Anderson Connectors - seem to be having issues?

john61ct

Adventurer
SAE connectors are all over the place in ampacity tolerance.

Mouser et all, Tecmate Optimate and PowerWerx I would trust up to 10 maybe 15A continuous.

Cheap Chinese unbranded maybe half that.

Genuine Anderson with proper crimpers and quality wire, much better to adopt as a standard
 

mechengrsteve

Most time adventurer
Two experiential examples with genuine and 'fake' Anderson connectors;
1) I designed and built a commercial product that had 2-100Ah Powersonic SLA batteries that powered an 850W inverter that usually ran about 250-300W. I did some pretty serious study on the heating of all the terminal connections using thermocouples directly attached to the contacts. The two batteries were on an exchangeable 'sub-pallet' so used a 75A (genuine) Anderson connector. The typical ~30A continuous current heated them to ~50-60C? I can't remember exactly but I recall being surprised that at less than 40% of its continuous rating, the contacts were already almost at the limit of the 60C wire insulation that was crimped into them. Mind you, Anderson does not specify high-temperature insulation for these connectors.
2) I put a 2nd alternator (24V) on my F350 that charged a 17kWh lithium battery (35 Nissan Leaf packs) in a 2" spacer under my truck camper. I thought I was safe using a 120A alternator with a thermostat that cut charging when the alternator reached 120C. Power ran through 4AWG/105C (rated 135A in engine spaces) wire with a Chinese copy of the Anderson 175A connector between truck and camper. I was expecting to see an absolute maximum of 120A and even then, not for long. Well, it turns out, the batteries can cause the 120A alternator to put out 150A @~29V when at a low SOC.....and it can go on for a long time with the thermostat cycling the alternator on at 110C and back off at 120C.... The clone nickel-plated (not silver as the Anderson are) contacts produced enough heat to melt the connector pretty significantly on two separate occasions. The first time, with the cable soldered into the contacts, it deformed the connector and burned the 105C insulation (rubber welding cable) enough that I had to cut the cable back several inches. I do believe the soldering of the cable instead of crimping contributed to the heating. The second time, with the cable hex crimped onto the contact, it melted the connectors completely fusing them together. I opine that when the contacts may not be lying completely flat against each other due to heavy cable forces (mostly twisting). At any rate, they got hot enough that the spring that retains the contact and forces them together conducts heat into the plastic body, softening it which then causes deformation due to cable twist, further misaligning the contacts, causing even more heat generation. My new camper is not dismountable (so no connector) and has 2AWG/105C wire (175A rated) on a 220A alt
SAE connectors are all over the place in ampacity tolerance.

Mouser et all, Tecmate Optimate and PowerWerx I would trust up to 10 maybe 15A continuous.

Cheap Chinese unbranded maybe half that.

Genuine Anderson with proper crimpers and quality wire, much better to adopt as a standard
SAE is the author of the specification and requires material thickness, plating, interference between plug/socket, etc. that will produce a contact set with an ampacity of 24A max (same as 1/4" QD terminals). The non-compliance of manufacturers notwithstanding, reputable parts cab be reliable. Beware that even though they are rated to receive up to 10AWG wire, that does not increase the rating of the contacts above 24A. I concur, that there are so many copies flooding the market and since the US automotive manufacturers no longer use this connector, I believe it to be difficult to actually find terminals that meet the spec.
 

john61ct

Adventurer
Two experiential examples with genuine and 'fake' Anderson connectors

Experience with fake ones? why? To be avoided at any cost.

I also hope you didn't try soldering wire to any Andersons, only use with crimping.

And again, unless the alternator is fitted with a special current-limiting VR

a DCDC charger is required to protect the alternator and intermediate infrastructure

and of course the battery.

LI is not self-limiting like Lead
 

mechengrsteve

Most time adventurer
You were meant to take away from my experiential examples, two things, 1) That NEITHER genuine Anderson nor their copies will support the load they claim. I do not believe that the copy I used was significantly lower in ACTUAL capacity than the genuine. By allowing excessive alternator output, I precipitated its destruction. 2) That alternator's rated capacity is just one number on the whole performance curve of an alternator. Given a low enough battery internal resistance, you can drive it to higher than its rated capacity. This excessive capacity can be survivable if the excessive temperature is prevented (which I did and do).

I am a mechanical engineer with lots of professional experience with DC power, aerospace and high current commercial connector design and manufacturing. I have had several negative experiences with genuine Anderson connectors and decided to see how the Anderson copies compare. Other than the nickel plating instead of silver, they were very well made. Contacts were well-formed, thick and solid copper. The housing was also thick and well molded. Being 'copies' it also retains the inherent design flaws of the genuine.

Connector selection is about 'appropriate use' and not about getting the highest performance when it is not needed. By the same token, connectors need to meet their specifications. I believe that my copy was very comparable to the genuine Anderson. I was anticipating current in the 100-120A range because, at the time, I knew that the 120A rated alternator would likely only produce 60A continuous duty but I expected it to peak at its max rated 120A till it overheated and the thermostat began to cycle it on and off (based on temperature bot approximately 10' on/5' off cycle). A 175A continuous duty rated connector seemed appropriate (and still does if either actually performed to spec). I made a mistake in my initial telling of the story. The first failure was crimped and the second was soldered. The first failed when I was in a campground, 2000 miles from my crimper so had to repair it with solder. I am fully aware of the undesirability of solder. I attribute the first (crimped) failure to cable twisting the contact, which is an Anderson design issue (part of my prior Anderson failure analysis), and the high amperage, precipitated failure.

Its worth noting that my prior failures with the genuine Anderson arose from a fundamental design problem which it still has. The contact surfaces are flat and must remain in perfect contact in order to support the load, but they rarely do. The housing does not prevent the contact from turning other than by the 'spring finger' used to retain the contact. If the cable is twisted in operation, or if it is not VERY FINELY STRANDED, the contacts turn relative to each other and even a tiny amount, resulting in their high contact temperatures (yes, Anderson admitted this to me).

At the time of this incident, My dedicated 120A/29.4V alternator (then a pre-set voltage output) had a thermostat in line with the excitation of the external regulator that limited the alternator case temperature to 120C. This system works well and has been tested extensively both before and after. While not explicitly controlling current (in this iteration of control), I enforce a duty cycle on the alternator by directly monitoring temperature. I also had voltage, current and temperature monitoring of the alternator in the cab so I was fully aware that it was producing 150A (undesirable but by itself is not a problem) into a cable/connector system that should have been able to handle it. What I did not have (and have now) is the ability to easily dial back the voltage output to reduce the amperage output.

In the latest control setup, I have 'operated' on my Transpo Voyager adjustable external regulator and replaced the embedded 10k adjustment potentiometer with a panel-mounted 10k precision potentiometer. (see attached photo). When I want to charge, I turn on the master switch (upper left) which supplies excitation, I check the voltage and current output (top row left) of the alternator (now a 220A), based on the current battery (last picture) SOC and use the trim pot (upper left) to adjust the alternator voltage set point to achieve a corresponding (max) current I want. At this point, I can hit the road and pay no attention. With a fixed voltage output, as the battery charges, the current will decline naturally. I periodically increase it if I want. If I'm producing more current than the alternator can sustain continuously, the thermostat will cut charging at 120C and restart at 110C.

I have several video clips on my Instagram feed @maximum.4x4.camper that show it in operation. There are more detailed writeups on my blog, workingonexploring.com and some videos on YT (search 'workingonexploring').
 

Attachments

  • Console12-24Vcontrols.jpg
    Console12-24Vcontrols.jpg
    213.1 KB · Views: 21
  • IMG_20210305_184515.jpg
    IMG_20210305_184515.jpg
    260 KB · Views: 22
  • IMG_20210318_160453.jpg
    IMG_20210318_160453.jpg
    207.4 KB · Views: 22
  • IMG_20210321_175909.jpg
    IMG_20210321_175909.jpg
    162.6 KB · Views: 22
  • WireDiagramAlternatorChargingV2-26FEB21.jpg
    WireDiagramAlternatorChargingV2-26FEB21.jpg
    512 KB · Views: 24
Last edited:

john61ct

Adventurer
NEITHER genuine Anderson nor their copies will support the load they claim.
And I say that is false.

A genuine 45A APP connection properly crimped using the right wiring will exceed their claims on the datasheet by quite a bit.

Part of that is no strain on the connection, not just pulling tension but also no torque twisting. And yes fine stranded tinned, UL1426 rated105c "boat cable" is all I'd use.

Same with all their connectors. What you call design flaws I believe are just you failing to observe the above qualifiers, or not reading the detailed spec sheets.

Most alternators cannot put out their current output rating for more than a very short time. If continuous load, used for charging a deep cycle bank - not what they are designed for - they must be derated with a current limiter to maybe even half the rated current.

Exceptions exist, large frame units designed for fire trucks / ER vehicles military etc.

But even these need the current limiting device with LI banks.

Experimental testing to destruction is fine, but reliable longevity requires prevention first.

Your Transpo VR rig sound interesting, but my preference would be the Wakespeed / APC units that grew out of Al Thomason's FOSS project, set the profiles once and forget it.

Derating based only on a couple temp sensors is fraught since so many other hotspots can develop.
 

cody c

New member
There is an alternative high end connector to these, one of the related companies to the one I work for is building and recently distributing:


Not cheap as chinese anderson knock offs but one and done, and I apologize for the shameless plug (pun?)
 

Attachments

  • eaxtron snapshot.png
    eaxtron snapshot.png
    161.1 KB · Views: 21

john61ct

Adventurer
> chinese anderson knock offs

No one here is suggesting using these, should not even be part of the conversation except to say "avoid"
 

DaveInDenver

Middle Income Semi-Redneck
1) I designed and built a commercial product that had 2-100Ah Powersonic SLA batteries that powered an 850W inverter that usually ran about 250-300W. I did some pretty serious study on the heating of all the terminal connections using thermocouples directly attached to the contacts. The two batteries were on an exchangeable 'sub-pallet' so used a 75A (genuine) Anderson connector. The typical ~30A continuous current heated them to ~50-60C? I can't remember exactly but I recall being surprised that at less than 40% of its continuous rating, the contacts were already almost at the limit of the 60C wire insulation that was crimped into them. Mind you, Anderson does not specify high-temperature insulation for these connectors.
You don't say which series or wire size so I assume the PP75 in a two-way configuration, e.g. a battery positive and negative, as multipole.

At 30 amps the PP75 multipole will experience between 10°C and 40°C temperature rise over ambient depending on wire size. So if it's normal room temperature it should not be surprising the connector would range from 35°C to as much as 65°C. This chart is right in their datasheet.

Screen Shot 2021-12-17 at 7.36.59 AM.png

The rest of the product specifications. They are rated it looks like for 75°C but Andersen says they can supply higher rated temperature options that aren't in the catalog.

Screen Shot 2021-12-17 at 7.38.14 AM.png

There's nothing wrong with things getting warm as long as they stay under their design envelope. In this case using a single pole PP75 with 6 AWG wire at 75 amps with 25°C ambient will be fine. Change any variables, the number of poles, wire size, etc. then the capacity has to be derated.

Screen Shot 2021-12-17 at 7.53.41 AM.png

I use the SB50/PP75 connectors all the time for charger connections at around 45 amps continuous on 8 AWG. In this use they don't rise noticeably above ambient just like the datasheet predicts. The PP45 (smaller size) in this case gets nearly hot enough to melt the housings. Also just like the datasheet predicts.

It's also important to assemble PowerPoles or any connector correctly. They give you instructions and criteria to test to make sure your process yields a connection that will meet the specs. It's entirely possible that a bad crimp ends up causing the blades to make poor contact, that can easily happen using the wrong crimpers.

Screen Shot 2021-12-17 at 8.04.09 AM.png

Screen Shot 2021-12-17 at 8.04.20 AM.png
 
Last edited:

mechengrsteve

Most time adventurer
(Summarizing this case from ~20 years ago) I was using SB50 which are the two-pole housings with ~2' of 8AWG AWM wire on both sides and crimped with a hex crimp. They share the same contacts as the PP75. I instrumented all connections (terminals and connectors) with a 12 channel HP thermocouple meter and subject it to a 450W constant power load over the discharge of 2-12V/100AH SLA parallel-connected Powersonic batteries. A constant power load causes amperage change (increase) while battery voltage drop but is ~40A. The current was within the spec but temp rise measured was well over APP's spec. Conversations with them led to the root cause being 1) APP relies on the cable to sink heat from the contacts. The wire was short (on both sides) and did not offer as much radiative capacity as anticipated. 2) The bigger issue was the short wire also caused cable stress to twist the contact such that the flat faces experienced some misalignment which turned the line contact into point contact which was the proximate cause of heating. The two factors combined to lead to the overall problem, which practically required longer, more flexible cables. We changed to welding cable and decide to live with it. It still heated more than spec but did not cause product problems.
 

DaveInDenver

Middle Income Semi-Redneck
Conversations with them led to the root cause being 1) APP relies on the cable to sink heat from the contacts. The wire was short (on both sides) and did not offer as much radiative capacity as anticipated.
This is fundamental knowledge that any engineer who's worked on cable harnesses would know.
2) The bigger issue was the short wire also caused cable stress to twist the contact such that the flat faces experienced some misalignment which turned the line contact into point contact which was the proximate cause of heating. The two factors combined to lead to the overall problem, which practically required longer, more flexible cables. We changed to welding cable and decide to live with it. It still heated more than spec but did not cause product problems.
The mechanical stress is interesting. Wasn't aware there's a minimum pigtail length for Andersen connectors. The SB50/PP75 datasheet does caveat about mechanical stress (perhaps it was put in as the mechengrsteve-discovered clause) but no way to qualify or measure it. The terminals are supposed to align when mated but sounds like there's a limit.

Screen Shot 2021-12-17 at 2.41.54 PM.png
 

Forum statistics

Threads
186,339
Messages
2,884,663
Members
226,200
Latest member
eclipse179
Top