DIY Moon Phase Dial

[saac]

I feel supremely under qualified to comment on that diagram Gina so all I will say is that I like the colours   I'm glad you know what your'e doing.

 

Jim

[Gina]

Well, I hope I do

[Gina]

I think I have now finished the new circuit board for the moon phase clock except for the N channel power MOSFETs to drive the RGB LED strip which haven't arrived yet.  This is a new piece of stripboard with all new components with RTC module, arduino Nano and stepper motor driver module having sockets for plugging in.  The plug in components are also new ones.

I shall have to set the RTC to real time first by using the USB and Serial Monitor to set the RTC to PC time.  Then I can try running the clock.  Of course, I can't test the LED strip variable coloured lighting until I get the MOSFETs.  They should have arrived today but hopefully tomorrow now.

Here are photos of the circuit board followed by the latest layout diagram.

[Gina]

Testing Time.

RTC and Nano plugged in, not motor driver module, and RTC read test performed (not set to real time yet) with Nano connected to PC via USB (no 12v power).  This worked giving the following display.

[Gina]

Real time set by uploading time via serial Monitor - test and time setting successful

[Gina]

Square wave test.  The following code snippet was used to check that the RTC square wave was working (a 1Hz period).

    int val = analogRead(sqwPin);  // read voltage level of 1Hz square wave
    digitalWrite(ledPin, (val > 500));

Result :-  Test successful

This concludes testing of the RTC module connected to the Nano for all functions required by the moon phase clock.

[Gina]

Well...  I connected 12v PSU and checked that I had correct polarity again before plugging in Nano and RTC module.  BUT I made two errors.  I had the +5v track from the Nano next to the +12v power line.  That was dreadfully silly!!! I should have known better...   The other error was to have the RTC module plugged in as well.  Result - two dead items instead of one due to a short between 12v and 5v tracks

I have now removed the two wires connecting to the bottom strip and replaced them with one wire that goes straight from the +5v pin on the Nano to the +5v logic power on the motor driver.  At least I didn't have the motor driver plugged in.  The wire from Nano +5v pin to RTC is nowhere near the +12v bus.

A new Nano and a new RTC module now tested without 12v supply plugged in and RTC set to current time successfully.  RTC square wave function also tested successfully. 

I won't have an RTC module plugged in next time I connect 12v...

[Gina]

Connected 12v PSU and checked that there were no volts where they shouldn't be.  Unplugged 12v PSU, plugged in Nano and plugged 12v PSU back in again.  Nano working OK.  12v off, RTC module plugged in and 12v back on - Nano LED flashed at half secong on and off correctly   Next tried with motor driver plugged in and board connected to stepper motor and applied 12v.  I set the sketch to output pulses to the motor normally for running the clock.  Motor ran with half second spurts as designed

In between working on the control circuit I had replaced the ABS motor gear with a TPE printed one and now the clock runs with just a quiet tick every half second - RESULT!

[ChrisLX200]

Amazing stuff that TPE, but it can produce 'stringy' prints - and the strings don't come off easy because they stretch instead :-)

ChrisH

[Gina]

Yes, unlike ABS you can't file it off - it has to be cut.  A nice sharp pair of side cutters work quite well.  Fortunately, being soft and flexibly, a fit bits sticking out don't matter, they just bend out of the way.  Makes nice gears

[Gina]

That didn't last long - the Nano died!   I don't know why or how.  The RTC draws just 300μA and the motor driver just 8mA max.  The Nano draws 20mA or so.  I know it's been suggested that 12v is perhaps too much for Vin.  28mA and 7v means 196mW - the internal voltage regulator chip is rated at 1.5W.  Even chips without any heat sink can dissipate 200mW easily enough or even 300mw.  RTC and logic power for the motor driver should not affect the Nano when run off 12v.

I guess I can try running the Nano off 7v derived from an LM317T.  I could set a current limit of say 100mA - that should stop the destruction in event of a short or something.

[Gina]

If my variable voltage and current bench PSU hadn't died on me I could have run the clock circuit off that and set a current limit.  Maybe I should wait until I get a replacement bench PSU or repair my current one (which I don't think I shall be able to do) or rig something up in the meantime.

[Gina]

Here is a circuit diagram for the LM317T to supply +7v for the Nano with current limiting at around 60mA though I might reduce that as the load is not expected to exceed 30mA.  eg. if the series resistor were increased to 15Ω the current limit would be about 40mA.  In the latter case the maximum load dissipation would be 280mW which shouldn't cause damage.

[Gina]

Here's a slightly amended diagram for 40mA current limiting with a red LED to indicate when current limiting is occurring.  This would be for testing.

[Stub Mandrel]

My philosophy would be that if the current is excessive, you need to find the problem, rather than mask it with current limiting, and for protection, fit a 100mA quick-blow fuse.

[Gina]

I agree - that's why I added "This would be for testing."  I won't be using the current limiting in the "production" units but a fuse in the circuit is a good idea.  Once I have all this properly sorted out I plan to use (maybe slightly amended versions of) this for several clocks I have in mind.  Even the pendulum clock will use this circuit to drive the winder motor and possibly an electromagnetic pendulum syncronous drive.  I might replace the A4988 driver with the higher power version but time will tell

Coming back to the testing rig, I think I might make up a proper supply unit.  I could add this to an ATX PSU or maybe make up an analogue supply with transformer, rectifier block and reservoir/smoothing capacitor.  It could have variable voltage and current limiting with voltage and current displays.  I have made numerous PSUs in the past.  With a floating power input I could use the simpler current limiter circuit shown in the LM317T datasheet.

[Stub Mandrel]

These seem to be very good, you can set the voltage and a current limit, and get a readout of voltage and current:

http://www.ebay.co.uk/itm/5A-CC-CV-Adjustable-Display-Step-Down-charge-Module-LED-Panel-Voltmeter-Ammeter-/272097473122?hash=item3f5a45f262

[Gina]

Thanks Neil   Shame about the long delivery time...  Could be useful for future projects though

Had an idea for a simpler very crude current limiting - just use a resistor in series with the LM317T input to drop the input voltage to a point where the Vin-Vout causes it to drop out.  eg. an 82Ω resistor would give a current limit of about 44mA.  Not for the "production" version though.

[Gina]

Here is the latest layout diagram using the LM317T to reduce the 12v to 7v to supply the Nano.  Shown with current limiting resistor but this will be replaced with a fuse in the final version.

[Gina]

I've had a thought   I wonder if there might be a problem with the stepper motor driver if the motor supply (12v) is off with supplying 5v to the logic supply input.  I had no problem with the driver not plugged in and it was alright at first with it plugged in and 12v applied but then I unplugged the 12v PSU leaving the logic supply fed from the Nano 5v pin.

[MarkyD]

The ones in my CNC box (A4988) are Ok without the 12V supply when the Uno is powered up via usb - no magic smoke has appeared   The 12V is just to supply the motor coils via the MOSFETS - it doesn't effect the switching.  The thing that does effect/damage them is if you disconnect the steppers whilst power is on

[Gina]

Thanks for that Mark - put my mind at rest

[Gina]

LM137T added to circuit board and tested.  Input voltage 12.24v output voltage 7.02v with no load.

[Gina]

Plugged in new Nano and powered up.  Working   Vpsu = 12.24v, Vin = 6.96v, LM317T input = 10.38v series resistor = 68Ω (couldn't find any 82Ω).  Calculated Nano supply current = 27mA.

[Gina]

RTC module plugged in and power applied.  Test sketch showing Nano LED blinking at 1Hz correctly from RTS square wave.  Vpsu = 12.24v, Vin = 6.90v, LM317T input = 10.04v/9.94v (depending on LED state).  Calculated Nano supply current = 32/34mA.  So far so good