LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[OceanSeaSpray]

Originally Posted by rvlandlubber

The use of latching relays for load/ battery disconnect function seems to be preferred over conventional normally open relays with a lot of the systems mentioned here.

From the fail-safe design viewpoint these devices can be problematic.

I would be interested to know what the thinking is when deciding to use them.

Is the low power consumption the most important factor?

How are users managing the possible situation where the latching relay is faulty and won't change state?

Thanks in advance.

When you look into it very closely, the other type is in fact even more problematic because the coil switching relies on power MOSFET transistors and those virtually ALWAYS fail short-circuited.
This means that the contactor can be closed as it should, with a failed transistor, and there is no way of telling that it is impossible to open it any more...

With fail-safe transistors, this is essentially taken care of, because if one transistor fails, the other one can still open the circuit and the BMS won't close it again afterwards as it will see the failure.

A latching control arrangement allows the hardware to self-check after each transition and it offers the same redundancy level.

So, at this point, latching or non-latching is a decision that needs to be made case-by-case. A latching relay cannot be released without power, but if there is no battery power, there is not much to disconnect. It is important to be very clear about what you are trying to protect against and look at the whole system, not just the component.

I can control standard contactors just fine, the unit detects which type is connected and deals with it. In some cases, it can make sense to hold the circuits closed. For me, the unwanted power draw of simple contactors is a big negative, because the installation is on 24/7.

[thomasow]

Originally Posted by SV THIRD DAY

One thing to keep in mind here is that we are talking about a battery and control system for a Cruising Boat, not a remote unmanned cell tower. You can tie yourself up in knots with the assumption that your bank and control system needs to function completely without human interaction, monitoring, and oversight. With any system, if you think you will be able to "set it and forget it" like the Ronco Rotisserie Chicken Cooker, then you could have a surprise coming....not just with your battery bank/control system but with life aboard a cruising boat in general.

You know Rich, I think you are right – any automation approach will not always get it right. And there will always be exceptions. Example with our lifestyle: during the few months we come into port for winter would be great to have the batteries in LifeExtender mode (SOC <50%). However if we know bad weather is approaching, and the likelihood of power outages, I would want to top off those batteries using shore power the day before.

Another example. While we are out for most of the year, if I know we are going to be heading into port, say for a monthly heavy-provisioning, I would be just as happy to leave the batteries at a rather lower SOC while underway - saving that upwards of 0.5GPH burden the large house alternator adds to the main engine.

And should recognize a fully manual-only control is also problematic, people forget. Does seem to me a good faith effort towards automation, combined with an ability to override is one answer.

[missourisailor]

Originally Posted by OceanSeaSpray

All latching relays need to be pulsed indeed and it is what I do with the Tyco. The BlueSea 7700 needs to be pulsed on the positive side with a lot of current. At the moment, it wouldn't be compatible, because I switch the negative side, which is a lot more efficient from an electronic point of view.

Changing the design around could address that, but the drivers for these outputs include monitored redundant transistors to be fail-safe and there is a bit in it.

I prefer the Tyco unit, which is industrial/automotive instead of marine-branded, can be wired on the high or low side, and also costs less than half the price - unsurprisingly.
The BlueSea is also a "violent" device to control, with a much higher inrush. I don't think it was quite intended for doing this kind of thing, i.e. electronic control.

Still, interesting query, thanks.

Glad to hear that someone is designing/building a BMS solely for marine. And now we don't have to use a modified automotive and/or off grid BMS.
Once you get out of Beta testing, I'd like to buy one of your first production units.

As to the BS 7700's, I'm currently, slowly modifying the ships wiring for the conversion to LiFePo batt's. The two BS 7700's are already in place. I knew from the onset that they would need special driver circuits to control them if using one of the "off the shelf" BMS's.

Have a guy building a few surface mounted "one shots" to drive master relays. Which in turn will control the 7700's. So in using your BMS, it looks like I'll be able to eliminate the "one shots". But would still need to retain the master relays.

[OceanSeaSpray]

You are right, controlling BS 7700s will require something like four small interposing relays to switch the positive when the BMS switches to ground.
I will make the pulse timings configurable to allow accommodating different latching relays and situations like those.
I would still rather use the Tyco BDS-A disconnectors, more reliable because driven directly, but you will certainly be able to make it work if you want to.

The boards shipped yesterday, they might turn up here within a week or so. A fair fraction of the components has arrived and I have just started laying out the foundation of the code.
Some documentation and a configuration utility will be needed before it becomes reasonable to ship a few units.

I will be hauling out and painting within the next 3-4 weeks and this is going to create a diversion.

[exMaggieDrum]

Originally Posted by missourisailor

Glad to hear that someone is designing/building a BMS solely for marine. And now we don't have to use a modified automotive and/or off grid BMS.
Once you get out of Beta testing, I'd like to buy one of your first production units.

As to the BS 7700's, I'm currently, slowly modifying the ships wiring for the conversion to LiFePo batt's. The two BS 7700's are already in place. I knew from the onset that they would need special driver circuits to control them if using one of the "off the shelf" BMS's.

Have a guy building a few surface mounted "one shots" to drive master relays. Which in turn will control the 7700's. So in using your BMS, it looks like I'll be able to eliminate the "one shots". But would still need to retain the master relays.

Could you describe your "one shots". I have to plead ignorance on the terminology. I was thinking that I could use an automotive style NO/NC relay to use the BMS going to ground to drive the BS 7700 to positive. I must be missing something (and probably am). Why you would need four relays, as OceanSpray says, is really confusing to me.

Thanks.

[OceanSeaSpray]

Originally Posted by exMaggieDrum

Could you describe your "one shots". I have to plead ignorance on the terminology. I was thinking that I could use an automotive style NO/NC relay to use the BMS going to ground to drive the BS 7700 to positive. I must be missing something (and probably am). Why you would need four relays, as OceanSpray says, is really confusing to me.

Thanks.

In order to drive a latching relay, you need two outputs instead of one: one to latch it on, one to latch it off, using short pulses. So you would need to add a small automotive relay to each one to convert the pulses to positive.

With two latching relays, for the load bus and charge bus, it makes it four interfacing relays.

Alternatives would be driving on the high side from the BMS, or making a solid-state add-on module to perform the conversion. I would favour the latter, because driving on the high side is less efficient, more costly and requires more board space.
Of course, if nearly everybody wanted to use BS 7700s, then there could be a case for converting all the power outputs to positive, but it seems to make no economical sense to me for the majority of installations.

A solid-state add-on interface module would be quite trivial to make, so it is a very real possibility.

[missourisailor]

Originally Posted by exMaggieDrum

Could you describe your "one shots". I have to plead ignorance on the terminology. I was thinking that I could use an automotive style NO/NC relay to use the BMS going to ground to drive the BS 7700 to positive. I must be missing something (and probably am). Why you would need four relays, as OceanSpray says, is really confusing to me.

Thanks.

exmaggiedrum:
As Oceanseaspray stated, the BS7700's need a control pulse to change states. One pulse for each state. BS tech tells me that the 7700 can not tolerate any long duration charge. Hence a pulse is needed.

Most current BMS's upon an event change, just either output a current or supplies a ground to control the OV and UV disconnects. Once the event is done, the BMS just returns the output to the normal charge (+or-).
This will not work with the BS 7700.

So you need to modify the output to a pulse. A "one shot" does this. It outputs a single pulse for so a set length of time, upon a change in the control circuit polarity. Thus opening the contactor.

However the circuit has to be smart enough so that when the control circuit changes back to the normal polarity, another "one shot" fires a pulse, to energize the close side of the contactor.

To the best of my knowledge, no "relay only" circuit will do this. It requires solid state logic.

This is why I have a guy who is trying to modifying some of his off the shelf components to meet my needs. Also why I'm excited about Oceanseaspray's new BMS. It should remove at least one headache from my life.

[SV THIRD DAY]

I've been talking to Mark Grasser over at DC Power Solutions who had been into LiFePO4 batteries for years. He is working on a Plug/Play LiFePO4 battery with BMS system tailored for the boating world with inputs for Solar, Wind, Alternator, Battery Charger. He's the one who has given me and several other cruisers guidance on our LiFePO4 systems so it will be interesting to see what he puts out.

A catamaran came through Morro Bay a week ago to pick up a water maker and while aboard they were telling me about their LiFePO4 system and I'm seeing more and more "real life cruisers" going with the technology. They went with a drop in system (Battery and BMS) from Elite Power Solutions and couldn't be happier. It's just a matter of time...lead is dead....

[mbartosch]

Regarding the discussion about latching relays: some months ago I decided I wanted to replace the charge bus relay with a latching relay. In addition I also intend to replace the all-in-one relay in our portable dinghy battery bank with a latching relay which will substantially reduce idle consumption of the pack.

Hence I have designed a discrete circuit to do exactly this. You can find the blog post including schematics and a demo video here:

Saving Energy: Discrete edge detector for driving latching relays

[OceanSeaSpray]

Very nice. I would just do a little more to make sure the output MOSFET can't take a hit switching into the coil. The NCR relay is a friendly device to drive, but others are not.
If the transistor gets nuked, it also burns the relay coil out.

[mbartosch]

Originally Posted by OceanSeaSpray

Very nice. I would just do a little more to make sure the output MOSFET can't take a hit switching into the coil. The NCR relay is a friendly device to drive, but others are not.
If the transistor gets nuked, it also burns the relay coil out.

Good point, a 24 V Z Diode or a varistor protecting the MOSFET DS might be a good idea for larger relay coils. Better safe than sorry.
I will probably update the schematics accordingly, thanks a lot for pointing this out!


Sent from my iPhone using Cruisers Sailing Forum

[OceanSeaSpray]

Originally Posted by mbartosch

Good point, a 24 V Z Diode or a varistor protecting the MOSFET DS might be a good idea for larger relay coils. Better safe than sorry.

Yes, this kind of stuff. The spike can also travel through the drain-gate parasitic capacitance and even if you clamp VDS to 24V, it could be several times over the maximum allowable VGS rating, so it can be a good idea to also add a small Zener between G and S, unless it is built into the FET already.
Another way is making the FET turn back on very briefly to short the spike out.

I often find that, between the basic functional circuit and its final hardened, industrial strength version, the parts count can climb up quite significantly. A lot of these components do nothing unless the circuit is wired wrong or something else unexpected happens, but it is where robustness and long-term reliability come from.

[JBPohle]

I am wrestling with a bit of the same questions as OceanSeaSpray and Entropy. I've used the HousePower BMS and terminal (E) of the BMS grounds the relay unless a protection event is triggered. I really like Stan Honeys approach using a P-Fet to provide +12 Volts to a BlueSeas ML-RBS 7713 battery switch. Unfortunately I've got a 24V bank and it looks like this voltage exceeds the Vgs of most, if not all, P-Fet's. Do you think a voltage divider (correct term?) with a resister between terminal E and the P-Fet gate that is the same as the resister between the gate and source as I've shown on the attached sketch would work?

Second question, while Stan Honey did use it without diodes, would a couple of diodes as shown to the right in the picture help protect the P-Fet from the relay when the P-fet switches off? If so what size diodes do think would work. All of these little parts are inexpensive, so why not put some protection in place?

Please understand I was a structural engineer, not electrical, but I've really found this project has peaked my interest so I am learning. I agree with Entropy that the solid state switch is the way to go vs a small relay. Over time the relays consume 24/7 a remarkable amount of energy when compared to this solution.

Any help would be appreciated.

Warmest Regards
John

[OceanSeaSpray]

John,

Of course, you can use a divider to keep VGS within specs. To be completely safe, it is a good idea to also clamp it.

Something like this?

Regards,

Eric

[pacific_voyager]

I have been unable to find the field descriptions for the CellLog 8s output.
The default string looks something like this, with fields separated by semi-colons. I have it mostly figured out. Here is a sample from a 6-cell bank of 2V AGM cells (not LiFePO4 yet).

$1;1;;2064;2068;2071;2066;2074;2056;0;0;12399;30

Fields 1-3: ??
Fields 4-11 are cell voltages in millivolts. Only the first 6 have values.
Field 12 is the pack voltage as reported by the CellLog 8s.
Field 13 is the difference between the pack voltage and the sum of Fields 4-11.

My question is: What does the pack delta in Field 13 mean in the context of logging. It is probably small enough to be ignored, but I'm curious.

Thanks

Jeff