Prints too small

I have finally got my printer to the point where it will actually do something since fitting it out with the ramps electronics.
To my surprise,the 20mm square hollow cube came out at around 16mm.
It was actually less than 12mm high,because the nozzle knocked it off the bed
I know this was caused by inconsistent levelling.
I can actually see the whole x carriage moving in a slight arc when it lifts between layers,it's obviously rather heavy,and a bit unbalanced.
But it's the actual size of the print that bothers me.It definitely showed 100% in Sli3er,so don't know where the fault lies.
Any thoughts,anyone?

ps.Now that I think about it,the printer did this previously with a 5.5mm spanner I was trying to print.
That came out at 3.2mm,so was even worse.

Just had a rummage through wiki,and it seems this may still be connected to those damned end stops.
Presumably,if the min end stops are wrong,the max ones will be out too,since they are controlled by the software only.
Either way,it's likely to be a calibration issue,so back to the drawing board.
I tried another print,and now the (mechanical) x and y stops seem to be MIA again!Z is working fine.
Some deep thought is called for.That's gonna hurt

Edited 1 time(s). Last edit at 12/15/2014 02:24AM by clogs51.

If the prints are consistently smaller then it sounds like its more of a steps per mm setting. This is in your config.h file

For my inital setup I don't print anything but use the software to move each axis and a ruler to measure the distance actually moved then change the steps until they are close to being correct. I then print a calibration cube and use verniers to fine tune the steps.

I used the formula move distance in mm / actual distance in mm * steps per mm for each axis. This gives the new steps / mm to achieve the correct length. I suggest using a large distance of 100mm and the EEPROM for tweaking as it is quicker before finalising it in the configuration.h.

For example, using your 5.5 print resulting in a 3.2 spanner with say 1234.5 steps per mm on the z axis
5.5 / 3.2 = 1.71875
1.71875 * 1234.5 = 2121.8 new steps / mm. Eg you need alot more steps to achieve the required z length.
You need to do this for each axis and assumes your motors are not skipping steps.

Edited 2 time(s). Last edit at 12/15/2014 04:27PM by AgentX.

Thanks for that explanation,Mart,I'll have to re-calibrate the whole machine I think.

Re bed levelling,I found the springs supplied were simply too weak and short as well.
This leads to the thumbscrews loosening with any movement.I had already replaced one corner screw with a 3x30mm hex screw,
and added a nyloc nut on the bottom.But the other 3 screws still needed constant fiddling.
I went to Bunnings,and bought some 1/4" springs,which are quite firm.
When i fit these,this should put enough pressure on the adjusting screws to keep them in place.
At least,that's the theory.If that doesn't work,I'll put some thin fishing line in the threads.
That'll stop movement,but may make adjustment harder.

I nixed the heavier springs,they make the bed bend down in the corners.The acrylic y sliding frame is the culprit there,methinks.
Getting one cut out of aluminium could be a mite expensive though.
I played around with the steps,and from an initial setting of 55.5,I now have them up to 88.8.
This seems to give the correct 200mm travel in x and y axes.
I'm back to that horrible grating noise when I home all axes though.
I looked hard at the x drive pulley,and it's not actually moving when it gets to the endstop,just sits there graunching away for 3-4 seconds.
The Y axis goes even longer.
To add insult to injury,it won't print again now.Rats!I'll probably have to reset the z-endstop in config.h again,because m119 now shows
both z min and max endstop triggered.That can't be right.
My extruder had the longer 44(?)mm motor,which meant i had to devise an adapter to fit the hotend.
This resulted in a rather wobbly extruder,so i have ordered a 34mm long motor from makershop.
This should allow me to fit the motor in the standard position,and get rid of the adapter.
So it should reduce the overhang by about 20mm.My hotend also wobbles slightly in it's socket,so some fiddling is required there as well.
But one thing at a time.

Quote
clogs51
I'm back to that horrible grating noise when I home all axes though.
I looked hard at the x drive pulley,and it's not actually moving when it gets to the endstop,just sits there graunching away for 3-4 seconds.
The Y axis goes even longer.
To add insult to injury,it won't print again now.Rats!I'll probably have to reset the z-endstop in config.h again,because m119 now shows
both z min and max endstop triggered.That can't be right.
My extruder had the longer 44(?)mm motor,which meant i had to devise an adapter to fit the hotend.
This resulted in a rather wobbly extruder,so i have ordered a 34mm long motor from makershop.

Check your axis and end stop sensors match, that the plugs are where they are supposed to be.

Your extruder probably had a longer motor to give extra torque to push the filament with sufficient power. Check the voltage of your new motor, I'd be looking at something rated at 6 V or under. I use a UD$45 digital clamp on meter from dx.com that can measure AC and DC current (zero it on DC) to check the motor current, I have mine set to around 250mA when not moving but powered. You could fry your stepper drivers if you use too low a voltage motor without adjusting the voltage down first on the driver.