routercnc_MK3

[routercnc]

Quick update.

Got the main panel on the wall and wired up all the stepper motors. I have changed over from series to parallel connections (on the X1,X2,&Y motors) to take advantage of the higher current available from the new drivers. I decided to solder and heat shrink the cables at the stepper motor end and dispense with the choc-block. One less terminal to come loose. Here it is before the final large heat shrink was placed over the cables:


At the other end I re-did the connnectors including new earth cables to the outer screen. It all looked rather neat and tidy. Panel on wall (box sides and front to follow):

(note stepper motors are un-plugged in photo)

Here is the loom I made up to join the BOB to the steppers:


I then spent (wasted!) many hours this weekend trying to get the breakout board to talk to the stepper drivers. I checked the output from the direction pins and got around 4V or 0V depending upon direction. I wondered if this was a bit low so unplugged the USB lead and found a 5V source in the PC and made up a cable to connect to the 5V and GND pins instead. Got 5V output this time on the direction pins, but still no stepper movement.

I checked and double checked all the wiring, tried common anode and common cathode connections but still nothing. Checked all the Mach3 settings and they were all correct. Tried various things but to no avail. Borrowed an oscilloscope tonight and connected to the BOB pins. Direction still gave 5V or 0V but no pulse-train out of the step pins.

This was the BOB I was using - it was about £8 from ebay:


In frustration I wired up the one which came free with the drivers:


5V and GND from the PC, then pins 2 & 3 for step and direction and got a pulse-train! Plugged the driver back in one axis and the stepper moved ! So this will have to do to get it moving again with the new steppers. From that quick test it sounds much better than my all-in-one board if that is anything to go by.

I need to make up some proper looms for the bits I lashed up to get the new board working, but that should be straightforward.

The new board is very simple with no relays, speed control etc. so once I'm running I'll be looking into better boards with more features. There are a couple on my list such as the PMDX126, but not rushing into things.

[routercnc]

The saga continues!

On Sunday morning I managed to finish the last bit of basic wiring and set up all the motor tuning etc. Managed to achieve 2500mm/min - 3000mm/min jog speed before stalling, so settled on a safe 2000mm/min on X & Y to get it all going. Old machine ran at 1000mm/min so a great step forward.

I set out the machine to cut the 3mm aluminium front panel out to house all the controls. Fired up the spindle, hit cycle start and it started to cut the first circle. About 3 seconds into the cut there was a large bang and blue flash from the 60V 10A SMPS. This was the supply to the X, A and Y drivers (Z is only its own driver and supply).

After a bit of head scratch I realised that in my haste to get things going I'd left the 3off drivers on the factory default current setting of around 7.8A peak (5.6A RMS). As the power supply is rated at 10A I guess that was too much! Annoyingly the fuse inside the power supply was not blown, so I guess the whole thing is dead. I bought the PSU early last year, and I can't see another one of that spec on ebay, so that is that.

So, been doing a bit of research and will most likely build my own linear power supply for the X, A, and Y axes. Here is where I'm at with it:

Stepper motors - M60STH88-3008DF x 3off, 3.1Nm, 2.1A per phase (4.2 total), 5.46V
Drivers - DQ860MA 3off, 80V max

Toriod calculations
Nominal current draw = 3off x 2 phases x 2.1A per phase = 12.6A
Actual real draw = 0.7 x 12.6 = 9A
Required output to drivers = 68V (with margin for back EMF etc.)
Secondary voltage in coil = 68 / 1.4 = 48V AC
Toroid wired in series = 24 + 24
Power requirement = 48V x 9A = 432 Watts (432 VA)
Power with overhead in case of 4th axis = 750 Watts (750 VA)

Model selected: 750VA 230v to 2x24v CM0750224 from airlink at £48.00


Power regulator
For simplicity I'm looking at these power regulator boards at £29.00+£6.99 postage ebay. I know 60,000uF 100V is more than required but it was not much more than the 40,000uF 100V system.


Input fuse
I'm not completely sure on this but am looking at a 6A Type D MCB double throw fuse to switch the Live and Neutral into the toroid. Do I need to match the 9A real current draw or do I go lower? This 6A one is from Chalon at £7.18:


Output fuse
Is this required, and what would be the best type and rating to use?

Any thoughts on the above spec? thanks.

[routercnc]

I've redrawn the circuit diagram with a few planned updates. What I'm stuck on is:

Question 1: Choosing values and types of fuse for the input mains side of the power supplies:
3 power supplies requiring input protection are:
1. Planned new linear supply for X, A, Y axes, @750VA, pulling up to 9A (4.1A parallel 3.1Nm steppers x 3off x 70%)
2. Existing SMPS for Z axis is 350W pulling about 1-2A (2A series 1.8Nm stepper x 1off x 70%)
3. Existing SMPS for ancillaries is 12V, 60W, 5A

What to choose for each of the 3 above power supply mains inputs:
> Use MCB din rail mounted double throw?
> What type (A,B,C,D) to choose to avoid trip on inrush?
> What current level to choose?

Question 2: Choosing fuse values for DC output side of supplies
1. Linear supply on X,A,Y could pull up to 4A per motor - choose 5A blade fuse to each stepper?
2. SMPS on Z could pull up to 2A - choose 3A blade fuse?
3. SMPS for ancillaries rated at 5A - just go with 5A blade fuse?

If it helps here is the updated schematic, with the required new fuses / MCBs shown in RED (ignore ratings and type, those are some ideas which I'd like confirming):


Question 3: Setting current rating on DQ860MA drivers
Stepper motors are rated at 2.1A per phase so 4.2A total. Current setting options are either PEAK or RMS. Which to choose on the driver dip settings? Is it 4.2A PEAK or 4.0A RMS?


Question 4: Do the toriod and power regulator calcs in the previous post seem OK?

Sorry for all the questions . . .

[routercnc]

I ordered the toroid (750VA, 2x24v), power regulator board (60,000uF 100V capacitors, 20A rectifier), double pole MCBs for each power supply of various current rating with type D trip, DIN rail, DC fuse box holders.

Whilst waiting for this stuff to arrive I looked again at the blown SMPS and noticed a tiny glass fuse tucked away in the corner and soldered direct to the PCB (F2 - photo shows it removed).




It was clearly blown with black on the glass. I managed to cut the legs and get it out to read the rating and it looked like 250V 15A (printing was faint so not sure) about 3.5mm wide and 10mm long. Bit of searching revealed it was called an axial lead fuse and only they cost a few pounds for a pack of 4. Started to wonder if I could repair it and get the machine running for now until the linear supply was here and built. So I took the board out but discovered alot of damage underneath, some of them were charred surface mounted components. Even if I replaced the fuse and it worked would I trust it? Decided not.




So I need to wait a few weeks for all the new linear bits to arrive. No doubt I'll post any questions I have at that time so I don't get any more blue flashes. All for now.

[routercnc]

All the bits arrived in double quick time. Now I need some help confirming the wiring colours and general connections.

Here is the toroid: 750VA 2x24 from airlink


Here is the power regulator (60,000uF, 100V, 20A rectifier):



Here is a circuit I've drawn up but need help with:
1. Toroid blue and brown I assume are mains input?
2. 4 output wires are from the pair of windings? - but which to connect to the regulator and in which order?
3. I'll mount all this on an aluminium plate which will go to earth - does the centre GND output on the DC side need to be earthed as well?


I'm away from the parts at the moment otherwise I would at least give you the resistance readings across the toroid wires, but hopefully the above is enough.

Thanks

[Boyan Silyavski]

I had similar parts wired like that http://www.mycncuk.com/threads/7659-...9481#post59481

Hope that helps

[Jonathan]

Here is a circuit I've drawn up but need help with:
1. Toroid blue and brown I assume are mains input?

Likely yes, but it should say on the transformer itself.

2. 4 output wires are from the pair of windings? - but which to connect to the regulator and in which order?

The capacitor arrangement you have is intended to obtain a positive and negative voltage, not the single supply that you require. To use the board as-is, you could connect both secondary windings in series to AC1 and AC4, but you'll then have 3 capacitors in parallel with two in series, so only 10000*3/2=15000uF. That may be sufficent ... but it's a bit of a waste when you could modify the board to have all 6 capacitors in parallel to get 60000uF.

[routercnc]

Silyavski,
Thank you for that link - very good reading. Based on Jonathan's comments and your selection of the 40,000uF model then based on Neil's suggestion then you might only be getting 10,000uF (10,000uF x 2 /2). Any issues with running like that on your machine? If I use my board as-is then I'll only get 15,000uF.

Jonathan,
Thank you. I briefly looked at the toroid when it arrived to make sure all was intact and didn't notice the sticker on the side!


As for modifying the board to get the full 60,000uF - the tracks are 'painted over' ; either that or it is a multilayer board because apart from the solder points there are no exposed tracks on the top or bottom. I'd probably have to demount the capacitors and run flying leads to them from the output of the rectifier.

Had a couple of minutes spare today so checked the output from the secondary coils:

Mains was 243V AC and each secondary coil was 26.4V AC

I need to wire the secondarys in series but cannot find any markings on the cables to show which is the start of the winding. I've read that it does not matter how they are connected (unlike parallel which is vital). Can someone confirm this is the case?