Bed Heater Maxing Out and Unable to Maintain Temp

Hi Everyone,

A new problem recently appeared on my i3. It seems of late that my printer can't hold/keep the bed temperature steady. This has never been a problem through 3 rolls of filiment but seems to happen 3/4 prints.

Here's an example of what the Temp Monitor in Repetier looks like for the bed. You'll see toward the end of this print(luckily) the power to the bed MAX OUT and the bed loose(and not regain) temp.


While this print survived, the MAX-OUT has happened in the early part of the print causing the bottom of the print to pull away due to the large temperature difference.

I've identified the following possible causes:
- bad wiring/broken wire between board + hotbed : everything looks good here, solder joints are solid
- power supply, I am using a generic OEM-type power supply that came with my kit. I replaced the fan with a super quiet 50mm one but that was about 50 prints ago and this only started in the last 10.

Any ideas would be most helpful as I continue to investigate on my own.

Thanks!

Make sure that you have adequate power from your power supply. An inexpensive volt meter will tell the story, if voltage drops, then your power supply isn't able to keep it up.

Measure the voltage at 4 places: 1. At the power supply terminals themselves, 2. At the connections to the control board from the power supply, 3. at the output to the heat bed ,and then at the heater bet itself. A drop in voltage from the one side of the wire to the other indicates that the wire itself causes problems. The wire that came with my i3 kit was terrible, and got quite hot while things were in operation, so I replaced it with good heavier gauge wire, and that solved a lot of heat problems for me. With better wire, the heat bed got hot faster, stayed hot better, and put less load on the power supply.

This can also identify problems with the wire where there may be a strained part that has poor contact when the wire itself gets hot. I'd be willing to bet on a wire issue between the control and heat bed.

Also, are you using any insulation on the bottom side of the bed? that can help a lot, too.

Thanks for your thourough suggestions. I'll test the wires thouroughly tonight, I forgot about the insulation, I read about that previously and will put some cork down after checking all my wires.

Standby!

I completed the voltage measurements as you recommended. Here's what I saw:

12 @ power supply output
12 @ power in to mainboard
9.5 on the mainboard output to heated bed
9.2 at the solder joints to the heatbed

I bought an IR thermometer over the weekend as well, which tells me that 70 on my sensor = 55 actual degrees... Something I more or less figured out by trial and error but I'll add that to the list of issues to resolve. Will a new 100k thermistor fix that issue? I assume the bed is supposed to be getting 12v right? Is that mainboard problem?

Either way I think I'll plan to replace the wires to the bed. I've replaced almost every other part of my kit thus far why stop now

Thanks for your post!

Edited 1 time(s). Last edit at 04/22/2015 10:37PM by atheimer.

You are losing too many volts in the main board. Check the power in terminals to the main board are tight first. If they are OK, and the main board is RAMPS, then either the large square 11A polyfuse or the heated bed mosfet will be getting very hot. If it's the fuse, replace it - some people use an automotive-type blade fuse instead. If it is the mosfet, then check the voltage of the 5V rail on the RAMPS, it may be too low.

If you are using a glass bed, it's entirely normal for the top surface to be 10C cooler on top @ 50C than the thermistor reads. Also an IR thermometer will under-read when looking at glass.

Edited 1 time(s). Last edit at 04/23/2015 03:20AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Should I be reading 12v on the mainboard heatbed outputs?

I just checked mine.

When my heat bed is on, I'm only dropping 0.5V from the supply. The supply to the printer is at 11.7V, and I measure 11.3V at the heat bed.

So I would say you are dropping too many volts on your RAMPs board (if that's what you have).

I suspect the MOSFET. I found on my RAMPS board the MOSFET was an STP55NF06. It should have been an STP55NF06L. The L indicates it can be driven from low voltage. The non-L version needs about 10V to turn on fully, it is only just starting to turn on with a gate voltage of 5V. So at 5V, its drain-source resistance is too high and you get voltage drop across it (too much)... and it gets stinking hot! You have an infrared thermometer now, how hot is your FET getting when your heat bed is heating up? Mine got to 185 degrees C!.
I know this has come up here before. If the FET is your problem, do a search and you will find a common replacement mentioned in some of the threads. I think a genuine STP55NF06L (note the 'L' !!!) would probably be ok. I used an FDP8896 because I had some handy (now the FET stays below 60 degrees C).

While we are on the topic, IR thermometers are not very good at measuring glass temperature as dc42 has already mentioned. I find anything reflective will give unreliable results. When I have to measure the temperature of objects like that (metals), I often stick some masking tape the object and measure the temperature of the tape, it has a much higher emissivity.

After digging into the FET's as recommended, I came to the conclusion that I need to replace the mainboard (I have a knock-off Sanguinololu clone that came with my original kit so replacing the FET is not an option). The other alternative is to just relay the heated bed power directly from my PS.
So that's what I'm going to try first, connecting a solid-state-relay up... and seeing if that solves my problem for the short-term. I suspect I'll be replacing the power supply next but it can't hurt to see if the juice is there. (230w 20a supply)

I probably would've gone the new mainboard route but it appears Smoothie's and the other popular 32bit boards are sold-out so I'll try and hack it together for a while longer.

I am amazed though, all of the problems and the subsequent fixes have really improved my machine.... Print times are down, accuracy is up, and hopefully the heated bed will be speedier with this addition of bed insultation, 12ga wire, and relayed power.

Thanks for all the help (happy to hear continued opinions on this issue as well)

Cheers!

Quote
atheimer
After digging into the FET's as recommended, I came to the conclusion that I need to replace the mainboard (I have a knock-off Sanguinololu clone that came with my original kit so replacing the FET is not an option). The other alternative is to just relay the heated bed power directly from my PS.
So that's what I'm going to try first, connecting a solid-state-relay up... and seeing if that solves my problem for the short-term. I suspect I'll be replacing the power supply next but it can't hurt to see if the juice is there. (230w 20a supply)

I probably would've gone the new mainboard route but it appears Smoothie's and the other popular 32bit boards are sold-out so I'll try and hack it together for a while longer.

I am amazed though, all of the problems and the subsequent fixes have really improved my machine.... Print times are down, accuracy is up, and hopefully the heated bed will be speedier with this addition of bed insultation, 12ga wire, and relayed power.

Thanks for all the help (happy to hear continued opinions on this issue as well)

Cheers!

Just for interest Replikeo.com have Duet boards at $79.99 plus carriage Duet board

Doug

Just wanted to post an update....

- I successfully installed a SSD 20A DC-DC relay between the PS and the heated bed
- Upgraded wiring from PS to heated bed/relay with 12g wiring

After all of this, my heated bed DOES NOT heat up faster and even more disappointingly DOESN't hold its temperature on long prints.... So.... what's next... replace the PS I figure?

I'm currently using all 3 COM and V+ outputs on the power supply (2 of each going to generic Sang.... mainboard) and 1 of ea. to the heatbed/relay setup. I figure I probably don't need the 2nd wire to the mainboard since I'm not powering the heated bed from it anymore, can I disconnect it and make more power accessible for the bed? Is there a firmware setting perhaps to lower the mainboard -> heated bed output since I'm only using it as a signal wire now? Voltage reading on the relay when the bed is heating up (full on) is 10V. Lastly I am using PID mode in the firmware currently, would switching to Bang Bang help? Also, I've been reading that on the MK3 type beds the

Last Last, my power supply is the cheap chinese kit kind, rated @ 230w 20a... Should I ditch it? If so would anyone be able to recommend a good all-in-one PS?

What voltage are you getting across the heat bed terminals now?

I still think you should have found and fixed the excessive voltage drop in the RAMPS. No need to use a relay unless the heat bed needs more than 10A.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for the idea, problem is that with the bed being a generica Sang.... clone what I suspect to be the Mosfet is glued under a big heat sink....not quite sure what to do about it.... As with everything on my cheap-o kit fixing the RAMP = buying a new control board.

It's pretty bad though, I'm printing now and am 15d below the bed's target temp 20m in.

If the voltage drop really is in the mosfet, the simplest fix is to install a better mosfet, if it's a through-hole component on your board.

Edited 2 time(s). Last edit at 05/10/2015 02:16PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I had this same thing happen when I updated to the latest version of Marlin, was working great before. My solution was to insulate the bed. I glued a piece of corkboard under the heated bed. I glued it to the y carrage so it sits about 10mm under the heated bed itself, leaving a small airgap. Now my heated bed runs about 50% of the time and keeps temp nicely.

Good luck!

Hi everyone,

Thanks for all the advice, however I'm still fighting this problem.

- I replaced my power supply, 200w 20amp -> 275w 30a = bed heats up about twice as fast
- I tried adding a 35a solid state rely into the mix and powering the bed directly but the temp still wouldn't hold (i removed the relay and am back to powering the bed via the mainboard).
- I added cork as suggested by DriftyDave but this doesn't help either
- The mosfet on my generic mainboard isn't a "hole through" component and I have not been able to identify any way to replace it (if it is the problem)

At this point I figure my problem is one of the following:
1. mainboard/mosfet combo -> solution: replace the mainboard..(probably upgrade to one of those fancy 32bit boards if they're ever in stock again
2. could it be the actual MK3 heated bed? Mine isn't quite like the one's shown here: [reprap.org] per the instructions that came with it, it only has 1 positive and 1 negative connection from the 6 available solder points... does it need a second position lead soldered on perhaps? Here's a picture of the connection:




As always thanks for all the help!

Hey all... I'm having a very similar issue and trying to determine if it's the Power supply, or something going on with the RAMPS or Heater...
The PS is only outputting 11.2V @ idle... only the Hothead fan is on.... When the bed heater kicks on, the PS drops to 9.04V at the supply,
Same where the power goes into the RAMPS, and 8.94V coming out of the RAMPS at the heater terminals...8.89v at the heater itself.
I'm guessing I've got a bad supply, maybe a weak cap or something. It's a brand new printer kit, but with one of the cheap 12V/35A Chinese supplies.

I'm going to contact the kit seller and hopefully he'll replace it. I'm also going to go down the an electronics reclamation place and see if I can score a surplus
medical or industrial supply... the output from those things is crazy smooth and solid. If I can, I'll keep the replacement as a backup.

This damn thing drove me nuts for hours last night, not being able to figure out why the table and head wouldn't get hot enough. I thought it
was a firmware or software setting, but had those covered. At least now I know!

R

Most of the cheap Chinese PSUs have a potentiometer at the low voltage end of the terminal block, which adjusts the output voltage. However, a drop of 2V under.load seems excessive.

Edited 1 time(s). Last edit at 06/02/2015 03:26AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

@atheimer
The wires to your heated bed look a bit thin.
With the new PSU, can you measure the v-drop over the cable? Do they get hot? Is it silicone or PVC insulated?
People who print at high bed temperatures should use silicone insulated wires or at least split these PVC-twin wires apart.
The heat creeps in the cable pretty fast and can make the PVC soft and cause shortcuts. Even more when the wire gauge isn´t sufficient.
-Olaf

@Olaf:

Thanks for the suggestion, I have actually tried using 12g wire when I was powering things through the SS Relay but I'm back to using the stock wire that came with my kit. Also of possible note, mesauring the voltage on both wire ends is fine except that the loss in temp doesn't occur until the middle of a print. IE I have let the printer sit at full bed heat for 60 min with no fluctuation in temp. It only appears to be affected when other elements of the printer are running, furthermore I logged 20+ hours of great printing before this issue started occuring, 4+ hour prints with no issues then one day.....

I ordered a Duet mainboard to replace my generic Chinese board... [www.think3dprint3d.com] lots of re-wiring to do but hopefully this finally resolves my problem.

Then I can focus on building myself a nice enclosure.

@AZ_Ron:

A couple thoughts from my experience (also purchased a Chinese kit although I have now replaced 95% of it, only the frame, motors, and heatbed remain)
1. Really make sure the 110/220V switch is in the right place, I read this tip many times but finally out of desperation realized it wasn't. = this didn't help me though..
2. Try turning up the Potentimeter as suggested above, I had already replaced the PS when I learned about its existance. In hindsight I don't feel the PS was ever my real problem.
3. Measure the voltages between all the heat bed connections as recommended above but based on my experience I would look at replacing the mainboard (or the MOSFET if possible) first. It seems the Chineese generic MB's don't always do well with the voltage regulation part of the puzzle.

Hope that helps!