Bed is showing spikes in the graph

Ask the MakerGear community for assistance...
Post Reply
markb
Posts: 180
Joined: Tue Apr 15, 2014 11:06 pm

Bed is showing spikes in the graph

Post by markb » Wed Aug 13, 2014 2:40 am

After doing PID tune it look goods again but the bed is showing spikes in the graph.
Why would the bed temp show spikes? They seem false because it could change that much in that short period of time.
Attachments
BedTemp.JPG
BedTemp.JPG (39.43 KiB) Viewed 4521 times

User avatar
ednisley
Posts: 1073
Joined: Fri Apr 11, 2014 5:34 pm
Location: Halfway up the Hudson
Contact:

Re: Bed is showing spikes in the graph

Post by ednisley » Wed Aug 13, 2014 2:54 am

markb wrote:Why would the bed temp show spikes?
That suggests the thermistor wires are:
  • making intermittent contact in the connector or
  • have a break between the bead and the connector
The usual test is to wiggle & jiggle the wires from one end to the other to trigger the fault. The simplest fix boils down to getting another thermistor-with-connector, as the chromel/alumel wires don't take solder very well at all.

markb
Posts: 180
Joined: Tue Apr 15, 2014 11:06 pm

Re: Bed is showing spikes in the graph

Post by markb » Wed Aug 13, 2014 2:58 am

Thanks I will look at it when the print finishes.

Dale Reed
Posts: 376
Joined: Thu Apr 10, 2014 1:39 am
Location: Cleveland Heights, Ohio USA

Re: Bed is showing spikes in the graph

Post by Dale Reed » Wed Aug 13, 2014 3:17 am

Interesting that all the spikes are downward. Most thermistors (and other conductors, for that matter) have higher resistance at higher temperatures. Now, it may be that the firmware is reading the temperature as zero when the thermistor circuit is open, and the graph is plotting a smoothed temperature trend, so you don't see the spike go all the way to zero. (Might have a look at the communications from the printer when the software is polling the temperature -- see if the number returned is zero.zero or something else....) Anyway, what I'm getting to is this: also look for some little bit of wire or other conductive stuff getting into the circuit --- a short rather than an open. Granted this is highly unlikely --- a loose or intermittent connection is most common --- but it wouldn't hurt to inspect for this other possibility...

Hope I'm wrong. But I do enough electronics projects that I get little strands of wire into places they shouldn't be when I'm not careful. Getting one in the electronics enclosure could lead to a bigger problem if you don't get it out of there!

Ask me what smoking 2-watt resistors smell like!
Dale

markb
Posts: 180
Joined: Tue Apr 15, 2014 11:06 pm

Re: Bed is showing spikes in the graph

Post by markb » Wed Aug 13, 2014 4:00 am

Strange after some time the spikes are gone.
Attachments
BedTemp2.JPG
BedTemp2.JPG (36.73 KiB) Viewed 4513 times

Josh
Site Admin
Posts: 91
Joined: Thu Apr 10, 2014 3:32 pm

Re: Bed is showing spikes in the graph

Post by Josh » Wed Aug 13, 2014 2:05 pm

Dale, we actually use NTC (negative temperature coefficient) thermistors - the resistance starts at 100kΩ(±1%) at 25°C, and drops as the temperature increases (35kΩ@50°C, 6.2kΩ@100°C, etc.). So a spike to 0 is actually an intermittent open circuit (and a short is generally read as 359.9°C).

markb
Posts: 180
Joined: Tue Apr 15, 2014 11:06 pm

Re: Bed is showing spikes in the graph

Post by markb » Wed Aug 13, 2014 4:29 pm

Thanks for all the help!
It was the thermistor connector coming apart. It was about ¼ the way apart. I put it back together and used some black elect tape to help it stay together.
I am doing a few hour print right now and it is working great again.

Post Reply