Posts made by DjDemonD

Just to say the corexy has now been dismantled and the parts are available separately:
https://www.ebay.co.uk/sch/therealdjdemond/m.html

Okay so I'm going travelling and won't be doing any 3D printing for a while, there's not much point in storing these machines as they might be fairly out of date by the time I get back.

I've listed them on ebay so here are the links:
https://www.ebay.co.uk/itm/202386854013
https://www.ebay.co.uk/itm/202386853581

Both Duetwifi powered, the kossel XL has a lot of decent parts on it.
Working capable, machines but a great source of parts as well if you were looking to build something yourself.

I'm in the UK, you would need to collect from Bristol.

I have to agree with David. H300 will instruct the machine to try to begin 300mm above the probing point, which is unnecessary, might be impossible given the z height available and will be time consuming. Sure setting it to 50 (or 20) might make sense if you've just changed something which affects the max Z height, such as slid in a new bed of differing thickness or changed delta arms etc... and this will ensure head crashes are less likely. Also probing from 300mm above relies on your machine to be extremely accurate and repeatable over that 300mm.

Also consider changing G31 P1000 to P100 this is for digital probes a debounce interval and it need not be as long as that.
Many are getting better results with M558 P8 which removes the debounce filtering altogether.

I have found using the latest builds, the multitouch probing options offer a greater sense of accuracy, in that it will throw an error and abandon calibration if multiple probes of each point are not consistent. This prevents a problem such as a loose probe or wobbly bed, generating a nonsense calibration and the resulting craziness that can ensue.

M558 P5 I1 R0.4 F1000 X0 Y0 Z0 H7 A5 S0.02 B1

F1000 - 1000 mm/min travel speed between probing points
H7 - dive height 7mm, enough if things basically stay the same between probing sessions
A5 - probe each point 5 times at most
S0.02 - Tolerance so if probe 2 is within 0.02 of probe 1 then the firmware moves to the next probing point, otherwise it tries again up to 5 times (A5)
B1 - turn off bed heater when probing. Seems to make a difference for some - I haven't made my mind up yet.

On the reprap forum you said that removing the deploy and retract files made the problem worse, this is very illogical since these files are only needed when a probe needs to be deployed or stowed, one feature of piezo probes is that they are permanently mounted and do not require any deployment ot stowing. Did this change help or not?

When deploy/retract was first introduced into RRF my printer did some odd moves, until I deleted these files. Then it probed normally. Post your M558 and G31 as you have them now please.

Hi have you read this?

https://forum.duet3d.com/topic/2671/precision-piezo-z-probes-guide-for-duet-users-reference/99

Just double check the wiring into the duet, as the Orion and latest products come with standard endstop wiring, i.e. +,-,sig so as long as these connections on the duet z probe connector are correct the config should be fine.

Remember its an active low switch so it triggers when the voltage drops, use switch I1 in M558, other than that just set it up as per the guide above ( see the 1st post there).

HiCraig, please see www.precisionpiezo.co.uk or contact us via https://www.precisionpiezo.co.uk/contact

My colleague Idris Nowell now runs the company and can provide support.

That flat cable looks great.

Network cables are problematic they do work harden and the insulation fails, probably in my case exacerbated by having slightly too tight a bend at one point in the cable, but a short from hotend heater (was using 6 cores for heater) and 5v fan kills 2 duetwifi and 2 paneldue boards. Second one shouldn't have gone, but I presumed the first one was a dropped nut which caused the short and installed a second board. That'll learn me.

I can answer some of this as I have upped the voltage on my CR10 to 13.8v.

I'm not sure I can answer your questions exactly as you've put them so,

Your heated bed has a fixed resistance, so a voltage increase will increase its wattage (output), and the gains are fairly impressive.

If you take a 12v 1ohm bed - 144w output
13.8v and 1 ohm bed - 190watts output

and I can confirm on my cr10 before doing this I could not get past 90 deg C on the bed, now I can hit 110.

Peak power demand - when heating up the machine before printing will be 2-3 times greater than the demand maintaining temperature. So you need to cover bed at max, hotend at max and all motors running at maximum speed, when determining what wattage PSU to buy, fans and LEDs can be factored in but unless they are massive don't use much power.

Adequate cooling (extra never harms the electronics - might offend your ears though), is a good idea.

The other issue is that the voltage regulators on your electronics will have to work a bit harder if your PSU is outputting more than 12v. On duet with many machines running on 24v and no issues there - it will make no difference if you go from 12v to 15v. On lesser controllers like arduino/ramps the 5v reg was notorious for failing.

EDIT- what I don't know is whether the total wattage of your PSU will increase or remain the same.

Hi yes, so its all on our website www.precisionpiezo.co.uk and on our reprap thread https://reprap.org/forum/read.php?1,767998 it a fully open source project so its not a secret.

If you connect a piezo disc to an arduino and ask it to monitor the voltage on an analogue pin, you will see that they are incredibly sensitive, any slight vibration or certainly any deformation of the disc will generate a large voltage. Hit one with a hammer and you'll see 50v! (don't do this with your 5v arduino analogue pin please). This is how electronic lighters work, by striking a piezo hard and generating an arc.

In the hotend sensor systems, the piezo disc is deformed when the nozzle contacts the bed. It only needs to be around 0.1m (which is repeatable so you can set this as a z offset) to give a signal, and can be very gentle when correctly tuned.

There is no electrical contact between the Precision Piezo sensor and the hotend, this would be very undesirable as a short or wiring failure would create damage to Piezo PCB or 3d printer controller board.

Autotune won't affect heat up times.

You can check your mosfet board, just connect the bed directly to the PSU, start a timer, and measure the temp, don't leave it unattended when it hits whatever temp your aiming for disconnect it. It should heat up just as fast with the mosfet.

WHat bed (amps or watts) and PSU (24v yeah? what amps or watts) are you using?

I take your point @pro3d that if you want a simple controller, then marlin running on one of the plethora of cheap controllers that exist now, is probably a simpler and cheaper solution. I have one machine still running on arduino/ramps and marlin (3 year old version at that) because it just works. I agree to some extent that if I was an OEM, it might be easier to just send a hex file to customers when I updated the firmware on their machines. It seems to work for Creality etc.

Duet is a product aimed at those who want to utilise a broad array of very useful features but not everyone needs them all (or even most of them), if the machine they are operating is fairly simple. I have a cr10 and wouldn't bother with a duet upgrade as it works out of the box. However my delta is complex and my corexy a platform for developing sensors, so the feature rich landscape with duet works there. I can't see an easy way to have a simple mode and an advanced mode, but maybe this is the way to go.

We are not talking about any of the BED power wires or wires to/from the PSU, we are talking about where to connect the signal wire, the answer is to the duet's heated bed output. My understanding is that polarity is usually important but not harmful if wrong, so try it whichever way around you wish, if it works you're good.

See the diagram post Vaei 5 Jul 2018, 12:51 above. The CTRL in goes to the duet bed heater + and - which are the only two connections in that diagram that Vaei was unsure about it.

I have the same setup (2 PSUs and a small mosfet the same as shown above). If you want me take a photo of it when I get home this afternoon I will do.

Okay so the heated bed normally has a + and - output. You need to connect them to the + and - of the CTRL-in on the mosfet board (sometimes labelled bed + & - or signal + & -) otherwise, the mosfet board cannot tell when you've turned the bed on. It's electrically isolated from the second power supply, so for the control side of the mosfet board to be able to tell if there is a signal present it needs a circuit of some sort, connecting one wire isn't going to create a voltage it can detect and then switch on the optocoupler which in turn opens the large mosfet to power the bed up.

What is your reservation about doing this? If it's that it seems wrong to connect the bed output to a signal input, remember a heater only draws the power its resistance allows it to. So if you connect a bed heater to the bed heater output and its 12v and has a resistance of 0.72 ohms for example, then it will draw 16.66amps. If you connect a small transistor and a few resistors (the low current side of the mosfet board) then it will draw a few milliamps.

Those of us using mains heated beds are using SSR's which is just a version of a mosfet board than can handle AC mains 240v on one side, but accepts 2 tiny wires from the duet heated bed output as the control signal.

Its a shame someone can't help you iron out the issues, my corexy when I swapped to duet just worked better than ever. True it was my second board and I had a handle on it at that stage, but if we look at your points in detail:

The extrusion is uneven and inconsistent
What microstepping are you using and what extruder? What motor current (is it 75%-80% of the motor's rated current?)

The BL-Touch has been a hassle and still is non consistent in its behavior
That's BLtouch for you, but I have managed to run one for testing purposes from duetwifi without any apparent problems.

Layer adhesion problems
That has nothing to do with the firmware, unless you are under-extruding or the nozzle isn't staying the correct height above the bed - probably related to bltouch issue.

Problems with using my 12V fans (as I do need for the blower!)
I presume you are running the duet on 24v, fans and different voltages are a challenge and the range of 24v fans is much more limited. I was using 5v fans until it killed 2 duets and 2 paneldues from a wiring short - won't make that mistake again. Get a few buck converter modules, there are detailed instructions to use them to get 12v PWM's fans from 24v supply.

The layerlines are very prominent
Got me on that one... could be a million reasons.

PID tuning problem - Warning??
What warning?

Thermistor issues - Beta value?? Do we really need to calibrate this as well?
Well no, I'd dump thermistors and go to PT1000s absolutely the easiest way to measure temperatures, just plug them in, add the correct line in config.g and you get known, predictable, (near) linear response to temperature changes.
I have calibrated thermistors using k-type thermocouples and measured resistance at known temperatures then used online calculators to get the beta value - but PT1000's are the way forward.

G29 S2/M561 Not clearing mesh? What is the difference and why do we have two options in the GUI??

I run this macro to do a mesh:
M561 ;disable compensation
G1 X135 Y97 Z5 F2000 ;move to bed centre
G30 ;set Z=0 level
G1 Z5 ;lift nozzle
G29 ;grid compensation
G1 X140 Y100 Z50 F2000 ;centre head

Then load it in config.g with
G29 S1.

I suspect any mesh issue you have relates to the BLtouch - have you considered a slightly more accurate probe?

If I had more time on my hands and you were not too far from me I'd come and help you iron it out and take a look at your printer which looks very interesting.

That would be good. I had my pump and fan stop working yesterday and presumed I might have overloaded the P1 fan output, but as usual just a loose connection.

I like the belts and motors on the outside, I was thinking about reframing my acrylic tube leadscrew delta (acrylic tube was a nice gimmick but isn't holding up too well), to have the screws on the outside of the towers, as you have your belts. Looks good.

Is there a way to remap a fan to a heater output, so if I want to use E1 heater connected to a fan/water pump rather than the P1 fan output? I would want to be able to thermostatically control it as I currently do with the P1 fan output. I am just concerned my fan and water pump combo might be a little too much for the fan output.

I could implement an external mosfet on P1 fan, but might be nice to know in case I need to do it later?

Everything there looks good, except you feed the CTRL in from the Duet Bed heater output - the big screw terminals. It doesn't matter that its massively oversized, the CTRL in will only draw a few milliamps from the duet bed heater. Make sure the + and - are correct, I don't believe you will do any damage if they are wrong but the system either won't work or will behave oddly. Usually the mosfet board will have an LED on it, which lights up when it is active, so you might see it flickering when using PWM control thats normal. If using bang-bang it will be either on or off. PWM is better but then for beds with reasonable thermal mass bang-bang will maintain a set temp without any problems.

Yep those are the things. If you have a 1 PSU system and your bed is below the 18A suggested limit, then they are probably an extra bit of kit you don't need. But a 2 PSU system is easily setup using one of these. Plus you get the highest current part of the system off the duet board - not a bad idea really.

Those smaller ones will work for most application they do make an even larger one

But I did buy one of these, they are massive could probably handle 50A+

I'd like to throw in a second way, just in case the method outlined above doesn't suit you.

Get an external mosfet board (they are £5) . Optoisolated, they run from your bed heater output on the duet (12v is fine for this "signal") when the bed comes on the 12v signal switches the mosfet on the external board. Your 24v PSU is connected only to the external board, and through it the bed. I have a machine in this configuration.

Advantages:
Complete electrical isolation of 24v and 12v supplies.
No high current switching on your duet (I know compared to lesser boards they can do it but I like not switching so many amps on my most expensive component)

Disadvantages
None that I can think of (except spending £5).