Another quick and deadly high voltage charger

[heathyoung]

<Warnings>
This produces an unisolated DC voltage that sits 200V above earth potential - both leads are deadly, your pack must be isolated from the chassis. This can kill you without even trying! You MUST have discharge resistors or you will get a belt from the pins on the power plug! Ouch!
</warnings>

Hey all - after I had yet another bms battery charger exit stage left in a spectacular fashion (sigh) I decided to knock up something to charge the 44S Lifepo4 pack on the vectrix.

I have seen a few people use very simple capacitive reactance to limit the charge current, but these don't limit the termination voltage, so can't be left unattended or very bad (tm) things happen to your battery.

So I decided to take another tangent and design one based on a reactive divider - where basically you use capacitive reactance to create a voltage divider, that has no heat loss and current limits by default!

I needed about 160ish volts, and have 240VAC 50Hz, so settled on a 100uf and 120uf capacitor combination, a bridge rectifier and a smoothing cap. This gave a current limited 165V DC @ about 450W to change with, and evolved very little heat (apart from the bridge).

The main problem with these chargers is they are highly reactive - expect a PF of no better than about 33, so you draw current spikes 3 X your charger output (ie my 450W charger was drawing 10A spikes) - the power meter in the house (which doesn't account for PF) was showing a 1400W draw - 1000W would be pushing your luck on a 16A breaker I think. Don't use it anywhere where you pay for power in KVAr, unless they use really big motors, in which case you will be doing them a favour

With no output loading, the charger simply acts as a large capacitor across the mains (ie. PF = 0) and consumes no energy (except for the discharge resistors)

Caps I used were ironically massive PFC units from a scrapped UPS, but motor run (not start) caps would work, and even polarised electrolytics run back to back may work as well (if they have bypass diodes) bearing in mind that their ESR becomes an issue.


Thought that this may be useful to someone out there, who needs a quick and dirty solution till their charger gets fixed (or they build something better)

 

[Tommy L]

Nice!

I posted this yesterday. Any thoughts?

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=48964

 

[heathyoung]

Well, 72V isn't too hard especially when you have 110V at your disposal.

72V would be 24S Lifepo4 right? You need a charge termination voltage of ~88V. Since mains is sinusoidal, your rectified voltage (minus diode losses), (assuming your mains when measured is 120VAC - I think this is right?) 120 * 1.414 = ~170V DC.

170/2 = 85V - so close it isn't funny. A pair of equal sized caps, with some trimming required with some X mains rated class mylar caps would get you pretty close to your target voltage.

To figure out your peak current - you need to know the reactance of the capacitor you use - ie. http://www.kusashi.com/reactance-c.php

You have 60Hz (rather than my 50Hz) - so a 100uf cap would have a reactance of 26 ohms. Its maximum current it can deliver is 4A.

Bear in mind that the final voltage is determined by your mains voltage. This can have a regulator applied (a zero crossing triac driver + triac with a typical SMPS feedback mechanism) which would short across the second cap when the voltage exceeded a preset threshold.

Oh and its very dangerous - I think I may have mentioned this before

 

[megacycle]

Subscribed Heath.
Like your style. Bad boy charging on the fly.
Whats the voltage regulation like.
Have you scoped the output waveforms? expect initially heavy charging smooths the pulses some.
What happens as charge tapers off?

We could work on a 3 phase car coffee break charger

 

[megacycle]

Hey your method could help the grid in general.
There usually allowing for higher kVA with larger cabling, equipment, cap banks, cetra
Could introduce charging stations with a large kVAr near industrial areas and help balance out the grid 8)

 

[heathyoung]

The output wave form is pretty peaky - voltage regulation out of the box (ie no aditional regulation) is what you would expect - dependednt on the ratios of the reactance - so if you have a high mains input voltage, you have a high output.

Regulation due to current draw - basically ignore the second cap and calculate peak current as a function of reactance.

At full load, power factor is 33, at no load its zero (as expected).

Three Phase - yeah easy, with the added benefit of less peaky output to filter/beat up the batteries with. Remember the plates take the current pulses, not the average, so at high currents they tend to 'sing' (vibrate) as they distort.

I'll sketch up the regulator section, its pretty simple. It just shorts out one cap when the voltage gets too high (on the zero crossing - so it doesn't have to dissipate the energy in the cap) and the load becomes purely reactive for part of the cycle.

 

[megacycle]

The output wave form is pretty peaky - voltage regulation out of the box (ie no aditional regulation) is what you would expect - dependednt on the ratios of the reactance - so if you have a high mains input voltage, you have a high output.

Regulation due to current draw - basically ignore the second cap and calculate peak current as a function of reactance.

At full load, power factor is 33, at no load its zero (as expected).

Three Phase - yeah easy, with the added benefit of less peaky output to filter/beat up the batteries with. Remember the plates take the current pulses, not the average, so at high currents they tend to 'sing' (vibrate) as they distort.

I'll sketch up the regulator section, its pretty simple. It just shorts out one cap when the voltage gets too high (on the zero crossing - so it doesn't have to dissipate the energy in the cap) and the load becomes purely reactive for part of the cycle.

Was wondering with regulation what happens to the caps because they have the different functions were one seems to just set the divider voltage and the second one also has to handle load, any issues with impedance,temperature, with load affecting divider ratio.
The 3 phase car coffee break charger 8) yeh those extra overlapping pulses in the same period would be much cleaner when its rectified and access to 500V+ dc, edit 550V+, oohhh that would be very burny, double disclaimer, bodily contact with rectified 3 phase ac voltage can cause death or severe injuries. advise only electrically trained, competent persons, under supervision will full body protection like this attempt to even think about making 3ph rectified dc.

 

[jonescg]

Hey that's my race suit!

 

[megacycle]

Hey that's my race suit!

No such thing as to safe when your legs are wrapped around 200kW+.
One of the extra options for ya with the cal suit, a fibre glass jockstrap, for extreme ebikers