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

[geoscash1]

Got my boards today, turned out nice. Need a few more parts and finish up another project then will get started on this.

[Technologyh]

Got my boards today, turned out nice. Need a few more parts and finish up another project then will get started on this.

[Technologyh]

Very beautiful. Can you picture it front and back? Excuse me

[geoscash1]

Very beautiful. Can you picture it front and back? Excuse me

Sure, front and back...

[van0014]

I’m looking forward to making this, and have a big bag of components. More parts to come, and BSP230 is hard to get.

i’m using an illuminated rotary encoder as a power on light. Also hopefully a 2004 LCD with a white backlight. Maybe even a potentiometer for audio and gain or SAT control

[van0014]

After having the PCBs and parts for a few days, I couldn’t help but start putting it together. It took me a while to get the hang of soldering. Mostly I solder car wires of about 1.5mm squared. I won’t be getting high marks for soldering. Though the IC sockets were easiest, and some of those joints might even have the recommended amount of solder. Most other joints have a bit too much. I chose lead free HASL for the PCB tinning.

All that’s left is the 100uF charge pump capacitors (I ordered 6.3v and it might not be enough. More to be ordered), the BSP230 (will a BSP225 be ok here?) and the coil socket and screen parts.

some of my components look odd. Resistors are 1%, and 1/4W or higher, though they are small. In hindsight, I should have looked into the voltage rating of these. 200v is the limit, though they handle more intermittently. Clock capacitors are C0G / NP0, and I used ceramic caps for the 1uF parts, X8R. The hope is that parts with better tolerances or less temperature variations may be beneficial.

The downside is, the more I stray from Moodz’ component choices, the more risk there is that some components may need changing. Or worse, scope testing. I haven’t got the talent he has. I just like to read datasheets. I work on car wiring mostly.

Moodz, thankyou for posting your design and project. I really enjoy reading through all your posts, and there’s many talented people on these forums. I’ve been itching to put something together on a PCB for a long time

[Technologyh]

Go over the topics from the beginning and you understand.

[Technologyh]

Did this project stop? Is it now ready to make it? Can you help us.

[van0014]

I've had a go at making this, and had some success. But also made some mistakes. I ordered an L79L05 instead of L79L09. It still seemed to detect things, but not quite the way it should.

Sourcing a genuine BSP230 is somewhat expensive ($20 - 30), though cloned ones are availiable they most certainly would be a compromise. I've came across DMP45H150DHE-13, and may try it as a cheaper alternative. It's unknown what difference it'll make and if it needs different gate drive characteristics. I can get BSP225 from digikey but it's at risk of avalanching. I am probably going to stick with BSP230

I'm looking for ways to improve on the design. The gain of INA125 seems slightly out of spec, risking non linear gain setting and unstable operation. Though admittedly, i'm not an engineer. I just like to read datasheets, and work on car wiring. I'm going to either try a 6.04 ohm precision surface mount resistor (RN73C2A6R04BTDF) or reduce gain as low as 1,281 (47 ohms). Previously I used CPF15R0000BEE14 as a precision 5 ohm resistor.
This gain change is based on some questionable advice from AI, which says gains above 120 db may be too unstable. This hopefully would give similar response by changing the code
//PR4 = (unsigned int)(ADCTarget * (original gain / new gain)) + audioTone + 128;
PR4 = (unsigned int)(ADCTarget * (12004 / 1281)) + audioTone + 128;

This change in gain is in conjunction with changes to NE5534 and LM6171, but I think I may be almost trying to over engineer it, when it may be working at it's best the way it has been designed originally.

INA125 Rg	6Ω (Gain ≈10,000)	24–60Ω (Gain ≈2,500 to 1,000)	Lower noise/offset, higher bandwidth
NE5534 feedback R	100Ω	220–470Ω	Lower gain, better stability
LM6171 feedback C	10pF	22–33pF	Lower bandwidth, less high-freq noise

[AhmatAhmatov1959]

I've had a go at making this, and had some success. But also made some mistakes. I ordered an L79L05 instead of L79L09. It still seemed to detect things, but not quite the way it should.

Sourcing a genuine BSP230 is somewhat expensive ($20 - 30), though cloned ones are availiable they most certainly would be a compromise. I've came across DMP45H150DHE-13, and may try it as a cheaper alternative. It's unknown what difference it'll make and if it needs different gate drive characteristics. I can get BSP225 from digikey but it's at risk of avalanching. I am probably going to stick with BSP230

I'm looking for ways to improve on the design. The gain of INA125 seems slightly out of spec, risking non linear gain setting and unstable operation. Though admittedly, i'm not an engineer. I just like to read datasheets, and work on car wiring. I'm going to either try a 6.04 ohm precision surface mount resistor (RN73C2A6R04BTDF) or reduce gain as low as 1,281 (47 ohms). Previously I used CPF15R0000BEE14 as a precision 5 ohm resistor.
This gain change is based on some questionable advice from AI, which says gains above 120 db may be too unstable. This hopefully would give similar response by changing the code
//PR4 = (unsigned int)(ADCTarget * (original gain / new gain)) + audioTone + 128;
PR4 = (unsigned int)(ADCTarget * (12004 / 1281)) + audioTone + 128;

This change in gain is in conjunction with changes to NE5534 and LM6171, but I think I may be almost trying to over engineer it, when it may be working at it's best the way it has been designed originally.

INA125 Rg	6Ω (Gain ≈10,000)	24–60Ω (Gain ≈2,500 to 1,000)	Lower noise/offset, higher bandwidth
NE5534 feedback R	100Ω	220–470Ω	Lower gain, better stability
LM6171 feedback C	10pF	22–33pF	Lower bandwidth, less high-freq noise

Hello. Please advise the sound amplifier circuit for an 8 ohm speaker with volume control. I made it on a 386 chip and didn't like it. Thank you.

[van0014]

The original sound circuit is basically a Class-D audio amplifier, and should be sufficient for an 8 ohm speaker. Please try other speakers. Also the volume is low by default, and needs a rotary encoder connected to turn it up, preferably also with a 2004 LCD screen with an I2C backpack..

If you really needed to, the source code can be modified from the most up to date file in these posts, to use a higher default volume. You will then need a pickit 3 if you want to compile the source code. But don't buy the pickit unless you are comfortable with the source code and have compiler software. This is not recommended if you haven't done it before. I highly recommended getting a rotary encoder and screen, and following the photos in these posts to see how to wire them up

[AhmatAhmatov1959]

The original sound circuit is basically a Class-D audio amplifier, and should be sufficient for an 8 ohm speaker. Please try other speakers. Also the volume is low by default, and needs a rotary encoder connected to turn it up, preferably also with a 2004 LCD screen with an I2C backpack..

If you really needed to, the source code can be modified from the most up to date file in these posts, to use a higher default volume. You will then need a pickit 3 if you want to compile the source code. But don't buy the pickit unless you are comfortable with the source code and have compiler software. This is not recommended if you haven't done it before. I highly recommended getting a rotary encoder and screen, and following the photos in these posts to see how to wire them up

Thanks for the reply. I'll have to buy an encoder and a screen.

[van0014]

That's ok. Here's the post: https://www.geotech1.com/forums/foru...007#post410007

[van0014]

After years of soldering, I have finally realised the importance of using flux. And although solder does have flux in it, it’s not enough to get good results with

These are before and after photos with flux and smaller solder

[moodz]

Hello. Please advise the sound amplifier circuit for an 8 ohm speaker with volume control. I made it on a 386 chip and didn't like it. Thank you.

Make sure you use a DC blocking low pass filter at the input to the amp. The Magpi puts out PWM but it need DC blocking to remove the DC remainder. ( ie connect speaker via capacitor )