Heater bed noise suppression

I just had a sudden thought as I posted another post.

A quick and easy thing to try is to use a crocodile clip on a length of flexible wire from the aluminium heater bed plate or the mounting nuts/bolts to the 0V from the PSU, I suggest picked up from one of the unused PSU leads. This will provide a screen and route for radiated noise back to the PSU.

---

Ormerod #007 (shaken but not stirred!)

Is there any evidence that noise radiated by the bed is a problem for the operation of the Ormerod?

---

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].

Quote
dc42
Is there any evidence that noise radiated by the bed is a problem for the operation of the Ormerod?

There may have been, prior to the slowing down of the FET switch turn off (which I haven't incorporated), but it is such a simple thing to try, I thought I would post it.

@Forum readers, I'm interested in how many have modified the FET driver to remove spikes? Post a reply YES or NO

I have been using my Ormerod with the supplied (correctly working) DUET without any problems related to this (as far as I'm aware, but no hanging in printing or resets), but only small prints so far, so if I wasn't aware of this forum I would not have been aware of any issues.

---

Ormerod #007 (shaken but not stirred!)

@Treth

Well I am still using the Duet with no modifications at all, only change on it is a firmware change (thanks dc) to use a different pin for the blue wire from the Z Sensor due to a short to ground.
Not yet used the Ethernet connection yet due to being lazy and not getting round to make a cable for it.
I too haven't yet got around to do a really large print as yet, the largest print that I have done so far was the coat hook

Paul

---

RS Ormerod No 436

Hi Treth,
YES I have modified the FET driver. The basic design overlooks the inductive nature of the load.
Unless you deal with this by slugging the FET, clamping its o/p with a schottkey diode and decoupling the 12V psu input properly
you will cause pwer supply transients and RF noise that will seriously confuse the downstream 5V & 3.3V regulators.
This in turn causes resets, software crashes and unexplaned funnies.

Been there, done it, wasted a week of my life on it

One day I may extrude plastic!

Dennis

Quote
Treth
@Forum readers, I'm interested in how many have modified the FET driver to remove spikes? Post a reply YES or NO

Yes, for both bed and extruder loads. Before I was getting lock-ups during something like 2 out of 3 prints. After, I have had 2 lock-ups in I guess around 30 prints. These may be due to something else like mains-borne transients or ground loop current. It has become harder to diagnose as the frequency of occurence has fallen.

---

RS Components Reprap Ormerod No. 481

Hi Treth

I will be writing in a lot more detail soon. I haste now: in my setup switching on the heated bed does not significantly affect the 3.3 rail and I have been unable to recreate a hang that I can attribute to the power supply (only hangs were firmware issues - now mostly resolved by updates done by Dc42 and Adrian). The 3.3V supplied is further regulated down to 1.62-1.95V by an on chip regulator for the core of the SAM. I have also done quite a bit of testing with modified firmware which uses PWM on the bed - this is noisey at the standard 1KHz and not currently recommended but in doing so the Duets (v0.4 and 0.6) in question have been subjected to many millions spikes without causing a processor hang.

I am not discounting the issues that Radian experienced but I think it is possibly not as clear cut as his fix suggested.

More to follow as soon as I have a free moment to get it all down coherently.

Cheers

---

DuetWifi.: advanced 3d printing electronics

Quote
T3P3
Hi Treth
... I have also done quite a bit of testing with modified firmware which uses PWM on the bed - this is noisey at the standard 1KHz and not currently recommended but in doing so the Duets (v0.4 and 0.6) in question have been subjected to many millions spikes without causing a processor hang.

Cheers

Thanks Tony, that is all of interest, but on your last point on PWM and the large number of spikes, do you recommend the slowing down of the power FET to reduce the spikes?

Do you actually recommend or use any additional spike suppression or do you consider the spikes to be within the safe operation of the FET?

In other threads it is combination the small voltage across the LVD linear regulator, the possible drop in Vin in combination with the heater bed turn off spikes.

Reducing the spikes by slowing the turn off of the power FET has been shown on several systems to reduce 'hang ups', this in combination with ferrite cores on the USB lead and the mains input lead have also been seen to help, so there does appear to be some degree of susceptibility. We would be interested in your view of these points.

---

Ormerod #007 (shaken but not stirred!)

Another thought re the bed heater circuit. It looks very similar to a boost converter. If you used PWM and(EDIT) added the diode, and used a resistor and capacitor as the load, there should be a smooth flow of the large current through the inductance of the heater..

regards
Andy

Edited 1 time(s). Last edit at 01/15/2014 06:39AM by kwikius.

---

Ormerod #318
www.zoomworks.org - Free and Open Source Stuff

Without suppression the flyback voltage peaks at around 50V with very high dV/dT. The combined surface area of the Heater Bed and ribbon cable carrying this spike are quite substantial providing plenty of scope for capacitive coupling. It just kind of gives me the creeps. WWND? I'm pretty sure they'd suppress it to prevent Feynman coming back (from Heaven) to pick holes in their designs.

---

RS Components Reprap Ormerod No. 481

@Radian, that's why I was suggesting grounding/earthing the aluminium plate (opening thread) as it seems to be easy to do and would provide a capacitive shield to the spikes in this area.
My scope currently not working so can't investigate. I don't want to make things worse!!!

---

Ormerod #007 (shaken but not stirred!)

Quote
Treth
@Radian, that's why I was suggesting grounding/earthing the aluminium plate (opening thread) as it seems to be easy to do and would provide a capacitive shield to the spikes in this area.
My scope currently not working so can't investigate. I don't want to make things worse!!!

Sure, I had your suggestion in mind - only there's a bit of distance between me and my Ormorod at the moment so I'm just spinning my wheels (and ideas) until I can try stuff out. If only it was as pocketable as a smartphone

---

RS Components Reprap Ormerod No. 481