liveforphysics
100 TW
Arlo1 said:
Yes to all.
I'm guessing the tubes are way too long and not poking much as a result. I will inquire about getting some extra short length material to try for his fairly uncommon application.
Conductivity improving grease project
- Thread starter liveforphysics
- Start date
Yes to all.
I'm guessing the tubes are way too long and not poking much as a result. I will inquire about getting some extra short length material to try for his fairly uncommon application.
It might be interesting to vary the pressure (or even know what it is). If the pressure is high enough it tends to be like a small weld spot with very low resistance.
The other crucial factor is aging. Most connections are good when you first mate them, then they go downhill from there from oxidation.
The other crucial factor is aging. Most connections are good when you first mate them, then they go downhill from there from oxidation.
liveforphysics
100 TW
fechter said:
Yep, this is why I still wouldn't rule out using dielectric grease on connectors even in high performance situations. This was going from brand new out of the sealed plastic baggy connector pin to getting mated for the first couple times during the dry test period. Had this been a connector that had served years of being in the weather, it may be that the greased unit wouldn't be effected while the dry unit would higher resistance from oxidation.
okashira
10 kW
Did you happen to try a copper antiseize? :lol:
http://www.amazon.com/Permatex-09128-Copper-Anti-Seize-Lubricant/dp/B000HBM8HU
[*]Temperature range: -30F to 1800F
[*]Provides good electrical conductivity
[*]Suggested Applications: Spark plug threads installed in aluminum, exhaust manifold bolts, engine bolts, oxygen sensors, knock sensors, thermostat housing bolts, fuel filter fittings and battery cable connections
http://www.amazon.com/Permatex-09128-Copper-Anti-Seize-Lubricant/dp/B000HBM8HU
[*]Temperature range: -30F to 1800F
[*]Provides good electrical conductivity
[*]Suggested Applications: Spark plug threads installed in aluminum, exhaust manifold bolts, engine bolts, oxygen sensors, knock sensors, thermostat housing bolts, fuel filter fittings and battery cable connections
liveforphysics
100 TW
Punx0r said:
No they aren't. CP #70 is the plain grease, and a better grease than every other type available at any price from my (years earlier) testing, even better than Krytox and the other fluropolymer greases.
megacycle
100 kW
I cant find the original document that had a variable calculation for the applied pressure on the joint, but was similarly exponential, there is a point where applying more pressure to a joint is a waste. http://apqi.org/PQ_Rescource/Docs/Copper_of_Busbars/c7-2.htm
Are you able to measure the a c. resistance of the joint, would be interested to see if the grease gives an improvement.
Are you able to measure the a c. resistance of the joint, would be interested to see if the grease gives an improvement.
liveforphysics
100 TW
Arlo1 said:
A proper hydraulic crimper flows the copper copper strands and terminal into a gas tight cold-forged solid copper cross-section chunk. They tend to be the coolest places when looking with a FLIR if you have a proper crimp setup. That said, I bet adding some grease in practice would avoid moisture intrusion from voids.
wb9k
10 kW
Punx0r said:
It can....it's all about strain relief....and keeping the solder from wicking too far up the wire away from the terminal.
johnrobholmes
10 MW
If I were trying to improve a crimped connection beyond a room temp method, it would be with a resistance brazed or soldered connection done during the compression operation . A 200a buzz box welder (old reactor coil type) run through a 10:1 transformer would be able to provide enough amperage to melt solder on 8g wires within a few seconds, and 20:1 to 25:1 should be able to bring the copper cherry red for brazing. Prince and Izant is an excellent source for powder and flux paste fillers http://princeizant.com/default.aspx
liveforphysics
100 TW
I think you guys may forget, solder or brass conducts much worse than just copper, and a good crimp will leave you with a continuous solid copper assembly. Any other metals you put in there is just in the way of what could have been solid copper.
johnrobholmes
10 MW
I would assume the filler to be silver and copper to make the time worth it, but you are right that a proper crimped connection is about as good as it gets when done right- Solid copper! Tin and lead would be contaminants unless copper oxide formation was proving to be a problem.
liveforphysics
100 TW
Arlo1 said:
Yeah, I don't disagree it might be possible to get silver particles all forged together in there with the copper, and if the disimilar metals interface didn't cause more resistance than it helped with, it may cause an infinitesimal decrease in resistance.
You gotta remember though, good hydraulic crimp joints are already the coolest and lowest resistance spot in the current path. If you were just trying to throw money after any drop in resistance, save your silver and go up a size in cable.
megacycle
100 kW
This paper discusses the effects of surface films on joining conductors and the effects of pressure on joints. https://www.google.com.au/url?sa=t&source=web&rct=j&ei=hWqJVNGvD4a4mwXHqIDQBw&url=http://www.ijcee.org/papers/423-E1118.pdf&ved=0CCkQFjAD&usg=AFQjCNGaRvjwgEIOXi6Uuira-1Mcp9VloA&sig2=pnEaRqJB-oXuLKLIIiLmRg
I think with the average currents draws, around 150A or less, we deal with, it would be purely academic discussing anything more than proper standard crimping and clamping techniques, to either similar metal, clean, well mated, film free surfaces, or mating dissimilar metal, thinly greased surfaces
I think with the average currents draws, around 150A or less, we deal with, it would be purely academic discussing anything more than proper standard crimping and clamping techniques, to either similar metal, clean, well mated, film free surfaces, or mating dissimilar metal, thinly greased surfaces
wb9k
10 kW
Well, I spent the better part of today taking measurements with another nanotube test. This time I used a power supply to deliver 10 Amps to a pair of heavy cables with ring terminals to a shunt. One rusty ring terminal was deliberately chosen to see if conductivity through the corrosion might be improved with the grease (it wasn't).
Measurements were taken across the shunt plus both of the ring terminals (the entire test system), and from each ring terminal to its corresponding shunt connection block (each individual joint under test). AC milliommeter was an Agilent 4338B with a pair of Kelvin probes. DMM was a Fluke 289. I took:
A series of 3 AC resistance measurements with untreated joints and no current flowing.
A series of 3 AC resistance measurement with untreated joints and 10 Amps flowing across the shunt (verified steady 2.50 -0/+3 mV drop across the shunt)
A series of 3 DC Voltage drop measurements across the same places (almost) with untreated joints and 10 Amps flowing.
I then loosened and retightened the joint hardware and repeated the above sequence two more times, for a total of three sets of measurements.
Then I mixed up some grease (this time with a much smaller amount of nanotubes, I'll explain why later) with the Tuball. I treated both sides of each ring terminal and the threads of the bolts that hold them in place on the shunt. I then took:
A series of 3 AC resistance measurements with treated joints and no current flowing.
A series of 3 AC resistance measurement with treated joints and 10 Amps flowing across the shunt (verified steady 2.50 -0/+3 mV drop across the shunt)
A series of 3 DC Voltage drop measurements across the same places (almost) with treated joints and 10 Amps flowing.
I then loosened and retightened the joint hardware and repeated the above sequence two more times, for a total of three sets of measurements.
Probe resistances and current flow were very regularly checked to verify nothing in the measurement system was drifting.
I have pics of the test setup and the data I collected, but for right now then numbers are so small and there is such an absence of clear trends as to not be worth the time just yet. The only thing worth reporting here is that AC resistance measurements did appear to improve appreciably with the application of the grease, but you can only see this indirectly. Because of the ACR measurement points for the individual joints, a portion of total system resistance is not captured in either of the individual joint measurements. This means we can't add our individual joint measurement to see if they add up nicely to roughly the same value shown in the whole system measurement. (My DC V drop measurement points plus the 2.5 mV "confirmation voltage" did allow for this and correlation was quite good.) Interestingly, in both the no current and 10 Amp measurements, the individual joints accounted for a significantly smaller portion of the total system resistance when they were treated with the grease. The effect was easier to see when current was flowing than not. I hesitate to make much of even this--I could be seeing nothing more than noise that stacks up pretty. I really think much more current is going to be needed to see anything clearly if it's there to be seen at all. I would much rather see an improvement in DC voltage drop than AC resistance. Hopefully a few hundred Amps will make this possible, but I'll have to do those tests at the office. I do agree with Megacycle that this is not something ebikers are likely to want or need to pursue. This is for bigger vehicles that pull hundreds of Amps, if it's for anybody at all.
The reason I used much less Tuball was because I noticed that in the sample kit they sent, the "modified materials" included used extremely small amounts of Tuball to achieve what they do in each case--usually less than 1%. The "conductive film" they include is transparent, so that gives you some idea of how much carbon there can be in there. It's a far cry from the pitch black you get mixing 50% with the grease. I also began to wonder if too much of the material would negate some if its properties. OCSiAl's literature talks a bit about MULTI-WALLED CNTs and how they are easier to make, but don't have the spectacular properties of true graphene. I worried that perhaps too much of this material would make it behave more like MWCNTs than SWCNTs. That doesn't appear to be the case, but who knows with so little signal to work with?
Measurements were taken across the shunt plus both of the ring terminals (the entire test system), and from each ring terminal to its corresponding shunt connection block (each individual joint under test). AC milliommeter was an Agilent 4338B with a pair of Kelvin probes. DMM was a Fluke 289. I took:
A series of 3 AC resistance measurements with untreated joints and no current flowing.
A series of 3 AC resistance measurement with untreated joints and 10 Amps flowing across the shunt (verified steady 2.50 -0/+3 mV drop across the shunt)
A series of 3 DC Voltage drop measurements across the same places (almost) with untreated joints and 10 Amps flowing.
I then loosened and retightened the joint hardware and repeated the above sequence two more times, for a total of three sets of measurements.
Then I mixed up some grease (this time with a much smaller amount of nanotubes, I'll explain why later) with the Tuball. I treated both sides of each ring terminal and the threads of the bolts that hold them in place on the shunt. I then took:
A series of 3 AC resistance measurements with treated joints and no current flowing.
A series of 3 AC resistance measurement with treated joints and 10 Amps flowing across the shunt (verified steady 2.50 -0/+3 mV drop across the shunt)
A series of 3 DC Voltage drop measurements across the same places (almost) with treated joints and 10 Amps flowing.
I then loosened and retightened the joint hardware and repeated the above sequence two more times, for a total of three sets of measurements.
Probe resistances and current flow were very regularly checked to verify nothing in the measurement system was drifting.
I have pics of the test setup and the data I collected, but for right now then numbers are so small and there is such an absence of clear trends as to not be worth the time just yet. The only thing worth reporting here is that AC resistance measurements did appear to improve appreciably with the application of the grease, but you can only see this indirectly. Because of the ACR measurement points for the individual joints, a portion of total system resistance is not captured in either of the individual joint measurements. This means we can't add our individual joint measurement to see if they add up nicely to roughly the same value shown in the whole system measurement. (My DC V drop measurement points plus the 2.5 mV "confirmation voltage" did allow for this and correlation was quite good.) Interestingly, in both the no current and 10 Amp measurements, the individual joints accounted for a significantly smaller portion of the total system resistance when they were treated with the grease. The effect was easier to see when current was flowing than not. I hesitate to make much of even this--I could be seeing nothing more than noise that stacks up pretty. I really think much more current is going to be needed to see anything clearly if it's there to be seen at all. I would much rather see an improvement in DC voltage drop than AC resistance. Hopefully a few hundred Amps will make this possible, but I'll have to do those tests at the office. I do agree with Megacycle that this is not something ebikers are likely to want or need to pursue. This is for bigger vehicles that pull hundreds of Amps, if it's for anybody at all.
The reason I used much less Tuball was because I noticed that in the sample kit they sent, the "modified materials" included used extremely small amounts of Tuball to achieve what they do in each case--usually less than 1%. The "conductive film" they include is transparent, so that gives you some idea of how much carbon there can be in there. It's a far cry from the pitch black you get mixing 50% with the grease. I also began to wonder if too much of the material would negate some if its properties. OCSiAl's literature talks a bit about MULTI-WALLED CNTs and how they are easier to make, but don't have the spectacular properties of true graphene. I worried that perhaps too much of this material would make it behave more like MWCNTs than SWCNTs. That doesn't appear to be the case, but who knows with so little signal to work with?
friendly1uk
1 MW
Crimp then solder.
The crimp makes a bond where the materials contact. The solder can't get between them. It just fills in the voids with something that conducts better than air, and stops ingress of pollutants such as salts.
The crimp makes a bond where the materials contact. The solder can't get between them. It just fills in the voids with something that conducts better than air, and stops ingress of pollutants such as salts.
liveforphysics
100 TW
friendly1uk said:
If you crimp properly, you already have a gas tight solid cold forged copper slug through the entire crimped area cross section. >99.9% of crimpers folks use are not capable of making a proper crimp.
Adding solder doesn't decrease the connection resistance of a properly crimped joint, but does add new failure modes unnecessarily.
wb9k
10 kW
liveforphysics said:
It's a double-edged sword. The professional world uses this method all the time, it can work just fine. HV packs seem more often have heavy cable that is ultrasonically welded to the terminals--better yet than a crimp but totally out of the reach of most hobbyists, as is (to a lesser extent) even a really good crimping tool.
megacycle
100 kW
Thanks wb9k for those test results, I found interesting, my thinking was from an a.c. current perspective, as personally I've only really used greases such as Alminox on a.c. HV terminations and its highly impregnated with suspension of disgusting zinc (if memory serves), for Al/Al use, i remember my colleague being given a different type, which was a light colored, murky brown grease.
Now we test all cables and joints for milliohms, (for use on 132kV substations) and this new grease failed every test and the resulting lugged cables were scrapped, my thinking was along the lines of what would the manufacturer of a conductive nanotube grease recommend it to be used on, in respect of parameters, voltage, current, ac/dc, freq?
Eg (might be over thinking here), but does the conductive grease work better for like micro-arc flash over points on the joint at certain voltage or could it introduce itself as a small negligible value capacitance points on the surface with ac current, allowing improved ac conductivity?
As professional sparkies, we're often more concerned with installation/ mechanical issues, so generally we'd follow the standard procedural methods for terminations of conductors.
With lugs, generally following the manufacturer's documentation.
For automotive purposes, most are multi-stranded cables for the higher degree of flexibility, (vibration), these are terminated in bell lugs with the viewing hole, then crimped using recommended crimper and pressure.
Sometimes if the lug is a fair bit longer than the crimp die, I personally like to double crimp a lug, with a space between the crimps.
The only time I've applied anything to a joint recently, has been silver soldering busbars, together.
Can't remember introducing solder into a lug for years, as I think Luke infers, it can cause a point of inflexibility where the cable meets the lug.
Back to your nanotubes, been reading up on the conductivity and find it fascinating, the conductive types, how they quantum channel the electron flow, amazing stuff.
Scuse the ignorance, but relating this to a lap joint, do the conductive nanotubes align for conduction in any plane, like a sphere, or is it just relying on the abrasive quality, as the latter I don't think would give an improvement in conductivity, just like other greases, it's mainly to keep oxidation away.
Hence in a fashion, it is a 'conductive' grease, by helping to maintain the conductive integrity of the joint.
Now we test all cables and joints for milliohms, (for use on 132kV substations) and this new grease failed every test and the resulting lugged cables were scrapped, my thinking was along the lines of what would the manufacturer of a conductive nanotube grease recommend it to be used on, in respect of parameters, voltage, current, ac/dc, freq?
Eg (might be over thinking here), but does the conductive grease work better for like micro-arc flash over points on the joint at certain voltage or could it introduce itself as a small negligible value capacitance points on the surface with ac current, allowing improved ac conductivity?
As professional sparkies, we're often more concerned with installation/ mechanical issues, so generally we'd follow the standard procedural methods for terminations of conductors.
With lugs, generally following the manufacturer's documentation.
For automotive purposes, most are multi-stranded cables for the higher degree of flexibility, (vibration), these are terminated in bell lugs with the viewing hole, then crimped using recommended crimper and pressure.
Sometimes if the lug is a fair bit longer than the crimp die, I personally like to double crimp a lug, with a space between the crimps.
The only time I've applied anything to a joint recently, has been silver soldering busbars, together.
Can't remember introducing solder into a lug for years, as I think Luke infers, it can cause a point of inflexibility where the cable meets the lug.
Back to your nanotubes, been reading up on the conductivity and find it fascinating, the conductive types, how they quantum channel the electron flow, amazing stuff.
Scuse the ignorance, but relating this to a lap joint, do the conductive nanotubes align for conduction in any plane, like a sphere, or is it just relying on the abrasive quality, as the latter I don't think would give an improvement in conductivity, just like other greases, it's mainly to keep oxidation away.
Hence in a fashion, it is a 'conductive' grease, by helping to maintain the conductive integrity of the joint.