Page MenuHomeAleph Objects Inc

Find more reliable and manufacturable thermistor connector solution.
Closed, WontfixPublic


We are still seeing issues related to questionable temperature readings from the cartridge style thermistors. These parts are currently soldered, but this is a labor-intensive and manual process, sometimes the solder will not flow into the terminal, and is not recommended by the manufacturer of the terminals. Part of the issues could be related to the fact that the connector is not mounted and susceptible to degradation due to vibrations.

An idea I had would be to make a tiny PCB with pads on one side and a connector on the other and one or two mounting holes. Then make the stencil so it puts paste on the pads and after reflow you would have a nice tinned pad that you could touch the thermistor lead to and the tip of an iron to and quickly get a good joint without having to add any solder. Depending on how big the board is and the weight of the copper we could make this tiny PCB to handle the heater circuit too. Small PCBs like this are easy to manufacture in panels with v-scores. The mounting holes would minimize flexing and vibrations on the connections which would be good.

Event Timeline

kent created this task.Aug 23 2018, 3:47 PM
kent triaged this task as Normal priority.
kent created this object with edit policy "LulzBot Hardware Products (Project)".
west added a subscriber: west.Aug 27 2018, 7:18 AM

@kent if we went to there stock thermistor we might have better results. I believe it uses 3.0 micro crimps

tutley added a subscriber: tutley.Sep 4 2018, 1:24 PM

@kent we might be able to get E3d to install the 2 thermistor pins and 2 heater pins in a 4pos microfit connector making it 1 simple connection

kent added a subscriber: marcio.Sep 5 2018, 11:41 AM

@tutley I like the idea of the 4 position connector since that will make it easier to replace the hotend which is a very common service item. We would want to verify two things: that the terminals are the right gauge for the wire and that they are rated for the current that is required for the circuit.

Note: the connectors are not the only parts of the assembly that pose issues. We have also seen it where the white "potting" material becomes degraded and the two leads of the thermistor will touch either each other or the casing.

Another idea that I had been thinking of is to use redundant temperature sensors in the hotend. Because accurate temperature readings are so critical to a good print, having increased fault tolerance is appealing. @marcio do you think using redundant temperature sensors would be hard to implement into marlin (it might be there already idk).

west added a comment.Sep 5 2018, 11:45 AM

The downside to that is if either the heater cartridge or the thermistor go out you have to replace both not one or the other

@tutely: Marlin already supports redundant thermistors. You can set a maximum temperature difference beyond which the printer is halted.

@kent we would have to manufacture a custom heater block (likely harder to source) to take two cartridges, unless E3d can put two thermistors in the same copper cartridge (also likely harder to source). This would also mean adding more cabling and require a redesign of the x carriage control board. Also the archim has 4 thermistor slots, but we would need 5 unless we use other pins. seems unlikely to get incorporated in quiver

@tutely: Also, Marlin only supports dual thermistor redundancy in single nozzle mode, by making T0 redundant to T1.

kent added a comment.Sep 5 2018, 1:19 PM

hmmm. I'm not sure what the solution here is, just that it seems to be a recurring issue. I do think that having the heater and thermistor on seperate connector housings liek @west says makes sense, especially because the current through the heater is going to be quite a bit more than the thermistor and no point in going complete overkill on the thermistor side of it. This is definitely a challenging issue but I'm convinced we can find a solution. I'll do some research and update the ticket with what I find, if anything.

tutley added a comment.Sep 5 2018, 1:30 PM

we can do microfit for the heater and the same molex we have been using on the therm

logan added a subscriber: logan.Sep 6 2018, 10:17 AM

@tutley @west For the record, from experience, e3d's microfit crimps on those thermistors aren't any better than the SL series Molex, they will still fail if not soldered and they're much more difficult to inspect/repair.

logan added a comment.EditedSep 6 2018, 10:23 AM

I think these thermistor cartridges as a whole could be better, but I don't have a viable alternative in mind at this time. Some are just bad internally, some fall apart like @kent said, and we've had connection problems the whole time. It would be nice if there was a better cartridge thermistor that could be used here

@logan how do you have so much experience with e3d's microfit crimps? Is this statement based on the few prototype angelfish and finch toolheads that had the microfit connector on the thermistors? Molex does make a 26AWG version of the microfit 3.0 pin which would work perfectly with the E3d thermistors. I imagine E3D is using these for their standard crimps.

@logan they may not be perfect, but they are much better than the bare bulb, reprap discount thermistors we have been installing in the hex hotends.

logan added a comment.Sep 6 2018, 12:48 PM

@tutley no, this was at home from 2 standard e3d kits that i built into my own taz aeros. Thermistors I installed that had the microfit crimps failed at the connector, exhibiting the same issues of intermittent connections and improper readings. I haven't searched to see if this has been reported by other customers.

Those thermistor leads are certainly multiple gauges smaller than 26 AWG so I'm not sure how that would work perfectly (without soldering)

The worst part about the reprap discount thermistors on the hex is installing them, if installed correctly, i think they are more reliable. The difficulty on install is our fault entirely for designing the heater block the way we did.

tutley added a comment.Sep 6 2018, 1:09 PM

@logan my bad i was thinking of the blower fan being 26AWG. I think they make a 30AWG microfit crimp

tutley added a comment.Sep 6 2018, 1:12 PM

@logan it looks like those are a .2mm diameter leads. so approx. 32 gauge wire.

logan added a comment.Sep 7 2018, 10:28 AM

@tutley, you probably have seen the same already but looking at the molex website it looks like the microfits only go to 30 AWG

logan added a comment.Sep 7 2018, 10:30 AM

The SL series molex go all the way up to 36 AWG. Has e3d tried using the correct crimp terminals?

Any updates on this one?

logan added a comment.Sep 11 2018, 3:09 PM

@nickp I think we should try answering the last question asked above given that e3d is the origin of this issue.

Agreed... @tutley can you answer the question Logan posted above?


@nickp @logan i cannot speak for what e3d has and hasnt tried. I can contact them about it and update the ticket when i hear back.

E3d is also looking at improving their thermistor cartridge and have come up with a good plan. They are going to be splicing the lead to a larger multistrand wire inside the cartridge. This will make the thermistor more robust, as well as improve the reliability of the crimp.

They will be contacting us when they have a few samples ready to test.

logan added a comment.Sep 12 2018, 8:01 AM

That is good to hear. Thanks for reaching out to them @tutley I'm glad we asked. If they're working on it, we can focus on other stuff.

kent closed this task as Wontfix.Nov 26 2018, 4:14 PM

this one is in e3d's court atm. We can open it back up if it continues to be an issue.

I'm opening this back up with an update on how frequently we see this issue.

Going back to the beginning of October we have had thermal issues on:

23 Aerostruder tool heads
4 Dual V3
2 Mini 2 tool heads

Results vary when we see this. Sometimes we can solve it by turning down the cooling fan or having it turn on later on. Sometimes we get it back and MER will find an issue with the thermistor on the tool head. But sometimes customers will report issues with the tool head after getting it back from MER, and short of double checking settings we don't have a good solution for customers who still report issues with these tool heads.

So I do think we should consider this an on going issue we have not completely solved.

west added a comment.Dec 10 2018, 10:55 AM

@lansky are the issues seen on the toolheads listed from the connector and the soldered thermister pins or something else?

kent reopened this task as Open.Dec 10 2018, 11:30 AM

I would like to add that we are seeing a lot of fallout of the E3D thermistors in the tool head building process. The thermistors seem to be breaking at the pins/ pin housing. One of my assemblers has had 5 break in the past two days. Another one of my assemblers has 5 fall out so far out of an order of 20 today. here are some examples of what we are running into.

oliver added a subscriber: oliver.Dec 27 2018, 11:58 AM

There is a large order of about 200 thermistor coming in that we can use as replacements, the bulk of which will be going in to Pre-sub so they can easily be swapped out. Once they arrive, pre-sub can start red tagging the thermistors and it will be easier to track using the current QC tab on OCA. Talking to @MichaelM, it doesn't seem that this issue affecting one hot end more than any other.

@west We can't say, we really just see that the user has thermal issues, the wires are not visibly breaking.

Steven added a subscriber: Steven.Jan 15 2019, 8:17 AM

@tutley have we heard anything more from E3D on this? If not, can you please check in with them on this?

In T3431#64770, @tutley wrote:

E3d is also looking at improving their thermistor cartridge and have come up with a good plan. They are going to be splicing the lead to a larger multistrand wire inside the cartridge. This will make the thermistor more robust, as well as improve the reliability of the crimp.

They will be contacting us when they have a few samples ready to test.


FWIW as of last Wednesday they were still working on this, they know there are some issues.
We believe that a lot of the issues we are seeing internally are from poor handling. We have started to add some measures to keep these assemblies from being damaged.

@Steven @tutley

The majority of the issues I see with red tagged hot ends is for the thermistor issue. The majority of the thermistor issues is due to either the insulation coming loose and exposing the wire. Currently they are removed from the bags to test the thermistors in QC and then put back into the bag to be moved to pilot. Pilot has been removing them from the bag and then transferring them to Pre-sub which is where we are seeing the majority of the fall out. This specifically is the part that will be changing and Pilot will be required to keep the each individual tool head with their included bags and a new storage to keep them from falling out of the bags and getting tangled.

On the other side E3D did request that we increase our tolerances in how were checking the thermistors.

@Steven i have not heard anything and just contacted Rory requesting an update on this.

Update from E3D is that this is still in the works. They have run into some complications and progress has been slow.

DaniAO added a subscriber: DaniAO.Mar 22 2019, 8:18 AM

@tutley is there any more updates on this?

@DaniAO E3D Online are working on changing the design of their thermistors to have a stronger connection. They said that once they get some on some of our hot ends they will send us several samples to test. One of the big improvements are crimps that are actually correct for the wire gauge they are using.

kent reassigned this task from kent to Steven.Apr 1 2019, 3:25 PM
Steven reassigned this task from Steven to kent.May 2 2019, 7:51 AM

@kent I think having this ticket on the Quiver board is the wrong place as it would affect other printer models. We can either move this or recreate it but I think having it in the change order process would be best.

kent closed this task as Wontfix.May 2 2019, 8:06 AM

It made more sense for this ticket to be here when I created it during the development phase than now that quiver is in production. I would prefer to not change other products that are in production and instead focus on designing new products which do not have the same issues that the existing ones do. Along those lines, I would prefer not to change this for quiver or other products which have already been released to production, but I highly recommend not using this exact part on future designs.

kent added a subscriber: samantha.May 2 2019, 8:10 AM