DIY Electric Car Forums banner
1 - 11 of 11 Posts

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #1 ·
Hi All: my ElCon just blew up. $1000 CDN down the toilet, after 7 yrs of service. Fan runs, no blinkenlights, no current.

No worries- my original E-Fire pack, after years of faithful service in first the E-Fire and now in the ER-6, is getting long in the tooth anyway. One day I'll have to come up with a new system- possibly some Leaf or Volt batteries etc. But for now, I just need to charge the car so I can have a fun pandemic grocery-getter.

Details: it's 33S 1P LFP 180 Ah Sinopoly cells, with a miniBMS (celltop board on each cell)

So I bought a variable autotransformer (variac) on Amazon- cheap, no certifications, but a good price for 20 amps capacity. And I have a high current bridge rectifier kicking around, so I thought I'd just try to full wave rectify the output of the variac, stuff it into the pack and see what happens.

Yeah, I know- a variac isn't isolated, so I'm tempting fate a bit. But my pack IS well isolated from the chassis on both the + and - sides, and I thought I'd give it a go.

I thought I'd need a resistive ballast between the pack and the variac to make the coarse knob adjustment low enough in sensitivity to avoid blowing fuses, but it's actually no big problem. The knob adjustment dials in a peak current quite nicely and stays there. Dialing in currents between 1 and 10-12A (my desired charge current) is easy enough to do.

My question is this: my assumption was that because I'm only charging at around 10-12 A from 120V just like I was with the ElCon, that the pack would act like a giant bank of electrolytic caps and not require any filtering capacitors- which obviously are a bit of a nuisance because if I had them, I'd need to rig up a way to precharge them etc.

I dug out my old osciloscope and measured the pack voltage ripple- as expected, it's very minor- the pack keeps the voltage very steady. But when I put the scope across my current shunt, I did see what looks like a pulse of current for half a cycle, then flat line (no current) for half a cycle- which I found weird, given I'm using a bridge rather than a single diode. Both my clamp on DC meter and my E Expert Pro Ah meter in the car give a nice steady current reading, and the two match- so I'm not sure if they're averaging this, or both messing up in a similar way- or perhaps my scope measurement is lying to me.

So: whadayathink? Do I need filtering caps? Am I at risk of wrecking something? Given that I'm charging at ~ 1/20 C, I figured I'd be fine, but am I at risk of wrecking something? The old rule of thumb I used in past was 1000 uF per ampere, so I'd need 10,000 uF of caps to flatten this thing out.

The intention is to have the BMS trip a solid state relay on the AC input to the variac. My pack always terminates charge when one cell goes high anyway. I'll back it up with a timer set to terminate charge prematurely so I can be around during the final stages of charging anyway.

Your thoughts as always are very welcome. The ER-6 has been a blast, but I'm staying close to home during COVID anyway and since the crash with the E-Fire I keep the ER6 off the highways anyway. It's just really an ultra-cool grocery getter at the moment, so no danger of being stranded somewhere with the need to charge.
 

·
Registered
Joined
·
150 Posts
Hi All: my ElCon just blew up. $1000 CDN down the toilet, after 7 yrs of service. Fan runs, no blinkenlights, no current.

No worries- my original E-Fire pack, after years of faithful service in first the E-Fire and now in the ER-6, is getting long in the tooth anyway. One day I'll have to come up with a new system- possibly some Leaf or Volt batteries etc. But for now, I just need to charge the car so I can have a fun pandemic grocery-getter.

Details: it's 33S 1P LFP 180 Ah Sinopoly cells, with a miniBMS (celltop board on each cell)

So I bought a variable autotransformer (variac) on Amazon- cheap, no certifications, but a good price for 20 amps capacity. And I have a high current bridge rectifier kicking around, so I thought I'd just try to full wave rectify the output of the variac, stuff it into the pack and see what happens.

Yeah, I know- a variac isn't isolated, so I'm tempting fate a bit. But my pack IS well isolated from the chassis on both the + and - sides, and I thought I'd give it a go.

I thought I'd need a resistive ballast between the pack and the variac to make the coarse knob adjustment low enough in sensitivity to avoid blowing fuses, but it's actually no big problem. The knob adjustment dials in a peak current quite nicely and stays there. Dialing in currents between 1 and 10-12A (my desired charge current) is easy enough to do.

My question is this: my assumption was that because I'm only charging at around 10-12 A from 120V just like I was with the ElCon, that the pack would act like a giant bank of electrolytic caps and not require any filtering capacitors- which obviously are a bit of a nuisance because if I had them, I'd need to rig up a way to precharge them etc.

I dug out my old osciloscope and measured the pack voltage ripple- as expected, it's very minor- the pack keeps the voltage very steady. But when I put the scope across my current shunt, I did see what looks like a pulse of current for half a cycle, then flat line (no current) for half a cycle- which I found weird, given I'm using a bridge rather than a single diode. Both my clamp on DC meter and my E Expert Pro Ah meter in the car give a nice steady current reading, and the two match- so I'm not sure if they're averaging this, or both messing up in a similar way- or perhaps my scope measurement is lying to me.

So: whadayathink? Do I need filtering caps? Am I at risk of wrecking something? Given that I'm charging at ~ 1/20 C, I figured I'd be fine, but am I at risk of wrecking something? The old rule of thumb I used in past was 1000 uF per ampere, so I'd need 10,000 uF of caps to flatten this thing out.

The intention is to have the BMS trip a solid state relay on the AC input to the variac. My pack always terminates charge when one cell goes high anyway. I'll back it up with a timer set to terminate charge prematurely so I can be around during the final stages of charging anyway.

Your thoughts as always are very welcome. The ER-6 has been a blast, but I'm staying close to home during COVID anyway and since the crash with the E-Fire I keep the ER6 off the highways anyway. It's just really an ultra-cool grocery getter at the moment, so no danger of being stranded somewhere with the need to charge.
Nope, you don't need no stupid caps. You need a big iron core inductor in series with the output. If this reduces the current too much, place a 30 or so mfd ac motor run (not start) capacitor in parallel with the DC output. This will increase your power factor and stretch that narrow spike of AC current way out.

That was easy.
 

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #3 ·
Thanks for the reply. Got to admit I'm a little out of my depth here.

By "big iron core inductor"- would a transformer secondary work? I may have something lying around with a secondary with wire heavy enough for a 10A DC current without getting too hot, but otherwise I'd be winding some 14 ga wire around a chunk of iron and trying my luck...Of course the primary would still be there and it would be a danger, because it would be generating quite high voltages- so maybe not...Maybe a field coil out of an old power tool? Again, there's a significant DC current flowing so I'd want a significant chunk of wire or else the DC resistance would be too high.

I understand, I think, what the inductor would do- it would resist the rising current by "charging" its magnetic field, and then resist the falling current by discharging that field, broadening the current pulses- which is what I want- I think...

I'm puzzled by the suggestion of the AC cap between the + and - of the DC side (i.e. in parallel). The voltage is DC and ain't moving much up or down- it's being held pretty constant by the pack. It's the current waveform that is pulsile, but it never goes negative, ie. current direction never reverses because the diodes in the bridge stop that. So why would an AC cap be any different than a 30 mfd DC cap, i.e. doing not much of anything?.
 

·
Super Moderator
Joined
·
1,354 Posts
I don't think you need an inductor either.

If you're getting a half cycle with no current flowing, I can't think of any cause other than that your FWB is busted or connected to the wrong terminals.

Current pulsing isn't an issue for you, nothing's going to overheat if the RMS of the waveform is below the Variac's peak. Yu don't pay for power factor on a residential power meter, so, I wouldn't bother.

Are you able to just connect the variac and let it charge on its own, or are you having to dial it up yourself, manually?

I didn't really have plans for a charger for my own thrifty build, was just going to use a variac until I had everything else working, good to hear it's doing the job just fine.
 

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #5 ·
I'm going to check, but I'm pretty sure I got the timebase wrong. What I'm pretty sure I'm seeing is that I get a peak of current for about 1/2 of each HALF cycle, then no current flow for the other half. As the voltage output of the bridge rectifier rises, it eventually reaches the voltage of the pack and forces current into the pack. It reaches its peak at the peak of the input AC waveform, and then symmetrically falls again to the point where voltage matches that of the pack and current flow stops.

Sketch attached. The top graph is the voltage output of the fullwave rectifier before it's connected to the pack. The real voltage waveform is just a very small ripple of around a volt or so, which my scope can't actually quantify.

The lower graph is what I assume my scope was showing me when I watched the voltage across my shunt.

The full wave bridge rectifier is working fine and it's pretty hard to screw up the connections, even for me!

I'd more or less concluded the same thing- just don't worry about the power factor.

I have the bridge on a good heatsink and it needs it- it gets pretty hot. And I have a fan directed at the side of the variac which is delivering the current. An optoisolated solid state relay connected to my BMS will turn off the AC supply to the variac when any one cell reaches HVC.

All in all I'm pleased- it was a cheap solution rather than going out and buying a Thunderstruck etc.
 

Attachments

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #6 ·
@MattsAwesomeStuff - I'll let you know how it goes, but I did shut off the variac's AC switch, let it sit and then turn it back on again- current starts and stops without problems. No need to wind the variac back down to zero and wind it back up again each time, if that's what you were worried about. So I'm imagining that I'll be able to set a maximum current and then just initiate a charge and let it run until the BMS trips it. As the voltage doesn't rise much between low and high SOC, current might fall off a bit through the charge but I doubt it will be much.
 

·
Super Moderator
Joined
·
1,354 Posts
I'm pretty sure I got the timebase wrong. What I'm pretty sure I'm seeing is that I get a peak of current for about 1/2 of each HALF cycle
Ahh, yes, that makes more sense.

You could throw a cap or inductor on there, sure, but, the cap is just going to be extra greedy during the peak times, so that it can supply some current during the valley times. So you'll still see it be quite spikey. Again for you, it doesn't really matter. The peak current could be 25 amps, as long as the average current is 10 amps or whatever you want it to be, I'd say good enough. You're charging the batteries so slowly it's not like they can't handle micro-bursts of 3x current.

I'll let you know how it goes, but I did shut off the variac's AC switch, let it sit and then turn it back on again- current starts and stops without problems. No need to wind the variac back down to zero and wind it back up again each time, if that's what you were worried about.
Indeed. I somewhat figured that as a vehicle-sized lithium pack is functionally a bottomless current hog, that to charge a 400v pack from say, 300 volts, you'd have to set the variac to 301v, then 302v, then 303v, and very slowly inch it upwards, else with too large a voltage difference you'll get it trying to suck in 1000 amps like a dog in front of a firehose.

How big of a voltage difference is your pack allowing, to keep the current at reasonable levels?

I'm not concerned about the reverse, i.e. running out of enough voltage difference to make enough current flow. I'm worried about too much current blowing a breaker.

On the variac, the coil itself I'm not worried about, air cooling can take care of that, it's the arcing to the wiper brush that scares me.

Plan B if a fixed-ratio transformer works, is to just rewire a pair of microwave oven transformers and let that be my charger. It'll be enough to max out a 120v circuit anyway.
 

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #8 ·
Right now my pack is happily sucking 10 A from the DC side of my bargain basement charger. It's been steady for a couple hours so far- my pack was down about 93Ah from as close to top balanced as it gets these days, when I had to open my master switch and hence lost my previous Ah reading (my E Expert Pro doesn't carry the current Ah value in nonvolatile memory- it rests to flashing 0 Ah when you shut off power to it, and it doesn't count backward (i.e. it won't count + Ah values) which is very annoying. So now I have only my manual monitoring of current, and the timer on my phone, to keep track of how many Ah I've stuck into the pack. So I go out every hour and check on it, just to make sure the outdoor GFCI outlet I have it plugged into hasn't tripped due to fatigue (they tend to trip at less than their full rated current if you run them too close to the full rated current for too long).

I can sneak up on 10 A very easily without being too careful with the dial, and once set, the current seems to stay steady despite small rises in pack voltage. I'll let you know what it looks like over a full charge though- how much current droop I get by the end of charge and whether it necessitates opening the hood and cranking the knob up a bit. But then again, let's be clear- I'm generating 109 V right now from 120 VAC which, if the variac were dialed all the way up, would be making 120 x 1.41 = 183V peaks. My variac dial is set somewhere between 1/2 and 3/4 of its full setting- I don't really look at the setting, just at my clamp-on ammeter while I'm changing the dial position.

Once the dial is set, so far I've just left it there. The brush isn't moving so there's no arcing. It's sitting nicely on two or three turns of the variac's windings.

Could you get away with a fixed transformer? Probably, though it would take a bit of fiddling to get the precise number of secondary turns required- and you'd need something with a very heavy secondary and quite a large core to transfer that much energy. The variac is just easier- and at $120 CDN from Amazon, with free Prime delivery... it didn't break the bank.
 

·
Super Moderator
Joined
·
1,354 Posts
I'm generating 109 V right now from 120 VAC which, if the variac were dialed all the way up, would be making 120 x 1.41 = 183V peaks.
Yeah, I would never trust it blindly, without an overvoltage cutoff elsewhere (or two, for redundancy). I think I'd even insist on manual restart (i.e. if voltage hits peak once, it trips the relay and the relay never re-engages automatically).

You could, I suppose, throw a small cap across the output and then with no load see what the voltage hits, and have that be your absolute ceiling, but I don't trust it.

Could you get away with a fixed transformer? Probably, though it would take a bit of fiddling to get the precise number of secondary turns required- and you'd need something with a very heavy secondary and quite a large core to transfer that much energy.
Actually, in my case, I'd need something with thinner secondary wiring, as I'll be charging 400vdc from 120vac. In your case, you've almost just got an isolation transformer, you're not actually altering the voltage much.

I once had a plan for a 170vdc pack on an E-bike, so I didn't need a transformer at all, just a power cord with an FWB and a capacitor :p. Was going to ride it coast to coast, and weight and size mattered. Could just park behind any fast food place and plug in while I ordered. Actually that kind of still is my plan for that bike, just never got around to finishing the battery pack.

The variac is just easier- and at $120 CDN from Amazon, with free Prime delivery... it didn't break the bank.
1000 watt transformers are free, all day long out of old microwaves. Just have to rewind a new secondary. Heavy and cludgey for a charger, but, on a budget build, price is right.
 

·
Registered
Joined
·
1,469 Posts
Discussion Starter · #10 ·
Just letting everyone know that I've been using my bargain basement Variac based charger consistently now and it works very well and reliably. I don't need to adjust the variac knob between charges, and the current droop between empty and full isn't enough to bother me. The solid state relay connected to my BMS takes care of terminating the charge.
 

·
Registered
Joined
·
916 Posts
It would be a tremendous contribution to the species to write up a step by step HowTo for noobs on this!

This would be a great site for hosting it:

 
1 - 11 of 11 Posts
Top