Meanwell PS creates a load?

[brulaz]

Recently realized that my solar systems daily output had increased by ~50Wh per day when in storage with no shorepower.

It was hard to believe that the parasitic loads or the battery self-discharge had increased by that much as the only change I had made was to install a Meanwell 27V 500W adjustable power supply to charge up my 24V battery bank when on shorepower. I assumed that when it's unplugged from 120VAC the Meanwell would have no load on the batteries, so I hard-wired the DC output through a fuse block to the 24V battery bank.

But after pulling the DC fuses in the Meanwell's output, that new 50wh per day load has disappeared. (Someday I'll have to measure its actual current draw with my clamp-on.)

Do other PS/charger/converters do this? I don't recall such behaviour with the WFCO or Prog. Dyn. when boondocking, but never really looked at it closely.

Anyway, it's not that big an issue as I don't need the Meanwell that much. Usually solar supplies everything we need, so I'll just leave the fuses out until it's needed. But it is unexpected.

[westend]

I've used these cased IXYS bridge rectifiers with pretty good success: IXYS Bridge rectifiers
You can search around at Digikey for the one that best fits your use but IXYS makes a FRED device that runs cool.

[MEXICOWANDERER]

"Thermal Runaway" THIS... 🙂



500 Amp - 30 V Schottky Rectifier - CPT50130 - 2 Diodes

http://www.ebay.com/itm/500-Amp-30-V-Schottky-Rectifier-CPT50130-2-Diodes-/282361292454?hash=item41b...

This is what I have in the BORG. Only the 60-volt version. Yes the plate is the cathode, so it needs to have a heat sink and the cathode and/or the heat sink needs to be kept insulated. A 5/16" tap can be used to thread the mounting holes. I used 5/16" x 24 silicon bronze bolts. Threads cut off, screwed flat into rectifier plate, then flipped over and soldered. Berquist Sil-Pad 2000 insulator pad to isolate rectifier plate from heat sink. Arctic Silver Epoxy to glue everything together. Meaning both sides of the Sil-Pad

Working in tandem the Megawatts at max dump 38.8 amperes, and 39.6 amperes through the schottky devices. .44 vf at max I. Anode temp Delta T shows 11F and 12F respectfully. Almost zip.

This rectifier shows .22 vf at one ampere. PIV leakage of the 60 volt device maxes out at 3ma at 55 volts.

The caveats are: Price, size, work involved to create.

[brulaz]

Finally read the manual of my clamp-on ammeter. For repeatable readings I need to zero it each time. It has a 1ma resolution. Anyway, 70-75ma seems about right for my Meanwell's back current when unplugged.

So, just for fun, I ordered a few, very cheap Schottkey Rectifier type MBR60100PT (60A, 100V, TO-247 case) from China. The MBR60100PT has two diodes with a single joined cathode lead but separate anode leads. I sandwiched the TO-247 case between two pieces of copper angle with the cathode folded over and soldered to one angle and the anodes to the other. So 2 diodes in parallel.

And sure enough, the Meanwell's back current dropped to 0-5ma when unplugged.

When I then tested the Meanwell under max load charging the batteries, it put out 24A forward current at 25.2V (605W!). Using an infrared temperature gun, the copper mounts got up to 80C right at the TO-247 case leads. You have to wonder how the small leads of the TO-247 case can handle up to 60A! I didn't attempt to measure the V drop across the diodes as that can be compensated for by adjusting the Meanwell's output.

So it worked, but just not sure how long this will last with these lead temperatures. We'll see. It will only rarely be used with this high a load, and the Meanwell fan does draw air across the device.

A more robust (and expensive) diode like Mex suggested earlier would certainly be a better solution.

Some notes:
The amount of current Schottky diodes pass thru increases with their temperature. So when in parallel you can get thermal runaway with only one diode handling all the current. I mounted two TO-247s (4 diodes, 2 per case) together in the copper sandwich. But only one of the devices heated up and, presumably, was passing all the current. Dunno if both diodes in the TO-247 were sharing the current equally, as I couldn't measure the temperature of the individual anodes.

The metal backing on the TO-247 devices I have is connected to the cathode. So I had to be sure the cathode was soldered to the copper angle touching the metal back. I had that reversed initially. 😞

[MEXICOWANDERER]

capacitors resist changes in voltage. So they are going to spark even when disconnected for a few seconds. The BORG has 00.000 amperes draw and .24 across the schottky at full song - both power supplies yelling at each other.

There are SO MANY WAYS to isolate a power supply it isn't funny...

Use a 120va tiny relay on the AC input to switch DC to a White Rogers, relay on the output.

Use a schottky rectifier and heat sink

Use a HD switch.

Well, time for bed. Early rise and shine. Tomorrow cross that line and face thousands of newly stoned California drivers...

[brulaz]

Landyacht's post persuaded to get out there and try to measure the actual current.

My system has two built in ammeters but they're low resolution. The Rogue charger has a 0.1A display, and the 100A ammeter with a shunt off the neg battery post has 0.01 display at low currents, but I don't think it can resolve better than 50-100 mA. My clampon is bit better than those, with a 1 mA display at low currents, but tends to wander a lot in that range. Prolly 50ma is its max resolution.

Anyway, my two built in meters were reading stable currents before I plugged in the fuse for the Meanwell RSP-500-27. I plugged it in and removed it several times. It sparked each time I plugged it in. The built in ammeters showed no consistent change, but the clampon showed roughly 100mA change plus/minus 50mA each time. At 26.66V that works out to ~64Wh per day. Reasonably close to what the Rogue daily log was telling me when I had the Meanwell fuses in.

FWIW

[landyacht318]

I just got around to clampmetering my Meanwell rsp-500-15 with it plugged into DC, No AC.

I have two Noctua fans attached to DC outputs, so IF I have it hooked to DC, the fans spin. I use 45 amp anderson powerpoles toconnect/ disconnect. I also have a GTpower wattmeter inline, and another brighter easier to read voltmeter which also illuminate when hooked to DC.

As far as I can tell, these are the only loads when I am plugged into DC with no AC.

The 60mm Noctua nf-a6x25-flx is listed at 0.12 amp draw
The 80mm NOctua Noctua NF-R8 is listed at 0.12 amp draw

My clampmeter over the wires to these two fans is reading 0.24 amps

The wattmeter draws 0.05 amps, not sure of the green voltmeter as its wire cannot be easily clamped, but I will assume it is somewhere between 0.04 and 0.001 amps

Total draw according my my clampmeter is 0.33 amps. I know these clampmeters are generally not really accurate down this low, but I have done side by side testing before with other ammeters hooked inline on the circuit and found it fairly close at a single t10 LED at 0.08 amps draw.

So as far as I can tell, the only loads when I plug in my meanwell rsp-500-15 to ~12vdcDC are my additional fans, wattmeter and voltmeter.

[brulaz]

I've learned a lot, as usual on this forum.

For now anyway, I'll just leave the DC fuses out until the Meanwell is needed, which is pretty rare actually. And even if I left them in, it's not that big a deal.

But someday I'll prolly try one or more of these suggestions, more out of curiosity than necessity.

Thanks all for the info.

[MEXICOWANDERER]

The secret is to find an isolation device that does not act like your banker: Charge you a fee to count and use your money.

Except for inevitable questions about lifespan, it is tough to argue against a top quality totally-automatic battery solenoid-separator. It has negligible voltage drop no issues when equalizing a battery. When the power supply circuit voltage drops below a certain value (cessation of charge) there is total and absolute isolation of the power supply. Again no thinking no sweat, automatic.

I was "encouraged" to use a Schottky based dual rectifier because of the twin power supplies of the Borg. No way was a pair of battery separators going to fit in the Borg's case. Why Schottky? Why 500 amperes rating? A typical automobile full wave bridge rectifier is going to see .8 - .9 volts drop at 40 amperes output. Paralleling all three phase rectifiers The Schottky has .22vf. Each Megawatt is connected to a separate anode. 40 amperes versus 250 amperes capacity.

The online full wave bridge rectifiers a person sees on eBay have superb components, commonly rated 1Kv PIV. But with those high PIV values come higher forward voltage drop. An eBay rectifier commonly develops 1.1 - 1.2 volts drop at or near full ampacity. Compare that to the 500 amp Schottky.

When building alternators I trusted nothing. Even new rectifiers got tested, graded and sometimes rejected. Selectable 25, 40, or 50 amperes and always 1Kv reverse leakage. Renard and Traanspo bragged they sold 200 PIV stuff. In reality about 98% of what I purchased passed 1Kv and 1ma PIV.

The 1200 ampere rectifiers I use in Quicksilver and the Delco 50DN are rated 1,675 PIV. These monsters have 2/0 braid leads with 1/2" terminal holes and 1-1/8" studs with extremely fine threads. Made in Vermont and used in the metal plating industry. Purchased through Barney Kaplan Supply, Detroit.

Isolating a power supply via rectifiers is more involving then doing it with a simple and relatively inexpensive battery separator.

Because of the 28 volt question I would tend to gently urge Mr. Brulaz, to consider the battery separator. Dinner for 3 at a nice restaurant would cost about the same.

[Gdetrailer]

wa8yxm wrote:
I would like to think the output is diode protected and thus the device wont resent any load when not powered.. But alas that is how **I** would build them (in fact I do build power supplies)

I know Progressive Dynamics the wizard is still on line and the fan, far as I know those are the only loads (The Wiz is basically a voltmeter load, not much).

Can't talk about others

IF you can adjust the output voltage of that Meanwell... I'd put a heavy duty diode in the line (Say a set or half set of alternator diodes with heat sinc and fan, hook the fan to the Meanwell side) And increase the voltage by the amount indicated.. or a diode isolator and again increase the outut by the diode drop.

PD converters ARE "designed" to be able to accept and live with "12V" connected to the output of said supply..

Meanwells and pretty much any other power supply which has not be specifically designed to charge a battery (or have 12V at the output terms at all times, not so much.

Many switching power supplies will have some sort of feedback network for the regulator to "sample" the output voltage.. Some may also have current limiting circuits present on the output section or some may even have a drain resistor to drain off the residual voltage when the power supply has been powered down. All of those scenarios can "sneak" current when the power supply is turned off.

When using ANY power supply not specifically designed to charge a battery or not designed to be paralleled it IS very wise to include your own diode in the output of each supply.. It will save you hassle and possible smoke packets from being accidentally released..

[wa8yxm]

I would like to think the output is diode protected and thus the device wont resent any load when not powered.. But alas that is how **I** would build them (in fact I do build power supplies)

I know Progressive Dynamics the wizard is still on line and the fan, far as I know those are the only loads (The Wiz is basically a voltmeter load, not much).

Can't talk about others

IF you can adjust the output voltage of that Meanwell... I'd put a heavy duty diode in the line (Say a set or half set of alternator diodes with heat sinc and fan, hook the fan to the Meanwell side) And increase the voltage by the amount indicated.. or a diode isolator and again increase the outut by the diode drop.

[MEXICOWANDERER]

http://www.ebay.com/itm/2PCS-6-24V-50A-Toggle-Switch-On-Off-On-High-Quality-50-AMP-Trailer-Truck-RV-...

One to Eat Now, the other one saved for later. Good switches.

AND..........................................................

http://www.ebay.com/itm/2PCS-6-24V-50A-Toggle-Switch-On-Off-On-High-Quality-50-AMP-Trailer-Truck-RV-...

I like the Sure Power separator best. No voltage drop and TOTAL disconnect is, well, it's it's ------ SMART

[MEXICOWANDERER]

Hmmm, lessee, I ran my (2) 500 amp units for an hour at 32 amperes and the cathode rose to 142F. Free standing copper has a surprising amount of thermal radiation. But yes, a heatsink would be nice. I found a junk finned isolator to be perfect for the job. I pass 100 amperes for hours with such a setup and at the end, the isolator, and the cathode are barely warm. These Schottky units are quite the item.

Brulaz, sir. Make up a bridge bypass AROUND the Schottky unit, with a fifty amp toggle switch.

REMEMBER the rule with rectifier ratings - the ratings are for full amperage conditions at an elevated temperature. A 30-volt rating will survive 30 volts continuously for years if the rectifier P/N junction temperatures are kept mild.

I tested quiescent leakage of my units at forty point zero volts. Nine milliamps at around 26c. This test does not utilize any load. Does not imply the rectifiers will survive at load with 40 volts.

There are 50 volt Schottky plates but forward voltage drop is increased and cost is higher.

[Gdetrailer]

MEXICOWANDERER wrote:


http://www.ebay.com/itm/500-Amp-30-V-Schottky-Rectifier-CPT50130-2-Diodes-/282361292454?hash=item41b...

Mount it up high in a bay. The base the cathode is energized. The anodes have 1/4" bolts which allows size perspective.

• The 10 amp diodes will fail. Please don't ask why
• You need absolute minimum voltage drop across the diode and this baby does it at .17 volt at 40 amps
• Mounted with 1/4" couplers. Standoffs on an non-conductive cargo bay surface
• The base of this is COPPER and serves as the cathode junction - meaning it has electricity.

The "base" ALSO needs mounted to a heat sink (IE chunk of copper, aluminum or even steel)!!!

Operating high current diodes without providing a way to dissipate heat from the diode junction will soon let the smoke packets out of them (and when that happens you can't put the smoke back in) :E

[brulaz]

MEXICOWANDERER wrote:


http://www.ebay.com/itm/500-Amp-30-V-Schottky-Rectifier-CPT50130-2-Diodes-/282361292454?hash=item41b...

Mount it up high in a bay. The base the cathode is energized. The anodes have 1/4" bolts which allows size perspective.

• The 10 amp diodes will fail. Please don't ask why
• You need absolute minimum voltage drop across the diode and this baby does it at .17 volt at 40 amps
• Mounted with 1/4" couplers. Standoffs on an non-conductive cargo bay surface
• The base of this is COPPER and serves as the cathode junction - meaning it has electricity.

Thanks, that 500A rating is certainly better than the 3x10A rating of the ones I linked to.

But the 30V rating is too low. I equalize the batts at ~32VDC.

How about 3 of:

MBR30100CT 30A 100V Dual High-Voltage Power Schottky Rectifier TO-220

One on each of 3 leads fused at 7.5A?
Quite a bit cheaper too, although there'll be a long shipping time from the Orient.