Gate Driver Spikes

Thread Starter

Stuntmaster16

Joined Apr 22, 2020
1
Hi I am trying to simulate a very basic Class D audio power amplifier. My problem I have is there appears to be shoot through current between my two swtiching nmos mosfets. I tried avoiding the shoot through current by implementing the dead time control, but it doesn't seem to stop the shoot through currents. An example is when output A is and output B pulses high, output A tends to have a tiny spike has output B pulses high. This tiny spike causes the two nmos mosfets to turn on at the same time causing a shoot through current. Is there anything I am doing wrong that I have done that would cause the low output to have a tiny spike when the other output goes high?
1587578943092.png
1587579119894.png
Red is the Shoot-through current, White is the low-side gate drive voltage and Green is the high-side gate drive voltage from the Outputs A and B. As you can see with even dead time implemented, the lowside voltage has a spike as soon as the highside voltage pulses high. Is there something I am doing wrong that is causing the gate driver output to behave weirdly. This is my bachelors thesis topic and I need help with this because the shoot through current is causing very bad power efficiency of around 40%-50%. I am trying to make sure it stays in the 80%-90% region.

Below I attached the Gate driver spice model and the Amplifier schematic.
https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/3312.Spike-.JPG
 

Attachments

ronsimpson

Joined Oct 7, 2019
680
There are three different pieces to the UCC21222.
There needs to be a 10 to 15 volt supply(s) from Vdda to Vcca and a different one from B to B. (third one Vcci_0 to GND)
V7 44.5 will kill a real part.
Keep what I drew in BLUE and RED separate. Do not connect red and blue together.

1587646668747.png
When you make one out of real parts there must be capacitors on all supplies! Right on the pins. No long traces.
 
Top