Page 2 of 2
Posted: Fri, 10 Jan 2014, 16:03
Just to eliminate any confusion:
Yes I understand, and had the same confusion on this point. In this case we are talking about conventional current flow, not the direction of electron flow in physics. So the conventional current flow is from +/- If you are putting the contactor on the positive wire between the source and the load, then A2 is closest to the battery or source and A1 is towards the load
Please let us know if you have further questions.
Office: 805-684-8401 x 222
Posted: Fri, 10 Jan 2014, 19:18
No that does not eliminate confusion. It applies to gigavac (A1,A2 part)
The generic rule is connect the designated + terminal to battery +. Simple .
It saves people pretending to understand current flow
I certainly don't. I can't tell green electrons from black ones.
Posted: Fri, 30 Oct 2015, 17:56
I have an interesting problem with reverse current through contactors.
I have a 50 V 400 Ah Lithium-ion battery connected to inverter/chargers that can charge or discharge the battery at 500 amps. I'm using two GX-14 contactors in parallel, bolted directly to 32 x 6 mm busbar on both the battery and the load sides. Also in parallel with these is a precharge circuit consisting of a 33R 100W resistor and a MiniTactor. I have the GX-14s oriented so that discharge current flows in the recommended direction for the magnetic blowouts. I understand they are not intended to interrupt current flowing in the reverse direction (charge current in my case). But I assumed arc-blowout would not be a big issue at 50 V, with very little inductance (only 3 m of cable) between the battery and inverters.
However, when I turned off the coil drive to the contactors while there was a 100 A reverse current flowing, they did not drop out! I checked with a multimeter that the voltage across their coil wires was zero, and there was only 40 mV between busbars (across their contacts). They were not welded, as they dropped out when I switched off the AC power to the inverters. So it seems that not only do the blowout magnets blow the arcs inward when the current is reversed, they also "motor" the conducting bar against the terminal contacts too.
Does anyone have a solution for bidirectional DC currents like this? Perhaps a contactor with no blowout magnets and hence lower voltage ratings? Or do I need to use GX-14s back-to-back in series?
Posted: Fri, 30 Oct 2015, 20:07
I don't have the answer but I'm following the question with interest.
Posted: Fri, 30 Oct 2015, 22:01
I have a feeling that it's going to be cheaper to use two series contactors.
http://www.schaltbau-gmbh.com/en/Produc ... rectional/
(I googled: DC contactor for bidirectional current)
Posted: Sat, 31 Oct 2015, 06:59
Thanks Johny. Those things are huge. And as you suggested, I've decided to go with pairs of contactors in reverse-series, and so have ordered two more GX14s from Rod Dilkes at EV-Power Australia.
I also emailed the question to Gigavac and will let you know what response I get. But I couldn't wait for it.
Coulomb pointed out that, because it's "only" 500 A, I could have replaced the existing two parallel GX14s with one EV200 and then put another EV200 in reverse-series with that. I wish I'd used EV200s from the start, but the prospect of having to remake busbars and lengthen cables (because the holes for the GX14 studs are 10 mm while the studs on the EV200 are only 8 mm and closer together) was too daunting.
At first I thought the Gigavac MX14 must be a no-blowout-magnet version of the GX14 since it's the same physical size and its contact voltage rating is only 48 V (as opposed to 750 V for the GX14). But the datasheet and application notes make it clear that the MX14 has exactly the same polarity requirements as the GX14. Sigh.
Another annoying requirement of both the GX14 and the MX14, and hence a reason to prefer the Kilovac EV200s, is that the Gigavacs aren't supposed to have freewheel diodes across their coils (as you would typically have to protect a MOSFET that drives them) because this slows their release too much. This means the MOSFET voltage rating must be about twice what is usual, as the maximum back-EMF from the GX14 or MX14 is specified as 55 V, which appears on top of the 12 V of the coil power supply.
I've said in the past that I would not design a system that required more than 250 A continuous (which can be cabled with two 50 mm^2 in parallel). Any system that requires more current, has too low a voltage. But the decision on this system was made before I was brought on board.
Posted: Sat, 31 Oct 2015, 07:12
BTW, apart from the apparent "motoring" of the contact bar preventing it from dropping out (if this hypothesis is correct), the reason that reverse currents might not be interrupted is that the blow-out magnets become "blow-in" magnets, and the two arcs that start at the ends of the moving contact bar may actually meet in the middle and then continue to arc indefinitely. Although this is unlikely to occur at 48 V with contactors of these dimensions.
An alternative to the motoring theory is that the contact bar was lightly welded, possibly due to the slow release caused by the freewheel diodes. But broke away the next time it was operated. I may have mis-remembered it dropping out as soon as the AC was turned off. But this seems unlikely.
Posted: Sat, 31 Oct 2015, 14:05
Either way it's not good to have them closed when you want them open. Disturbing stuff.
Some relay manufacturers recommend resisters across the coils to minimise back EMF where diodes would slow opening down.
Posted: Sat, 31 Oct 2015, 14:41
Yesterday I sketched two contactors on a notepad started typing a suggestion of what I'm assuming is (reverse-series) then deleted it thinking I had overlooked some thing.
So (reverse-series) That's Series connection but tapping into the center and feeding from both ends?
Posted: Sat, 31 Oct 2015, 15:37
If we use a diode symbol -->|-- to represent any simple two-terminal device that behaves differently depending on the direction of current or voltage, then the following diagram shows the four ways we might usefully combine two of them to make a single compound two-terminal device.
-->|----|<-- anti-series or reverse-series
--| |-- parallel
--| |-- anti-parallel or reverse-parallel
In all cases the terminals of the compound device are on the right and left. There are no centre-taps.
Posted: Sat, 31 Oct 2015, 17:45
There's some discussion in this 2010 thread re contactors not turning off.
viewtopic.php?title=contactors-gigavac& ... 988#p28695
Posted: Mon, 02 Nov 2015, 22:48
If the ac is turned off that is supplying the charge current and the dc from the solar then the current from the battery can only be out going I assume the battery management system is trying to save the battery from over charge if the current is going in. Fusing will protect from over current going out .
Posted: Tue, 03 Nov 2015, 00:10
DavidC wrote: If the ac is turned off that is supplying the charge current and the dc from the solar then the current from the battery can only be out going I assume the battery management system is trying to save the battery from over charge if the current is going in. Fusing will protect from over current going out.
The BMS must be able to disconnect the battery to protect cells against overcurrent going in or out -- overcharge or overdischarge. Fuses are mainly there for short-circuit protection which, as you say, will only be due to outgoing current. And of course the BMS has to disconnect the battery when someone presses the emergency-stop switch.
The fourth paragraph of this excellent post from 2008, in the DIY Electric Car forum, by a Gigavac person, makes my second theory (slight welding and too-slow release) sound far more likely now than my first theory (contact bar motoring due to reverse current).
http://www.diyelectriccar.com/forums/sh ... stcount=10
GIGAVAC GX contactors have been designed to withstand some tack welding. Some tack welding is good because the weld reduces the contact resistance. To break the tack weld, our GX contactors have a built in impact break mechanism that acts as a hammer that breaks the tack weld when power is removed from the coil. When a diode is put across the coil, the release time becomes slow & sluggish and the impact break does work property. It’s got to be a quick break.
And I foolishly thought it would be fine to close the main contactor when there was still 16 volts across it (about 25% of the battery voltage). I will reduce the value of the precharge resistor to 3R3 so the voltage across the main contactor can get down below 5 V (10%) despite the standby current of the inverter.
But I still need to connect contactors in anti-series, since they are not rated to interrupt reverse current at all.
It seems someone in Perth bought up the entire Australian stock of GX14s on Friday. Neither EV Power, EV Works nor Elmofo/Solar-online have any. So I'm paying crazy FedEx fees to get two from the US in a reasonable time (I hope).
Posted: Tue, 03 Nov 2015, 04:57
Sorry Webber I may have not got my thoughts clear. If the Bms starts protecting the battery by turning off solar charge current and then turns off the ac into the inverter / charger then the only current flowing will be load current from the battery. If voltage is going high from over charge there is no need to drop load ac or DC . If battery voltage is going low ac load should be disconnected first if voltage still drops ie fault in inverter still drawing more out than in the DC will need to be opened . This is all controlled switching by a clever BMS I am assume the system has 4 contactors ( DC solar in , Ac charge in , ac load out of inverter , DC out of battery ) The estop being pushed should open circuit the control supply in both legs at the source This is supplied to seperate relays that have a monitoring function to indicate if one is not in the same state. Ie both open or closed There are specially safety relay for the use in estop circuits. This may or may not solve the problem of having to switch off a contactors with current flowing in the reverse direction. Some estop relays will even give a 1 to 3 second delay on some contacts so the current to the battery is disconnected before the battery contacts are opened .
Posted: Tue, 03 Nov 2015, 06:15
Ah! Yes I can't blame you for assuming it would be like the Black Monolith. But this is a demand charge management or peak load shaving system with no solar input and no separation of load and AC input. The inverter/charger only uses a single AC port that operates in parallel with the grid and the loads and is bidirectional. It can be commanded to put any amount of power in or out (up to its maximum in either direction).
At present the E-stop and BMS do not control a contactor on the AC port. That is handled by separate manually-operated circuit-breakers on a board just above the E-stop.
But if the cells get stressed the BMS tells the system controller, which should command the inverter/charger to back off the current. If it does not do so, the only thing the BMS can do is disconnect the battery.
Do you think we need to add AC contactors under the BMS' control too?
[Edit: "controlled" -> "controller"]
Posted: Tue, 03 Nov 2015, 13:33
An Ac contactor should be cheaper than the DC ones for the same power rating and if included would island the equipment in the estop event. Ie smoke. The standard shut down for inverters is to isolate the ac first this stops the DC current and saves arcing at the DC isolation device. The 100 amps ac from the inverter is easier to break than the 500 amps from the battery.