Bipolar PI design

[Detectorist#1]

Hi pechkata,
Тhose who know doubt, those who do not know are certain.

[pechkata]

Greetings to all readers of this forum.

[pechkata]

For the half-sine the TX clocks are identical to the bipolar PI:



For the Impulse waveform the TX clocks are this:

Thanks Carl-NC
Whoever can, helps.

[pechkata]

Serious conversation is how the difference in Rds on and Cds between P-ch and N-ch transistors will affect the real work in practice. It is good to make this:
1. Adding of external resistors (this reduce the difference). The resistor in drain of N-ch transistor have to be with higher resistance.
2. Adding of capacitor in parallel to TX coil.
3. Use serially connected fast diodes with very low capacitance in series of the transistors
All these thinks have to seriously reduce imbalance. Maybe some one have possibility to analyze these ideas with good software simulator.

Hello detector #1
This is taught in first grade.
You're posing like a fairground vendor.

[Detectorist#1]

Serious conversation is how the difference in Rds on and Cds between P-ch and N-ch transistors will affect the real work in practice. It is good to make this:
1. Adding of external resistors (this reduce the difference). The resistor in drain of N-ch transistor have to be with higher resistance.
2. Adding of capacitor in parallel to TX coil.
3. Use serially connected fast diodes with very low capacitance in series of the transistors
All these thinks have to seriously reduce imbalance. Maybe some one have possibility to analyze these ideas with good software simulator.

[Altra]

You will never be able to match the P and N drive of a half-bridge driver. To get the Rds to match, the PMOS needs to be much larger than the NMOS. Then the parasitic capacitance of the PMOS will be much larger and the damping will be different for the P & N. You can fake it somewhat by adding series resistance or extra capacitance to the FETs.

Very true. You can get close enough to be usable.

[Orbit]

You will never be able to match the P and N drive of a half-bridge driver. To get the Rds to match, the PMOS needs to be much larger than the NMOS. Then the parasitic capacitance of the PMOS will be much larger and the damping will be different for the P & N. You can fake it somewhat by adding series resistance or extra capacitance to the FETs.

Series resistance didn't prove to be an ideal solution maybe mosfet capacity!?

[Altra]

Mikron hybrid use IXTP26P20 too.Any idea how demodulator are controlled on it.

No, I have only seen Youtube videos. But judging from the Tx circuit, it's probably more like a BBS detector than a PI?

[Carl-NC]

You will never be able to match the P and N drive of a half-bridge driver. To get the Rds to match, the PMOS needs to be much larger than the NMOS. Then the parasitic capacitance of the PMOS will be much larger and the damping will be different for the P & N. You can fake it somewhat by adding series resistance or extra capacitance to the FETs.

[Orbit]

There are smoke signals now all you need are hallucinogenic mushrooms and there you have it.

[N@SKO]

Mikron hybrid use IXTP26P20 too.Any idea how demodulator are controlled on it.

[Altra]

Hi Altra,
Do you really need in this design MOS FETs with so low Rds on? What is the inductance of your TX coil and TX max current?

It's not so much about low RDSon, but matching. Low RDSon has benefits. The best test is the flyback voltage comparison between n and p. Coils are 300-600uH, 1R or less. The tx current depends on a many factors like PW, Tx freq, Tx voltage coil resistance etc..

[Carl-NC]

Hello Carl-NC
Thanks for the clarification.
What do we feed to the points?
What kind of impulse?
NCLK1
PCLK1
NCLK2
PCLK2

For the half-sine the TX clocks are identical to the bipolar PI:



For the Impulse waveform the TX clocks are this:

[Detectorist#1]

Hi Altra,
Do you really need in this design MOS FETs with so low Rds on? What is the inductance of your TX coil and TX max current?

[pechkata]

In your three photos. I think you are confusing the Tx and demodulation signals. Carl's drawing shows the demodulators sampling during the halfsine. The I sample is taken during the complete HS period and the Q sample is inverted for the first 90 deg of the HS and alternated between pulses. In the hand drawing and Johnson patent those are the Tx signals and not the demodulators. To make all this happen you need to concentrate on programming micro processors and learn how to generate the timing signals. Carl's book has all the info you need for the rest. I use the IRF640, IXTP26P20 and a TC4428 driver for the TX. The IXTP26P20 is expensive and the ones from China are fake

Hello Altra
Exactly about Johnson's patent - I want to build his Tx signals.
Not for demodulators.
Thanks for the encouragement.

Google doesn't translate correctly.