Power supply help please

[Matt Sherman]

I bought the icharger 4010 duo and the website says quote *** This charger MUST be used with a stable power supply, it is not recommended to be used with an automobile as a power source, or using multiple server power supplies in series as a power source ***
I was going to use 2 700 watt computer server power supplies together but after reading that I am a little worried. What would happen if I use 2 power supplies together and is anybody using 2 power supplies with this charger, does it work alright?

All help is appreciated, thank you!

[al93535]

If one power supply fails, the voltage drops and the other supply tries to ramp up the amperage. Usually the single PS will not be able to supply the required amps. The charger will try to keep charging your batteries using the lower voltage while drawing many more amps. This is where you get into trouble.

If both supplies shut down, no problem. If the I charger would just shut down, no problem.

Single PS and charger is indeed alot safer.

Apparently alot of chargers are damaged to the point of having to return them. And that's why there is the new warning.

[RedOctobyr]

Like he said.

I was going to use 2 server supplies for my 308Duo, then the series-supply discussion started. I went with a good deal on a single supply instead.

If you use two, I'd do my best to ensure that neither will shut down during use.

- I'd overbuy on capacity. The DPS-1200FB can put out 900W, 75A, I think. Two in series can ensure you won't over-draw from the supplies, triggering one to shut down. The 4010Duo can draw 65A max.

- I'd test each supply individually first. Do a charge pulling 65A, if the supply can manage that. If it passes, great. Test the other. You don't want one supply solid to 65A, but the other shutting down at 60A. The first time you do a powerful charge, and you hit 60A, you will trigger the bad situation.

- If they can't supply 65A, set up the charger limits appropriately to avoid over-drawing anything. You can leave a buffer, limiting the charger's draw something less than the supplies can provide. This will help reduce the risk of a supply shutting down.

- You might consider 3 supplies, vs 2. Higher voltage (36V vs 24V) will keep things running cooler, and you'll stay further-below the max amps for each supply.

Many people run series server supplies, I have no doubt. With some care, you can minimize the risks.

You could also buy this, for peace of mind with series supplies. A circuit developed to help ensure you don't just have 1 supply drop out, while still providing voltage to the charger. This helps ensure you get all-or-nothing, which is the safer situation.
http://www.rcgroups.com/forums/showthread.php?t=2070862

[Matt Sherman]

Can I really use 2 of those DPS-1200FB power supplies? Each one of those puts out 75 amp and the icharger 4010 says it can only handle a maximum of 65 amps. Won't that burn the charger out?

[RedOctobyr]

Nope. The charger will only draw what it needs.

Supplies only "offer" current, you could say. They won't "force" more than the device requests.

Don't use a supply with a higher voltage output than your device can handle, however!

My 106B+ can only take about 22A max. I've run it for several years with a 47A supply. That's perfectly safe. My big supply for my 308Duo can provide 70A, the charger can only pull 60A.

I'd prefer my supply be capable of more current than the charger can draw. That way I don't have to worry about setting lower limits in the charger, to avoid overloading the supply.

Sent using Tapatalk.

[Matt Sherman]

deleted

[TheBum]

Originally Posted by RedOctobyr

You could also buy this, for peace of mind with series supplies. A circuit developed to help ensure you don't just have 1 supply drop out, while still providing voltage to the charger. This helps ensure you get all-or-nothing, which is the safer situation.
http://www.rcgroups.com/forums/showthread.php?t=2070862

If you want to do something equivalent on the cheap, and power on/off can be controlled by connecting/disconnecting two control pins, you can use two relays w/ 12VDC coils, a couple of diodes (to shunt the coil charge when power goes off; some relays come with built-in shunt diodes), and a momentary pushbutton. The relays only need to be SPST. Radio Shack has everything needed.

I'll try to post a schematic later, but the basic concept is that you use the relay powered by one PS to switch the other PS's control pins. Of course, this won't work when neither supply has power, so you need a momentary pushbutton to initially connect the control pins on one PS, which will cause that PS to come on. When that happens, the control pins on the other PS will be connected and it will come on. The two PS will sustain each other's power when you release the button.

This approach is also easily extensible to any number of power supplies. For example, with three PS: PS1's relay could switch PS2's control pins, PS2's relay could switch PS3's control pins, and PS3's relay could switch PS1's control pins.

[Gregor99]

Originally Posted by TheBum

If you want to do something equivalent on the cheap, and power on/off can be controlled by connecting/disconnecting two control pins, you can use two relays w/ 12VDC coils, a couple of diodes (to shunt the coil charge when power goes off; some relays come with built-in shunt diodes), and a momentary pushbutton. The relays only need to be SPST. Radio Shack has everything needed.

I'll try to post a schematic later, but the basic concept is that you use the relay powered by one PS to switch the other PS's control pins. Of course, this won't work when neither supply has power, so you need a momentary pushbutton to initially connect the control pins on one PS, which will cause that PS to come on. When that happens, the control pins on the other PS will be connected and it will come on. The two PS will sustain each other's power when you release the button.

This approach is also easily extensible to any number of power supplies. For example, with three PS: PS1's relay could switch PS2's control pins, PS2's relay could switch PS3's control pins, and PS3's relay could switch PS1's control pins.

It won't be fast enough to prevent damage. The threshold set by Revolecrtrix is 100ms response time from the point the supply starts to lose power.

[TheBum]

Originally Posted by Gregor99

It won't be fast enough to prevent damage. The threshold set by Revolecrtrix is 100ms response time from the point the supply starts to lose power.

Not even a relay with a 0.5ms response time? The ones I ordered from Digikey are rated at that speed. I'd be willing to recant my statement about getting the relays at Radio Shack because I'm sure those relays aren't that fast.

[RedOctobyr]

Alan, I floated what seems like a similar idea last fall in the Duos thread on RCG.

My thought was a few small relays powered/driven by the 12V output of each PS. Connect *their* outputs in series, used to power/drive a *big* relay which is wired into the main output from the group of supplies to the charger. If any PS shuts down, its relay opens, killing power to the big relay, which opens and kills the circuit to the charger.

I was told relays are too slow for that, and that FETs would be much better, as they can respond extremely quickly. This is apparently why I'm not an EE

I eventually found a good deal on a single, high-wattage industrial power supply, and went that route instead. Much simpler, and I paid about what people charge for a pair of already-modified server supplies.

[Gregor99]

Originally Posted by TheBum

Not even a relay with a 0.5ms response time? .

Its not just the relay, its the whole system. The first supply must lose enough power for the relay to trip. Then the second supply must power down. Just because you put the supply into standby does not mean the voltage instantly shuts off.

On the Powerlabs, the charge cycle is terminated within 100ms of supply issues. A scope may be required to see how fast end result is with the dual relay setup.

[TheBum]

I can speed things up with a FET feeding the relay coil. That will allow relay cut-off at a higher voltage. I'll try it without first and see what kind of response times I get without a load (worst case scenario).

Does anybody know how long an iCharger can tolerate this condition before damage occurs?

[RedOctobyr]

If you're talking about using it for testing, I would strongly propose using something else as your load on the supplies

Some people have Duo damage when a supply shuts down, some don't. I don't think there is any more concrete data than that.

Last year Junsi posted some additional explanation of why the sudden voltage drop is a problem, in the Duos thread on RCG. Their hardware cannot react fast enough to manage the condition safely, unfortunately.

But there's nothing like "2 seconds is ok, don't go beyond 4".

Sent using Tapatalk.

[TheBum]

Originally Posted by RedOctobyr

If you're talking about using it for testing, I would strongly propose using something else as your load on the supplies

Note I said that I'd test it under no-load conditions. Under load, the voltage will drop faster. If I determine that everything shuts off within a few milliseconds with no load, I'll be reasonably certain that it's safe. If not, I'll measurement a typical non-charging load (charger on but idle) and substitute an equivalent load for further testing.

[Coolice]

Hey guys.

I've got to write carefully as the product spoken about is mine, developed in association and with the expert electrical skills of Xygax, the developer of SimStick.

My original idea involved a relay, but their reliability is an issue and I was better advised this is the best route.
The only other thing I can say is that I've saved a few chargers now and offer a simple plug and play safety feature for the DPS-600's.

I have a new item I will pop in the vendor pages soon for you to look at.
.

[TheBum]

You mentioned on RCG that you have one for DPS-1200 power supplies. I'm setting up a couple of those, but instead of soldering wires to the power pads, I'm using modified card edge power connectors. I don't like the idea of having something as massive as 10AWG wire pulling up on those pads; the pads are not designed for that sort of stress.

Would your DPS-1200 solution work in that case?

[Coolice]

Originally Posted by TheBum

You mentioned on RCG that you have one for DPS-1200 power supplies. I'm setting up a couple of those, but instead of soldering wires to the power pads, I'm using modified card edge power connectors. I don't like the idea of having something as massive as 10AWG wire pulling up on those pads; the pads are not designed for that sort of stress.

Would your DPS-1200 solution work in that case?

Hey Buddy.

Yes I do, same idea and circuit just without the connector block and with the required resistor in the switching lines.

I've recently switched to the edge connector setup, a couple of my charge case builds experienced a PSU failure (to be expected with the odd second hand supply, even after extensive testing) and so I've chosen to provide the faster means of PSU replacement.
But yes, put simply you can wire my circuit to the edge connector pins no problem.

If you'd like to know more PM me your email and we can chat without my infringing the forum rules.

Ian

[TheBum]

I think I figured out one of the main issues with using relays: the specs show that there's significant hystersis between the turn-on voltage and the turn-off voltage. The voltage would have to drop pretty low before the relay would turn off, which could take some time.

I went ahead and ordered one of Coolice's IceMeters for a permanent solution, but since I have the parts, I'm going to do a modification of my relay idea in the interim, mainly because I have a fun fly to attend this coming weekend. I should be able to address the hysteresis issue by driving the relay with an N-channel MOSFET switch. The input to the MOSFET will be a potentiometer acting as a voltage divider between V+ and Gnd, which can be used to calibrate the shut-off voltage.

[RedOctobyr]

Bum, I just saw this thread on another site, about trying to make a relay turn off really fast.

It is pretty much over my head, but maybe it means more to you, perhaps it can be useful:
http://www.eevblog.com/forum/project...urn-off-relay/