0

Two separate questions:

1) What is the typical value of contact resistance in electrical connectors? I am interested mostly in a banana connector for delivering power supply from the source to the PCB.

2) What is the typical value of contact resistance of a "good" handmade solder joint? This is assuming the contact was soldered properly (in particular, enough solder was used).

Appendix 1. My particular situation
My particular situation: 1.8 V, 1 A DC are delivered from the power source to the PCB and I am going to use a thicker copper wire to decrease IR drop. I calculated 1 mm diam copper wire, 50 cm length has a resistance of about 10 mOhm (so, 10 mV IR drop at 1 A). So, if I buy a thicker copper wire, solder it to banana jacks which are to be inserted into the power source and the PCB, I'll have four parasitic contact resistances: two for the banana connectors and two for the two solder joints. So, I wonder, does it make sense to fight for 10 mOhm (if thicker wires are used) if contact resistances could be greater. This is a test PCB (to test a chip), no VRM on it.

Appendix 2. Some contact resistance values for banana connector found in the Internet
URL1, page 2: 10 mOhm maximum.
URL2, page 5: 0.5 - 1 mOhm maximum.
URL3, page 4: 0.5 - 1 mOhm maximum.
URL4: 1 mOhm maximum.

Appendex 3. Questions about conact resistance found at Stack Exchange
Relay with contact resistance less than 10mohm
How the relay contact resistance is measured?
What kind of effects does a relay have on signals?
What causes contact resistance?

Community
  • 1
Sergei Gorbikov
  • 1,723
  • 3
  • 25
  • 29
  • 1
    (2) Work it out from the dimensions of the joint and the conductivity of your chosen solder. If you can't find data on that, unroll a metre of it and measure resistance and diameter. It'll be a slight overestimate thanks to the flux core but you want a conservative number anyway. –  Jan 19 '17 at 11:49
  • 1
    It seems you have answered question 1 by yourself. Now, regarding solder joint resistance, I'd expect it to be much less than this, and really really (like, *really*) negligible in regards to wire/trace resistance. – dim Jan 19 '17 at 11:52
  • Have you considered putting the power source on the PCB? See PC graphics cards: the chip needs 3.3/1.8V, but the card takes 12V and there's a buck converter really close to the chip. – pjc50 Jan 19 '17 at 13:12
  • @pjc50 yes, I considered the option with VRM (as I c, this is a type of a buck converter). Frankly, I don't have experience with VRMs, but I guess, since this is a test board (not a mass production) and I have an external power source (and can arrange a good power delivery network, including capacitors and ferrite beads) - I hope, I'll be fine with an external power source too. no doubt, VRM should be usually better if located in a close proximity. – Sergei Gorbikov Jan 19 '17 at 13:19

4 Answers4

3

10 mΩ for a full-size banana jack sounds really high to me. However, the obvious things for you to do are:

  1. Measure it. Put a known current thru the connection and measure the voltage across it. This will tell you what you are actually getting, not the worst case it could be.

  2. Look at a datasheet. Find a reputable company that makes banana connectors and see what they rate the maximum contact resistance as.

These things should have been obvious, and basic research you should have done before asking.

As for the solder joint, there is no mechanical "contact" there, and the problems associated with it. It's just a solid metal connection. Most likely the weak link is the copper trace on the board after the solder joint.

For starters, analyze a solder joint as if only the copper were present, but directly connected. In a proper solder joint, the copper parts are touching anyway. Solder has higher resistivity than copper, but there is also a lot more of it in parallel around the joint. The gap between the solid conductors is also very small.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
2

Measuring is a good idea. But please consider to use the kelvin (4-wire) measurement. Otherwise you will influence the measurement with the resistance of the measurement leads + contact resistance and so on.

Explanation of the kelvin measurement can be seen in this link.

Lundin
  • 17,577
  • 1
  • 24
  • 67
Johannes
  • 207
  • 2
  • 11
2

The range if resistance values you've found quoted for banana connectors instructive. The resistance is very variable in detail, it depends on surface finish, contact pressure, and surface contamination.

You will get the most consistent, and lowest, contact resistances from connections where both the plug and socket are gold plated. This is because they do not get contaminated like base metals. However, connecting banana plugs 'wipes' the contact surfaces, so even with base metals, a fresh connection will be good.

As you want to connect a relatively low voltage high current power supply, it's worth considering the resistance of the wire-to-connector connection, and the length of wire itself, both of which are in series with the connector. It would be quite easy to end up with an order of magnitude more resistance end to end than was present in the mating connector surfaces themselves.

Neil_UK
  • 158,152
  • 3
  • 173
  • 387
  • Yes, resistance does creep up with time where wire attaches to banana body. Many are crimped, some tightened with a screw, few soldered. "Exercising" a banana (in-out-in-out...) can reduce contact resistance after a period of non-use. But the wire-to-body contact is often hidden and not accessible...sometimes you can reduce banana-to-body resistance by "stretching" a patch-chord. Grab banana body ends and pull. – glen_geek Jan 19 '17 at 18:07
1

Typical solder joints are in the range of a few hundred micro-ohms (or at least they were every time I have made a measurement).

But of course it varies with the size of the joint, type of solder used, thickness of the solder, etc.

user4574
  • 11,816
  • 17
  • 30