Duet Wifi - Board ratings limits and design tolerances

yngndrw

Hi All,

The Duet Wifi is advertised as supporting a maximum input voltage of 25V and a maximum heated bed current of 15A. How close to these limits (And beyond) can you go while still being supported ?

I'm guessing that the maximum voltage is not an absolute maximum as some people are using 24V and that wouldn't give very much headroom. Meanwell do a nice 27V power supply which would improve warm-up times for a standard 24V heated bed and bring the current draw closer to the limit of the board, so would this setup be usable ?

Another example is with my current 12V heatbed (One of the older Mk2 heatbeds) which seems to draw a little over 16A - I'd like to upgrade to a 24V setup at some point but in the meantime, would that kind of draw be supported ?

deckingman

I think it might be a bit unreasonable to expect the makers to state that it is OK to exceed the recommended maximum ratings. By all means do so if you feel you have to, but it must be at your own risk.

dc42

The absolute maximum rating of the TMC2130 drivers is 30V when they are driving stepper motors. However, stepper motors generate power when you decelerate them, so the VIN voltage can go higher then the voltage from the PSU, especially if you are not powering a heated bed from the same power rail. That is why we recommend a nominal VIN of 25V maximum. If the power monitor detects 29V or more, it will shut down the stepper drivers in an attempt to protect them - which is likely to ruin your print.

The M122 command will show you the maximum and minimum supply voltages measured since you last ran M122, along with the number of these overvoltage events.

We are hoping to increase the rated bed heater current on the blue production Duet WiFi boards beyond 15A when the new thermal test results come in. The white pre-production boards will remain rated at 15A.

yngndrw

@deckingman:

I think it might be a bit unreasonable to expect the makers to state that it is OK to exceed the recommended maximum ratings. By all means do so if you feel you have to, but it must be at your own risk.

While I agree that it's unfair to expect a warranty for damage caused by exceeding specifications, different designs will use different tolerances and design in different amounts of headroom. Also current limits are generally related to cooling and this can vary, see David's response for a good demonstration of this.

@dc42:

The absolute maximum rating of the TMC2130 drivers is 30V when they are driving stepper motors. However, stepper motors generate power when you decelerate them, so the VIN voltage can go higher then the voltage from the PSU, especially if you are not powering a heated bed from the same power rail. That is why we recommend a nominal VIN of 25V maximum. If the power monitor detects 29V or more, it will shut down the stepper drivers in an attempt to protect them - which is likely to ruin your print.

The M122 command will show you the maximum and minimum supply voltages measured since you last ran M122, along with the number of these overvoltage events.

We are hoping to increase the rated bed heater current on the blue production Duet WiFi boards beyond 15A when the new thermal test results come in. The white pre-production boards will remain rated at 15A.

Thanks David, I appreciate that. I thought that might be the reason why the drivers had a higher limit in the datasheet but wanted to confirm - I wasn't sure if the deceleration issue was only valid for DC brushed servos although thinking about it, it also AC servos which are very similar to steppers. I was thinking that something like the Geckodrive returned energy dump circuit (http://www.geckodrive.com/returned-energy-dump) could help safely increase the limit slightly - At least enough to be able to push the heat bed voltage higher.

I look forward to the thermal test results - I did notice that the traces did heat slightly but I don't know how hot they would have gotten as I only heated to 30C. I'm hopeful that a fan behind the board could increase the rating by a fair amount.

Edit 1: I tried a fast move on the X and Y axes to measure the voltage change: (This is on a Prusa i2 with Nema 17 motors at 1A - I didn't include Z as for some reason G0 seems to do a coordinated move ?)

G0 X100 Y100 F36000
...
Supply voltage: min 11.9, current 12.0, max 12.1, under voltage events: 0, over voltage events: 0

Edit 2: Using a voltage of 10V reduces my heatbed voltage to a much more manageable 13.5A. This obviously reduces the maximum speed of my steppers, but this shouldn't affect me as the Prusa i2 is about as rigid as a piece of wet pasta. As an aside, when trying to perform PID tuning on my heater bed, the motors were clicking at the same rate as the heater. (~5Hz ?)