Using an EUC to charge a laptop

[Kenzi]

Hey I'm an electronics engineer!

My EUC has an 800Wh battery, and I have never used more than 50% in a single trip.  I love going to coffee shops and parks to get work done, so why not tap into that enormous potential to power my laptop all day? That'd cover me for more than 8 hours, even on ridiculous workloads like VR gaming.

So I bought a cheap power supply that can drive my laptop using 30-90V, which is more than the range of the battery pack.

I'm using an MCM5, and I've found that charger pins 1 and 2 go directly to the battery for charging, but it appears that something (possibly around the BMS) doesn't really like power going out of the charging port, so now I'm considering alternatives.  Before I pull this apart, I'm wondering if anyone has suggestions or guidance - perhaps someone who's seen and understood the insides, like @RagingGrandpa?

[Planemo]

Why don't you just tap directly into the battery outputs (via the board to save having to tap into both packs further up the chain), and screw a panel mounted socket to the EUC case somewhere.

I must admit, I wouldn't like to draw any significant current out of the charge port. 

[mike_bike_kite]

No idea about electronics but couldn't you use the USB output from your wheel then put that through a USB to USB-C converter then use that to power a laptop? Obviously you'd need a fairly new laptop to power it via USB-C but I assume you'd have that if you're doing VR Gaming. If not you could use this as an excuse to buy a new laptop - the new 5800u processors from AMD are coming out soon as are the new mobile RTX 30 series GPU's. 

[Planemo]

I don't know what the current limitations of the wheel USB ports are, but I can't imagine it's a lot. 1A at most? Will a laptop run on that?

[DangerDan]

1-2A at 5V, so you get 5-10W at best out of usb port. Normal laptop chargers usualy have around 60-100w+ (look at the label of your charger). Can you tap 12-20v somewhere in the wheel, maybe from the lights?

 

Edited January 16, 2021 by DangerDan

[Kenzi]

5 minutes ago, Planemo said:

1A at most? Will a laptop run on that?

Unfortunately no. My laptop (which is on the beefy side) tops out at 220W (11A, 19.5V) and good USB ports generally don't produce more than 10-15W (2.6A, 5V).

17 minutes ago, mike_bike_kite said:

[...]use the USB output from your wheel then put that through a USB to USB-C converter[...]?

The wheel's USB output just can't produce the power even if I do boost it to USB-C's standards, but USB-C output maybe directly from the battery is definitely a better long-term solution! If I do this, I'm sure lots of other people would be able to use the build plans.

22 minutes ago, mike_bike_kite said:

[...]you'd need a fairly new laptop to power it via USB-C [...] RTX 30 series GPU's. 

I've got a dual-screen ASUS ZenBook Pro with an RTX 2060. It can't charge from USB-C - probably because it's too much of a power hog!

21 minutes ago, Planemo said:

Why don't you just tap directly into the battery outputs [...] ?

Yeah, looks like this is what I'm going to have to do. Just want to digest some of these excellent suggestions before taking it apart.

[Tawpie]

Reminds me a little of Volkswagen before the 70s. They repurposed “power” sources so the windshield squirters use the air in your spare tire, and on their early campers the camping heater used the gas from your tank. So it was pretty easy to end up with a flat spare after driving in misty moisty. Camping in the woods on a cold night with the heater on was a good way to wake up frozen, with no gas. At all. Always a pleasant surprise.

It’s a fun thought, but do you want to have to Uber home?

[Kenzi]

1 minute ago, DangerDan said:

Can you tap 12-20v somewhere in the wheel, maybe from the lights?

I expect any intermediate 20V, 12V or 5V that might be inside the wheel require it to be turned on and aren't made to handle the current of a laptop. Only the 84V rail direct from the battery is made to be high-power, and it'll be more efficient to step-down directly from that anyways.

But there are 4 pins on that charger port and I'm wondering if the other 2 are used for anything?  If yes, could they be used?  If no, could they be repurposed?

[Kenzi]

1 minute ago, Tawpie said:

It’s a fun thought, but do you want to have to Uber home?

Thanks for the warning of discouragement, but I do actually know what I'm doing.

Also, shown in the picture is a meter that I'm including in the supply case that will alarm if the battery gets below a threshold of my choosing.

[RagingGrandpa]

You can definitely yank 400w out of a Gotway charging circuit. I discharge my batteries this way occasionally, for testing purposes. Prove it to yourself with a simple resistive load, like a 120V incandescent lightbulb.

p.s. It was reported that ewheels modified some older Gotways by adding diodes in the recharging harness to protect the port. So if you can't get any current out of your port even with simple resisitive loads: inspect the harness for warts.

Modern Gotway BMS should 'trip' and interrupt the port if there is excessive current flow. This 'trip' could be a nuisance that happens during the inrush when you connect your power supply, but I bet you could workaround it by manually cycling the connector twice, or through use of a precharge resistor circuit, inrush thermistor, etc. Don't connect your load (laptop) until your power supply is 'on.'

My MSX trips at a current around 20A; your MCM5 pack is smaller.

Try not to exceed whatever power level that you would recharge the wheel at. E.g. if you think the cables and connectors are good for 5A recharging, then don't use them above 5A for discharge either.

33 minutes ago, Kenzi said:

there are 4 pins on that charger port and I'm wondering if the other 2 are used for anything?

Some Gotway EUC's use pins 3 and 4 as redundant Bat- and Bat+ connections, in order to enable 10A recharging using all 4 pins. In response, some AC chargers are wired this way, too.

So, I don't recommend repurposing pins 3 and 4, since there are many GX16 EUC cables out in the world that have pins 1-4 shorted, and pins 2-3 shorted. If you ever (unknowingly) encountered one of those recharging cables, your modified pinout would have to tolerate this bridging.

Enjoy

Edited January 16, 2021 by RagingGrandpa

[mike_bike_kite]

19 minutes ago, Tawpie said:

Reminds me a little of Volkswagen before the 70s. They repurposed “power” sources so the windshield squirters use the air in your spare tire, and on their early campers the camping heater used the gas from your tank.

I still have a Beetle that does that, it's a 72 Cabriolet and it was my first car.

[Kenzi]

1 hour ago, RagingGrandpa said:

You can definitely yank 400w out of a Gotway charging circuit.

This entire reply is exactly what I was looking for. I imagine there is considerable inrush, so I'll try with a resistive load.  The charger is around 1A, so if the protection on the port is 1A that will work for the laptop under most circumstances but definitely not for the inrush... The incandescent will tell if it's reverse flow protection or current limiting, so thanks for the suggestion.

I'll check the other two pins to see if they're wired in parallel (and protected independently?) so I can maybe use them to distribute the load.

Worst case, I'll take it apart and replace the protection with something that can handle the inrush to my custom laptop power supply... or ... I don't know... put a soft-start on the input caps?

Thank you so much for the reply!

[Kenzi]

1 hour ago, RagingGrandpa said:

I bet you could workaround it by manually cycling the connector twice, or through use of a precharge resistor circuit, inrush thermistor, etc. Don't connect your load (laptop) until your power supply is 'on.'

Just checked it out.  It powers a 60W incandescent fine... so it's inrush.

I haven't put any load on the supply yet, so its definitely inrush. It flickers on then shuts down again. Cycling the connector might be getting me closer but it's still not enough.

Whatever is tripping in the wheel is fast; it resets fast too. Probably not a PTC?

I have no idea what IC my supply uses, otherwise I might be able to add/tweak a soft-start somewhere in there, since the output capacitance is twice as big as the input and is probably contributing to the inrush.  So I guess my only remaining options are to remove or alter the charge port protection or build a precharge resistor circuit.

[RagingGrandpa]

Try the 60W bulb in series... then bypass it once the power supply is on?
If this method works, you're probably just one carefully-chosen NTC thermistor away from a solution.

(Sorry I don't have component details for the modern BMS like ours which trips... but I agree, almost certainly not a PTC.)

Edited January 17, 2021 by RagingGrandpa

[Kenzi]

1 hour ago, RagingGrandpa said:

Try the 60W bulb in series... then bypass it once the power supply is on?
If this method works, you're probably just one carefully-chosen NTC thermistor away from a solution.

That worked for getting the power supply started. I'll test with the laptop shortly.

I don't know how to choose an appropriate NTC, much less do it carefully.  Can you help?

The laptop's original supply is rated for 19.5V 220W, but inrush aside I've never seen it consume more than 110W. The EUC is 20-cell, so operating range (that I'm interested in) is 60V to 84V.

The total capacitance on the buck appears to be 200uF, with another 400uF on the output.

[Kenzi]

I'm writing this post from an EUC-powered laptop! Now to solve the inrush problem...

[RagingGrandpa]

14 hours ago, Kenzi said:

I don't know how to choose an appropriate NTC

I was thinking: try a resistance that gives ~10A at 84V... 

https://www.mouser.com/Circuit-Protection/Thermistors/Inrush-Current-Limiters/_/N-axfxj

[Kenzi]

The inrush protection NTC worked perfectly!

Now suddenly the laptop doesn't like the supply... keeps flipping between charging and not.  There might be some kind of dip/surge, but it's hard to see on the meter, and no amount of capacitance is helping.  I think I need to work on a higher quality supply now. 

[Kenzi]

By the way - I did finally achieve my goal!  24h of laptop work, or 8 hours of VR gaming, plus enough power to ride it back home.

[supercurio]

15 minutes ago, Kenzi said:

By the way - I did finally achieve my goal!  24h of laptop work, or 8 hours of VR gaming, plus enough power to ride it back home.

Connected directly via an XT60 connector instead of a charge port: makes sense!

A couple questions:

• Is the XT60 is connected to all packs at the same time (I suppose)
• Did you design a 3D printed enclosure for the step-down converter(s)?

[RockyTop]

Awesome!   

    Your MCM5 looks kinda funny though? 

[RagingGrandpa]

3 hours ago, supercurio said:

Connected directly via an XT60 connector

Uhoh... so when a short-circuit happens in the laptop power supply, what makes the circuit safe again?

 

Edited September 2, 2021 by RagingGrandpa

[RockyTop]

She added a fuse at the battery. 

[Kenzi]

22 hours ago, supercurio said:

Connected directly via an XT60 connector instead of a charge port: makes sense!

A couple questions:

• Is the XT60 is connected to all packs at the same time (I suppose)
• Did you design a 3D printed enclosure for the step-down converter(s)?

Unlike my old MCM5, the S18 charge ports don't expose battery voltage, so I had to tap into the main PCB.

The S18 has a weird pack configuration - two 42V packs in series, which are in parallel with two 84V packs.  I connected to the sum 84V.  Using just one 42V pack would unbalance the packs and cause problems.

Yes, I designed and printed the enclosure.  It's too big.  The voltage status isn't necessary (the KingSong app works fine), and neither is the fan.  I'll make a smaller one eventually.

18 hours ago, RockyTop said:

She added a fuse at the battery. 

Yes, I added a 5A fuse.

Edited September 3, 2021 by Kenzi