Conductor clearance for 60V + soldering heavy wire

[paulca]

I assembled a DPS5020 last night.  It mostly went well, but I had small issues with soldering the heavy wire, I think it's 14 or 16 AWG.

So the main DC power in went really well as the board had through holes that would take the wire, so they were easy to hold in the hole and feed solder to get a nice shiny blob around the wire on both sides and trim the back side.

The switch send/return was more problem matic.  The board only had surface pads with a tiny hole.  For this I had trouble keeping the wire on the pad and stable while the solder hardened and I had to have 3 goes at one of the wires.

The result was that this wire flattened out when the solder melted and the pressure of the iron squished it.

The contact is solid, but the wire overruns the pad on one side.  It is about 1.5mm to 2mm from the second wire.  These are switch send and return, so I assume the DC will have no interest is arc'ing across when the switch is closed.  When the switch is open, I am (possibly unnecessarily) concerned it could potentially arc across.

I will be powering it with 48V DC up to 10A, but as the device is rated for 60V and 20A input there is always the possibility I could later getting a higher power DC brick for it.

So...

Two questions. 

Is 1.5mm enough gap to prevent arc'ing at 48V or 60V DC?

What tips do you have for soldering heavier gauge wire?

I have a weller 25W non-thermistatic iron.  (I also have a 40W cheapo iron).  The trouble I have starts with tinning the wire.  It takes ages to get the solder to wick into the wire and by that time the wire is too hot to hold.  When I do get it to wick the solder has a tendency to wick down into the cable making it stiff and inflexible.  Usually as I have to use pliers to hold it, it also ends up squashed and sets in that shape.

Then when I apply the tinned wire to the tinned pad or contact it takes ages to make sure the solder in the wire melts along with that on the pad.  I might be overdoing it here, as I assume as soon as the pad melts, placing the wire into the molten solder for a second or two will melt the tinning solder and create a solid joint.  Then there is the issue of holding the wire steady enough while the solder cools slowly with the large amount of heat capacity in the wire.

I'm made worse messes than last night doing this, so all tips welcome.

[paulca]

Might have found my answer to the arc'ing question.

Breakdown voltage of air is apparently 3MV / Meter.  So 3KV per millimeter.  If this is true I should be absolutely fine.  I could get down to 0.02mm  (which does sound awfully small to me).

I know that gap rises dramatically once the arc has started as it creates a path through ionized air.

[ChrisLX200]

I'd say you need a heavier wattage iron - there's no substitute for power when tackling jobs like this. Same goes for working with big ground planes. A short/stubby/wide chisel tip helps dump the heat into the part quickly, but if the initial heat tranfer is still insufficient for the job then you're in trouble. You could pre-heat using hot air which would lessen the load on the iron, but really - get a bigger iron (at least 60w).

[paulca]

I'd say you need a heavier wattage iron - there's no substitute for power when tackling jobs like this. Same goes for working with big ground planes. A short/stubby/wide chisel tip helps dump the heat into the part quickly, but if the initial heat tranfer is still insufficient for the job then you're in trouble. You could pre-heat using hot air which would lessen the load on the iron, but really - get a bigger iron (at least 60w).

I assume the idea being to get enough heat into the stripped end of the wire quick enough to melt the solder before the heat gets to spread down the wire and pull the solder with it?

I think I might have bought a cheap soldering kit years ago when I failed to solder 12AWG wire.  I will hunt it out.  It's not a good iron, but I believe I choose it due to it being high power, just for jobs like this. 

My weller has a blunt conical tip on it which might not be helping but it works find for normal PCB soldering of components or header pins.  I might see if I can find a small wedge tip for it.

[ChrisLX200]

You have to be able to dump heat into the part much quicker than conduction can take it away, otherwise you end up heating much more area - and possibly not even reaching critical temperature at all. That happened to me trying to desolder a TCXO in the middle of a big ground plane using my 45w Weller iron - and that was after heating with a hot air gun (clearly not enough but I was worried about causing damage to a wider area). After that failure I bought a 60w iron, but even this is under-powered IMHO and perhaps I should have gone for something bigger. We're talking special case here - you only need high wattage for big heatsink jobs and the smaller irons are fine for general use. Just don't bother even trying to use a small iron where you need the extra power though, waiting a bit longer won't help and you will damage nearby components. You need a big highly localised thermal gradient.

[Ian.M]

I thought the DPS5020 had screw terminals?

For low voltage DC stuff, if the feed is fused or otherwise current limited, that sort of clearance is plenty, as long as there is absolutely no chance of anything bending and touching.

!4AWG is only 1.6mm dia.  That's nowhere near what I'd call 'heavy'.   However with thicker stranded wire, you do need to prepare the end carefully, especially if your iron is a bit on the wimpy side (low power or low heat capacity).  Strip, immediately gently twist before the strands get disturbed from their original lay, flux the very end if its bare copper or old oxidised tin plated wire (not too much - you don't want excess flux running up under the insulation and possibly causing corrosion), and tin it,  remove excess solder and trim to desired length.  Then, and only then,  you can solder it to a PCB pad. If you try to solder it without tinning it first, you'll have an unholy mess, probably with burnt or melted insulation for a fair way up the wire.  For difficult wire tinning cases a wet tissue held round the insulation with a clothespeg can be used to heatsink it so it doesn't melt.

IMHO those blunt conical bits are only fit for pokerwork!  A chisel style (flat) tip for the Weller will make a fair bit of difference, and seldom needs to be changed for fine work - just use the corner.  A hoof bit is even better for delivering heat to the joint, but can be a PITA if oversized.   If the cheapo iron isn't so daggy its useless, maybe see if you can find one for it as the capability to dump 40W into larger joint will be useful.   

For difficult joints in heavy wire, or with multiple wires, I reach for my Weller 8100 soldering gun.  100W of heat delivered right at the tip.  However there's no temperature control and the amount of heat put into a joint is only dependent on the dwell time and how long you hold the trigger down, so until you develop a fair bit of skill with one, work on delicate boards is best avoided.

[paulca]

I thought the DPS5020 had screw terminals?

It does but the case version has a fan controller board at the back.  DC in, Switch out/return and the fan solder on to it. 

I should be fine with the gap, I might hot-snot a bit of heat shrink tubing in between if I'm feeling paranoid.  Nothing can bent or move about while it's in the metal case.

The DC brick (LED 48V, 10A) supply I got for it hasn't arrived yet.  But I'm powering it with 30V from my bench supply and nothing bad has happened, yet, I haven't load tested it yet.  Might try it on 24V 5A for a quick check tonight.

[Gregg]

There is no need to solder the large wires to the board or to the switch.  Use crimp ring and spade terminals.  A small wire from the switch to the fan board to power the fan can be added as shown in the enclosed picture.

[paulca]

There is no need to solder the large wires to the board or to the switch.  Use crimp ring and spade terminals.  A small wire from the switch to the fan board to power the fan can be added as shown in the enclosed picture.

Nice.  Didn't think of that, I just followed the instructions.

Anyway it's all wired up now.  The 48V DC brick arrived and so it's happily running on 48V without fireworks or smoke (yet).

I have it powering a LiPo charger with 4A and while both the DC brick and the DPS5020 are making an awful fan racket, they both seem fine.

I don't actually have anything to load test it up to 10A or to find out if the 48V 10A supply will give me 20A at 24V (it should right?).    5A is about the maximum load I have right now.  Probably for the best.

EDIT: I probably shouldn't have realised this, but the LiPo in question is 5000mAh and rated for 2C charge, so I do have a 10A load.  It would mean trusting the DPS5020 with a fairly nasty LiPo to put out if it screws up though.

[Jwillis]

There is no need to solder the large wires to the board or to the switch.  Use crimp ring and spade terminals.  A small wire from the switch to the fan board to power the fan can be added as shown in the enclosed picture.

Never been a big fan of crimps.But I'm bias because of the horrors I've had to deal with in automotive and industrial applications.Crimps tend to be a breeding ground for corrosion and bad connections. I've always soldered uncoated ,or removed the plastic,  crimps and heat shrunk. 
Most of the soldering I've done over the years for heavy gauge wire between 6 and 12 awg has been done with a Weller 260 watt or the 200 watt .For gauges down to 2 awg I used a propane torch.For gauges 14 to 28 awg I've just used a standard 60watt .For gauges higher than 28 and very fine point tip and fine solder
If the concern is arcing ,just don't leave wires bare.Always shield.Cables running 48volt even 92 volts with close to 2000 amps have been bound together without any problems.