Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
astralhighway, Wed Feb 17 2021, 05:59PMHi folks,
I recently completed a first DRSSTC using an H-bridge of FGA60N65SMD (TO3P) devices. It uses the well-discussed and much-loved UD topology and PRF modulation.
The aim of this project was to have a portable rig weighing as little as possible, resilient and transportable, and the meta-purpose is to photograph the plasma. I'm happy with the results. The inverter weighs less than 2lbs.
I actually scaled-down the power side of things after realizing that a design based around an IGBT brick isn't suitable for the environment I'm living in - the one I've built already sounds like a small helicopter and I don't want to bother the neighbors.
So I'm not chasing spark length and don't want to destructively test the limit of my bridge.
At the moment, it runs cool with no fan needed. The switches get to about 30 degrees C after quite long runs.
However...
I've tuned the coil quite well and nudged it up to 190VDC on the bus. Max on-time I'm prepared to risk is 100uS.
Alex Yuan reported that these IGBT's will manage 300A.
Based on scope traces from my 1000:1 cascaded CT, the primary peak current is 155A at 190VDC on the bus - pic shows around 15ins of spark to free air. Not risking strike to grounded object.
Q1.Sanity check: when we talk about 'peak current', we do mean 'peak current', and not 'peak-to-peak' current, which is clearly double? Right? This seems quite fundamental, but I've never seen a reference to peak-to-peak current.
Q2.Because the rig is well-tuned, plasma length increases measurably with even 10V increments on the bus, so I'm wondering if anyone has modeled what I would expect in terms of increasing current if I were bold enough to nudge above 200VDC on the bus. (Remember, we're talking a full bridge here. I'm in the UK so 240V is the wall socket voltage and I'm running it down through a variac)
Q3. What are the specific advantages of the H-bridge compared to a half-bridge, in terms of reliability/resilience? Clearly, each IGBT is working only half as hard as if it were in a half-bridge, but is that it?
Re: Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
astralhighway, Wed Feb 17 2021, 08:43PM
Just realised this should be in DRSSTC's. Mods, can it be moved please?
astralhighway, Wed Feb 17 2021, 08:43PM
Just realised this should be in DRSSTC's. Mods, can it be moved please?
Re: Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
astralhighway, Thu Feb 18 2021, 05:59PM
Photo update.
I tuned the coil by adding 1/3 turn. Pix show:
1) Modulation envelope. Voltage ('scope probe in the air), ch.1 and current, ch2.
Note with 190VDC on the bus, the IGBT's are still way within their limits.
2) Approx 45uS pulse.
3) Approx 65uS pulse.
Primary L=9.41, C=40nF -- F(selfres)=259KHz
Sec + toroid -- F(selfres)=277KHz.
What's noticeable now is that significant breakout happens@ just 6uS pulse-width, when 60A of primary current is flowing.
When I ramp up to just 70uS, spark length leaps up dramatically. 158A of primary current flows.
At pulse widths beyond this, primary current drops again to around 140A.
astralhighway, Thu Feb 18 2021, 05:59PM
Photo update.
I tuned the coil by adding 1/3 turn. Pix show:
1) Modulation envelope. Voltage ('scope probe in the air), ch.1 and current, ch2.
Note with 190VDC on the bus, the IGBT's are still way within their limits.
2) Approx 45uS pulse.
3) Approx 65uS pulse.
Primary L=9.41, C=40nF -- F(selfres)=259KHz
Sec + toroid -- F(selfres)=277KHz.
What's noticeable now is that significant breakout happens@ just 6uS pulse-width, when 60A of primary current is flowing.
When I ramp up to just 70uS, spark length leaps up dramatically. 158A of primary current flows.
At pulse widths beyond this, primary current drops again to around 140A.
Re: Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
RoV IW3IPD, Thu Feb 18 2021, 09:31PM
I don't see photos in the messages...
RoV IW3IPD, Thu Feb 18 2021, 09:31PM
I don't see photos in the messages...
Re: Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
astralhighway, Thu Feb 18 2021, 11:06PM
Trying to upload pix again.
astralhighway, Thu Feb 18 2021, 11:06PM
Trying to upload pix again.
Re: Successful DRSSTC build with H-bridge of TO3P IGBTS. Some Q's!
Alex Yuan, Sun Apr 04 2021, 04:27AM
My oneTesla ran around 250A with those 65N60s, 300A was the upper limit for those IGBTs in that case.
Q1. Amplitude
Q2. Is that 9.41uH for your primary? Way higher impedance than the oneTesla, so your max current should be less than 250A, which is what you are seeing. Set your current limit to like 250A and just crank up the voltage, while watching the output of the H-bridge and ensuring you don't have any significant transients that will exceed the 650V rating of your IGBTs. oneTesla runs a 370V bus.
Q3. P = IV, so if you are driving your IGBTs the same and have the same current flowing through them, the power dissipation of each IGBT, in either half or full bridge configuration should be the same. Reliability should be the same assuming both are well tuned. However, with the full bridge you will get twice the voltage across your primary.
Photos are still not showing up.
Alex Yuan, Sun Apr 04 2021, 04:27AM
My oneTesla ran around 250A with those 65N60s, 300A was the upper limit for those IGBTs in that case.
Q1. Amplitude
Q2. Is that 9.41uH for your primary? Way higher impedance than the oneTesla, so your max current should be less than 250A, which is what you are seeing. Set your current limit to like 250A and just crank up the voltage, while watching the output of the H-bridge and ensuring you don't have any significant transients that will exceed the 650V rating of your IGBTs. oneTesla runs a 370V bus.
Q3. P = IV, so if you are driving your IGBTs the same and have the same current flowing through them, the power dissipation of each IGBT, in either half or full bridge configuration should be the same. Reliability should be the same assuming both are well tuned. However, with the full bridge you will get twice the voltage across your primary.
Photos are still not showing up.
Print this page