The ultimate onboard charging solution

[cal3thousand]

I can't wait for the satiator. Absolutely. That's a real charger and I certainly trust the reliability.

I'm only spewing about it's awesomeness everywhere because I'm currently playing with one

 

[Make More Pi]

Grindz145,

I had noticed in your trek videos that you were reading charge current on your CA. I had figured that you must somehow be sending the charge current through the controller shunt. (My charger connects directly to the battery.) In an effort to find out more I found this thread and saw this

“After an afternoon of hacking a tap to the shunt in my controller ...”.

Can you provide more details on how you did this?

BTW, congratulations on completing your trek! Cant wait to see more videos. (Up to day 18 is posted at this point).

 

[grindz145]

Grindz145,

I had noticed in your trek videos that you were reading charge current on your CA. I had figured that you must somehow be sending the charge current through the controller shunt. (My charger connects directly to the battery.) In an effort to find out more I found this thread and saw this

“After an afternoon of hacking a tap to the shunt in my controller ...”.

Can you provide more details on how you did this?

BTW, congratulations on completing your trek! Cant wait to see more videos. (Up to day 18 is posted at this point).

Just posted a few more to my website, Almost done with editing now

What you have to do is run the low side of the charging current through the shut. This is achieved by adding a tap to the shunt (I literally soldered to the high side of the shunt) and using this wire as the negative wire to the charger.

 

[Make More Pi]

(I literally soldered to the high side of the shunt) and using this wire as the negative wire to the charger.

That makes perfect sense. (... see what I did there?) And the high side of the charger taps to the high side of the battery inside the controller I assume. Do you bypass the switch so that you can charge whether or not you have the CA powered? Do you run everything through the shunt (as in lights) so that you are tracking power use for these other devices? Or do you consider that just noise?

 

[grindz145]

(I literally soldered to the high side of the shunt) and using this wire as the negative wire to the charger.

That makes perfect sense. (... see what I did there?) And the high side of the charger taps to the high side of the battery inside the controller I assume. Do you bypass the switch so that you can charge whether or not you have the CA powered? Do you run everything through the shunt (as in lights) so that you are tracking power use for these other devices? Or do you consider that just noise?

I don't run everything through the shunt, although I should. I can charge when the CA is off, it just won't get measured.

 

[michaelkramer11]

Any chance you can post a "schematic for dummies" drawing on how to run these safely in series and explain what the diodes do and what they are needed for? And also maybe a link to the diodes so I can purchase some?

I've been using one Meanwell and having to break the pack in half but it would be awesome to be able to run two in series and leave everything hooked up!

 

[Alan B]

Just ordered three of these 240W supplies in the 24V versions to charge my 18S setups (I'll set them to 25V for 4.167V/cell). I've had perfect service from a set of three 320W units for a couple of years already, this is an excellent solution to reliable charging. The 240W units are the highest in watts per cubic inch and close to the lowest cost per watt of the Meanwell LED supplies from 150 to 600 watts, so a good choice from several perspectives. Three will make a 700 watt charging system which is unlikely to blow a breaker but still produces a good charging rate (about 10 amps).

In my particular case I don't use the diodes, instead bring a 2 conductor cable out from each 6S bank and plug a charger into each of the three banks. The supplies see the batteries directly, not each other. Takes a couple more cables and connectors but avoids the issues of series connection and the extra components.

 

[michaelkramer11]

Just ordered three of these 240W supplies in the 24V versions to charge my 18S setups (I'll set them to 25V for 4.167V/cell). I've had perfect service from a set of three 320W units for a couple of years already, this is an excellent solution to reliable charging. The 240W units are the highest in watts per cubic inch and close to the lowest cost per watt of the Meanwell LED supplies from 150 to 600 watts, so a good choice from several perspectives. Three will make a 700 watt charging system which is unlikely to blow a breaker but still produces a good charging rate (about 10 amps).

In my particular case I don't use the diodes, instead bring a 2 conductor cable out from each 6S bank and plug a charger into each of the three banks. The supplies see the batteries directly, not each other. Takes a couple more cables and connectors but avoids the issues of series connection and the extra components.

Awesome! Thats one way to do it. I guess the only thing I wouldn't like about that method is that the separate banks may charge unevenly and you'd have to monitor two "packs" instead of one. Although one should be able to get the charge currents and voltages pretty close.

 

[Alan B]

I set the voltages and currents precisely with one good meter, and they stay very well matched. What happens is the current drops off with time as full charge is approached, and when it gets quite low (less than 200mA on a 20AH pack, so 1%) on the one section I'm metering I stop the charge. The meter itself adds some resistance to this bank so it is slightly behind the others, so it is the worst case and all the others are at lower current. The differences between the banks is very slight, below the differences in individual cell capacities. Many hundreds of cycles have been done this way without problems. I also use a 6S Battery Medic to check the balance pretty often, and charge to 4.16 so there is a long way to 4.2, it has never reached 4.2 on any cell that I've ever metered.

A bigger risk would be if one supply failed to charge at all. I connect their primaries together with one power cord, so they tend to all work or not.

Wiring it the way I do means I can use it on a 12S or 18S system, so it gives me some flexibility, and it keeps the voltages at the supplies at 25V so I don't have a voltage over 50V at an unplugged connector. That way I can legally stick a meter into the connector at work without breaking work safe rules that require PPE and work authorization for anything over 50V. (Good thing these supplies are UL listed or I could be prohibited from using them at work at all).

I don't know what diodes he used, but they don't carry charge current, so a 1N4007 would probably do the trick, or even a lower voltage power diode. If a supply quits there will be no charging current through the diode since the reduced voltage would be insufficient, unless the pack is shorted...

 

[michaelkramer11]

Do you have to plug in the connectors from each charger to each bank every time you charge?

 

[Alan B]

I use three powerpole pairs, they can easily be connected into a single 6 pin connector block. I don't bother, just plug the three connectors each time, but if that's an issue it is easy to make it one connector. It hasn't been much of a problem.

If you mount the supplies on the bike they can remain connected to the batteries, they don't draw power when they are themselves unpowered.

 

[grindz145]

I use three powerpole pairs, they can easily be connected into a single 6 pin connector block. I don't bother, just plug the three connectors each time, but if that's an issue it is easy to make it one connector. It hasn't been much of a problem.

If you mount the supplies on the bike they can remain connected to the batteries, they don't draw power when they are themselves unpowered.

This kind of defeats the purposes a little bit, but it does work. I just get off and plug in the AC plug. It's a beautiful thing.

 

[danielrlee]

I've just been offered two Mean Well HLG 240w 48v supplies for a reasonable price. Can they be used in parallel to charge at 480w?

 

[jkbrigman]

I plan to repeat Troy's/Alan's deal with two HLG-340-48's .

I like the flexibility of connecting one LED supply to a 12S portion of the whole battery. I'll split the 24S into two 12S and put a DC-DC converter into each half. That'll give me two 12V outlets I can run lights and charge USB (using a car adapter) from. So dividing the battery like Alan's talking about has more benefit than just charging: it enables use of easier-to-find DC-DC converters.

One thing to add about Alan's configuration: Splitting the pack and chargers up still triples your aggregate charging current toward the end of the charge, when everything slows to a trickle. 3 trickles always beats 1 trickle in my book!

And finally: if you build the pack right, you can bring out the 12S balance wires and balance charge each section (either 6s "one third" or 12S "one half") and easily balance charge, using the same bulk charging connector plus the balance connectors.

Also, if you lose a supply somehow, you can still charge at 1/2 the rate. Looks like a win-win-win to me.

JKB

 

[danielrlee]

Anyone know whether the Mean Well HLG power supplies can be used in parallel?

 

[heathyoung]

Yes. But voltages must be exact or they won't current share evenly.

 

[riba2233]

Yes. But voltages must be exact or they won't current share evenly.

That's true only for CV stage.

 

[danielrlee]

Yes. But voltages must be exact or they won't current share evenly.

That's true only for CV stage.

So no risks as such, just a slower end of charge at the CV stage?

My thought was that although it could be done with two 'half voltage' power supplies in series, a parallel arrangement would give some redundancy if one was to fail on the road.

 

[midwest mayhem]

This system has been working great for me and my 96v 15Ah headway pack. I charge through a BMS and use 2 Meanwell HLG-320-54-A's in series. Plug in and walk away, wherever there's an outlet.