diy solar

diy solar

My Push Button Pre-Charger Install for the SW 4024 Inverter

That kind of inrush might not harm the inverter, but it might blow fuses. class-t fuses @ $40 a piece make a $20 precharge circuit worth it.
Ehhh. I dunno. I think class-t fuses are usually pretty fast, but the inrush we are talking about is a small fraction of a second. I'm not convinced it would blow a fuse, but I guess it is a possibility. What I remember is that a *REALLY* fast fuse might blow if a 2X current lasted for more than 0.1 seconds. I believe that the times we are talking about for these front-end capacitors would be less than 0.1 seconds. However, I haven't done the analysis.

I'd reword your sentence. If you are putting in a new battery system, adding a pre-charge is EASY enough to make it worth it.
 
If nothing more, we have seen videos where Will triggered the over-current protection of the BMS and then had to reset the BMS. That is a PITA I would gladly avoid.
I figure using a simple pre-charge circuit keeps the current very low and the whole system is happier for it.
That's a good point. I don't have enough experience yet with LFP and a BMS to know what it takes. I know that it seems like the JBD / Overkill reset themselves once the current drops down to an acceptable level for a few seconds. Sounds like kind of a brute-force pre-charge via the BMS! ;)
 
the cost of replacing a T-class fuse does motivate me.

if i can invest 20-30 usd on even a simple timed pre-charge circuit, that pays off on the first time it prevents a “spurious” fuse burn ?

the current sensing to me is superfluous, yet also represents another utility.

current sensing on the precharge circuit could be a perfect place to check for diagnosis. if the power profile of the pre-charge changes drastically, that could help as a clue perhaps when things go weird. ?
 
Ehhh. I dunno. I think class-t fuses are usually pretty fast, but the inrush we are talking about is a small fraction of a second. I'm not convinced it would blow a fuse, but I guess it is a possibility. What I remember is that a *REALLY* fast fuse might blow if a 2X current lasted for more than 0.1 seconds. I believe that the times we are talking about for these front-end capacitors would be less than 0.1 seconds. However, I haven't done the analysis.

Curve for 400A class T shows that it should blow in < 0.01 seconds at 3000A.
I think dead short on LiFePO4 (assuming 0.17 milliohm per cells) will deliver 20,000A.
Whether it blows the fuse should depend on resistance of wires, PCB, capacitors.
0.6F capacitor or greater could sustain 3000A for 0.01 seconds charging to 50V

 
Ehhh. I dunno. I think class-t fuses are usually pretty fast, but the inrush we are talking about is a small fraction of a second. I'm not convinced it would blow a fuse, but I guess it is a possibility. What I remember is that a *REALLY* fast fuse might blow if a 2X current lasted for more than 0.1 seconds. I believe that the times we are talking about for these front-end capacitors would be less than 0.1 seconds. However, I haven't done the analysis.

I'd reword your sentence. If you are putting in a new battery system, adding a pre-charge is EASY enough to make it worth it.
There's a youtube video floating around on this site (documenting chargery BMS) where video op blew multiple class t fuses trying to figure out chargery's precharge. Is that conclusive evidence? No, but it's pretty strong though.
 
Cool idea.

I'm not sure the little ADC board is needed if you can round up an ESP8266 or ESP32 board, both of which already have an ADC. The current sensor is another animal. If you want current sensing, that looks like a good solution.

Personally, I think we (on this board) have kind of over-played the whole "pre-charge" thing. I certainly don't know the whole population of inverters, but I think many - maybe most - inverters would be pretty hard to harm with the inrush to their front-end capacitors. I am very convinced that SCC's don't need to mess with this.

Having said that, I put a pre-charge circuit into my new LFP box, just because I could. Rube Goldberg was a genius! :p
It isn't the inverter, it is harm to the BMS that causes problems.

More than one member here seems to have bricked their BMS when hooking up the inverter.
 
Curve for 400A class T shows that it should blow in < 0.01 seconds at 3000A.
I think dead short on LiFePO4 (assuming 0.17 milliohm per cells) will deliver 20,000A.
Whether it blows the fuse should depend on resistance of wires, PCB, capacitors.
0.6F capacitor or greater could sustain 3000A for 0.01 seconds charging to 50V

Well then, I stand corrected! Good data @Hedges .
 
There's a youtube video floating around on this site (documenting chargery BMS) where video op blew multiple class t fuses trying to figure out chargery's precharge. Is that conclusive evidence? No, but it's pretty strong though.
I'm trying to locate this video. Can you offer any help?
 
I've got other videos on that channel plus my build thread where I go into the nuts and bolts but the summary is that I went with Orion for BMS and a Midnite 250 amp DC breaker. I manually precharge for now but Orion has a precharge feature coming in the next few months ...
 
I've got other videos on that channel plus my build thread where I go into the nuts and bolts but the summary is that I went with Orion for BMS and a Midnite 250 amp DC breaker. I manually precharge for now but Orion has a precharge feature coming in the next few months ...
Thank you for video. I will take a look at others. Yea. I wasn't too impressed with Chargery after what you pointed out. That solid state switch was horrible. Did you try a low pass filter on the current shunt signal? I've seen inverters with very large filters to provide average bus current. They work fine.

I've got a 6kW inverter on lead-acid that I am getting ready to plug into the EVE280 cells. I was concerned that the inrush would take out the BMS or fuses when connected to LiFePO4. I simulated charge event using 6800uF x4 bus caps - dealer thinks that is a good number. It looks like the 200A class T fuses will survive a power event with a small amount of stray resistance. 0.005 ohms looks good. I'm assuming about 5uH of stray inductance. The batteries present a large loop that drives that number. I was considering doing a precharge circuit. I found some NTC thermisters that look like they will work. I've used these parts before without a contactor and had good success. They run warm and a little bit lossey. Most of the time my inverter is loafing at about 500w, so not terribly concerned with losses.

I haven't picked out exact part yet, but this is a contender.


Can you tell me what the bus capacitance you were trying to bring up was? I would like to simulate and see what the inrush current was and compare to the fuse melting energy. I'm trying to check simulator with a few data points.
 
Thank you for video. I will take a look at others. Yea. I wasn't too impressed with Chargery after what you pointed out. That solid state switch was horrible. Did you try a low pass filter on the current shunt signal? I've seen inverters with very large filters to provide average bus current. They work fine.

I did, but I'm not sure I got the calculations right because it didn't help. After the other issues, I gave up on that exercise.
Can you tell me what the bus capacitance you were trying to bring up was? I would like to simulate and see what the inrush current was and compare to the fuse melting energy. I'm trying to check simulator with a few data points.

I didn't do any calculations. I have 2 5kVA inverters plus two SCCs and the DC-DC (all 5 devices have precharge situation) so it's a LOT of capacitance. I started with a larger resistor based on what others were saying about their setups. I ended up buying a smaller one later to speed up the precharge. Hopefully I listed that in my parts list in my sig.
 
I am looking for clarification about precharge circuits.
Most solutions recommend a "manual" delay or with a relay.
What is the problem with an on-off circuit: a resistor bridging the relay contacts, isolation switches or BMS fet (with a 0-I switch that allows the circuit to be shut down for maintenance)?
-After the main circuit is switched on, by the relay, switch or fet, there is hardly any current flowing through the preload circuit! But by leaving it in service, wouldn't we protect the main contacts from the induction choke effect of an untimely cut-off, while benefiting from the precharge?
Two birds with one stone.
Thanks for the answers.
 
Personally, I want 'off' to mean 'off' and not 'kinda off, but still on through a resistor'. Keep in mind that if something goes wrong with the inverter (broken FETs) and the main switch/relay is off, you now have a 5 or something Watt resistor directly coupled over the battery through the broken (shorted) FETs: it will heat up and could burn up since there are no charging capacitors to taper this current off.
If your inverter has a physical on/off switch as well, and you turn it on without the main relay/switch being on, you get the same thing where the caps never really charge (because the inverter is trying to turn on and switch FETs) and your resistor can become a hot mess again.
 
Thank you for the answer,
I was thinking more of a circuit managed by a contactor.
- If we insert a very high value resistor, the capacitors theoretically always remain charged and the residual current seems to me insignificant to represent a danger, am I wrong?
- Do you have an answer concerning the protection that this assembly can bring, for the contacts, at the time of an untimely cut?
Greetings
 
I found this suggestion from Mr @FilterGuy . I might change my 50 ohm on my 24v inverter... But maybe it doesn't matter?

View attachment 45079
Some folks have described using light bulbs to provide the needed pre-charge resistance... just wait until the bulb dims and stop pressing the momentary switch. Easy. So if instead you use the wire wound resistor method how do you know when to let up on the switch?... what acts as the dimming bulb indicator of a fully charged capacitor bank?
 
Back
Top