Moodz' Awesome Gold Pulse Induction Version 3 - MAGPI V3 Project

[moodz]

Ok will take a closer look at the source code although normally a rotary transition should be used to generate an interupt either from A or B and the routine will read the state of the other pin to determine the direction.
Still trying to think how direction would work with only A as the input signal and have a hunch it may have somthing to do with what is happenning on the common pin.

Hi Dean ... yes I had a look at the rotary wiring and you are right I have wired the C ( common ) to the B switch Good Pickup !
So I changed the wiring to AB and C to common and its still works OK ... so I guess the code and the wiring is foolproof by accident.

I have never used interrupts for rotary encoders ...if you have some other more critical loop task ( like the TX interrupts ) then if the loop is fast enough to sample you get a rotary encoder or multiple REs with very little code overhead plus free debounce.

The less interrupts the better I always say.

moodz.

[KingJL]

... Its more of a safety feature for those who might be probing around on the board whilst it is operating...

Yeh, I once killed a VHMS3 digital board that way!! Luckily, Eric Foster had a spare and sent it to me.

[moodz]

OK so looking at the output without the zener I gather we should leave the zener diconnected would you agree..?

Sorry Dean .. my explanation was not clear ... the yellow trace is the TX control ... the pink trace is the X sample ... the blue green trace is the Zener fitted / not fitted trace exactly on top of each other .. the zener has no impact on the ouput.

The zener has no impact on the output.

However if people dont want to fit it ... OK. Its more of a safety feature for those who might be probing around on the board whilst it is operating.

moodz.

[Dean Sarelius]

That board that was just built up on the CRO with a minelab 10inch mono ...

Zener installed ( stored trace green ) and Zener uninstalled ( live trace blue ) overlayed ...


... I dont think I can resolve a meaningful difference ???

The pink pulse is the X sample pulse from the CPU ... ( bit of ringing there .. but no biggie ... the hyst on DG411 is not bothered. )

moodz

OK so looking at the output without the zener I gather we should leave the zener diconnected would you agree..?

[Dean Sarelius]

Regarding the display ... the notes indicate the backlight runs fine of 3.3 volts .. The serial to parallel converter module will run down to 2.5 volts so my guess is that 3.3 volts to the LCD logic just sneaks in.

The Alibaba add confirms it ( not that Alibaba is an oracle )

3 or 5 volts supply.

6 Dollars ...

3.3V 5V 2004 204 20x4 Character LCD Display Module 2004 LCD Monitor Module Yellow Blue Blacklight for Arduino Printer 3 D


Description:

1. The module is a low-power consumption character LCD Module with a built-in controller.
2. The module can be easily interfaced with a MCU.
3. Display Format: 20 Characters x 4 lines.
4. It is fantastic for Arduino based project.
5. Supply voltage: 3.3V / 5V.
6. Fully assembled and tested Serial LCD 20x4 Module.
7. Yellow / Blue backlight.
8. Analog Interface.
9. Size: 98 x 60 x 12mm.

moodz

Thanks Moodz,
I couldnt find this display so glad that it exists and will layout one of the headers to match...cheers

[Dean Sarelius]

On the contrary ... that rotary picks up direction no problem ... otherwise I would have a problem ... have a look at the source code ( Its in one of the interrupt blocks - TX I think ... but the rotary does not generate interrupts ) so the TX runs at about 15 Khz ... so that is the sample rate of the rotary.

moodz.

Ok will take a closer look at the source code although normally a rotary transition should be used to generate an interupt either from A or B and the routine will read the state of the other pin to determine the direction.
Still trying to think how direction would work with only A as the input signal and have a hunch it may have somthing to do with what is happenning on the common pin.

[moodz]

Yeh I was a bit concerned at first that I would have to rejig the board to add pull up resistors or solder them in midair or something .... but the PIC takes care of the pullups.
These 2004A displays ( cheapies of ebay ) seem to run fine on 3.3 volts. The display does look a little washed out on the photo but that is because the LCD update process is strobing the phone camera.

moodz.

Regarding the display ... the notes indicate the backlight runs fine of 3.3 volts .. The serial to parallel converter module will run down to 2.5 volts so my guess is that 3.3 volts to the LCD logic just sneaks in.

The Alibaba add confirms it ( not that Alibaba is an oracle )

3 or 5 volts supply.

6 Dollars ...

3.3V 5V 2004 204 20x4 Character LCD Display Module 2004 LCD Monitor Module Yellow Blue Blacklight for Arduino Printer 3 D


Description:

1. The module is a low-power consumption character LCD Module with a built-in controller.
2. The module can be easily interfaced with a MCU.
3. Display Format: 20 Characters x 4 lines.
4. It is fantastic for Arduino based project.
5. Supply voltage: 3.3V / 5V.
6. Fully assembled and tested Serial LCD 20x4 Module.
7. Yellow / Blue backlight.
8. Analog Interface.
9. Size: 98 x 60 x 12mm.

moodz

[moodz]

Hi Moodz,
Just looking at the encoder connections and it appears you have A connected with Orange and C (Common) connected with Red with B going to ground via the Green.
I gather that you are only detecting a low to high transition on the A connection and using this as your count so are likely unable to determine the direction..?

On the contrary ... that rotary picks up direction no problem ... otherwise I would have a problem ... have a look at the source code ( Its in one of the interrupt blocks - TX I think ... but the rotary does not generate interrupts ) so the TX runs at about 15 Khz ... so that is the sample rate of the rotary.

moodz.

[moodz]

Hi Moodz,
I gather you have enabled internal pull ups in the PIC for the encoder...?
Typically you would use a 10K resistor in series with the A and B signals with a capacitor to buffer the output.
Refer to a schematic which I have used below on an AVR.

For the display interface a lot of these require that the VCC = 5V primarily to drive the backlight and there are pull ups to VCC connected to the SDA and SCL pins.
I gather you are driving the display with 3V3 for the VCC so this should have no issue but I guess the display will be dimmer as a result..?

Yeh I was a bit concerned at first that I would have to rejig the board to add pull up resistors or solder them in midair or something .... but the PIC takes care of the pullups.
These 2004A displays ( cheapies of ebay ) seem to run fine on 3.3 volts. The display does look a little washed out on the photo but that is because the LCD update process is strobing the phone camera.

moodz.

[moodz]

Just observed something... added a 9.1 zener to the sim (I generated a sim using models of the actual parts to the TX and FE circuit (except for the analog switches) and noticed a 4 usec increase in the minimum sample delay. I don't know if all the 9.1 zeners exhibit the same effects... need to check this more. Right now I need some sleep!

That board that was just built up on the CRO with a minelab 10inch mono ...

Zener installed ( stored trace green ) and Zener uninstalled ( live trace blue ) overlayed ...


... I dont think I can resolve a meaningful difference ???

The pink pulse is the X sample pulse from the CPU ... ( bit of ringing there .. but no biggie ... the hyst on DG411 is not bothered. )

moodz

[Dean Sarelius]

Hi Moodz,
Just looking at the encoder connections and it appears you have A connected with Orange and C (Common) connected with Red with B going to ground via the Green.
I gather that you are only detecting a low to high transition on the A connection and using this as your count so are likely unable to determine the direction..?

[Dean Sarelius]

Hi Moodz,
I gather you have enabled internal pull ups in the PIC for the encoder...?
Typically you would use a 10K resistor in series with the A and B signals with a capacitor to buffer the output.
Refer to a schematic which I have used below on an AVR.

For the display interface a lot of these require that the VCC = 5V primarily to drive the backlight and there are pull ups to VCC connected to the SDA and SCL pins.
I gather you are driving the display with 3V3 for the VCC so this should have no issue but I guess the display will be dimmer as a result..?

[KingJL]

... and the 9.1 volt zener across the bsp230 bias cap.

Just observed something... added a 9.1 zener to the sim (I generated a sim using models of the actual parts to the TX and FE circuit (except for the analog switches) and noticed a 4 usec increase in the minimum sample delay. I don't know if all the 9.1 zeners exhibit the same effects... need to check this more. Right now I need some sleep!

[moodz]

That would be a good idea!


I would be interested in the outcome!


Thanks... will change my schematic accordingly.

Smoke test -- passed. ( no smoke escaped )




detail of the LCD connection




detail of the rotary encoder connection ...




Detail of the menu ...


The ">" is the current menu selection. Pushing the rotary encoder button will move menu selection // rotating will change value. Volume is home position ( ie boot up )
Note the detector will work without the LCD and Rotary .... I use it that way.

Have fun ... moodz.

[moodz]

Hi Moodz,
In addition to the errata list just got a couple of questions.
Q1 - Label DRV pin 5 of U11 (OUT_B) doesnt appear to have a connection within the schematic..?
I think this may have somthing to do with your comments "U11 is not needed if you link pin 7 U11 to pin 7 of U6"..just not sure what it all means..?
Cheers
Dean

yeh the mosfet driver chips are complementary outputs ... only need one for the TX mosfet ... however since the voltage multiplier uses the same chip and is driven by the TX line from the CPU then you could "steal" the mosfet drive from here. Some purists might want to keep a separate mosfet driver chip though ... I have tried it both ways cant really notice a difference and you can save a driver chip. OUT_B is not used.
moodz