Multiple buck converters using a shared ground

Posted by: @byron

↑

A simple request, a good and thorough answer, and many many many more posts from the topic unsubscriber ranting on about something.  I think a special sub forum called Rons Rantings is called for.  BTW it is quite possible for two circuits in a camper van to be connected for comms thus the desirability for a considered grounding scheme to be warranted.  Simple answers can be very misleading. 

@byron, Indeed, in electrical or electronics, there is no such thing as a simple answer, just incomplete, and misinformed ones...lol The thing is, no knowledge is wasteful, and if you don't need it, you can always file it away for a time you may need it, or ignore it completely.

I'm quite sure many more benefit from @davee "sermons", than are repelled, but then again I state that the purpose of this forum, or any forum for that matter, is to exchange information, enlightenment, and share progress that we have made on our projects. Learning is a life long endeavor for some of us who haven't risen to the "know it all" realm yet!.. 😀

regards,

LouisR

Hi Ron @zander,

   Sorry, I can't comment about your suggestion to Bill, but on the technical side, what you are claimed seems to be flawed more than once, much as I like to encourage you to help others.

---------

I am not clear what a "CAMPER" signifies from an electrical viewpoint. In the UK, practical reasons based on our roads, touring vehicles with sleeping accomodation tends to mean smaller caravans or vehicles than US/Canada, but they  all are normally equipped for both 12V and 230Vac, implying multiple power sources are the norm. (This is before solar power or other batteries are even considered.)

---------------------

The original question included " to monitor the BMSs. " -- i.e. more than one BMS.  I accept one might have more than one BMS for a single battery, but the conclusion I came to is that there are probably multiple batteries, and hence multiple power sources.

---------

It is true that many buck converters, have a GND or 0V line which simply connects GND IN to GND OUT, and so if two (or more) of these converters are connected to a single power source it follows that they will have the virtually the same potential, at the respective GND OUTs.

If that is the case, then there could be a de facto 'star point' where the battery and the two GND IN wires meet. The consequences of understanding this is tied with understanding the parasitic inductance and parasitic resistance in the discussion.

--

The original question did not state that their intended circuit included (or did not include) such an arrangement of one battery supplying two converters). The question was should the grounds be joined together, which seems an unlikely question to ask if this is scenario, because it should be obvious that the GND INs are connected together, and  connecting the GND outputs will not achieve anything useful, given that they are already wired together.

-----------

It should  noted that not all buck and/or boost converters have 'GND input' connected 'GND output'. e.g. Some do current monitoring in the return line, so 'GND input' and 'Gnd output' will be connected via a small value resistor, say 0.1 Ohms. These need even more careful handling ... but that would be a different question and another sermon to explain!

------------

RE: It also will not hurt to make even more connections. (just to keep Dave happy)

There are many Daves in the world, so it might keep one happy, but it wouldn't keep this one happy, because the second half of my discussion was the importance of avoiding unnecessary Ground Loops.

---------

Best wishes all, and lets all relax a bit, Dave

@davee Indeed, those pesky ground loops can be a nightmare, especially when dealing with dc power sources that are derived from an AC source.We encountered this many times in the plant environment where instruments or controllers were AC powered, and converted to DC so that the output would be, say a 4-20 mAdc signal into a 250 ohm load. If the device didn't have an isolated output, then we had to provide a buffer called a current converter that was either transformer coupled or opto-isolated. The reason being that the earth ground for the AC power had a separate grounding grid system than the DC power system for the instrumentation.

Each system (AC and DC ) had their own ground grid system, so as a result the grounds could be, (and most often were)  at different potentials. Hence, the converters modules.

I have seen ground loops with an excess of 200mA!..

Grounding systems technology, as you apparently know, is far more complex than it appears, and thus required a bit more than just casual knowledge. We had many study courses on proper grounding techniques as related to instrumentation and electrical systems in general,  as much of the industrial controls world have to rely on this to make analog, digital equipment work seamlessly. I for one, appreciate your concise explanations and understand what you are trying to convey..

Regards,

LouisR

@davee, I made the dumb mistake of assuming too much. Having just recently moved from a camper/RV/coach/5th wheel/trailer, call it what you will, I immediately 'knew' a lot about the OP's situation. Also, seeing a mention of BMS made me assume he either built his own or had an external BMS, but in any case, made me think he was a serious full-time RVer who lived at least some of the time off-grid.

What that means to me a recent RVer is he had a battery bank of unknown size but most likely at least 4 in parallel (I had 6).

This means he would have 12VDC available ALL the time, mains power when plugged into shore power (yes, we call it shore even for dry land yachts) and very likely mains power via an inverter.

I also assumed, since he was talking about buck converters, that he was building an IoT-type system that always ran off the 12V battery bank as opposed to any connection to the AC system.

Given all those assumptions, I knew that there was a single feed of 12VDC. Think of the RV as a breadboard. The 'house' battery is connected to the red + bus and the black - bus. All the little projects then get connected across that common power rails. (Only motorized coaches have a second battery of a different type to run the gas/diesel engine)

Returning to the original question, he asked if it was OK to connect all the grounds. The answer is yes; in fact you have already done that as an RV has only one ground otherwise, the battery measurement system (not to be confused with BMS) would not work. The system is based on a shunt. The shunt MUST be the only connection to the battery bank's most negative post. ALL other 'return' or negative/ground (means the same in this case)  MUST be terminated at the other end of the shunt.

Knowing the above, I was guilty of assuming too much since I built my own solar off-grid electrical system.

As far as Dave's corner, I meant that as a compliment. You have demonstrated many times that you have a vast amount of information at your fingertips, and I see it repeated, so why not put it somewhere as a resource for new members? I would say the same for Dennis @inq for robotics, Will @will for steppers and even cranky old Byron for Python.

To summarize, I am 100% sure everything you have said about 'ground' is 100% accurate, but in the case of a home on wheels, all the 'grounds' have to come back to the business end of the shunt to properly measure battery capacity or more to the point battery remaining capacity and hours to go. I did assume the OP had a properly connected shunt since he knew what a BMS was.

Hi Ron @zander,

  Thanks for that last note .. it is appreciated.

  I tend to agree that having a 'Wiki' style area for reference material, principally indexed by subject rather than contributor 🙂 , that could be updated as and when, and could include the synopsis of some past discussions, would be great.

But Bill (@dronebot-workshop) is already providing a great resource at his own expense and effort (and for which I thank him), so I would not impose to ask him to extend his hospitality still further.  

Best wishes and take care my friend, Dave

@davee Agreed re Bill. The Wiki approach may be better. If I still had my server farm, I would gladly host it, and I agree we should not impose upon Bill; he is busy enough. Maybe somebody reading this knows of a willing host? I'm sorry, I'm assuming you and others are willing to move/copy some of your great posts to such a mechanism. I am a simple user of an occasional WiKi, so I have no idea what is involved other than it must need a server and most likely open-source software to make it into a WiKi.

OR, is github a possible resource? I sometimes download stuff from there, but I had never seen it before joining this forum. I am mostly card decks and a little bit of what came later, so github is foreign to me.

It seems a shame that we can't leverage your ability to produce such well-thought-out articles rather than you having to type it all in every time, or am I being foolish, you no doubt have it somewhere that you can do a simple copy-paste, duh! The card deck mind strikes again!

@zander Wow this thread has spawned a lot more replies than I thought was even possible.  Here is a little back-story: my original plans went sideways when I tried to build a 2P4S LiFePO4 battery bank (400AH 12V) for my camper.  After struggling with this configuration for some time, such as balancing, capacity, and high-loads, I found one cell had a much lower capacity than the other 7 cells.  So I decided to order another set of 4 cells from a different Aliexpress vendor.  And guess what, they shipped me 180AH cells, and insisted that they shipped me 200AH (which were clearly marked with stickers that said 180... Arrgh!).  So to satisfy my curiosity, I capacity-tested the 4 newer cells with some of the older cells.  Yep, the test results do not lie... the new batch had a lower capacity of around 20AH on average per cell.  Hmmm... (200-180=20), go figure.  Oh well, I guess you get what you pay for when you order from Aliexpress.  So rather than order yet a fourth set and get ripped off again, I decided to just take the top 4 capacity-tested cells from the 200AH cells, and the 4 newer 180AH cells and build two 4S batteries then wire them in parallel.  I know not ideal, but probably safer than building a single battery with mis-matched cells.  So fast forward to now, I have two batteries and each battery has it's own fuse, disconnect switch, shunt, BMS, and active balancer.  They are connected in parallel with heavy duty busbars which gives me around 350AH with a nominal voltage of 12.8V.  I know, 180AH+200AH should be closer to 380AH, but my actual capacity testing results were about 10% lower per cell.  Now in addition to the battery bank, I also have a 200W solar panel and MPPT charge controller, along with a DC to DC charger and an AC to DC converter.  My next goal is to monitor/control all these devices and push the data into an Influx DB.  Then build a control panel with a Nextion display to monitor the capacity and control things such as the 1500W inverter.  Some of the devices will need 3.3V, some will need 5.0V, and others will need 12.0V.  And that's where my original question came from.  At some point, all these devices will need to be powered, connected, and communicate with each other.  I'm still trying to work through the initial design, but it looks like maybe a raspberry pi, arduino, nextion display, relay shield, a USB expansion hub, some USB to UART controllers, a USB to RS485 adaptor along with a USB to SSD adaptor.  Oh, and some software that can talk to all these devices, looks like maybe Node-RED.  Wish me luck!

@iwannastout Wow, some adventure. I am unsure re the two shunts. Is that needed? I am not sure what benefit it has and how to 'manage' with two.

Worst case solution is to just treat the entire bank as a 350AH 12V bank, or you could also treat it as 12V 400AH and just let it self regulate. As long as each battery has a BMS and balancer it will be ok.

As far as your original question I think you know by now that in your system like mine there is one ground and it is the negative side of the battery bank. As I mentioned earlier, you have a giant breadboard with a positive rail and negative rail. In my RV I use small busbars connected to the big busbars for connecting the several low power circuits (12VDC -> 7VDC using buck converters followed by 7VDC -> 5VDC using LVR) If 3.3VDC is also needed, then adding another 5VDC to 3.3VDC is simple enough. Of course any cross connection (I didn't have any 3.3) has to be done with a level shifter.

Hi Ron @zander & @iwannastout,

Re:  I am unsure re the two shunts. Is that needed? 

  Without a proper schematic diagram reflecting physical wiring aspects, I can't even begin to do a cursory analysis of what is being proposed, but if your (Ron's) comment refers to shunts in the ground return, then I tend to share your concern that care is needed, as it can be easy to end up in a mess.

When there is more than 1 source of power (e.g. two batteries), independently measuring current flow from/to each battery by means of a shunt in the 'ground'/'0V'/'low side' wiring can become more tricky, because the 'ground point' is no longer a battery terminal.

Measuring current on the 'high side', is technically a little more complex for the actual measuring circuit, which must be designed for 'high side' use, but can make life simpler when the overall system is considered, especially as the overall system becomes more complex.

-----

My previous comments, in which I mentioned a star point earth, become more important, as 'accidentally sharing' a conductor (wire or part of the vehicle chassis) with both circuits could give unexpected (and unwanted) results. ('Accidental sharing' was what the 'Incorrect' wiring diagram I showed previously was attempting to illustrate.)

I accept that what I was describing is rather difficult to understand, and may have looked academic, but current flows etc. have their own 'rules' and do not make exceptions to make life simple for humans.

Note, I am not saying there is only 'one way' to do something, or anything prescriptive, but rather it is a situation that needs to be carefully thought through, imagining the current flows, etc. to ensure any control and monitoring systems (home brew or commercial) work correctly.

----

This message is primarily aimed at the main battery/vehicle load current flows ... it will affect the internal wiring of any monitoring systems, with buck converters or whatever, but they are "add on"s to the main system, which will also need to be correctly placed in the total wiring strategy, if they are to work correctly.

The current levels, wire lengths, etc. of this project mean that there will be places that two ends of the same wire, or even two points on the vehicle chassis, may be (momentarily at least) at different potentials, and any monitoring or control systems may see a distorted picture of what is happening, and will react accordingly.

Best wishes and good luck, Dave

@davee @iwannastout Without a wiring diagram I assume the ground buss has a connection to a shunt for each battery. The benefit of doing it that way is to have correct coulomb counts for each battery. That way each battery has it's own SOC, %remaining etc. It isn't necessary but since the OP mentioned using an Influx DB I can tell he is a detailed numbers guy. In NA we ONLY wire the shunts on the negative side. The only thing connected to the battery most negative negative (yes that's a thing) is the shunt, all the return lines go to the other side of the shunt including frame. I never used or advise the use of the frame for an RV and of course VERY bad for a boat. That means there is a bus (often two as I mentioned above) where ALL the negative wires are run back to. In my case some of those wires were over 30ft long.

So I think if I use your language, this is a star connection, all the circuits connect to a bus and the bus connects to the shunt. In my case that last wire was 4/0 or 107mm^2.

The fact he is using 2 shunts so he can measure two different batteries is a little unusual but not a concern.

@davee Just an FYI. You mention tiny voltage differences in a wire being a concern. What you do not know is the positive 12VDC actually varies daily between 14.x (where x is 2 to 8) and 12.8 IF the batteries are healthy, with age the lower range will drop further. So worrying about millivolt differences along a wire is pointless.

@davee We NEVER put the shunt on the high side! Remember, this is an RV in North America. I can't speak for Europe and it's standards.

@davee In a RV at least in North America, the battery post is NOT the ground point, the negative buss is the ground point. The connections to the negative buss are shunt, solar charge controller,  inverter, frame, and usually a smaller bus for all the DC circuits in the RV. The frame is not used for a conductor although it could be, it's to guarantee a frayed positive wire if it touches the frame will blow a fuse rather than electrifying the frame.

Similarly the positive bus has connections to the battery most positive positive, solar, inverter, heavy use house circuit (hydraulics) and a feed to the small bus to feed all the 12VDC circuits.

Hi Ron @zander,

PS I know nothing about RV or other vehicle regulations/standards in any country ... this discussion only relates to electronics, so please ensure you conform to local regulations/standards.

  I totally accept there are ways of wiring with low side shunts, and you may well have adopted wiring your own RV with something similar to the star point method. But I think you generally have a good feeling for the dangers of failing to ensure that an entire circuit loop must be considered, especially when high voltages and/or currents are involved. For that reason, you would naturally ensure that part of that loop did not include conductive parts which were ill suited to the task. Others are unlikely to have your level of experience, and hence my concern and aims to explain what to look for.

Hence I have been trying to explain the strategy, which is widely used, but less well understood, even by many 'electronic' engineers, than it needs to be, if mistakes are to be avoided.

--------

I can readily accept the possibility and plausibilty (I haven't checked) that 'low side' placement of the current shunt is the de facto standard for commercial equipment used in the particular field this discussion is revolving around. It is not uncommon for one manufacturer to be a 'leader' in a particular market, and the rest make 'compatible' products, especially when the alternative can be a little more difficult or expensive. For direct control of charging and discharging of the batteries using commercial equipment, this may limit the choice.

-------

For 'home brew' requirements, where the limits of commercial availability may be less limiting.

Technically, 'low side' is easier, because one side of the shunt is always '0V' and the other side, the offset voltage, which can be measured directly. But it can make the overall system more complex in SOME cases.

'High Side' is a more difficult measurement, because it is measuring a small voltage offset, at a voltage based on the incoming voltage, which may itself be significantly variable, depending upon the source, and in many applications will be 10s or 100s of Volts above the '0V', whilst the control and monitoring systems are typically referenced to '0V', so that a mechanism is required to transfer between the voltage levels.

For low voltage systems, say 12V or 24V, chips to enable high side sensing are readily available (providing it is not a chip drought).

Higher voltage systems, involving 100s of Volts or more, are feasible, but may be more challenging.

Alternately, devices based on Hall effect (and also current transformers for AC) are available, which are intrinsically isolated from the supply, so can readily be inserted in to the high side feed, providing insulation requirements are observed.

------------------

Best wishes, Dave

Hi Ron @zander,

Re You mention tiny voltage differences in a wire being a concern. 

Current measurement by shunt is always about measuring a small voltage drop across a shunt ... this is the voltage drop I was considering. I was not concerned about normal battery voltage changes.

In the 'better' circuits, the shunts will have Kelvin connections, to sample the voltage drop across the shunt directly, AND the voltage measurement circuit is truly differential, so neither side is strictly at 'zero volts' and the input circuit is able to deal with common voltage offsets without loss of accuracy.

Ideally, all circuits would conform to this model; but whether they all do, is another matter that can only be answered by checking each individual case.

Best wishes, Dave