed, I'll just put my little pancake compressor in the back of the truck and run it off my Ecoflow power station. I think I paid around $89 for it at Lowes.

 

[Canyon Trekker]

@Canyon Trekker

Finished a 1 hour test of just the inverter and it drew:

53mA max
2mA average
negative 38mA min (this is odd). Wonder if there's some reverse leakage somewhere in a FET intrinsic body diode or regular diode.

Well, I attempted making a measurement on mine, the clamp meter I have is too small for the cable size I have, so I will have to disconnect and hook up a smaller wire. Can't do that today or possibly tomorrow, the inverter is in the bed at the rear, we have one of those atmospheric river storms happening right now and I would fill the bed up in about 30 minutes or so. If we get a break in the storm, I'll try again. Don't want to get rain in the inverter.

Meantime, to address your -38ma issue, that is strange indeed. So, there's a few things I would look at, mainly, connections. If you had any kind of corrosion in the connectors anywhere in the distribution path, this could cause similar response. The reason is that the EMF (voltage) may be slightly changing causing the meter to respond to the changes. I don't know too much about the Fluke 117, but I would imagine it is a high sensitivity meter with a fast response, slow readout type meter. In other words, any slight change may be reported as a signal, inconsistently. This happens because a digital meter can't respond fast enough for a quick change but can load it in memory and report it as the data is processed. In other words, there may be fast changes not being captured by the meter readout as they happen. Causes would vary; it could be the connections, your batteries in the truck changing value slightly, (again connections) or a circuit causing it either internal to the truck or the inverter. (Or the meter itself) Something is varying to get + and - readings.

One other thing to consider is that if you are running the meter in series with the inverter input voltage, (long leads) a phenomenon known as "body capacitance" could play a role. Body capacitance is when you have leads hanging during the measurement and anything changes with respect to that measurement plane. If you wave your hand over wires or move with respect to the measurement, while making low current measurements, or if the wires move, the change in capacitance can cause a change in the reading. (Too complex to explain why this phenomenon exists here, but it does.) In other words, the change is a false indication in the measurement. This is the primary reason a lot of low voltage and current meters will have shielded cables.

I don't know if you have access to an oscilloscope or analog meter, but this is probably the easiest way to see if the voltage is changing randomly or if it's caused by one of the other effects mentioned. The scope would show instantaneous changes in voltage. (Which would affect the current demand) An analog meter would as well, if the meter is sensitive enough with an AC input. (Such as a Simpson 260 with an "output" terminal.) You make the measurement on the low voltage scale with the 12V DC input on the output terminal. Sounds backwards, I know, but this is what that jack is good for. You would see instantaneous changes on the meters analog scale. The 12V DC would be blocked. (AC input (output terminal) is through a capacitor)

Not to say you don't have a component breaking down somewhere, but a scope would be the fastest way to see this. If you see the trace varying while measuring the DC voltage, disconnect suspects until it stops varying.

Here's an oscilloscope, basically a cheap "throw away after you use it" scope that will work just fine for this measurement. (May come in handy for other projects as well.)

Amazon.com

For reference only, here's what a Simpson 260 looks like.

Amazon.com

 

[BroncOh]

Well, I attempted making a measurement on mine, the clamp meter I have is too small for the cable size I have, so I will have to disconnect and hook up a smaller wire. Can't do that today or possibly tomorrow, the inverter is in the bed at the rear, we have one of those atmospheric river storms happening right now and I would fill the bed up in about 30 minutes or so. If we get a break in the storm, I'll try again. Don't want to get rain in the inverter.

Meantime, to address your -38ma issue, that is strange indeed. So, there's a few things I would look at, mainly, connections. If you had any kind of corrosion in the connectors anywhere in the distribution path, this could cause similar response. The reason is that the EMF (voltage) may be slightly changing causing the meter to respond to the changes. I don't know too much about the Fluke 117, but I would imagine it is a high sensitivity meter with a fast response, slow readout type meter. In other words, any slight change may be reported as a signal, inconsistently. This happens because a digital meter can't respond fast enough for a quick change but can load it in memory and report it as the data is processed. In other words, there may be fast changes not being captured by the meter readout as they happen. Causes would vary; it could be the connections, your batteries in the truck changing value slightly, (again connections) or a circuit causing it either internal to the truck or the inverter. (Or the meter itself) Something is varying to get + and - readings.

One other thing to consider is that if you are running the meter in series with the inverter input voltage, (long leads) a phenomenon known as "body capacitance" could play a role. Body capacitance is when you have leads hanging during the measurement and anything changes with respect to that measurement plane. If you wave your hand over wires or move with respect to the measurement, while making low current measurements, or if the wires move, the change in capacitance can cause a change in the reading. (Too complex to explain why this phenomenon exists here, but it does.) In other words, the change is a false indication in the measurement. This is the primary reason a lot of low voltage and current meters will have shielded cables.

I don't know if you have access to an oscilloscope or analog meter, but this is probably the easiest way to see if the voltage is changing randomly or if it's caused by one of the other effects mentioned. The scope would show instantaneous changes in voltage. (Which would affect the current demand) An analog meter would as well, if the meter is sensitive enough with an AC input. (Such as a Simpson 260 with an "output" terminal.) You make the measurement on the low voltage scale with the 12V DC input on the output terminal. Sounds backwards, I know, but this is what that jack is good for. You would see instantaneous changes on the meters analog scale. The 12V DC would be blocked. (AC input (output terminal) is through a capacitor)

Not to say you don't have a component breaking down somewhere, but a scope would be the fastest way to see this. If you see the trace varying while measuring the DC voltage, disconnect suspects until it stops varying.

Here's an oscilloscope, basically a cheap "throw away after you use it" scope that will work just fine for this measurement. (May come in handy for other projects as well.)

Amazon.com

For reference only, here's what a Simpson 260 looks like.

Amazon.com

Good info @Canyon Trekker appreciate it. I need to buy another scope; sold mine before moving as didn't want it to get damaged by movers. Had an old Tektronix.

If the inverter is fine, it leaves me to suspect the BJB under the hood, and I may just order one, as I can use all the extra fuses for spares from the factory BJB. Heck I may even pot the circuit board in the new BJB to fix the issue for good.

 

[BroncOh]

@Canyon Trekker I emailed Samlex as well, asking why my remote control panel displays a weird artifact when booting.

I also confirmed that disconnecting the inverter from the truck battery bus bar left my truck battery voltage at 12.5 volts in the morning, which is higher than it was before with it attached.

This screen goes away and the inverter still works as usual, outputting 120VAC.

 

[Canyon Trekker]

@Canyon Trekker I emailed Samlex as well, asking why my remote control panel displays a weird artifact when booting.

I also confirmed that disconnecting the inverter from the truck battery bus bar left my truck battery voltage at 12.5 volts in the morning, which is higher than it was before with it attached.

This screen goes away and the inverter still works as usual, outputting 120VAC.

Odd. I recall that when I first got my inverter installed, that remote box also had some odd behavior. I took it apart and found the "LAN" cable connector was causing it. I fixed it I can't remember if I did anything to the connector, but I replaced the cable with another CAT6 cable, (Shielded) and it was working OK the last time I used it. The original cable is very flimsy. I generally do not have it plugged in because I operate it from the bed access only. I made a portable box with the meter in it so I could run it to my RTT when attached.

I wonder if the remote box is energizing your inverter when you don't want it to.

 

[Canyon Trekker]

@Canyon Trekker I emailed Samlex as well, asking why my remote control panel displays a weird artifact when booting.

I also confirmed that disconnecting the inverter from the truck battery bus bar left my truck battery voltage at 12.5 volts in the morning, which is higher than it was before with it attached.

This screen goes away and the inverter still works as usual, outputting 120VAC.

FYI, Got a message back from Samlex. I know this is not 100% accurate, but the 2ma you were getting is reasonable in that the Samlex is always on, else it would not be able to respond to the remote operation command.

But here is what they said:

Thank you for reaching out to us on your inquiry. The unit has no current draw while in the OFF state, but there is a No load draw while it is ON. If your unit has a draw from your battery while in the OFF State even with the NTX-RC connected, there might be something wrong with your unit as such we can process it for warranty.

Kind regards,

Francis Natividad

Technical Specialist

103-4268 Lozells Ave

Burnaby, BC V5A 0C6

www.samlexamerica.com

www.samlexsolar.com

 

[RedZilla]

Bumping this thread up after installing my ARB twin this week. Got the bracket made and mounted my compressor to the side of the driver's side battery box. Install is clean and plenty of room for connections/wiring, etc. Hooked it up to upfitter #6 and fired up the compressor. Sounded good and pumped up the two gallon tank pretty quickly. Once it reached 145psi the compressor shut down and immediately I could hear air hissing from the area of the compressor. Soap mix produced no bubbles at any fittings and I could tell placing my hand around the housing that the air was coming from internal locations.

Well that's a shame and another black mark on ARB for producing pour quality products. This compressor has been in the box since I purchased it many months ago and it's clearly flawed in assembly and design. Guess I'll be pulling it apart and figuring out the seals since I'm unfortunately out of warranty on a brand new install. I doubt ARB will be any help anyway. Thankfully @BroncOh shared his experiences and solutions to date. I'll give an update when I get the problem solved. In the mean time I'll throw an in-line check valve into the system to keep the tank from leaking down. Otherwise I'm very pleased with the install and I'll be adding this project to a soon to be started build thread I'm composing for the forum in the near future.

 

[Churchkey]

I installed a check valve and replaced the poppet valve, which is where your compressor is probably leaking. Don't spend the money on the poppet valve. It was over $30 and didn't change a thing.

With the check valve installed my tank will hold air for a couple of days. Although, in the winter, not so much. I am assuming that the airline shrinkage creates loose seals and the system airs down.

 

[RedZilla]

I installed a check valve and replaced the poppet valve, which is where your compressor is probably leaking. Don't spend the money on the poppet valve. It was over $30 and didn't change a thing.

With the check valve installed my tank will hold air for a couple of days. Although, in the winter, not so much. I am assuming that the airline shrinkage creates loose seals and the system airs down.

I'm currently communicating with ARB, hopefully they will help but not holding my breath. I already have a check valve enroute to accomplish the work-around for now. I know I won't satisfied as the compressor will quick cycle with the check in place. But overall I'm pretty happy with the install of the overall system including air tank, air bags and regulator system in the bed of the truck.

 

[MiTremor906]

During our trip to glacier this summer our compressor wouldnt shutoff. Pulled over and turned the truck off, still ran. Had to pull the fuse to get it to stop.

Finally had time to rip it apart today and one of the relays was stuck closed. Might as well replace both of them well I'm in there.

Make sure you guys know of a quick way to turn yours off in case this happens to you, like I said I just pulled the fuses. It over pressurized the tank and the relief blew as I was pulling over but wasnt sure what that was in the moment and had a minor panic attack.