Resistance controlled track (heater pad)

[DutchGert]

Is it possible to have Altium draw a pre defined pattern of tracks in a certain board area with a defined resistance (not impedance(!))?

I am designing a little heater pad for one of my projects and want to fill a lets say 8x5cm area with a 10 Ohm trace so it heads up quite nice when applying 12V.

Ofc I can take the plating thickness, trace thickness and calculate a length and then route it manually but then takes a very long time to figure out a good pattern.

[DerekG]

I am designing a little heater pad for one of my projects and want to fill a lets say 8x5cm area with a 10 Ohm trace so it heats up quite nice when applying 12V.

Not that I'm aware of. The problem is that you will test this under varying ambient temperatures & the copper is not designed to heat to such a temperature on a regular basis. It will eventually lift away from the pcb & in doing so will fail.

Install a wire wound resister that is designed to regularly run at 100C & you will have a reliable heat source.

[Whales]

As Derek asked, what sort of temperatures are you after?  It might be easier to ballpark the resistance and then use a variable switch regulator -- this would also allow feedback to be added in easily.

I'd make a long snake with lots of feeders, so I would not have to make a 'test' board to see if the final resistance seems right.  Keep cutting the feeders in order until the desired resistance is achieved.

[DutchGert]

I need to defrost a glass cylinder when a camera system boots up in winter. The idea is to cover the non-visible half of the cylinder with a semi-flex or flex pcb with the 10 Ohm track on it and use the transferred heat to defrost the glass.

[free_electron]

Copper changes resistance as it heats up . So there is no way to create a '10ohm resistor' using copper foil. It will drift from 5 ohm to 20 ohm going from -25 to + 100....
Wrong material

[Whales]

Copper changes resistance as it heats up . So there is no way to create a '10ohm resistor' using copper foil. It will drift from 5 ohm to 20 ohm going from -25 to + 100....
Wrong material

This feature could actually be useful -- negative feedback.  Its smooth rolloff does not really matter, as long as the glass is defrosted and things are not taken to damage temperatures.

[T3sl4co1l]

It would be very easy to make a temp controlled circuit: serpentine shaped heater, thermistor and a couple transistors (who even needs op-amps, let alone anything worse) to get something very robust.  Hint: you want the *worst case highest* resistance to be *lower than your worst case requirement* so you can guarantee being able to do the job.  The temp / power control (however you arrange the circuit) will regulate down from 100%.

Don't expect better than 20% tolerance on trace resistance (at a given temperature) from the PCB manufacturer.  Even if you request impedance controlled traces (an AC thing), I wouldn't think things would be much better (at DC).

I wonder if you could convince a PCB manufacturer to do printed nichrome...

Tim

[DutchGert]

Copper changes resistance as it heats up . So there is no way to create a '10ohm resistor' using copper foil. It will drift from 5 ohm to 20 ohm going from -25 to + 100....
Wrong material

I know, let me put it this way: I need a approx. 10 Ohm resistance track at 25*C. It will be heated up for a short period of time to defrost a glass cylinder so the exact amount of current passing through it is of no concern at the moment (it is derived from a at that moment only for the heater used supply capable of delivering 40W at 12V) so there is enough headroom.

In our company we use this technique in a lot of camera's and it works very well, especially if u add a temp sensor as feedback and regulate it a bit so the copper doesn't burn of the PCB.

My question is if there is a easy way to draw the track in Altium .

[T3sl4co1l]

My question is if there is a easy way to draw the track in Altium .

Besides copy and paste, the Trace Length Tuning tool might do the job.  You'd need to set up a basic path (a few segments back and forth, like a squared up "W" or something), then set the 'amplitude' of the tool to a bit less than the distance between sides of the initial shape.  Then 'paint' it along each segment.

It probably won't be regular -- another good reason to go with copy and paste.  Also, to use length tuning, I believe the net has to be part of a bus, so it has something to compare to.

Tim

[Rufus]

My question is if there is a easy way to draw the track in Altium .

If you want a regular pattern copying and pasting is quick and easy. Set the grid to the spacing you need. Draw an L shape trace copy selecting the right end and paste with X or Y flip to extend the existing track. Then copy the 2 tracks and do the same then 8, 16, 32, 64....

An 80 x 50mm patch filled with 0.4mm trace on 0.8mm pitch will give about 5m of trace and around 8 ohms at 25C in 1oz copper.

Not too sure about the reliability of 5m of trace when you are hammering it and you are relying on uniform contact with what you are heating to prevent any hot spot.

[DerekG]

I need to defrost a glass cylinder when a camera system boots up in winter. The idea is to cover the non-visible half of the cylinder with a semi-flex or flex pcb with the 10 Ohm track on it and use the transferred heat to defrost the glass.

Remember to open up the solder mask so that the heat can get out (this will reduce the life of the copper however if you overheat it).

Good engineering practices would say "why not use a number of low resistance resistors at regular intervals to generate the heat." At least you know that the resistors are designed to reliably dissipate the heat on an on-going basis.

I know I would definitely prefer to buy a camera that was designed this way.

[T3sl4co1l]

I need to defrost a glass cylinder when a camera system boots up in winter. The idea is to cover the non-visible half of the cylinder with a semi-flex or flex pcb with the 10 Ohm track on it and use the transferred heat to defrost the glass.

Remember to open up the solder mask so that the heat can get out (this will reduce the life of the copper however if you overheat it).

Wouldn't recommend this.  Solder bridges or extra conductor thickness (HASL or otherwise) or just shorting that'll later require an insulator anyway, are all ideal reasons to keep it.

We're not talking vaporizing heat here.  If we were, FR-4 would obviously be the wrong substrate to do it on!

Good engineering practices would say "why not use a number of low resistance resistors at regular intervals to generate the heat." At least you know that the resistors are designed to reliably dissipate the heat on an on-going basis.

I know I would definitely prefer to buy a camera that was designed this way.

Agree, this would be something more of a 'gold standard' approach while still involving PCBs.  The uneven surface would be annoying, but a thermal pad could fill in, for example.  There's also the possibility of flex PCB, using traces or resistors (but do mind the trace version is flexible while the resistored version is limited).

Tim

[wraper]

Why not just use flexible heater? The same thing made from the right materials.

[DerekG]

Why not just use flexible heater? The same thing made from the right materials.

Actually this is the best idea.

Install a length of nichrome wire which will easily bend around the circular lens. Remember that you will probably need to use some form of connector as nichrome does not solder well.

[DutchGert]

It is not getting designed to get really hot. There will be a feedback loop to keep the tracks itself from getting hotter than 80-100 degrees C to prevent failure. It will be on for very short periods of time (<1min) when the system is booted up below 0 degrees C to help warm the sensor housing quick enough to defrost in time. After that the heat from the system itself will keep the unit temp above 0 degrees C

Routing wire is not an option. For now I designed a proto 1 layer Flex Rigid board with 0,42mm tracks that will be folded in the cylinder. Proto arrives in a week so we'll see how it goes.

[DutchGert]

I need to defrost a glass cylinder when a camera system boots up in winter. The idea is to cover the non-visible half of the cylinder with a semi-flex or flex pcb with the 10 Ohm track on it and use the transferred heat to defrost the glass.

Remember to open up the solder mask so that the heat can get out (this will reduce the life of the copper however if you overheat it).

Good engineering practices would say "why not use a number of low resistance resistors at regular intervals to generate the heat." At least you know that the resistors are designed to reliably dissipate the heat on an on-going basis.

We also did that with a little board and some MELF resistors on it and a current source. Works very well but in this case it's not an option.
I know I would definitely prefer to buy a camera that was designed this way.

[DerekG]

It is not getting designed to get really hot. There will be a feedback loop to keep the tracks itself from getting hotter than 80-100 degrees C to prevent failure.

80 to 100C is pretty hot for copper substrate.

It will be on for very short periods of time (<1min) when the system is booted up below 0 degrees C to help warm the sensor housing quick enough to defrost in time. After that the heat from the system itself will keep the unit temp above 0 degrees C

Be aware that condensation often occurs above 0C (it's called the dew point). If someone was on an airconditioned bus in the tropics & they get out to take a picture, the cold lens is very likely to fog up if the humidity is high.