oldjohnno wrote:I've tried all sorts of (mostly ratchet type) crimpers over the years, and with all of them found that occasionally I could pull a joint apart with a bit of effort. So a couple of years ago I bought some Thomas & Betts WT111M pliers and you could tow the Queen Mary with the joints that these make. They might look crude and old-fashioned but they certainly make a tight and secure crimp. I'll never go back to ratchet crimps.
Far as I can tell, SK Tools is selling the same crimper with green handles and "SK" printed on it. P/N SK15011, about $35 + shipping.
Proper signal connections are "gas tight", and that includes water vapor. Water molecules are much smaller than air molecules and have their own difficulties.
Soldering achieves this well because the solder dissolves a bit of copper and alloys at the boundary. If you do much of this at all, pre-tinned and silver plated wires are available and should be used. Gold is a contaminant in solder joints, but is generally tolerated at the thickness of plating normally used. Mostly you see this on circuit boards which have been gold plated. Gold plated circuit boards resist the environment better and they also look pretty.
Crimped or swaged terminals also work very well, they must be performed at enough pressure to cause the metal to flow and weld. Without this the joint is not gas tight and will fail for small signals. Remember those small water molecules which are just a single oxygen atom with a couple protons hanging on the side. I think the bulk of connections in aviation and rockets are crimped and that would be for reasons of production. On a production line you can use a really nice crimper which is more like an arbor press, not a hand tool.
At my last job we had a big drawer full of crimp tools, 10's of thousands of dollars worth because each and every connector and it's pins seems to have slightly different dimensions I suppose. Carrying one of those tools from the lab to a production area would have been a very serious offense.
Oxides in connections are bad because they form diodes in the circuits. If you remember old car battery chargers, they typically used metal plates on the frame of the unit which had an oxide layer to rectify the current. If you use your trusty little meter to check wiring on a car you might notice one time the connection looked OK and then later it was clearly bad. Did you swap the leads checking the resistance? That's why you should use the point on the meter leads to scratch the surface you are testing, no oxide. Scopes will show this stuff easily.
Automotive connectors have a heavy tin plating and the contacts are designed to gouge when they engage. That's where the gas tight junction is formed. If you don't see a bright streak on the contact metal when you pull them apart - it was a bad connection. They are only rated to be cycled maybe a dozen times, I knew last year but can't remember right now.
The price on good connectors is so high that if you're making ECU's, it might make better sense to think of yourself as making wiring harnesses...
I’m the quality manager at a company that builds wire harnesses. We build harness and cable assemblies for everything from bubble gum machines to trains, helicopters, and tanks. Our largest customer builds construction vehicles, cranes specifically. 99.99% of our connections are crimped. As mentioned, a gas-tight crimp that withstands the required tensile pull test will perform well for decades of use.
In the automotive harnesses we build about the only places we see gold contacts specified are for certain sensor connections and CAN signals.
As far as splices are concerned we use either ultrasonic welding (bare copper only) or parallel crimp splices where the wires lay side-by-side in a metal barrel and then are crimped together into a solid mass. For environmental reasons some customer have us solder seal the ends of the barrels. This is the method I use on my own car. I’m an old solder guy myself, was certified to teach hi-rel assembly and soldering by the Army. But working here I’ve come to appreciate what a good crimp connection can do.
The only time we use butt splice connections is when required on a customer’s drawing.
Something to consider, SN 63 solder melts at around 371⁰ F, sometimes your under hood temps can get higher than that.
j-c-c wrote:
I think you covered most of the high points well, maybe in my original question I should have referenced specifically the sensor circuits, cam, crank, IAT, etc ie the low current stuff, that has significant impact on proper EFI operation, and is seldom mated, ie no self cleaning, which means i'm a solder guy.
And on the side issue of solder vs crimp, I just find it hard to comprehend how a crimp connection can maintain a great connection with any length of exposure to moisture, with dissimilar metals in likely a slightly closed/trapped space, and a street driven car at speed in weather is a worse case scenario. Vibration issues I can resolve, starting with adequate sized/layer heat shrink, etc.
the engine sensor circuits are not low current or low voltage in the 'needs gold' sense. In fact most of them in use today are Tin plated. Some use Silver if temps are >125*C continuous
pamotorman wrote:connections have to get pretty bad on a car before things stop working. this is the chassis ground connection from a C-5 corvette.
Haha... one of those things caused me FITS on my 2005 Chevy Colorado.
The heater / defroster blower motor would occasionally cut out. It ended up being a chassis ground exactly like that one that was mounted way up near the radiator support / passenger side headlight. Not the first place you go looking when chasing an issue under the dash!
[quote]However for my hotrod applications I will usually remove the plastic cover on the crimp terminal if so equipped, strip and tin the wire ends (apply minimum solder), slide on the appropriate size heat shrink tube, crimp terminal with best tool available, solder the assembly, and then shrink the tubing.[/auote]
That's what I do as well, except I start with uninsulated terminals. I also don't tin the wire first, which probably reduces the effectiveness of my soldering on the crimped connection. The solder still makes me feel better though.
j-c-c wrote:
I think you covered most of the high points well, maybe in my original question I should have referenced specifically the sensor circuits, cam, crank, IAT, etc ie the low current stuff, that has significant impact on proper EFI operation, and is seldom mated, ie no self cleaning, which means i'm a solder guy.
And on the side issue of solder vs crimp, I just find it hard to comprehend how a crimp connection can maintain a great connection with any length of exposure to moisture, with dissimilar metals in likely a slightly closed/trapped space, and a street driven car at speed in weather is a worse case scenario. Vibration issues I can resolve, starting with adequate sized/layer heat shrink, etc.
the engine sensor circuits are not low current or low voltage in the 'needs gold' sense. In fact most of them in use today are Tin plated. Some use Silver if temps are >125*C continuous
It seems this whole thread centers around "needs", which as reading the replies so far, "needs" looks to be very subjective. My thinking on the gold for sensor circuits, is they I thought, and correct me if I am mistaken, they are the lower current/lower voltage circuitry that can be a significant headache with efi in poor connections exist for whatever reason, and that gold would help in taking them farther from the head scratching scenario when trouble shooting an issue, didn't mean to maybe imply they were the be all/end connection solution. Silver is nice, but it does tarnish.
j-c-c wrote:
It seems this whole thread centers around "needs", which as reading the replies so far, "needs" looks to be very subjective. My thinking on the gold for sensor circuits, is they I thought, and correct me if I am mistaken, they are the lower current/lower voltage circuitry that can be a significant headache with efi in poor connections exist for whatever reason, and that gold would help in taking them farther from the head scratching scenario when trouble shooting an issue, didn't mean to maybe imply they were the be all/end connection solution. Silver is nice, but it does tarnish.
Low current isn't really a problem...as I said I use gold when I'm trying to send more current though a given connector and need the lower contact losses.
On voltage most all sensors are 0-5V these days so lower voltage but a few mV noise/shift at the connector isn't really much of a concern. A NBO2 that is 0-1V with most of the action in the 0.4-0.8 range is about the biggest issue I've seen....shielded wire and not being near the plug wires and its usually good to go. I guess EGT sensors are more trouble but they have standardized plugs and usually go to a dedicated conditioner/logger
mk e wrote:
Low current isn't really a problem...as I said I use gold when I'm trying to send more current though a given connector and need the lower contact losses.
As per my first reply here.......its the difference between low voltage AC only circuits and DC circuits.
The oxide layer will clip the signal because it acts as a semiconductor. Any DC bias eliminates the problem.
The term "contact wetting" explains everything.
Yes, NB o2 sensors can be a problem because it's pretty much impossible to solder the wires. Reworking wires on those, it's often best to use screw connectors.
