Failed kapton insulation on thermocouple blew up RS485?

[Jetty]

The piece of Kapton tape that covers the thermocouple and insulates it
on my MK7, broke over time and came in contact with the heater block
yesterday. After some time diagnosing with an oscilloscope I'd found
it had taken out both RS485 transceivers (SMD SN75176) (extruder and
motherboard) and I lost extruder comms.

The weird thing is the transceivers weren't completely bust, they just
weren't transmitting true 0's and 1's anymore, I was getting a logic
low of about 2 Volts, and a logic high of 2.5Volts.

I've replaced both transceivers and everything is working again. Can
anyone shed any light on why the thermocouple shorting like this
should blow up the comms? And as a piece of tape is the only thing
insulating it, why there isn't some protection built in somewhere? Is
there anything I can add to fix this issue? Cos I detest SMD
soldering :-)

Thanks

[Txoof]

Wow, I just posted on this subject literally a minute ago:
http://groups.google.com/group/makerbot/t/53a62690506e0e94

There's some suggestions that grounding the core can help protect the
extruder board. When I ground my extruder (via the plate it is bolted
to) the thermometer readings drop down to 0. Removing the ground
cable completely rectifies this problem. I checked the insulation on
the thermometer and it appears to be intact. In fact, after removing
the thermometer completely from the extruder, I still see the same
problem when the extruder is pulled to ground.

Any ideas?

[Luis E. Rodriguez]

I've enjoyed reading this article/blog and have heard of folks shorting their EC boards by touching the nozzle of the extruder and ZAP! I've had the static shock but didn't destroy the boards. I just touch the PSU before monkeying with stuff.

http://softsolder.com/2011/01/09/thing-o-matic-mk5-extruder-static-control/

Luis E. Rodriguez

Are you a Maker?

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.

[Jamesarm97]

Would this type of thermocouple type K make more sense to be using
than the standard one that comes with the kit? Seems like everyone is
trying to figure out how to isolate and mount the sensor. This one is
on Grainger and is already mounted in a 1/4" ring that bolts on:

http://www.grainger.com/Grainger/TEMPCO-Ring-Thermocouple-3AEZ6

[Dan Newman]

Looks just like mine after I mounted it to a solderless lug with
Omega CC High Temp (an insulating epoxy intended for this application
with t/c's). My underlying t/c was from Makerbot so obviously
compatible.

A key question is whether or not the t/c junction is electrically
isolated from the mounting ring OR if you're supposed to provide an
insulating mica washer and hole insert/spacer to keep the screw head
and shaft away from the ring. (You see these all the time when mounting
FETs and whatnot to heatsinks.) It seems sensible that the t/c
would be insulated from the ring, but that's just an assumption.

Other than that, it should work. I'm tempted to order one just to
see for myself and to have a spare on hand if it does work.

Dan

[Jeremy Green]

I got one from somewhere else (mcmaster I think) that wasn't electrically isolated from the ring.
It'd be nice if it was…

-Jeremy

[Jamesarm97]

I'm trying to understand what exactly is getting shorted out when this
happens? I apparently fried my brand new extruder controller.
Everything I checked looked ok. What is the T+ / T- getting shorted to
that draws too much power and opens the T- trace and blows the
MAX6675? I measure no resistance from the toolhead heater core power
leads to the metal block the t/c mounts to. I was thinking it shorts
to +12 on the heater but since I am measuring no continuity from the
either heater lead to the block I'm now not sure. Just trying to
understand and prevent it in the future. I was also thinking about
adding a piece of those thermal heat transfer pads used to mount a
TO-220 style transistor to a plate and transfer the heat off thinking
it would add another layer of protection yet conduct heat.
http://search.digikey.com/us/en/products/53-77-9G/HS419-ND/1625645

[Jetty]

On Jan 16, 1:07 pm, Jamesarm97 <armstrong.ja...@gmail.com> wrote:
> I'm trying to understand what exactly is getting shorted out when this
> happens?

I'm no authority on this (I'm a new makerbot owner), but it's happened
to me, I fixed it
and I think there's a couple of potential issues that can happen.
Also, my extruder and HBP are now grounded.

The issues all seem to originate from the thermocouple breaking
through the kapton tape over time. I
re-insulated mine the other day, and it broke through again in about 5
days, then it contacts the block.

Once it's contacting the block (and also potentially when it's not
where static is involved if it can arc across the kapton), then
from that point one of 3 things can happen, which result in nasties
going via the thermocouple into the extruder board:

1. Nothing

2. Your heater fails and puts 12V onto the block

3. (the most likely). Static is conducted to the extruder block, via
you touching it, or what happened to me,
static buildup from extruding hot plastic. (worse in dryer
atmospheres)

Grounding the block to the PSU case solves this issue. If you get
erratic or low thermocouple temperature readings after grounding,
your kapton tape is likely pierced, re-cover it, or better still, pot
it in a lug and high temp potting compound.

Also, just because you can't see the thermocouple through the tape
doesn't mean it hasn't pierced it, on mine, the hole
is pin prick size and can't be seen until you pull the tape apart.
The test to see if it's pierced is to disconnect the thermocouple and
check
continuity to the block, you shouldn't have any continuity. But, it
could also contact only when hot. So the real test is once grounded,
it's pretty obvious from the readings.

If the heater fails, I would guess there's enough current to nuke the
trace. If it's static, anything is game, but from reports, most
likely is the A to D convertor or the RS485 chips.

[Txoof]

I followed your advice, and re-insulated my thermocouple.
Unfortunately, I get wildly erratic temperature readings when I ground
my heater. This even happens when I pull the thermocouple completely
out of the head and leave it dangling in the air, completely
disconnected from the core.
The grounding wire goes from a solderless lug on the extruder mounting
plate down to a screw on the outside of the PSU. As soon as I touch
the ground wire to the mounting plate, the temperature jumps up or
down to the limits. I'm certain the wire is continuous and I've
checked for shorts. It seems odd that the plate should have any
electrical connection at all!
Any idea what could be causing this odd behavior?

Thanks!

[Jetty]

Strange, this is what happens to me once the kapton is pierced and the
thermocouple is touching the grounded core.
It reads about 5C below what it should be, and keeps dropping to 0
intermittantly. Insulating it, solves this issue (and not screwing it
down too tight),
for a while until it eventually pierces again.

Try this:

1. Disconnect the thermocouple completely at the extruder board. Use
an Ohmeter. How many ohms?

2. Whilst still disconnected, measure the resistance between one of
the thermocouple leads and the extruder block.
Should be infinity. If it's <1K than you've got a short (normally
reads about 26 ohms), retape the thermocouple and test again.

Also check you have your thermocouple wires in the correct places.

If you are 100% sure you're not shorting to the block (check with an
ohmmeter), and the thermocouple is reading 20--30 ohms disconnected,
then there's no where for
it to short to, so you have to come to the conclusion that the
temperature is being read correctly, and it's an issue on the extruder
board. So I'd guess the next thing to be looking at would be the A to
D convertor.

[Jamesarm97]

Would it help with the grounding mods to also add some Zener ESD /
over voltage protection to the t/c lines and maybe the other lines on
the extruder controller? I would think the MAX6675 is running at
either 3.3 or 5v so I added a 5.6v zener across the T+ / T- lines and
am also thinking about a 15-20V zener across the fan and heater lines.
Any thoughts? I think I have some of these sitting around work.

[Z LeHericy]

You don't want to put anything in the T/C lines, since they work by dissimilar metal junction. So unless you have a zener with chromel or alumel leads, it may cause some error in the temp reading.

-Zeno LeHericy

//((=:Z:=))\\
INVENTIONS
Technologies
zinventions.com

[Dan Newman]

On 16 Jan 2012 , at 7:59 AM, Jamesarm97 wrote:

Drove over to a nearby Grainger's at lunch as they had some of these
in stock. The short answer is that you cannot use these without figuring
out how to electrically isolate the log AND sheath. Long story is that
it is nicely built with a nice braided metal sheath as well. The t/c
junction is not electrically isolated from the lug ring nor the
sheath.

Manufacturer's part number is TRW00114 / 3AEZ8. Made by Tempco,

Dan

[Dan Newman]

On 16 Jan 2012 , at 7:59 AM, Jamesarm97 wrote:

I seem to recall that Omega used to list a kapton insulated version of these,

http://www.omega.com/ppt/pptsc.asp?ref=WT

in their printed catalog a few years back. But, now they say to
inquire about kapton insulation. Might be worth a phone call? If you
go down to the bottom of the page, you can see how to order lengths longer
than 2 feet. (And you want glass not PFA) If you do find out about
ordering with kapton insulation, I'd be curious to know what you learn.

Dan

[Txoof]

Jetty,

I finally found some time to tear apart my TOM and run the tests you
suggested. Indeed, the kapton tape had failed around the
thermocouple. I removed the failed layers and added some fresh layers
and everything seems to be doing ok now.

Thanks for the advice!

On Jan 17, 8:54 am, Jetty <clelland...@gmail.com> wrote:
> Strange, this is what happens to me once the kapton is pierced and the
> thermocouple is touching the grounded core.
> It reads about 5C below what it should be, and keeps dropping to 0
> intermittantly.  Insulating it, solves this issue (and not screwing it
> down too tight),
> for a while until it eventually pierces again.
>
> Try this:
>
> 1. Disconnect the thermocouple completely at the extruder board.  Use
> an Ohmeter.  How many ohms?
>
> 2. Whilst still disconnected, measure the resistance between one of
> the thermocouple leads and the extruder block.
> Should be infinity.  If it's <1K than you've got a short (normally
> reads about 26 ohms), retape the thermocouple and test again.
>
> Also check you have your thermocouple wires in the correct places.
>
> If you are 100% sure you're not shorting to the block (check with an
> ohmmeter), and the thermocouple is reading 20--30 ohms disconnected,
> then there's no where for
> it to short to, so you have to come to the conclusion that the
> temperature is being read correctly, and it's an issue on the extruder
> board.  So I'd guess the next thing to be looking at would be the A to
> D convertor.
>

[Jetty]

Just for a heads up, you'll likely find it will punch through again
over time and it will continue to happen.

I ended up embedding the thermocouple in a copper bracket I bent set
in high temperature potting compound Dan recommended.

This is a much better long term solution.