Ideal diode bridge rectifier with GaN transistor and LT4320
Bunpei has replaced the 4-diode rectifier (bridge) circuit with “ideal diode type” 4-diode rectifier circuit prototyped by Mr. Hazawa.
LT4320+GS61008T（GaN)＿組立編: アナログ回路のおもちゃ箱 (In Japanese)
Here is a picture of his meticulous prototype.
The prototype use one LT4320 driver chip and four GS61008T GaN E type transistors.
Bunpei could perceive a very positive SQ improvement after the replacement.
Bunpei would like to express sincere gratitude to Mr. Hazawa’s effort.
Bunpei has replaced the 4-diode rectifier (bridge) circuit with “ideal diode type” 4-diode rectifier circuit prototyped by Mr. Hazawa.
LT4320+GS61008T（GaN)＿組立編: アナログ回路のおもちゃ箱 (In Japanese)
Here is a picture of his meticulous prototype.
An externally hosted image should be here but it was not working when we last tested it.
The prototype use one LT4320 driver chip and four GS61008T GaN E type transistors.
Bunpei could perceive a very positive SQ improvement after the replacement.
Bunpei would like to express sincere gratitude to Mr. Hazawa’s effort.
Attachments
By Bunpei’s request, Mr. Hazawa in Tokyo designed and assembled a new PCB of “ideal diode type bridge (4-diode) rectifier. The PCB incorporates one LT4320 and four GS61008P GaN E-mode HEMT.
Bunpei will apply it to DC power supplies for his DACs by replacing existing Si diodes and SiC diode with it.
AC input must be 9-75V and a capacitor must be placed at an output.
Bunpei will apply it to DC power supplies for his DACs by replacing existing Si diodes and SiC diode with it.
AC input must be 9-75V and a capacitor must be placed at an output.
Attachments
Hi, basreflex!
Would you like this application of the same kind of GaN device?
Direct PDM input to GaN E-mode HEMT Class-D Amplifier
Bunpei
Member
Joined 2006
a sigmadelta converter consists in it's simplest for of an integrator, followed by a clocked comparator, and a feedback path feeding back the digital representation at the output to the input to create a correction signal. the inegrator acts as "noise shaper" such that low frequency sampling noise is moved up to the higher frequency energies.
a power sigmadelta is the same structure, only the output logic is now + or - Vsupply, that could be like +/- 100V. that one needs fast and perfect switching, an ideal job for GaN.
a power sigmadelta is the same structure, only the output logic is now + or - Vsupply, that could be like +/- 100V. that one needs fast and perfect switching, an ideal job for GaN.
I’m sorry that I can’t catch the meaning of your words “why not”. Are you affirmative to apply GaN devices to a power delta-sigma converter?
What I reported on my post in Class D amplifier Forum is just “a power delta-sigma converter”. Are you negative to the application?
Anyway, in both applications, a rectifier for 60Hz and a switching device for 2.8 - 11.2 MHz delta-sigma modulated 1 bit audio signal(DSD), Mr. Ozawa and Bunpei perceive something very low-noise in its resultant sounds.
dear Bunpei,
I could not read from your description that the your GaN classD is actually a sigma delta converter with noise shaping. or said differently, measuring the power supply rfejection would reveal the difference.
I am a great fan of GaN. I use it for 6.78MHz classD power stages for wireless charging and active rectification. the very low gate drive requirements are a leap forward in active device technology. one caveat I found is that the gate capacitance has such a high Q, that any inductance in the gate path leads to ringing, such that always series resistors are needed.
I could not read from your description that the your GaN classD is actually a sigma delta converter with noise shaping. or said differently, measuring the power supply rfejection would reveal the difference.
I am a great fan of GaN. I use it for 6.78MHz classD power stages for wireless charging and active rectification. the very low gate drive requirements are a leap forward in active device technology. one caveat I found is that the gate capacitance has such a high Q, that any inductance in the gate path leads to ringing, such that always series resistors are needed.
I described as shown below, saying clearly a delta-sigma modulated 1 bit audio signal was directly input.
basreflex.
I can't understand your definition about DSM amplifier. One bit signal fed to an amplifier is already DSMed. Then 1-bit signal amplifier can be considered as powered DSM though it doesn't have delta-sigma modulator in itself. I can agree that bunpei's approach is not true classD, because classD generally accepts analog signal. But I can understand his project is DSM amplifier. I'm not familiar with classD. If I am wrong, please correct.
I can't understand your definition about DSM amplifier. One bit signal fed to an amplifier is already DSMed. Then 1-bit signal amplifier can be considered as powered DSM though it doesn't have delta-sigma modulator in itself. I can agree that bunpei's approach is not true classD, because classD generally accepts analog signal. But I can understand his project is DSM amplifier. I'm not familiar with classD. If I am wrong, please correct.
Bunpei applied Mr.Hazawa’s LT4320 ideal diode driver board with GaN E-mode HEMT GS61008P to +5V DC power supplies for a master clocking circuit in Chiaki’s ES9038PRO Dual Mono DAC and for a bipolar transistor driver circuit in NDK DuCULoN OCXO. Original four bridge diodes were four G-S tied GS61008T
The resultant SQ improvement is very satisfactory though no apple-to-apple comparison was not made because of an unavoidable transformer configuration change associated.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.