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

[moodz]

The rh control tap connects to the bt ...the lh control tap toggles the setting selector and I already found a

[moodz]

The posts are being chopped by the forum

[moodz]

They are being chopped off at the first emoticon ???

[Gunghouk]

The rh control tap connects to the bt ...the lh control tap toggles the setting selector and I already found a

What did you find and have you fixed it?

[Gunghouk]

I have paired the HC-05 now but the app still closes on rh press. PIC updated to latest software. Is the paired name required HC-05 ?

[moodz]

What did you find and have you fixed it?

The text highlighting was in the wrong state ... its OK now.

Here is the fixed APK.

app-release.zip

[moodz]

I have paired the HC-05 now but the app still closes on rh press. PIC updated to latest software. Is the paired name required HC-05 ?

Make sure your HC-05 is running at 9600 ... the code does not autobaud so the app exits if bad speed.

[moodz]

I have paired the HC-05 now but the app still closes on rh press. PIC updated to latest software. Is the paired name required HC-05 ?

I found the problem .... from Android 12 onwards they changed the default bluetooth permissions.

If you go into your phone settings / privacy / bluetooth / allow nearby devices ...

set the required app ( MAGPI ) to allow.

I checked this on a Pixel 8 Pro ( android 14 ) and that was the case. Pressing the rh control no longer closes the app.

This instruction is only for Android 12 upwards.

[Gunghouk]

Yay......it lives

HC-05 is in slave mode is that ok or is master mode ultimately required for 2 way comms ?

[moodz]

Yay......it lives

HC-05 is in slave mode is that ok or is master mode ultimately required for 2 way comms ?

The HC05 is left as it comes from the factory …ie default slave mode. For two way comms your phone is master and the HC05 is slave.

you should not have to config the HC05 at all unless you were using it for something else and changed something.

[Gunghouk]

Hi Moodz regarding R12, is your final value of 330 Ohms due to the characteristics of the ML mono coil you are using? The reason I ask is I have determined the nearest standard value of 390 Ohms for my homemade coil (320uh, 1 ohm, 400pf) but have instability and a ringing on the output of the first amp stage which perpetuates through to the input of the ADC.

Some tests I conducted with the opamp and quad switch removed showed a value nearer 200 Ohms is required to quell the ringing at the coil. Does this sound ridiculously low or entirely possible within the operating characteristics of the circuit? In the meantime I'll put a 330 ohm in to see what happens. Thanks

yellow = coil flyback
blue = U4b pin 7
magenta = junction R10,C8

[Gunghouk]

On closer inspection this looks to start at the beginning of the flyback pulse and could be ringing on the +5v or gnd supplylines

[moodz]

On closer inspection this looks to start at the beginning of the flyback pulse and could be ringing on the +5v or gnd supplylines

You must solder a ground from the tab of the 7805 direct to coil ground .... I think I may have mentioned this in a previous post.

The 7805 heatsink tab is "grounded" ..but via a small inductance. The flyback pulse gets on this "ground" via capacitive coupling and the regulator oscillates randomly as it tries to do its job.

Also ... your analogue switch looks blown ... the sample should not look like that on pin U4b pin 7. Additionally your flyaback is barely reaching 100 volts and looks like it is avalanching.

The flyback should look clean ( raised cosine ) and approx 200 volts for 300 microhenry coil. What mosfets are you using in frontend.

See attached traces .... The only glitch occurs as the flyback falls to zero and the damping current diode D1 turns off.

Yellow is TX gate
Light blue is Flyback
Dark Blue is R10 C8
Magenta is U4b pin 7

[Gunghouk]

Thanks moodz, the ringing disappeared with the connection to the 7805 tab. A new DG411 made no difference to signal stability but the coil placed in a bare bones breadboard TX cct showed even less voltage at 125v peak.
My mosfets were bought in China so they could be the issue so I'll purchase again locally. Methinks a decent coil is also on the menu.
Cheers

[moodz]

TURBO CHARGING THE MAGPI frontend.

Changing R7 to 1 Megohm lifts the preamp gain to 10,000 ( 80 db ) coupled with the gain of the Diff amp at 6000 ( 75 dB ) the MAGPI signal chain gain prior to the ADC is now 155 dB which is better than true 24 bit performance which would be 146 dB.

Even at this extreme frontend gain the preamp response is still sharp with the preamp recovering in less than 2 microseconds.

Trace below is the output of the preamp ( blue ) with the RX window pulse ( yellow ).