TX Driver

[Daves]

Hi guys,

What do you think the TX driving method should be like?
Garrett uses Lm386 most of the time.
There are people using Mosfets in different models.
There are people who drive with transistors.

What do you think is the difference between these?
What details would you suggest?
Can you share an example schema?

Considering the non-resonant coil driving method for Single Frequency and Selective Frequency, it seems more logical to drive with Transistors or Mosfets. Could you share an example diagram or a topic?

[Carl-NC]

There are many ways to design a TX driver, depending on exactly what you are wanting it to do. Which becomes the question.

[Daves]

There are many ways to design a TX driver, depending on exactly what you are wanting it to do. Which becomes the question.

I want to use a non-resonant coil.
I want to determine the TX frequency with the driver and ensure that it operates at the frequency I want.
I don't want to lose depth due to TX driving.
What method should I follow for this and what parameters of the coil should be like?

[kt315]

i can not to take for hand of every russian and bring him to here. look on schematics of Ciklon in a folder Thunderbolt_Groza_Verator
also i was banned in my forum and do not see last pics from the site. i am not in course what new there are. the access to pics was blocked by moderator.

second. if you want to see half-bridge or full bridge circuit just try to search the forum. for example, there will be a topic of Carl.
this is not new matter for us... but before... read the manual... ITMD book, link on the pages... and ask.

[Carl-NC]

The way to get an exact frequency is to clock the driver from a micro or a digital circuit. A non-resonant coil implies to me that you want to be able to change the frequency. I would use an H-bridge driver which creates a square wave voltage on the coil and a triangle wave current. A large cap across the driver supply will recycle the coil energy for good efficiency. A square wave driver is a little less efficient than a sine wave driver because some of the TX energy is lost in the harmonics, assuming you only demodulate the fundamental. However, you can also demodulate the third harmonic and do salt cancelation so that it works in salt water. Also, the depth loss due to the harmonics is very minor as it takes a large delta in TX energy to get a noticeable depth difference.

The coil design will depend on the frequency you want to run. At low frequency (1-5kHz), TX resistance is a major concern. At high frequency (50kHz & up) a low inductance and litz wire may be necessary. A lot of "it depends."

[kt315]

Inverter Project

---

This prject started with the idea to build a cheap inverter for our new house.
-- The bottom of the page will contain the newest updates. -- none yet
the basic requirments are:
1) Must be very very reliable, as it will run the whole house 24 hours a day.
2) Must be able to run a variety of loads... ranging from small motors up to 15HP wood working tools.
Preliminary Specs:

0) 15kw output power, may only be 8kw per 120v inverter, 16kw at 240v, not sure yet .
1) 48v battery bank
2) 120v RMS AC ouput -- two inverters can be connected in anti phase for 240v.
3) 48v from batteries will be converted to 170v DC at 50khz with ferrite transformers.
4) 170v will then be converted into 120v AC with H bridge
5) carrier frequency will be 19khz
6) the carrier duty cycle will be modulated (PWM) to create a sine wave on the output

The inverter also has to be very efficient, so as not to waste the 2kw (hydro) input to the battery bank. A PIC microcontroller (PIC16F62 will be used to create the pwm signal with its built in pwm module, which will be modulated with values from a sine table. External gates will be used to send the pwm signal to opposite sides of the H bridge to generate a complete sine wave.

Pin 9 of the PIC is the PWM output pin.
Pin 6 determines which half bridge the PWM signal goes to.
The two ground points in the schematic are isolated from eachother... the 170v dc supply will be floating.
The idea is that when one side of the H bridge is being sent the pwm signa, the other half will be tied to ground. This method of modulation should result in the below waveform.

Numbers above the pulses are duty cycle.
A high pulse means the lop left and bottom right side of the H bridge are ON.
The middle line means both the bottom mosfets are ON, shorting the ouputs to ground.
A Low pulse means the bottom left and top right mosfets are ON.
All of my ideas/theories are open to question... this is by no means a finished project, im sure many faults will be discovered along the road.
[email protected] -- all questions, comments, ideas, criticisms welcome
Home

[kt315]

[pito]

I have been using TC4420 for a long time, input = 5V pulses. In the case of the non-resonant variant; coil 80 turns, 0.5 wire, coil diameter 20 cm or larger, e.g. 35 cm,
To the output of IC4420 are connected the coil with 100uF capacitor connected in series.

[Daves]

The way to get an exact frequency is to clock the driver from a micro or a digital circuit. A non-resonant coil implies to me that you want to be able to change the frequency. I would use an H-bridge driver which creates a square wave voltage on the coil and a triangle wave current. A large cap across the driver supply will recycle the coil energy for good efficiency. A square wave driver is a little less efficient than a sine wave driver because some of the TX energy is lost in the harmonics, assuming you only demodulate the fundamental. However, you can also demodulate the third harmonic and do salt cancelation so that it works in salt water. Also, the depth loss due to the harmonics is very minor as it takes a large delta in TX energy to get a noticeable depth difference.

The coil design will depend on the frequency you want to run. At low frequency (1-5kHz), TX resistance is a major concern. At high frequency (50kHz & up) a low inductance and litz wire may be necessary. A lot of "it depends."

Yes Carl, I want to manage the frequency. I plan to do a strong drive with square wave and then import and process the rx signals, perhaps with analog demodulation or direct sampling.

Is there an example schematic, diagram or project you can share with me with the h bridge on the TX side? I would appreciate it if you could share it in a project I will examine.

[Daves]

I have been using TC4420 for a long time, input = 5V pulses. In the case of the non-resonant variant; coil 80 turns, 0.5 wire, coil diameter 20 cm or larger, e.g. 35 cm,
To the output of IC4420 are connected the coil with 100uF capacitor connected in series.

Thank you for your suggestion. Is there a scheme or project you can share with me with this TC4420 or another

[pito]

connect NE555 pin #3 to pin#2 TC4420 and do testings like here
https://www.geotech1.com/forums/foru...e17#post415377

[Carl-NC]

I don't have a project to share, but here is a simple H-bridge driver:



With the right switches, you can run VTX up to 44V and drive it from a 3V micro.

[pito]

It is not so simple, you need deadtime .......... ......in TXCL

[Carl-NC]

I've built and used this circuit and it works fine clocked from an 80MHz Cortex M4. I wouldn't clock it from an excessively slow micro. I also AC-couple the clocks to prevent a stuck state. Most times I now use a DRV8872 which makes for a very compact TX circuit.

[pito]

I'm not saying it doesn't work, I'm saying it's not the simplest one, it requires a microprocessor and software, and it's good for high power, I mean 10W or 20W, TC4420 can handle up to 10W which is more than enough and it's very simple.