Not enough "wetting" current for connector type?

[Faringdon]

Hi,
We are dividing down 390Vdc to 2.5V, then feeding this to a PFC control chip (UCC28070A) which is on a daughter PCB, through the below header and socket housing.......there is no current going through the header pin.....is this OK?....or should we put the lower divider resistor on the daughter PCB, so that at least a few 100 microamps flows through the header connector? (to "wet" the contacts).

Internet re-searching this reveals a huge number of different opinions, without concensus, eg...
https://www.eng-tips.com/viewthread.cfm?qid=107221

Pin header
https://uk.farnell.com/multicomp/mc34767/header-2-row-r-angle-8way/dp/1593450

Socket for pin header
https://uk.farnell.com/harwin/m20-7830446/socket-pcb-0-1-4-4way/dp/7991991

[jpanhalt]

Seems odd to me that if you are worried about "wetting" current you would pick a contact with tin finish rather that gold.  Gold is good for 300 insertion cycles; tin for only 50 cycles.

At least tin is what came up for that number.  Maybe you have specified gold.

[ejeffrey]

Wetting current is more often a problem for switches and especially relays.  The motion of a relay often can't provide a sufficient wiping action to really clean the contacts, and they are intended to be switched after installation.  Pin based connectors make a much more thorough wiping action on insertion.  Once they form a good gas tight metal-on-metal contact they are likely to be good for as long as the connector stays mated, often the life of the device.

I'm not going to say there is zero chance of problems, but it's not the top of my list of things to worry about.  If your device will be exposed to a chemically harsh environment, sit for a long time before being assembled, or be frequently disconnected it might be worth considering.

[Zero999]

I've not had a problem with wetting current on connectors, nor ever heard of it until this thread. As mentioned above, I would have thought the strong wiping action of mating would clear any oxide and ensure a good contact is made.

[Faringdon]

Thanks, about 10 or so years ago, i went for an interview with Ultra Electronics in UK.....The Chief engineer brought up the subject of wetting in switches during that interview...and stated that it was a real issue to be extremely concerned about.

Most of the PSU's ive reverse engineered, in fact all of them, have the daughter boards soldered in...not  with connectors.....this could be for cost...but also, could be because of wetting  issues.

We are only doing a prototype, but if we get sometimes_intermittent joints for the signals. it will really screw up everything..specially since eg a feedback signal being absent could cause mayhem.

[jpanhalt]

Try gold plated.  If that doesn't work rub a little indium on the male pins.  Indium cold welds great and dissolves gold.  Of course, if the design is bad, soldering won't work either. (Mercury does the same but some people fear it.)

[dmills]

I have actually been fucked over by a mix of gold plated pins and a tin plated socket, not a whetting current issue but an issue of tin fretting.
Had to recall over £200ks worth of kit because of it, live and learn.

Oh also, Multicomp? Fuck that! Creating a rod for your own back with the BOM going there in my view, single source distributor.

Whetting current absolutely is an issue both in relays and switches and it is telling that you can buy explicitly 'low signal' versions of both.

Trap for young players: You can buy a relay specified for 2A that is ALSO specified for low signal use, these typically have a thin gold plating (for low signal operation) over something that will actually survive an arc... The implication being that you can use these reliably in low signal service PROVIDING you don't use the part for power, but once you use it for power it is ruined for low signal service.   

[Doctorandus_P]

Indeed, don't mix tin and gold contacts. Use tin on tin, or gold on gold but don't mix them.

Back in the '90-ies I read a datasheet of a relay that is fit for both high and low currents. and it had dual contacts. The big current contact closes first, and after that a low current contact is also closed. It was specifically designed as an output relay for audio amplifiers.

[Gyro]

I've not had a problem with wetting current on connectors, nor ever heard of it until this thread. As mentioned above, I would have thought the strong wiping action of mating would clear any oxide and ensure a good contact is made.

Nobody else remember the problams problems with tin plated dip socket reliability in the '70s / early '80s?

[wizard69]

Well yeah it is an issue with contacts in switches and relay, a good quality connector though should produce a gas tight seal that lasts for years.   Also I was always under the impression that wetting was a function of applied voltage not current.   Basically you need a high enough voltage to punch through any oxide layer.    This is why we have a bunch of relays with "dry" contacts, gold plating and the like.   

As an aside; years ago I fixed a problem on a tool by simply replacing a conventional pushbutton contact block with a reed based element, contact block.   The signal was TTL level and the problem literally went away for 15 years.   It was a simple fix that no amount of contact cleaning and replacement would solve.   On a problem more similar to what you are describing I had a CNC controller and its EPROM board create problem and solved that with some CAIG Deoxit sprayed into the EPROM board connector.   Normal removal and reinsertion would not solve the intermittent problem but a quick spray and another reinsertion had the machine running for years.   The point of this story is that connectors, even connectors designed for the job, create reliability problems.

Thanks, about 10 or so years ago, i went for an interview with Ultra Electronics in UK.....The Chief engineer brought up the subject of wetting in switches during that interview...and stated that it was a real issue to be extremely concerned about.

Most of the PSU's ive reverse engineered, in fact all of them, have the daughter boards soldered in...not  with connectors.....this could be for cost...but also, could be because of wetting  issues.

Soldered in connections, properly done, can be far more reliable than just about any plug and socket connector.   This especially in the industrial sector where tools can shake the earth under you.   I highly doubt that the concern is "Wetting".   Frankly I could post dozens of cases (if I could remember them all) where connectors turn out to be the root cause of a failure, during my life in the world of automation.   That doesn't mean that solder joints don't fail by the way, as there have been a few of them.

So no I don't think cost is even a factor, I'd vote for reliability.    Think about it would you be happy if a transient across a bad connector trigger the crow bar on a power supply?

We are only doing a prototype, but if we get sometimes_intermittent joints for the signals. it will really screw up everything..specially since eg a feedback signal being absent could cause mayhem.

Exactly!   This is why in my opinion the critical paths need to be on one printed circuit board in a power supply.   I'm trying to think right now as to how many power supplies in the plant even have more than one board, not many that is for sure.   A power supply can do a lot of damage if it supplies the wrong voltages to the main circuitry.   Often power supplies are neglected in electronic designs which is sad because careful selection and implementation can eliminate a lot of issues.