Metal Plus heatbed is always on

[daveculp]

I have a metal plus I purchased about two years ago. It has been down for awhile as it had a failure during a long print. Afterwards the X axis acted bizarre. I tried re-flashing the firmware but that did nothing. I thought it was either a failed motor or the board went bad. After a few months of it being down I decided to work on it today. I purchased a new motor and a new Rev F6 board from Printrbot. I replaced the motor and still nothing so I replaced the board and it still didn't work. Stripping down the printer it turned out to be a failed endstop (I felt stupid once I tracked the problem down). A bit of solder later and the printer was working except for one thing:

The heat bed was always on.

As long as the power was applied the heat bed is heating up. With the new F6 board it was getting to 130 degrees!! I rechecked all connections and tried again - same behavior. I then put the old board in which originally came with the printer and fired it up - same behavior, as long as power is applied the heatbed is heating up. With the old board it only gets to around 105 degrees.

Any thoughts? Two bad boards somehow?

[AZBotr]

Check those two mosfets by the power terminal. If they're even barely touching that could be causing your problem. Hopefully it's that simple for you...

[daveculp]

Nope, not touching on either board.

Anyone have a part number on the FET? I can easily desolder and replace it.

[RetireeJay]

IRLB8743PBF-ND
The operation of both heater circuits is that one terminal is permanently connected to +12V, and the other terminal gets grounded through the N-channel MOSFET when calling for heat. So if you have some other pathway to ground that bypasses the MOSFET, your heater will always be on and it isn't the MOSFET's fault.

[daveculp]

IRLB8743PBF-ND
The operation of both heater circuits is that one terminal is permanently connected to +12V, and the other terminal gets grounded through the N-channel MOSFET when calling for heat. So if you have some other pathway to ground that bypasses the MOSFET, your heater will always be on and it isn't the MOSFET's fault.

Thanks, I just ordered some from Digi-Key.

The heating element on the hotend works fine, heats to the set temp and stays there and turns off when done. I will go back and check everything and also disassemble the bed to check the heating element and thermistor and look for any shorts.

I was reading in another post in which you told someone how to to test the MOSFET. It appears I simply need to test the resistance between the two terminals on the screw terminal correct? What readings am I looking for if the MOSFET is bad?

[RetireeJay]

It appears I simply need to test the resistance between the two terminals on the screw terminal correct? What readings am I looking for if the MOSFET is bad?

Actually, one of the screw terminals goes to +12V and the other to the MOSFET so measuring across the screw terminals is not the best approach. If the MOSFET is shorted, it's going to be zero resistance to ground, not +12V. So here is how to check it. The meter is checking between the MOSFET output (drain) and ground.

Printrboard Checking Heatbed MOSFET.JPG

[daveculp]

Thanks, looks like a bad MOSFET on the new PrintrBoard I received and I assume a bad one on the old one but have not checked it yet. Weird that both went bad. The new MOSFETS should be in the middle of next week so I will replace and see if that fixes it.

[daveculp]

Looks like it wasnt a bad MOSFET. I was measuring incorrectly.

I got the brand new Printrboard in today and installed it and had the same issue - the heat bed is always on and heating as long as power is applied and will not turn off. The hot end work fine, it doesnt heatup unless told to.

So what do I check now? Could it be the actual heat board?

[RetireeJay]

You must have a short circuit to ground somehow. The heat board itself can't heat up unless current is flowing through it, which means one terminal connected to +12V (which is normal) and the other connected to ground.
Is your board supported on insulating standoffs? Any chance there is an inadvertent connection to ground under the board? If everything there is OK, then check the cable between the board and the heater. Is it pinched against a metal chassis somewhere?

[daveculp]

Thanks for the help. I checked everything I could last night before giving up. The board is on metal standoffs, I dont believe they are insulated. I will try putting a nylon spacer between the board and each standoff and see if that works.

[RetireeJay]

You could disconnect the heater wires from the Printrboard, and with both wires "floating" in air, check for continuity from either wire to ground. They should be floating, with megohms to ground. To be very thorough, check for shorts both to chassis ground, and to the power supply's ground lead (black) when the power supply is plugged into the board. And, using the diagram I showed before, check for continuity from the MOSFET drain to chassis ground (when power is off); again, this should be in the megohms or at least kilohms, not in single-digit ohms.

[Mooselake]

Does the displayed bed temperature increase as it warms up, or is constant? If constant, what is it? Perhaps it's a problem with temperature sensing, like it fell off the board. That problem has lead to extruder fires, not sure if the heated bed would get hot enough.

Kirk

[daveculp]

Does the displayed bed temperature increase as it warms up, or is constant? If constant, what is it? Perhaps it's a problem with temperature sensing, like it fell off the board. That problem has lead to extruder fires, not sure if the heated bed would get hot enough.

Kirk

No, the displayed temp continues past the setpoint temp. With the new board revision it gets to over 120C when I have it set at 70C.

[daveculp]

Update:

I left my meter at work so I couldnt take any continuity or resistance measurements. However, I did experiment.

I removed the board from the standoffs and left it "floating in air" in other words, not touching the Printrbot at all. It works fine that way, it stays off when power is applied to the bot. When commanded to heat to 70C it heats and stays at 70C. When commanded to go to a lower temp it turns off and cools to that lower temp and maintains it. When screwed back on to the standoffs it goes back to its old behavior, the heatbed is on and heating as long as power is applied to the bot.

I didnt have any nylon washers of the right size so I ordered some via Amazon and they should be here tomorrow.

Edit: Another experiment I did. With only ONE of the two heat bed wires (doesn't matter which one) plugged into the first screw terminal of the heat bed (the one closest to the hot end plug) and the other wire left unconnected and floating the heat bed will still heat uncontrollably with the power on and the board screwed to its standoffs.

[RetireeJay]

Another experiment I did. With only ONE of the two heat bed wires (doesn't matter which one) plugged into the first screw terminal of the heat bed (the one closest to the hot end plug) and the other wire left unconnected and floating the heat bed will still heat uncontrollably with the power on and the board screwed to its standoffs.

Aha. Basic physics here: the heatbed CANNOT heat up unless current flows through it. Current must always flow in a loop, with a continuous path from one terminal of the power source, through the load, and back to the other terminal of the power source. If current flows through the bed with only one wire (+12V) connected, then that proves that there's a non-wired path for current to return to ground. But if it doesn't matter which wire you connect, then it means there must be a short to ground somewhere within the heat bed, maybe about halfway between terminal A and terminal B.

The heated bed is actually just a printed circuit board with a copper trace snaking around. The width, thickness, and overall length of the trace define the ohmic resistance of the trace, which in turn defines how much power it will convert to heat when 12V is applied. The circuit board is coated with one or two layers of insulation to isolate the copper trace from its external environment. But if the insulation has been compromised and the board is sitting on a grounded metal structure, then the copper trace will have a sneak bypass path for current to return to ground. The solution could be as simple as adding a piece of paper between the heat bed and the underlying metal bed.

The fact that you need to have the board installed means that the ground connection from the power supply to your metal chassis is going through the board - which may in fact be perfectly normal and proper. The through holes for the screws have copper-plated collars around them, and it wouldn't surprise me if they are connected to the power supply ground.

[daveculp]

I was thinking the same thing, that the short may be in the heat bed itself. I have gone through every wire on that printer looking for signs of a short, frayed wires, broken or burnt wires etc... I cannot find anything. However, I have not checked the heat bed at all. I actually ordered a new heater and it should be here today or tomorrow.

[daveculp]

The solution could be as simple as adding a piece of paper between the heat bed and the underlying metal bed.

And you might be right! I just disassembled the bed, inspected it, put a piece of paper under the bed heater and installed everything. The bed does not come on when not commanded. When commanded to heat to 70C it does so and stays at the set temp. When commanded to go to a lower temp it does so and holds the set temp.

Initial tests are positive, I have a new heat board coming tomorrow and I may go ahead and swap it out.

[cbxbiker61]

A thought would be to reverse the wires going to the heatbed....

It is a very common practice when it comes to switching DC circuits on and off to run the "hot" +12V lead to the load unswitched and switch the -12V side. It's just a lot easier to do it this way with MOSFET's. The frame of the printer is usually at -12V, if it's connected to the power supply in any way. So in effect if what the heatbed manufacturer considers the -12V side is instead connected to +12V and it comes in contact with the frame you have this sort of problem.