I know nothing about the failure modes of IGBTs. Are they likely to open instead of shorting when they fail.
I'm dealing (long distance, I have no direct access to the amps) but from two different techs, I'm told that the high-side is getting 12v of drive at the gates but the high-side IGBTs (IRGP4066) aren't pulling up to the positive rail. This is an H-bridge amp using two Si8244 driver ICs. The rail voltage is approximately 250v (single rail) and the square wave out to the output filter inductors is approximately 60v in amplitude.
I'm dealing (long distance, I have no direct access to the amps) but from two different techs, I'm told that the high-side is getting 12v of drive at the gates but the high-side IGBTs (IRGP4066) aren't pulling up to the positive rail. This is an H-bridge amp using two Si8244 driver ICs. The rail voltage is approximately 250v (single rail) and the square wave out to the output filter inductors is approximately 60v in amplitude.
Can you get some photos, scope traces, etc?
Normally the gate drive level is a bit higher, around 15V.
Normally the gate drive level is a bit higher, around 15V.
Hi. It's strange a bit, to hear about igbt used in a class-d. Mostly they are slower than MOSFETs, also require higher drive voltage. Benefit is just at high voltages and big current, but operation is up to 20Khz . Welding invertor as example. D-class amplifiers require high switching frequency, hundreds of khz, especially full audio band ones ( not only for subwoofer ) . So my guess would be they are chosen improperly, maybe someone repaired and put them in place of MOSFETs. Maybe they can't fully turn on, and next cycle begins, so big loss occurs.
Rayma:
I don't have drive signals I can post. I may be able to get some but one tech is moving so not available and the other is waiting on parts.
The IC's low-side is powered by a 15v supply (actually, from the regulator resistor values, the voltage will be a bit over 16v) so the high-side is going to be around 15v.
Even with 10v of drive, shouldn't they be able to pull up to the rail voltage with no load? Again, I know nothing about IGBTs so my questions may be ridiculous.
This is a subwoofer amp.
There are several class D amps that use IGBTs. The diagram for this amp shows the IGBTs that are installed. NDA so I can't post it.
My main question was about the failure mode of IGBTs. FETs short, almost without exception, when they fail. I don't know what the failure of IGBTs looks like.
I don't have drive signals I can post. I may be able to get some but one tech is moving so not available and the other is waiting on parts.
The IC's low-side is powered by a 15v supply (actually, from the regulator resistor values, the voltage will be a bit over 16v) so the high-side is going to be around 15v.
Even with 10v of drive, shouldn't they be able to pull up to the rail voltage with no load? Again, I know nothing about IGBTs so my questions may be ridiculous.
ximikas:This is a subwoofer amp.
There are several class D amps that use IGBTs. The diagram for this amp shows the IGBTs that are installed. NDA so I can't post it.
My main question was about the failure mode of IGBTs. FETs short, almost without exception, when they fail. I don't know what the failure of IGBTs looks like.
The output can go either open or short.
Same for the gate.
IGBT drive needs +15V for the ON state, and up to -15V for the OFF state.
They are not interchangeable with other device technologies.
Same for the gate.
IGBT drive needs +15V for the ON state, and up to -15V for the OFF state.
They are not interchangeable with other device technologies.
Last edited:
These are never driven negative and the positive drive is limited to 15v (thereabouts) by 15v Zener diodes.
Check the datasheet curves, Vge @ 12V gives just under 250V@25C. Faulty drive circuit not able force saturation negative drive will be inherent based on driver.
I'm told that the high-side is getting 12v of drive at the gates but the high-side IGBTs (IRGP4066) aren't pulling up to the positive rail.
Remote diagnostics is hard to debug and will lead to vague answers chasing tail.
This part exhibits high frequency switching characteristics.
Total switching time = td(on) + tr + td(off) + tf = 50ns + 30ns + 190ns + 60ns = 330ns.
Fmax = 1/330ns results in 3.03 MHz
The culprit maybe the bootstrap high-side, unable to drive enough gate charge? bad driver IC? bad buffer? Something
in the high-side drive path is faulty and needs careful analysis each step of the way. The datasheet curves will tell a story but the drive
waveform will unlock the truth.
Driver is capable of doing 4A peak my guess is no buffers to the IGBT.
Here are the important curves for an IGBT:If it's not asking too much, which 'figures' for the curves are relevant?
1. Output Characteristics (Ic vs. Vce)
- Description: This curve shows the relationship between the collector current (Ic) and the collector-emitter voltage (Vce) for various gate-emitter voltages (Vge).
- Importance: Helps in understanding the saturation region and the active region of the IGBT.
2. Transfer Characteristics (Ic vs. Vge)
- Description: This curve shows the relationship between the collector current (Ic) and the gate-emitter voltage (Vge) at a constant collector-emitter voltage (Vce).
- Importance: Provides information on the gate threshold voltage and how the IGBT turns on.
3. Gate Charge Characteristics (Qg)
- Description: This curve shows the gate charge (Qg) required to fully turn on the IGBT as a function of the gate-emitter voltage (Vge).
- Importance: Important for understanding the switching performance and the drive requirements of the IGBT.
4. Switching Characteristics
- Turn-On and Turn-Off Waveforms: Show the switching behavior of the IGBT, including turn-on delay, rise time, turn-off delay, and fall time.
- Importance: Critical for designing efficient switching circuits and for thermal management.
5. Safe Operating Area (SOA)
- Description: This curve shows the safe limits of current and voltage that the IGBT can handle without damage.
- Importance: Essential for ensuring the reliability and longevity of the IGBT in various applications.
6. Power Dissipation vs. Collector Current (Pd vs. Ic)
- Description: This curve shows the relationship between the power dissipation and the collector current.
- Importance: Helps in thermal management and designing appropriate heat sinks.
7. Thermal Characteristics
- RthJC and RthJA vs. Time: Show the thermal resistance junction-to-case and junction-to-ambient over time.
- Importance: Useful for thermal design and ensuring the IGBT operates within safe temperature limits.
8. Forward Voltage Drop (Vce(on)) vs. Ic
- Description: This curve shows the voltage drop across the IGBT when it is fully on as a function of the collector current.
- Importance: Important for understanding conduction losses and efficiency in the on-state.
9. Capacitance Characteristics (Cies, Coes, Cres vs. Vce)
- Description: These curves show the input, output, and reverse transfer capacitances as functions of the collector-emitter voltage.
- Importance: Critical for high-frequency switching applications and for understanding the IGBT's dynamic behavior.
- Home
- Amplifiers
- Class D
- IGBTs in the output of a class D amplifier