Amperage Draw Shuts off Generators

[live4fun]

Hello my friends, I have a TM2720 with 2 x 100 ah LiFePO batteries, Victron BMS & solar controller, Wildkat 4655 Main Board Converter, 300W PV panels on rooftop. I have a 1000w (7.2a) and a 2000w (13.5a) inverter generators. I carry either one in case we need the extra power which is very rare. The battery and solar systems are working great for over a year. I can be power independent for 4-5 days in heavily shaded forest areas. All is good and proven road worthy.

Here is the issue. I exercise my generators every couple months for other needs I have. This summer I turned off the TM solar system for a month after a trip and decided to charge the battery system with one of my generator to see how it worked. The amp load immediately shutdown the 1000w generator and did the same for the 2000w. I hooked it to 120v A/C power and everything charged right up.

My electrician happened to be here for other reason and he thought the charge converter (Wildkat) may have some setting to lower allow me to modulate the amperage draw input. After researching and speaking to Best Converter.com, they did not know of such setting on the converter.

So I recorded the initial amperage draw to be between 24-27amps. Certainly enough to shutdown my generators after 15 seconds. I have attached my BMS recording of the data 1. No input amperage, 2. & 3. Showing my 2000w generator trying to keep up, 4. My battery/solar system compartment.

What do people think is going on and how to you think I can remedy this situation in the future. Thanks in advance, for all the help. Kurt

[Bill]

For your system, there are others much more competent than I am, so I will leave the details and final answer to them. My first reaction, though, is that you say you shut down the solar system for a month, leaving no source of charge current. Like most RVs, a TM has a number of parasitic loads - devices which continue to draw small but non-zero current even when everything is nominally OFF. The sum of these currents will draw a Group 27 lead-acid battery to zero in less than a month. Perhaps it pulled your lithiums down pretty far, so they were hungry when you tried to charge them with the genny.

A month is 720 hours, so a discharge current of 0.2 amps amounts to 144 AH.

Were the TM shells up or down during the disconnect period?

Can you charge the lithiums, then measure the discharge current with no charging going on?

Bill

[Bill]

One other thought. When you hooked up your generators to the TM, did you plug the TM into the 120VAC output of the generators, or did you use the Battery Charging output of the generators? Most generators have a 12VDC Battery Charging port, but it is good for only 8 amps. A 27 amp load would certainly overwhelm it.

Bill

[rickst29]

The WildKat, a slightly modified PD converter "main board assembly", cannot be programmed to limit either it's input or output current to a value below its design limits. It will always try to "dish out" whatever the battery string (ansd other 12v loads) are willing to absorb, up to a limit of about 60A. I measured the maximum continuos capability of my own unit at 63A during an initial test when it first arrived.

However, the continuous input draw required to provide 60A output @ 14.2 volts is only about 1070 watts. (60A * 14.2 volts = 852 watts, with internal power losses at 80% efficiency increasing input power rating to about 1065 watts.)

But your Victron shows only 389 watts delivered at 14.26 volts (27.3A), shortly before even the "2000 watt" generator bails out with an overload error. That continuous output power rate SHOULD be pulling only about 500 watts (continuous) from the input 120-VAC side. The WildKat can pull much higher current to charge internal capacitors for a short period of time, but that that won't take very long. But I don't know the capacitor sizes and the maximum current which could be drawn, so this is a SWAG.
- - -
If the 120-VAC wiring between the generators and the WildKat is all in excellent condition, without short circuits leaking power away and also with no high-resistance problems at connectors or the circuit breaker, then the problem will be internal to the WildKat - most likely a failed capacitor.

But the WildKat 120-VAC input "hot" wire SHOULD BE on a circuit breaker limited to 20A (or, even better, just 15A): Your "27A" input current measurement should have tripped the breaker, if it continued for even a single second of duration. Because the breaker did not trip, AND because power from the rI suspect 2/3 of the measured input amperage is being lost in a ground fault, before reaching the WildKat.

Can charging run for more than a few second from the 30A cord with a household/grid connection? That test can be performed on a 20A circuit, but it should be on a standard breaker - not a GFCI.
- - -
You could also reduce the maximum output power delivery of the WildKat by changing the battery type DIP switch to "GEL" (that switch position is also used for AGM batteries), enforcing a maximum output voltage limit of 14.0 volts (instead of allowing more than 14.2, as indicated in your 3rd picture). But that change would only be effective for "power limiting" when the WildKat was delivering nearly 60 Amps, and you're nowhere near that level of current.

[rickst29]

Originally Posted by Bill

For your system, there are others much more competent than I am, so I will leave the details and final answer to them. My first reaction, though, is that you say you shut down the solar system for a month, leaving no source of charge current. Like most RVs, a TM has a number of parasitic loads - devices which continue to draw small but non-zero current even when everything is nominally OFF. The sum of these currents will draw a Group 27 lead-acid battery to zero in less than a month. Perhaps it pulled your lithiums down pretty far, so they were hungry when you tried to charge them with the genny?

Bill

Good question, but his "pre-generator" picture shows "12V" voltage of 13.12 volts in a state of slight discharge. That's between 60% and 70% State of Charge for LFP batteries at normal temperatures.

In spite of moderatately high SOC, however, LFP batteries ARE willing accept very high input current and power with only a small voltage diiferential, because they have low internal resistance. If you give them a good voltage differential, they're ALWAYS "hungry".

Internal resistance of the cells is utterly inadequate to protect them from an excessive rate of charge, a protective limit needs to be enforced by external BMS unit. The BMS units in the '100Ah" batteries should allow at least 25A into each battery pack, and a 50A value would be appropriate for some of the better kinds of cells (EVE cells, for example). If each pack allows 50A, the 'bulk' charging limit will be the 60A limit of the WildKat.

[Bill]

Originally Posted by rickst29

But the WildKat 120-VAC input "hot" wire SHOULD BE on a circuit breaker limited to 20A (or, even better, just 15A): Your "27A" input current measurement should have tripped the breaker, if it continued for even a single second of duration. Because the breaker did not trip, AND because power from the rI suspect 2/3 of the measured input amperage is being lost in a ground fault, before reaching the WildKat.

Be real careful of that underlined statement, as it can draw you to some wrong conclusions. Circuit breakers don't trip instantly upon overload. They have what is called a time-current curve. It shows amount of overload on the horizontal axis (shown as a multiple of the breaker's rated current), and time-to-trip on the vertical axis. They show that trip time gets shorter as the overload gets larger. Makes sense, right? At rated current (shown as 1x on the horizontal axis), it will never trip. At 2x overload, most household-type breakers take many seconds to trip. This is what enables the breaker to withstand a momentary overload, such as a motor-start surge. At 10X overload, trip time will be on the order of 1 second. And at 100x, a small fraction of a second. In this thread, we are talking about less than 2x overload, so trip time will be long, and other protective circuits, in the genny for example, may kick in first.

The attached picture shows a generic time-current curve. It shows that at 6x overload, this particular breaker will trip in 0.6 seconds. At 2x overload, the trip time is above the chart, but looks like maybe 10 seconds. Again, this is generic. As you can imagine, different kinds of breakers have different trip characteristics, which are set when the breaker is designed.

Bill

[live4fun]

Okay, i have to slow down and visualize what you’re saying. In the meantime, I charged the LiFePO batteries on my 20a. GFCI protected outlet. Here are the results:
Duration(min) SOC V Amps W
0 83% 13.1 -.16
1 83% 14.2v 26.4 376
2 83% 14.2v 24.5 350
4 83% 14.3v 22.8 326
6 84% 14.3v 22.4 322
9 84% 14.3v 21.4 306
120 99% 14.4v 10.3 149

So I find it interesting that circuit breaker didn’t trip nor did the GFCI which can be sensitive on this 120v circuit over 20a? I’m trying to determine what my next step should be to remedy this issue. Do i need to be charging only on 30amp 120v? Is there a way to regulate the input current at all? I did change the Wildkat dip switch from AGM to Li based on Best Converters recommendations and didn’t see any improvement in the input amperage side but I didn’t have the BMS working at the time but I know the AGM switch also shut down my generators? I appreciate the discussion but are trying to extrapolate the nuggets of information that will help me fix the situation. I am grateful for the information. Thanks

[rickst29]

On review, I agree strongly: A 'proper' circuit breaker at 15A might not trip from 27A current before the generator determines an overload situation, and a 20A circuit breaker would take substantially longer time to trip out. These are not like "fast acting class T" fuses.

[rickst29]

Originally Posted by live4fun

Okay, i have to slow down and visualize what you’re saying. In the meantime, I charged the LiFePO batteries on my 20a. GFCI protected outlet. Here are the results:
Duration(min) SOC V Amps W
0 83% 13.1 -.16
1 83% 14.2v 26.4 376
2 83% 14.2v 24.5 350
4 83% 14.3v 22.8 326
6 84% 14.3v 22.4 322
9 84% 14.3v 21.4 306
120 99% 14.4v 10.3 149

So I find it interesting that circuit breaker didn’t trip nor did the GFCI which can be sensitive on this 120v circuit over 20a? I’m trying to determine what my next step should be to remedy this issue. Do i need to be charging only on 30amp 120v? Is there a way to regulate the input current at all? I did change the Wildkat dip switch from AGM to Li based on Best Converters recommendations and didn’t see any improvement in the input amperage side but I didn’t have the BMS working at the time but I know the AGM switch also shut down my generators? I appreciate the discussion but are trying to extrapolate the nuggets of information that will help me fix the situation. I am grateful for the information. Thanks

Per above, DO NOT run the WildKat in "Lithium" mode unless you want it to run all the way up to14.6 volts (14.4 in newer models), quitting only when the BMS terminates charging.

The "GEL" setting imposes a lower maximum, charging voltage of 14.0 volts, which will provide less voltage differential at the battery terminals. Pushed less hard, they will absorb power more slowly. The "GEL" setting has an undesirable side effect of also reducing the float voltage and "boost return" voltage setting, you will probably need to use the pedant in order push the pack voltage high enough to equalize the cells. If the "GEL" setting caused your generator to quit, it was because the 120-VAC load on the generator vanished - AND your generator is set to shutdown at low load. The generator knows nothing about the battery charge process, it only sees draw from the Converter and othe r120-VAC appliances.

You don't want to stay at 99% SOC during storage anyway, and your solar can top it off during trips. When solar isn't available and you need to use a generator, can can either use the Wizard mode switch to "force" bulk charging, or you can let it let it terminate "bulk charging" at lower voltage (50-70% SOC) and then let solar top it off.

[rickst29]

then you will need to replace the WildKat with a lower-powered Converter, and you might want to use a programmable Victron for that job. IIRC, the cheaper Victron "smart" converters go up to either 30 or 35A maximum, and the efficiency is also higher (reducing 120-VAC input requirements). They don't need fans, but don't exactly fit the WFCO base box either. I sent a PM with my phone number, please call for details about your replacement options.