Charging and heat

NorthGuy · February 2013

I know that when I bulk charge batteries with the inverter/charger with 100A, the temperature of the batteries rises 1.5 to 2C per hour.

I noticed that when charging with similar 100A current from solar charge controllers, the temperature does not rise as much, it hardly rises at all. For some reason, same current causes much lower heat, and thus much better efficiency.

Why is the difference? The only thing that came to my mind is that the current produced by the inverter/charger may have ripple of significant magnitude, as if the charger rectified AC with diodes, but then didn't smooth it. The ripple would certainly increase heat losses compared to pure DC.

I don't have a scope to check the ripple, yet alone the scope that could read 10mV signal from the shunt. Is it common for charging current to have significant ripple?

inetdog · February 2013

Re: Charging and heat

NorthGuy wrote: »

I know that when I bulk charge batteries with the inverter/charger with 100A, the temperature of the batteries rises 1.5 to 2C per hour.

I noticed that when charging with similar 100A current from solar charge controllers, the temperature does not rise as much, it hardly rises at all. For some reason, same current causes much lower heat, and thus much better efficiency.

Why is the difference? The only thing that came to my mind is that the current produced by the inverter/charger may have ripple of significant magnitude, as if the charger rectified AC with diodes, but then didn't smooth it. The ripple would certainly increase heat losses compared to pure DC.

I don't have a scope to check the ripple, yet alone the scope that could read 10mV signal from the shunt. Is it common for charging current to have significant ripple?

I suspect that the problem is the duty cycle of the DC current rather than a voltage ripple (at least given that the ripple is filtered by the battery while the charger is in use.)
You would have to use a shunt or current probe to look at the current on the scope rather than the voltage. But just looking at the AC voltage reading on the shunt with a meter (using a series capacitor if your meter does not include one when on the ACV range) should give you a good idea.

If the inverter/charger or generator charger delivers only a 20% duty cycle (peak conduction only) then the heating for the same average (not RMS) current would be five times as much in terms of dissipation in the internal resistance of the battery as straight 100% duty cycle DC.
In this regard, either a PWM or MPPT controller fed from panel DC would be inherently current limited by the panels while in bulk, so the duty cycle could be 100%.

Do you notice a comparable difference in gassing when doing an EQ charge? Or has that not come up?

Blackcherry04 · February 2013

Re: Charging and heat

NorthGuy wrote: »

I know that when I bulk charge batteries with the inverter/charger with 100A, the temperature of the batteries rises 1.5 to 2C per hour.

I noticed that when charging with similar 100A current from solar charge controllers, the temperature does not rise as much, it hardly rises at all. For some reason, same current causes much lower heat, and thus much better efficiency.

Why is the difference? The only thing that came to my mind is that the current produced by the inverter/charger may have ripple of significant magnitude, as if the charger rectified AC with diodes, but then didn't smooth it. The ripple would certainly increase heat losses compared to pure DC.

I don't have a scope to check the ripple, yet alone the scope that could read 10mV signal from the shunt. Is it common for charging current to have significant ripple?

The question would be, is the Increase in heat indicating a greater acceptance of the current and a greater reaction in the battery ?? Is that a bad thing ??

NorthGuy · February 2013

Re: Charging and heat

inetdog wrote: »

But just looking at the AC voltage reading on the shunt with a meter (using a series capacitor if your meter does not include one when on the ACV range) should give you a good idea.

Thanks. I'll try that. My multimiter can measure AC and DC at the same time. Will do that next time I charge with inverter, although as things are going with solar, I would have to get a long stretch of cloudy weather before I can do that.

inetdog wrote: »

If the inverter/charger or generator charger delivers only a 20% duty cycle (peak conduction only) then the heating for the same average (not RMS) current would be five times as much in terms of dissipation in the internal resistance of the battery as straight 100% duty cycle DC.
In this regard, either a PWM or MPPT controller fed from panel DC would be inherently current limited by the panels while in bulk, so the duty cycle could be 100%.

I thing that close to full capacity, inverter/charger should have a duty cycle not far from 100% too.

inetdog wrote: »

Do you notice a comparable difference in gassing when doing an EQ charge? Or has that not come up?

No. I didn't notice any differences in gassing, and, for that matter, in heating during absorptions/equalizations.

inetdog · February 2013

Re: Charging and heat

NorthGuy wrote: »

I thing that close to full capacity, inverter/charger should have a duty cycle not far from 100% too.

Not if the charger drives its rectifier diodes from a transformer. That circuit can only deliver current to the battery when the transformer output is above the battery voltage plus the diode drop. That is what I was referring to when I mentioned "peak clipping". For a 14 volt AC transformer output, the instantaneous voltage will be below 14 volts for nearly 2/3 of the cycle.

But an inverter/charger may be using a DC to DC convertor instead, which could have a much better duty cycle and would have output filtering incorporated.

The term inverter/charger is a little ambiguous. In off-grid terminology it refers to a unit which can accept AC input and charge a battery or accept battery input and deliver AC. That is not at all the same as driving a conventional AC powered battery charger from AC produced by an inverter. Nor is it necessarily the same design as an RV power convertor whose primary purpose is to charge batteries from shore power, deliver shore power directly to loads, and also to provide AC (typically MSW) using the batteries when no shore power is present. It assumes that there will be some external means of charging the batteries when shore power is not there.
Units like the Xantrex use a circuit which can work bidirectionally rather than coupling a separate inverter and charger inside the same housing.

It might help to know what model inverter/charger you have.

NorthGuy · February 2013

Re: Charging and heat

Blackcherry04 wrote: »

The question would be, is the Increase in heat indicating a greater acceptance of the current and a greater reaction in the battery ??

Thanks BC. I didn't think of it that way. I think, if anything, the increase in heat would indicate worse acceptance of the current.

Blackcherry04 wrote: »

Is that a bad thing ??

I don't know. Just something new that I've found.

Blackcherry04 · February 2013

Re: Charging and heat

NorthGuy wrote: »

Thanks BC. I didn't think of it that way. I think, if anything, the increase in heat would indicate worse acceptance of the current.

I don't know. Just something new that I've found.

According to what I read, not necessarily, In your case it might be a good thing within some acceptable range. Thats about what I get when putting 250 amps into a 1100 amp hr bank @ 12V in Bulk.

"" As the battery accepts current, its internal temperature rises. The rise in temperature reduces the battery impedance, causing it to accept
more current. ""

"" As the temperature rises, electrochemical activity in a battery increases, so the charging voltage should be reduced to prevent overcharge. "" This statement was about charging limits.

NorthGuy · February 2013

Re: Charging and heat

inetdog wrote: »

Not if the charger drives its rectifier diodes from a transformer. That circuit can only deliver current to the battery when the transformer output is above the battery voltage plus the diode drop. That is what I was referring to when I mentioned "peak clipping". For a 14 volt AC transformer output, the instantaneous voltage will be below 14 volts for nearly 2/3 of the cycle.

I hope that's not how they do it. If you want 100A average with 20% duty cycle, then you need to output 500A for 3ms. If it's getting close to absorption, that would rise the battery voltage dramatically, so they would have to clip even less. Also, all the cabling etc. is not designed for 500A and there would be 5 times higher voltage drop than for 100A. May be some small chargers can work like this. I hope not XW.

inetdog wrote: »

But an inverter/charger may be using a DC to DC convertor instead, which could have a much better duty cycle and would have output filtering incorporated..

That's what I thought they were doing. If the converter is switching somewhere around 100kHz then smoothing the output should not be a problem even for 100A. So, I thought there was close to pure DC on the output.

I thought that the poor quality of the DC that is on the input to this DC to DC converter might cause some ripple.

inetdog wrote: »

It might help to know what model inverter/charger you have.

Xantrex (that is Shneider Electric) XW6048.

NorthGuy · February 2013

Re: Charging and heat

Blackcherry04 wrote: »

According to what I read, not necessarily, In your case it might be a good thing within some acceptable range. Thats about what I get when putting 250 amps into a 1100 amp hr bank @ 12V in Bulk.

"" As the battery accepts current, its internal temperature rises. The rise in temperature reduces the battery impedance, causing it to accept
more current. ""

"" As the temperature rises, electrochemical activity in a battery increases, so the charging voltage should be reduced to prevent overcharge. "" This statement was about charging limits.

I see. You're talking about the effect of heat after it has been produced. For me, most of the year that could be a good think, but when batteries are already at 35C, it probably isn't.

I'm mostly concerned about producing heat. We have two devices, which use the same parameters, and put the same amount of energy into the batteries. Inverter produces more heat, which means that more energy is spent on generating this heat, and consequently less energy is accepted by batteris. Solar chargers produce less heat, which means more energy accepted by batteries.

I'm thinking about adding more panels, and if I do they will be capable of producing a lot of current. From that viewpoint, the fact that solar controllers produce less heat is good for me, because even if they generate a lot of current, this may be Ok for batteries.

inetdog · February 2013

Re: Charging and heat

NorthGuy wrote: »

I hope that's not how they do it. If you want 100A average with 20% duty cycle, then you need to output 500A for 3ms.

So do I. :-)
The numbers were just an example and the important point is that voltage ripple, which is usually considered a figure of merit for a DC supply, is not the best thing to measure here since the battery being charged is such a non-linear load. The current ripple will be a lot higher percentage-wise (especially for a charger that does not use a series inductor for filtering) and will also be what directly affects the heating.
I would not expect a small amount of ripple to make a big difference either, unless something else is wrong which is producing an unexpectedly high ripple.

Blackcherry04 · February 2013

Re: Charging and heat

NorthGuy wrote: »

I see. You're talking about the effect of heat after it has been produced. For me, most of the year that could be a good think, but when batteries are already at 35C, it probably isn't.

I'm mostly concerned about producing heat. We have two devices, which use the same parameters, and put the same amount of energy into the batteries. Inverter produces more heat, which means that more energy is spent on generating this heat, and consequently less energy is accepted by batteris. Solar chargers produce less heat, which means more energy accepted by batteries.

I'm thinking about adding more panels, and if I do they will be capable of producing a lot of current. From that viewpoint, the fact that solar controllers produce less heat is good for me, because even if they generate a lot of current, this may be Ok for batteries.

You may be right, it's had to know if your measuring apples to apples. You'd have to have a very controlled test to ever know that. i do agree with you about adding more PV. My reason for the high current is economical, I charge with either a 12.5, 15 or 20 KW generator, I can cut the Bulk time in half and pay for a set batteries in 2 1/2 - 3 years on fuel savings alone. When I reach 14.2 volts I drop one of the charging sources and drop to 125 amps to absorb, but I only do a full charge once a week.

NorthGuy · February 2013

Re: Charging and heat

inetdog wrote: »

The numbers were just an example and the important point is that voltage ripple, which is usually considered a figure of merit for a DC supply, is not the best thing to measure here since the battery being charged is such a non-linear load. The current ripple will be a lot higher percentage-wise (especially for a charger that does not use a series inductor for filtering) and will also be what directly affects the heating.

Current ripple is certainly what I should look for.

However, if there's a current ripple, there has to be an associated voltage ripple. It definitely, will be much smaller in magnitude, but may be easier to detect and scope.

inetdog wrote: »

I would not expect a small amount of ripple to make a big difference either, unless something else is wrong which is producing an unexpectedly high ripple.

I made some guesstimates, and it appears that to explain heat differences, the current ripple has to be really dramatic. I wouldn't expect that from Xantrex.

May be there's something else, although I cannot think of anything.

Vic · February 2013

Re: Charging and heat

To me, the primary difference must be ripple current current from the XW interter's charger. Most quality TSW Inverters have a fairly large filter capacitor bank across the battery terminals. This is to reduce the ripple current to the battery bank. And it is customary for mfgs of Hybrid GT inverters to spec a minimum battery bank capacity, due to 60 Hz Split-Phase Ripple currents imposed on the battery bank. Believe that runing the XW inverter charger at/near its max charge current imposes a similar ripple into the battery bank.

N G, believe that your battery bank has a nominal Capacity of about 670 AH, so charging at 100 A for an extended time may be a bit above the Max suggested by Trojan (??).

Here, when charging from the largest generator, at about 10% of Capacity (about 130 A DC) for two hours, the temp rise is about 1.5 degrees C. This would be from about 50 V battery V at that current, to a point where gassing just begins -- about 56.5 V (temp compensated). This represents a bit lower charge rate than in your example.

Most of the Shunts that are included or available for systems of N G's size are 500A + 50 Mv, so this will be a small voltage at the shunt terminals.

One other thought is, if the charging from the XW inverter was from the Genset. In this case, perhaps the AC waveform from the genset is not that great, and that results in additional ripple current. Are you tired of folks like me PICKING on your Generac ??

The only other thought is that the test for the different temp rise for the XW Inverter's Charger vs the that from PV is that the battery bank should really be recovering from the same SOC in each case, as the battery impeadance does increase as Bulk progresses. There is some gassing before Vabs is reached (ie still in Bulk) in the case of many battery chemistries. This results in efficiency reduction and more heat developed in the battery.

Some will argue that both the XW Inverter AND the XW MPPT CC generate ripple current, but believe that the PV charger is a much more pure DC source (even if it is a Schneider XW SCC).

You are paying close attention to your battery bank and the rest of the system. This is good. These small differences in temp rise probably make NO real difference in battery life or system performance. Opinions. Vic

NorthGuy · February 2013

Re: Charging and heat

Hi Vic,

Vic wrote: »

And it is customary for mfgs of Hybrid GT inverters to spec a minimum battery bank capacity, due to 60 Hz Split-Phase Ripple currents imposed on the battery bank.

I think this refers mostly to inverting process.

For charging, I hope they have a choke and/or capacitor filter to smooth the ripple. Up to recently, I thought it was close to pure DC.

Vic wrote: »

N G, believe that your battery bank has a nominal Capacity of about 670 AH, so charging at 100 A for an extended time may be a bit above the Max suggested by Trojan (??).

I looked all over and I couldn't find any suggestions from Trojan about this. During our conversations, I told them many times that I use 100A XW charger, and they thought this was good.

Vic wrote: »

Here, when charging from the largest generator, at about 10% of Capacity (about 130 A DC) for two hours, the temp rise is about 1.5 degrees C.

Given the charge rate difference (14% vs 10%) and assuming similar batteries, I should get about twice as much heat - that is 3C in two hours. This is what I'm getting with XW. But you have SW, which is essentially the same. So, we should have similar ripple characteristics.

Do you get any heating with solar charging?

Vic wrote: »

One other thought is, if the charging from the XW inverter was from the Genset. In this case, perhaps the AC waveform from the genset is not that great, and that results in additional ripple current. Are you tired of folks like me PICKING on your Generac ??

No. Not yet

I don't think Generac has problems with waveform. And you do get the same heat without Generac

Vic wrote: »

These small differences in temp rise probably make NO real difference in battery life or system performance.

Based on heat capacity of lead and assuming acid has the same heat capacity as water, I estimated my heat loss to be around 300-500W. With 150 hours of genearor runs a year, this is 45 to 75 kWh/year. It's 1 to 1.5% of the consumption.

Cariboocoot · February 2013

Re: Charging and heat

Quick question:
Is there an RTS on the SCC and on the I-C?

NorthGuy · February 2013

Re: Charging and heat

Cariboocoot wrote: »

Quick question:
Is there an RTS on the SCC and on the I-C?

Yes, of course.

BB. · February 2013

Re: Charging and heat

I would bet most battery charges do have a 120 Hz ripple current that follows the I squared current that all 60 Hz inverters do... If they do not bother with DC filtering/intermediate energy storage for the inverters, then they are (most likely) not going to do it for the charging section (or simply running energy "backwards" through the inverter circuits) either.

If you have a "cheap" DMM, you can just put it on AC scale. Most of the inexpensive digital meters just have a capacitor to block the DC current and will read the AC ripple voltage on the battery bank. (if the meter reads around 12-15 volts on the AC scale when connected to a battery bank, then it will not read the AC section of the ripple voltage as I suggest trying here--True RMS meters will usually read 12 volts on the AC scale--the read both the DC offset and the AC ripple--which is the "correct" way to read the voltage for most applications--RMS includes both the DC and AC components of power).

-Bill

westbranch · February 2013

Re: Charging and heat

Bill your second paragraph confused me, so I took my cheap dmm and got .5v DC an the AC scale, on a 12V battery (disconnected). Is that correct?

NorthGuy · February 2013

Re: Charging and heat

I went ahead and measured it. My Fluke lets me measure AC and DC at the same time. It aso has a "Peak" mode - records min and max measuring 250 us intervals.

The charging current has been set in XW to 90A. XW reads 88.9A.

Voltage measurements of battery terminals:
DC: 54V
AC: 1V
min: 51V
max: 57V

Current measurement at the shunt:
DC: 90A
AC: 70A
min: -3A (that's minus 3!)
max: 200A

Fluke measures at 4kHz, so this cannot be related to high frequency switching. Looks like 120Hz ripple. Looks like no attempt has been made to smooth the ripple coming out of the rectifier.

Interesting, they advertise it as power-factor corrected charger. This doesn't look like power factor corrected to me.

BB. · February 2013

Re: Charging and heat

westbranch wrote: »

Bill your second paragraph confused me, so I took my cheap dmm and got .5v DC an the AC scale, on a 12V battery (disconnected). Is that correct?

I would have hoped that on a disconnected battery, that a simple DMM would read 0.0 Volts AC (set to 20 VAC full scale--You could so similar by adding a medium size blocking capacitor between the meter and the battery and then do the AC measurements). Any ripple current would be obvious from the chargers/inverters.

Remember that some "de-sulphators" appear to use high peak current and high edge rates (square wave/impulse wave) to help reduce sulfphation. So, in some circles, a pure DC charging current would not be the "ideal".

-Bill

BB. · February 2013

Re: Charging and heat

NorthGuy,

That 1 volt AC reading during heavy charging/discharging (via AC battery charger/Inverter) does not sound unreasonable.

One of the papers I read somewhere said that setting "float voltage" needed to be at 12.7 volts + ripple voltage to make sure you do not "micro cycle" the battery bank in the middle of the day when running loads with high ripple current (such as AC inverters).

-Bill

Vic · February 2013

Re: Charging and heat

N G,

Am fairly certain that there is a filter capacitor bank directly across the battery terminals of the XW Inverter/Charger.

On the SW+, the capacity is in the order of about 70,000 uf. Perhaps you have noticed that when the XW inverter has been disconnected from the battery for minutes to hours, that when the main DC circuit breaker is thrown to ON, there is a nice, loud "SPUTtt) noise? If so, this is that cap bank charging, and it is there to do SOME filtering of the DC coming and going. Batteries are fairly good capacitors, but thick plated Deep Cycle batts have a relatively high impeadance.

Opinions, Vic

NorthGuy · February 2013

Re: Charging and heat

Vic wrote: »

Am fairly certain that there is a filter capacitor bank directly across the battery terminals of the XW Inverter/Charger.

I don't think so. 70A rms AC ripple during 90A DC charging current is approximately what you would expect if there was no filtering at all. If there was a filter, the ripple would me much smaller and current wouldn't wonder between -3A and 200A.

It is possible that they had a filter in SW then eliminated it in XW as a cost saving measure.

Vic wrote: »

On the SW+, the capacity is in the order of about 70,000 uf. Perhaps you have noticed that when the XW inverter has been disconnected from the battery for minutes to hours, that when the main DC circuit breaker is thrown to ON, there is a nice, loud "SPUTtt) noise? If so, this is that cap bank charging, and it is there to do SOME filtering of the DC coming and going. Batteries are fairly good capacitors, but thick plated Deep Cycle batts have a relatively high impeadance.

I've never tried to turn it off. I can't recall how it sounded when I turned it on the first time.

BB. · February 2013

Re: Charging and heat

It is very difficult to put large capacitors in parallel with the battery bank and have the capacitors do much of anything.

Capacitors work by storing/delivering energy as a change of voltage (E=1/2 * cV^2, if I remember correctly).

If you could put enough capacitors to get zero ripple voltage, then by looking at the above equation, they have to be doing "zero" work.

Nominally, you would need to add a series inductor, capacitor, another series inductor ("T" filter) or similar... You need to have a storage element between the battery and the AC inverter (or battery charger) to effectively reduce the AC ripple current.

There are other methods, such as used by PFC (power factor correction circuits) where there is a DC to AC high frequency switcher which charges an intermediate capacitor bank, when then supplies energy to the main AC Inverter switching circuits. The high voltage capacitor bank is what will have the ripple current (added expense, possible issues with life--battery ripple vs capacitor ripple voltage/current).

-Bill

Vic · February 2013

Re: Charging and heat

Bill,

Have never seen the overall schematic of any of the quality Inverter/Chargers. After puling the cover from one of the SW+ 5548's, noticed about 11 ea 7,000 uf filter caps. Did not see, but did not look for any inductors. Capacitors are expensive, and subject to age effects etc. Am familiar with Pi, L and C filters ... dunno what Xantrex was trying to do with the Cs mentioned, but when switching on the main breakers on the 5548s, the breaker contacts make that SPUTt noise of a largish C bank. I don't know and have not been told. Somewhere, have a pic or two of the 5548 guts, but mostly one just sees a miriad of Flat-Ribbon cables ... Vic

Cariboocoot · February 2013

Re: Charging and heat

Confirming Vic's info, some of us who are old, brave, and foolish have hit the inverter input wires to the battery and experienced that big knock of large capacitors taking on juice - complete with sparks! Touch the wire again and the zap is gone because the caps are charged.

But as Bill said, probably not much done for ripple filtering anyway as the batteries don't really need it. The small temp increase difference observed isn't anything deadly.

Even so, might want to dial that charge rate back from 14.9% if it is sustained there long enough to cause noticeable heating.

BB. · February 2013

Re: Charging and heat

The resistance/inductance of the incoming power lines, in this case, are actually a bit of a help to reduce ripple current at the battery bus (and "noisier" at the input to the inverter).

The inductance of the cabling and the capacitors at the inverter can cause oscillation issues (makes a "tank circuit") may cause big issues. That is why, it is almost always better to keep the inductance of the power cables at a minimum (instead of trying to make it part of a filter circuit).

-Bill

mike95490 · February 2013

Re: Charging and heat

The XW inverters have a fault code for Capacitor Overheat, which is a honking big cap on the battery buss

NorthGuy · February 2013

Re: Charging and heat

mike95490 wrote: »

The XW inverters have a fault code for Capacitor Overheat, which is a honking big cap on the battery buss

Found it. F45. Recommends reducing AC loads.

May be something wrong with the capacitor on mine. Do you know where the capacitor is located physically. Is it behind the big door on the right side which says "No user serviceable parts inside"?

That's quite strange. I sent my numbers to Xantrex support and they told me that such ripple is normal:

"The current ripple it is expected - any single phase charger will have this behavior.
The charge current has a ripple of 2x line frequency (e.g. 2x 60Hz) and this is perfectly normal.
The system operates just fine."

Vic · February 2013

Re: Charging and heat

N G,

Personally, would not worry about your situation. Nothing is perfect, and as much as I DISLIKE Schneider, the XW is probably a fine inverter. Most of the bugs and issues that folks here report are when using it connected to the Grid. At some point, you have mentioned that you are exiting the grid.

YMMV, Good Luck, have a brew on me. Vic

boB · February 2013

Re: Charging and heat

NorthGuy wrote: »

Interesting, they advertise it as power-factor corrected charger. This doesn't look like power factor corrected to me.

120 Hz half sine current ripple is exactly what I would expect to see at your battery shunt with a PFC charger.
It tries to draw a sine wave of current from the AC input.

The charge controllers will most likely apply a fairly steady DC current to the batteries. After all, PV is DC (if that's what you're using)

boB

Charging and heat

Comments

Categories