bang-bang hysteresis 1.09m dc42

this is more a query than anything but i cant seem to find an answer to it. I have a duet 0.8.5 and the latest version of dc42's firmware but for the life of me cant find where to adjust the hysteresis for the heatbed control.

I have modified M304 B2 (the band to various values but with no effect)

This is only bang-bang or hysteresis control so this should be possible, no?

its a simple thing but very frustrating at the moment.

Also is there something with the heat bed output because it fried my DC-DC SSR but my standard 16A finder relay is fine? and yes i did check the polarity before connecting it

There is no hysteresis control in bang-bang mode. There doesn't need to be for two reasons: the thermal mass of the bed heating system normally provides some hysteresis anyway, and the heater is switched between on and off at most 5 times per second, which any SSR can handle.

Your DC-DC SSR fried because it was a cheap one with a large voltage drop and consequent high power dissipation. A massive heatsink would have saved it, but the power loss would probably have prevented your bed heater from reaching high enough temperatures to print ABS anyway. Cheap DC-DC SSRs are no good for switching heated beds. Whereas cheap DC-AC SSRs are adequate for switching mains voltage heated beds.

If you are using a mechanical relay to switch the bed heater, then I hope you remembered to connect a flyback diode in parallel with the coil. The energy released by the coil when the relay is switched off probably exceeds the avalanche rating of the mosfet.

Edited 1 time(s). Last edit at 12/18/2015 08:26AM 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].

Thanks for the quick reply, there might not need to be hysteresis but it is a nice to have and every other system i have ever worked on over the past 25 years has this functionality as standard because its easy to implement.

The DC-DC SSR i was using was not cheap it was part of my own stock i keep and was supplied by omega instruments. I have used the same relay with both ramps boards and smoothie boards with no issues.

Using mechanical relays for hysteresis temperature control isn't new and has been working successfully in industry for many decades just look at Siemens Honeywell Eurotherm ABB etc... And yes there is a diode.

So basically to have simple on/off control of a heat bed and use a relay without chatter i cant use your firmware? is there another version of the reprap firmware that has hysteresis built in?

@bobblejot, I have a long list of features to add to RepRapFirmware. Adding hysteresis will be very low on that list because AFAIK you are the only person using a mechanical relay with it. It would also need a new parameter added to M305 to allow the hysteresis to be enabled and adjusted, because nobody without a mechanical relay will want to add it. Some folks even claim that you need to control the bed with PWM to avoid banding caused by small temperature variations.

If your SSR fried, I can only think that you connected the input terminals the wrong way round, or you didn't notice that it was overheating.

I suggest you get another SSR (make sure it has a very low voltage drop), or perhaps a "power expander" from reprap.me. Or you could download the firmware source and add hysteresis yourself.

---

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].

Connecting things the wrong way is not something i do, i design and build industrial control systems for a living and none of my clients have ever complained about things being connected in the wrong manner. The reason i used the relay is because i wasn't risking another £50 SSR and driving the bed from the board itself is not something i would ever do especially when your paying £100 for it a cheap Chinese knock off i might risk.

I decide to try the duet after all the good reviews about the board itself and in particular your version of the firmware, but all i can say is i have now had it 3 days and not got a print out of it yet. This considering i have another 3 deltas 2 with smoothie and one with marlin which never took this length of time to get going.

None of this is actually anything to do with you more that the documentation and information for the reprap firmware is spread far and wide and some of it seems to contradict each other.

I have just a couple of questions?

i had a direct drive extruder with MK7 drive gear set under smoothie at 190.9 steps/mm but under v1.09 i had to change it to 93.4 steps/mm to get the same output.

when using the web interface pausing then cancelling a print followed by homing the carriages results in only the z axis homing properly and so requires another G28.

as above but then restarting with another print the print is allowed to start without the hotend being at temperature.

BTW none of this is a slight at yourself in anyway shape or form i am just trying to understand this new firmware to me, i can see the work you have put in and reading the change files you have put a lot of time and effort in which is very much appreciated.

And i have already modded the firmware i just need to try it out over the weekend, tonight its a bottle of red

Quote
bobblejot
...
i had a direct drive extruder with MK7 drive gear set under smoothie at 190.9 steps/mm but under v1.09 i had to change it to 93.4 steps/mm to get the same output.

According to [reprap.org], MK7 drive gears are about 11mm diameter at the hobbed part. Assuming you are using a 1.8deg/step motor, that's 11 * pi mm per 200 full steps, which is 200/(11 * pi) full steps/mm, or (200 * 16)/(11 * pi) microsteps/mm which works out at about 93. So the value you found is correct. I've no idea why you needed double that with Smoothieware, unless you were using drivers with 1/32 microstepping or a 0.9deg/step motor.

Quote
bobblejot
when using the web interface pausing then cancelling a print followed by homing the carriages results in only the z axis homing properly and so requires another G28.

I suspect you have not set the homing distance in the homedelta.g file large enough to cover the maximum distance that a carriage may need to be moved to home it. See [reprap.org].

Quote
bobblejot
as above but then restarting with another print the print is allowed to start without the hotend being at temperature.

My guess is that you didn't have a T0 command in the start gcode of your print. So no tool was selected and there was nothing to heat (and no extruder to drive). There was probably an error message on the Gcode Console tab of the web interface.

---

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 was using 1/16 and 1.8° steppers under smoothie as well which is why i found it strange.

my z height is 342mm and my homing distance is 380mm in homedelta.g.

i haven't tried my own gcode files yet was just trying out the sample files from the sd image.

Quote
bobblejot
my z height is 342mm and my homing distance is 380mm in homedelta.g.

When the nozzle is at the edge of the bed opposite a tower, how far does that carriage on that tower have to move up to home? I think you will find it is more than 380mm.

---

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].