Your 60-cell panel should contain bypass diodes. With these diodes a series configuration does better in shade than parallel.
Shut up! You're blowing my mind right now. I can have my cake AND eat it too? Best of both worlds? I don't have the panels yet, but I will have to do that. Dzl mentioned bypass diodes, but he didn't elaborate much. I'll look into that.
He is a really smart guy but he gets religous about his approach, and he has tunnel vision on use cases. For example he seems fundamentally incapable of grasping an RV use case. I don't care what alternator efficiency is nor what it costs in diesel fuel to charge my battery bank. I just want to charge it. You will never get through to him on that or on other topics out of his focus, so best just to move on. Evaluate the SBMS0 and DSSR and if it works for you, great, use it.While I think there are blindspots in his argument, and I think he tends to get laser focused on one particular use-case or factor and not fully acknowledge or consider alternative use-cases/design goals, I definitely don't feel that he is acting in bad faith or trying to hype his product. He is a very disarming and straightforward guy, and strikes me as very honest, even when I disagree with him or feel he is overlooking something.
Well I disagree with you on BMS purpose, but that aside--this is incorrect. There's no reason you can't add disconnects in addition to controls. There are sufficient control lines and you can use separate settings (the OV Lock and UV Lock) to drive the disconnect devices of your choice. That's one of the things I like about this BMS vs. others.BMS main purpose isn't to be a controller, it's to be a safety device when the other things crap out. If it can only control the thing that crapped out and can't cut power itself then it's pretty much useless. For example: mosfets in your SCC have short-circuited, battery is in over voltage, what this BMS will do? nothing; it'll try to turn the SCC off but it can't so the battery will continue charging until the cells are destroyed.
Hey @Cal, long time no speak. What are you looking for on the SBMS0? As @LB3 said, if you're interested in the SBMS0 then join his forum, he will respond as will others there.I actually got involved in this thread/discussion because I may purchase one of his units. I'm looking for info. No hype please!
Well I disagree with you on BMS purpose, but that aside--this is incorrect. There's no reason you can't add disconnects in addition to controls. There are sufficient control lines and you can use separate settings (the OV Lock and UV Lock) to drive the disconnect devices of your choice. That's one of the things I like about this BMS vs. others.
@Cal, I admire your economy of words, but how would panels with bypass diodes work better than in parallel? At most it should work the same, no?
You're right that the safety disconnect capability is not Dacian's primary targeted use case. In fact he argued with me about the need for it while simultaneously jumping on fixing a defect I found in it. While he doesn't value it that much, it is listed as a supported feature and he ensured it works. He feels that cutting power (actual power interrupt, not just a remote control line) to any chargers is sufficient, beyond that is corner cases.I didn't say the BMS can't control other things, but just that it's not its main purpose, you can totally have other things in addition to that
You're correct, I didn't thought about that. However this is not the primary use case intented, else the disconnect would be sold with the BMS.
Some basics... In a parallel string, panel voltage is the dominate factor, while panel current doesn't matter. In a series string these relationships are reversed. In series, current is king. This means every panel in the series string has the exact same current. Panel voltage is a lesser factor.
So now we got some panels (36 cells in series), containing 3 bypass diodes. The maximum power point of the panel is 18V. Each of the 3 diode sections produce 6V. If one of the sections is shaded, and the bypass diode conducts, then the panel outputs 12V. In a two panel series string the total voltage reduces from 36V to 30V. The current remains the save, let's say 5A. Total power reduces from 180W to 150W.
In the parallel configuration one panel is at 18V and the other at 12V. The 12V panel becomes a non-factor as battery voltage is above 12V. Total power reduces from 180W to 90W.
Sorry long reply but you asked a lot of questionsHey Airtime. Yeah, long time. Other than sub-par current measurements, I'm happy with Chargery. I would like to know if SBMS0 does a better job in AH counting. Chargery uses a 12-bit A/D which should provide close to adequate current resolution, but it shuts down current measurements in the +/- 1.0A region. My RV has a 0.7A quiescent load current. That's my load at night. Chargery registers 0A. That means in a 24 hour time span, Chargery misses about 0.7A * 10 hours = 7 AH. In a short time period the SOC display shows greater than 90% when in reality the battery might be at 50%. SOC is important to me as that's my gauge to how much shade is acceptable. This summer I camped in a lot of shade since SOC always showed greater than 90%. That was a mistake!
What's SBMS0 current resolution when using a 300A load shunt? It's already strange that it requires 2 shunts. Do you know if SOC is accurate? Usually these meters have a reset to 100% when battery gets fully charged. How accurate is SOC when battery doesn't get fully charged for a week? Have you noticed any current measurement stability issues when measuring small quiescent currents? I'm looking into the possibility of building my own BMS. I built a current monitor/counter which currently has a thermal offset error. The error is coming from the instrumentation amp. Thermal offset error might be in the +/- 0.5A region. That blows my quiescent current measurement out of the water. Is that why Charger doesn't measure current less than 1A?
Regarding DSSR: is that a simple FET on/off switch? That means the fet switch connects solar to the battery. Once any cell reaches max voltage the fet opens and charging is terminated. That's quite risky isn't it? The system isn't fault tolerant. A fault will overvoltage the battery. You may add an additional relay (controlled by OV Lock) between panel and DSSR. But that helps only if the DSSR fails. If something within SBMS0 fails then the battery will still overvoltage.
What's SBMS0 current resolution when using a 300A load shunt? It's already strange that it requires 2 shunts. Do you know if SOC is accurate?
Two shunts--that's just because he's really into logging and understanding solar input in his house. The separate shunt lets him monitor solar charging separately and also with higher resolution since you can use a lower current shunt for measuring solar.
You're right that the safety disconnect capability is not Dacian's primary targeted use case. In fact he argued with me about the need for it while simultaneously jumping on fixing a defect I found in it. While he doesn't value it that much, it is listed as a supported feature and he ensured it works. He feels that cutting power (actual power interrupt, not just a remote control line) to any chargers is sufficient, beyond that is corner cases.
My Winnebago Via has a Zamp Solar 3 port combiner box on the roof, which I learned contains a 12V 40A breaker (Bussmann 121A40-A2P). It's apparently rated for 12V±2, though no one has actually reported failures at the 18V most RV solar panels run at.
And I do think that shorted FETs in a charger are a corner case. If someone is worried about shorted FETs then they wouldn't use any of the BMSs that come with FET switches built in, use only relays. Of course mechanical relays have their own failure modes. I worked in test equipment design and we got rid of mechanical switches in attenuators decades ago due to reliability issues.
Lots of views on purpose of a BMS, my view is its purpose is to manage the health of the battery. And I prefer to control charging based on state of the battery including individual cell voltages, not just pack voltage. Actually it was @Cal who first argued that point with me and I came to agree. The SBMS0 enables me to do that, and Victron equipment lets me control actual power switch for charger and inverter separately. They work very well together.
Hey Airtime. Yeah, long time. Other than sub-par current measurements, I'm happy with Chargery. I would like to know if SBMS0 does a better job in AH counting. Chargery uses a 12-bit A/D which should provide close to adequate current resolution, but it shuts down current measurements in the +/- 1.0A region. My RV has a 0.7A quiescent load current. That's my load at night. Chargery registers 0A. That means in a 24 hour time span, Chargery misses about 0.7A * 10 hours = 7 AH. In a short time period the SOC display shows greater than 90% when in reality the battery might be at 50%. SOC is important to me as that's my gauge to how much shade is acceptable. This summer I camped in a lot of shade since SOC always showed greater than 90%. That was a mistake!
It's already strange that it requires 2 shunts.
That blows my quiescent current measurement out of the water. Is that why Charger doesn't measure current less than 1A?
Regarding DSSR: is that a simple FET on/off switch? That means the fet switch connects solar to the battery. Once any cell reaches max voltage the fet opens and charging is terminated.
That's quite risky isn't it? The system isn't fault tolerant. A fault will overvoltage the battery. You may add an additional relay (controlled by OV Lock) between panel and DSSR. But that helps only if the DSSR fails. If something within SBMS0 fails then the battery will still overvoltage.
His hardware is open source and he publishes schematics. They can be a bit hard to follow, personally I would organize it better but most of the info is there. He uses the STM32F373 MCU which has three 16 bit sigma delta ADCs and one fast 12 bit ADC. So it has resolution.
If the BMS itself fails--I don't know how you handle that in general, other than by having a separate redundant BMS driving the disconnects. For example if the BMS FW locked up and all control lines remained in last state, then nether OVD nor OV Lock would trip and so no switches would open.
Yes, in the google group. Here's a link or just search on "EXTIO defect"Did this conversation occur on Dacian's google group? It piques my interest, I would like to read up on it if you have a link. By safety disconnect capability are you referring to essentially a second line of defense with a slightly wider bandwidth (also controlled by the BMS) that triggers a relay or something similar that cuts all charging or discharging (or both)?
I did intentionally say resolution and not accuracy .What's the accuracy of the Vref tho? because I bet the internal one isn't really accurate and will drift with the temp.
Meant to comment on that as well. Keep in mind that the DSSR is optional. You are free to use an MPPT controller with the SBMS0 if you want, whether for performance or for a redundant and independent disabling of solar charging.Yep, especially when you have several of them in parallel, one shorting is enough to have problems. That's what bugs me the most about this BMS, it doesn't protect the battery against basic faults. And for the protections it actually has there's no redundancy.
...
Well in general you have a SCC and a BMS so if the first fails the second will save the day. Here the BMS is actually the SCC too and has no redundant disconnect (well you can add one yourself as you said earlier, but it's still the same non-redundant brain so not really redundant actually).
I did intentionally say resolution and not accuracy .
Good point on Vref, another temp drift mechanism along with current sense amp input offset voltage.
Meant to comment on that as well. Keep in mind that the DSSR is optional. You are free to use an MPPT controller with the SBMS0 if you want, whether for performance or for a redundant and independent disabling of solar charging.