Author Topic: picclock's modified i3 3d printer attempt  (Read 31564 times)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #25 on: February 25, 2016, 03:43:20 PM »
Looking good! Will be interesting to hear your experiences of the build. You have a number of neat solutions of which some will be on my list for improvements.

I hava also been thinking a bit about the extruder, I have the Wades version a bit like yours, but not completely. With the stepper it is a quite heavy bit to be carried by two 8mm bars. Moving the feed to the portal and having a PTFE tube leading down to the hot end would relieve the X carriage of some weight.

BR

/Peter
Always at the edge of my abilities, too often beyond ;-)

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #26 on: February 26, 2016, 04:35:12 AM »
Hi PeterE

>>With the stepper it is a quite heavy bit to be carried by two 8mm bars. Moving the feed to the portal and having a PTFE tube leading down to the hot end would relieve the X carriage of some weight.

Thats my opinion also. One of the issues with Bowden feeds is that the longer the feed and the more that it bends the greater the likelihood of the filament buckling/jamming. I have in mind to limit the range of bowden movement by supporting the extruder drive above the centre of the X carriage whilst connecting it to the Z bar supports. This would make it a constant height above the head enabling a much shorter bowden length.

Re the weight on all the axis. The heavier and stiffer it is to move, the lower the accuracy will be. That is why I am driving the motors closer to the rating. If you think of the motor as a spring, the further it deviates from its intended position the larger the force needed to move it. But this means that for small deviations only a low force is needed. By increasing the motor current the spring is made stronger and better accuracy/repeatability more achievable. Obviously, this only applies within any given motor true step, typically 1/200th of a revolution.

I will also incorporate a servo mount for auto bed levelling on the upgraded extruder, so at best what I have is very much a work in progress. One of the things not shown in the photographs is the X bearings. The top one is twice as long as the std prusa ones, and the lower single std one is allowed to float about the z axis, thus all things are true to the top bar.

Looking forward to seeing your printer up and running.

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #27 on: March 09, 2016, 04:12:18 AM »
Hi All
Things have moved on a bit since last post.

The bed wiring is now complete, with it passing through the belt attachment for the glass bed. When fully operational one of the first things I will print is cable chains for the X and Y axis. The cork is attached to the heater pcb with a high temp silicon adhesive, and this in turn is attached to the glass with clips as shown.

The Z limit switch is sorted, and the Y axis switch fitted. The Z limit switch is one I reused from a scrapped photocopier. Although it is push to make, (default is push to break so that wiring integrity is maintained) the accuracy and repeatability are better than I can measure. I put one of them in my mill vice, connected to a beeping continuity meter and used the Z digital readout to detect the point of switch actuation. The reading at the switch point was exactly the same, so its easily as good as 0.5 thou of travel. Picture with simple mounting and adjuster below.

The distance between the print head and bed is an important setting to ensure the correct adhesion of the first level of plastic. Having powered it up, I was able to set this to 0.05mm, by adjusting the screw and re homing the axis. Once the left side was complete I manually adjusted the right side to the same gap. I intend to automate the bed levelling process by using a servo and a limit switch, parts on order. 

Next, the filament feed calibration. This is achieved by simply marking a piece of filament and telling the extruder to extrude a given distance. For this I removed the extrusion nozzle and set the minimum extrusion temp to 1C in the Marlin firmware. I had an extrusion rate figure from the internet, taking account of the different gear ratios, and in an unusual stroke of good fortune this proved to be exactly correct. I extruded 100mm, and exactly 100mm was extruded :thumbup:

So for a trial print. I thought I would try the blue tape adhesion method first with some hair spray to help things along. I loaded up a new reel of blue ABS filament, sliced the design for a cable chain link in cura, and dropped the file into the octoprint browser window. Had a quick think to see if I had missed anything, then clicked the print button. The bed started to warm to its final temperature (100C), once reached the print head quickly warmed up to its operational temp (230C), and the process commenced. I was very impressed with the fine detail of the outline and its accuracy. All seemed to be going well, with the first layer going down but after 4 or 5 layers the parts started to warp and come off of the bed, at which point I cancelled the print.

I thought I would try to improve the bed adhesion using abs juice so attempted to make some from the scrap printed parts and filament dissolved in acetone. I left this for several hours but was surprised to find that it had not dissolved, in fact was not even sticky. I took a piece of known ABS and dunked it and it went sticky almost immediately.  :scratch: I was thinking that I had been supplied with the wrong filament type, so I checked out a roll of conductive abs filament, which I had ideas about using for plated items. In doing so I came across the original carton which clearly showed that my blue abs was in fact nylon, and further checking of the order indicated this was exactly as ordered. Its a terrible thing getting old .. . :loco:

I have ordered some other colours of abs and hope to show some first printed parts soon. I may have a go using the conductive abs, but this is far from typical due to the carbon loading, so will not be too good for calibration. The white roll of PLA I have may be useful for testing or prototyping but not very good for calibration or working parts.

I also think that it would be a good idea to fabricate an enclosure to maintain an even, elevated temperature throughout the printing process, as this would minimise warping.

Thanks for your interest

Best Regards

picclock



Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #28 on: March 09, 2016, 05:11:32 PM »
Looking really good picclock  :thumbup: :thumbup: :thumbup:

I was given the tip to use a dab of acetone at each corner of a beginning print to make it stick better to the build plate. That should be enough I believe. I will use one of those cottonbulbs on sticks (whatever they are called in english) to get that necessary small amount. Of course that means that I will be using ABS most of the time anyway.

The daughter is now nicely installed in her new apartment so time for shop activities is in sight  :drool:

/Peter
Always at the edge of my abilities, too often beyond ;-)

Offline awemawson

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 8966
  • Country: gb
  • East Sussex, UK
Re: picclock's modified i3 3d printer attempt
« Reply #29 on: March 09, 2016, 05:22:41 PM »
Cotton-bud  :thumbup:
Andrew Mawson
East Sussex

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #30 on: March 10, 2016, 03:22:51 PM »
Aha, thank you Andrew, good to know. In Sweden they normally go under their brand name which is Tops.
Always at the edge of my abilities, too often beyond ;-)

Offline tom osselton

  • Hero Member
  • *****
  • Posts: 1255
  • Country: ca
Re: picclock's modified i3 3d printer attempt
« Reply #31 on: March 10, 2016, 04:13:52 PM »
Q tips here.

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #32 on: March 11, 2016, 03:27:28 AM »
Hi All

Thanks for your kind comments PeterE. How would you know where to put the dabs of acetone to hold it down? With a heated bed I would imagine it would evaporate very quickly, and when printing multiple items would be very difficult. Supposedly, with abs juice (a mixture of acetone and abs) you can just wipe the bed over and the job is done.

Picture below of my first print (PLA). On the whole, not too bad. Its the right shape and reasonably concentric. Surface finish is very good, better than I expected. Its all slightly undersized, likely due to shrinkage.

Measuring the 1st step up from the bottom i get 19.75 at the narrowest point and 19.90 at the widest, so about 0.15mm out of round, and 0.18mm too small. Part of this may be because the slicers define a circle as a multisided polygon, and the number of sides will affect the circles concentricity. The steps narrow slightly in the middle before coming out again at the top. The step heights are pretty good, at 5mm +-.05, which is impressive for two bits of M5 studding.

The top pillar is nbg, as the plastic did not have time to harden before the next layer was added. The slicing was done with Cura, with its recommended temp setting of 210 for PLA. Other users recon from 160 upwards so it may be too high. Also Cura allows the setting of a pause time between layers for the plastic to harden, so I may well experiment with that.   

Unfortunately, PLA is just not suitable for the things I need to make, mainly because of its low temperature softening, so I will have to wait until the abs arrives.

Thanks for your interest.

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #33 on: March 11, 2016, 04:48:50 PM »
The tip I was given on putting a dab of acetone to "glue" the print to the bed was to do it when the "skirt" is laid and then dab in "four corners" of course regardless if the object is round or not. It would evaporate quickly, but I don't think quickly enough to not glue the part to the bed plate. My own experiments will tell.

When it comes to temperature I have underrstood it so that the range given by the maker of the filament is the appropriate range, and I would start at the lower end to begin with. This to get a good bond between layers but still easy enough to spread. I guess this is one of the things that even may be different from machine to machine as the machine placement could affect the surrounding temp.

Making a box around it would make that easier to control, and you would get a chance to make an "exhaust" to guide the print fumes out of the room to keep it healthier.

BR

/Peter
Always at the edge of my abilities, too often beyond ;-)

Offline Pete49

  • Sr. Member
  • ****
  • Posts: 353
  • Country: au
Re: picclock's modified i3 3d printer attempt
« Reply #34 on: March 11, 2016, 10:04:32 PM »
I have found the best glue to use is cheap hairspray :lol: It sticks ABS and PLA well and when it cools the items lift off. I put on 3-4 coats then refresh it every few prints. This is on a heated glass bed though.
Pete
oops..........oh no.........blast now I need to redo it

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #35 on: March 12, 2016, 02:36:53 AM »
Hi PeterE, Pete49. Thanks for your thoughts. I will certainly try those ideas next time I am printing.

Because my printer is in the garage, and I am in the office, I need a way to power up and program the printer remotely. I am using octoprint with a raspberry pi, and this can be modded to allow other control outputs, for remote power on/off and camera on/off. Then I will be able to print stuff remotely, check how its doing, and if necessary, power off the printer. To this end I have designed a very small pcb to attach to the pi to drive a solid state relay and camera on/off with a spare output for stuff I have forgotten. I will start a thread in the electronics section for this.

@ PeterE

Hope you manage to get your printer sorted soon.  Looking forward to seeing it in operation. Mine just looks like a pile of bits ATM because I am waiting for the abs to print the parts which will finish it off. I have just received a servo for bed levelling/probing, and I will incorporate this into the extruder assembly design when I print a new one. It does seem the best way to go.

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #36 on: March 12, 2016, 05:57:40 PM »
Hi picclock,

My printer will get its attention from tomorrow when all reamining small bits of supporting work is done to get my daughter happy in her new apartment.

/Peter
Always at the edge of my abilities, too often beyond ;-)

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #37 on: March 15, 2016, 03:14:51 PM »
 Hi. So I have designed and completed my little RPi addon pcb - see http://madmodder.net/index.php/topic,11358.0.html. Octopi now sports two extra on off options on the system menu to control the power for the printer and webcam. So all appeared to be going well. The abs filament has arrived and I made some abs juice to try out as per web instructions. Fired up the printer with the intent to print some cable chains to try to sort out the mess of wires. Dropped the sliced Gcode file onto octopi and hit print. Then about 5 mins later as the bed was warming up there was a loud crack and the mirror bed had self destructed. :scratch: From the pattern of the break it looks like it may have been caused by differential expansion of the glass - it occurred at around 85C - see picture. Was not expecting that. Glass was 2mm, which may be a bit thin.  No other mechanical stresses on the glass. Looking at the odd fracture pattern it could be that the centre warmed and expanded more quickly that the outer edge.

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #38 on: March 15, 2016, 05:33:40 PM »
picclock,

Ahh, too bad when such things happen ...

The glass for my bed is a 200mm square piece of borosilicate glass around 3mm thick, so thicker than yours but can also handle higher temperatures. Borosilicate glass is the same type used in some oven-proof bowls and dishes for pies etc. It may be so that the mirror glass is basically ordinary glass with less temperature range and together with the variation of temperature it could crack easier. This is just a guess though.

Another thing is that I hold the glass plate to the heatbed with just four clips, one in each corner, that will also slightly lessen the clamping forces if the bed warps or twists during heating.

My 2 cents ...

Good work so far, keep the good tempo up!  :thumbup:

/Peter
Always at the edge of my abilities, too often beyond ;-)

Offline Will_D

  • Hero Member
  • *****
  • Posts: 668
  • Country: ie
    • National Homebrew Club of Ireland
Re: picclock's modified i3 3d printer attempt
« Reply #39 on: March 15, 2016, 07:07:52 PM »
I have found the best glue to use is cheap hairspray :lol:

IIRC the largest/bigest aerosol can I ever did see was for a "cheap" Hairspray

Also good as a "non insecticida/non chemical" fly spray.
Engineer and Chemist to the NHC.ie
http://www.nationalhomebrewclub.ie/forum/

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #40 on: March 16, 2016, 02:48:48 AM »
@PeterE

Thanks for your interest. Hope your daughter is now settled and work can progress on your printer.

The glass that cracked on mine was only held by the 4 clips at the sides. The end clips hold the bed heater to the glass. The fittings that touch the glass underside are all fibreglass to act as a heat break. Borosilicate is without doubt an improvement, but have also read about that breaking, not sure what the mechanism for that could be. Several things spring to mind. It may be that adding an insulating layer to the top at warmup would cause the heat to be more evenly distributed, the layer could be removed prior to start of printing. An aluminium heat spreader under the glass could improve heat transfer. Alternately a slower warmup period would allow the heat to spread more evenly. Thicker glass would also help by increasing the heat conduction path to the outer edges. If I can get hold of some borosilicate glass I have no doubt that would also improve things, but I doubt it will be easily available.

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #41 on: March 16, 2016, 01:06:27 PM »
Yes, now the lass is in place in her first own apartment  :ddb:

I got a glass sheet similar to this http://www.ebay.com/itm/Geeetech-Borosilicate-Glass-for-Heatbed-MK2-MK2A-of-3D-Printer-Reprap-Mendel-/221661533512?hash=item339c0e9548:g:srgAAOSw-W5UtPEE and at about the same price but from a local seller in Sweden. It is 3mm thick which increases its strenght quite a bit and it evens out the temp spread as well.

Just have to get my foot in order after spraining the ankle carrying furniture bits down some stairs  :palm: Work will proceed during the coming weekend.

/Peter

Always at the edge of my abilities, too often beyond ;-)

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #42 on: March 21, 2016, 04:40:15 AM »
Hi. I have managed to print some usable bits but things have been difficult. I replaced the mirror glass with some 4mm glass liberated from a skip behind a window shop (with there blessing). I heated it to 115C and all seemed fine, but after the first print it had cracked again  :palm:. prior to the print I had been monitoring the temperatures with a thermocouple type thermometer. The nozzle temp was exactly correct - quite impressive really. However the bed temp is all over the place. The centre area, on the surface of the glass was about 95-98C measured, however an inch or two from the edge it had fallen to 55C or so. With that kind of temperature gradient over a couple of inches I am not surprised the internal stresses caused issues.  Also as the bed moves small distances cold air is pushed onto the edges but the centre always has warm air, exacerbating the thermal gradient.

So I have thought of several fixes which will hopefully improve the situation. Firstly, use an aluminium heat spreader directly above the heater below the glass (will try 1mm as I have it around). In the centre area of this I could put some kapton tape to further reduce the centre temperature and increase the heat to the outer edges. Secondly, I could fit an air dam around the bed, basically just a 25mm high strip of an insulating material, which will prevent cold air from reaching the outer areas so readily when the bed moves. Thirdly, a form of enclosure to prevent external drafts from having an effect. If I'm lucky this will bring things under control and resolve the issues.

Pictures of my first real parts follow. As you can see, the cable chain links were printed with a large raft to promote adhesion. The bed cracked part way through printing, though it was held together with the tape. Use of the tape likely increased the thermal differential even further, promoting the failure. The yellow stuff around the edge was from some attempts I made to print with ABS juice, which did not seem to work well. For printing, bed temp was set 110C, and nozzle temp 260C. Sliced with Cura, 0.1mm layer thickness. Layer adhesion was excellent, although it struggled with the bridge across the top of the links, as can be seen from the top right hand link in the 'chain links on raft' picture.

Hope you found this informative and of interest.

Onward and hopefully upward.

Best Regards

picclock 
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline RussellT

  • Hero Member
  • *****
  • Posts: 520
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #43 on: March 21, 2016, 05:32:34 AM »
Can you find an oven door window the right sort of size?

Russell
Common sense is unfortunately not as common as its name suggests.

Offline BillTodd

  • Hero Member
  • *****
  • Posts: 1096
  • Country: 00
  • Colchester Essex (where the lathes were made)
Re: picclock's modified i3 3d printer attempt
« Reply #44 on: March 21, 2016, 05:44:59 AM »
Does the bed need to be glass?

was glass suggested because it readily available and flat to a few thou?

the model is stuck to a layer of tape or lacquer so , would any flat surface work?

Tooling plate anyone?
Bill

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #45 on: March 21, 2016, 06:50:07 AM »
@ RusselT
Oven door glass would be good but I believe it is tempered. This occurs after it is cut to size and any further attempts to cut it will cause it to break.

@BillTodd
Thats the very question I have been asking myself. I have a sheet of 4mm Aluminium, T6 I think, which could easily be used as a bed, and made heated by the addition of some resisters. This would allow much higher power and placement of the resisters could be (skilfully ?  :palm:) made to even out the temperature differentials across the surface. However, from internet research it seems that this is not an approved solution, with various problems cited. I forsee a couple of issues but nothing major. One is that the top surface cannot be replaced, so once it is damaged its there forever. Another is that of surface warping/bending due to uneven heating. I think this could be largely eliminated by resister positioning and multipoint bed probing for auto levelling.

But the issue is that glass will happily withstand the bed temperatures providing they are equalised throughout the material, and this will need to be sorted for any solution to be effective. Will try a heat spreader - hopefully today and map the temperature differential.

Best Regards

picclock

Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline Sid_Vicious

  • Full Member
  • ***
  • Posts: 103
  • Country: no
Re: picclock's modified i3 3d printer attempt
« Reply #46 on: March 21, 2016, 12:43:34 PM »
Tempered glass can be cut with an anglegrinder if you are careful, so if you can pick some stove glass you could give it a try.
Nothing is impossible, it just take more time to figure out.

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #47 on: March 23, 2016, 07:35:55 AM »
Just thought I'd share my experiences with a very low cost webcam. Advertised from ebay :
http://www.ebay.co.uk/itm/Mini-USB-2-0-PC-Camera-HD-Webcam-Camera-Web-Cam-With-MIC-For-Computer-PC-Laptop-/311435955773?hash=item4883078e3d:g:ipYAAOSwd0BV5En1

With the original description original Mini-USB-2-0-PC-Camera-HD-Webcam-Camera-Web-Cam-With-MIC-For-Computer-PC-Laptop

So I did what most rational people would do, having checked its functionality, and not believing a word of the advert, I took it apart. Did a quick internet search for the pcb number, and came up with :

http://www.penguin.cz/~utx/hardware/USB_Camera_AX2311/#cheap_usb_web_camera_teardown_analysis_and_examples

from Stanislav Brabec. So my HD webcam is in fact a more sedate 640x480 1.2Mpixel device. However, I am not surprised, as this is pretty much exactly what I needed, and combined with the any position camera neck and suction base I think that at £3.66 it represents very good value for money, though I have since seen it cheaper. The only mod I made to the camera was to remove a plastic ledge from around the base of the lens to allow me to focus it more easily. I found this works in Octopi with camera_usb_options="./mjpg_streamer -r 640x480 -y -n" set in octopi.txt

You can see the quality in one of the pictures which show it in operation.

Best Regards

picclock

Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline picclock

  • Hero Member
  • *****
  • Posts: 613
  • Country: gb
Re: picclock's modified i3 3d printer attempt
« Reply #48 on: March 28, 2016, 08:26:54 AM »
A little more info. Earlier, i was having issues with bed glass cracking due to differential expansion. I measured a centre to edge glass temperature around 50C, hence the glass failure. I then fitted a 1mm aluminium heat spreader between the glass and pcb heater. This worked incredibly well, bringing the differential, centre to edge down to about 5C. After taping the edge the the difference over the bed is around 1 or two C, pretty perfect. However this has shown up weaknesses in the pcb heater, as it now reaches 100C/ with an ambient of 10C, which is a bit on the edge, but at least it allows abs prints to stick (with applications of hairspray, followed by ABS juice). I think it likely I will replace the pcb bed heater and spreader with a 2mm aluminium sheet with resisters attached, as this would allow double the heating power with much faster warmups. Because the sheet will be warmed evenly no warping should occur - well maybe.

So onto printing my first working parts - at least attempting to. In order to get parts flat I have found the use of a raft is essential. Picture shows raft pieces being broken out. However, if you look carefully at the topside picture you will see the top of the spur gear is missing. I thought this was a slicer area or software issue, but I later found the filament had caught in one of the notches on the supply reel, resulting in the filament being shredded in the extruder. Extruder maintenance now needed  :drool:

Best Regards

picclock
Engaged in the art of turning large pieces of useful material into ever smaller pieces of (s)crap. (Ferndown, Dorset)

Offline PeterE

  • Sr. Member
  • ****
  • Posts: 458
  • Country: se
  • Gothenburg
Re: picclock's modified i3 3d printer attempt
« Reply #49 on: March 28, 2016, 11:49:09 AM »
Hi picclock,

You are advancing very well indeed!  :thumbup: and solving the cracking glass plate matter as well - very good!

I notice that you are using a good skirt or raft which I also think is very useful when printing. All the parts I got to build my machine had the same type of raft. simple to clean out and it gives flat prints very well. Bugger about the filament catching the spool and shredding just at the final bit.

You are well on your way now.

/Peter
Always at the edge of my abilities, too often beyond ;-)