Voltage rise on gate
furnace, Fri Jan 30 2015, 04:16AMHi everyone!
Can someone please try and explain why I get a voltage rise on my gate, this only happens to the first mosfet that is switched on, on a normal half bridge circuit with 2 dc blocking caps for induction heating.
The circuit runs very smooth at 200volts ac through a variac which is rectified. Its only that the high side gets slightly warmer than the low side.
This is the signal on both mosfet with te variac on. If the variac is switched off than both signals rise and fall times are even, from the gdt of my pll driver.
this signal
Re: Voltage rise on gate
johnf, Fri Jan 30 2015, 06:57AM
Furnace
Your high side drive is too low---- could be caused by the bootstrap cap not being enough or insufficient drive through your GDT
johnf, Fri Jan 30 2015, 06:57AM
Furnace
Your high side drive is too low---- could be caused by the bootstrap cap not being enough or insufficient drive through your GDT
Re: Voltage rise on gate
Wolfram, Sat Jan 31 2015, 01:29AM
How do you measure the voltage on the upper gate with power on the half bridge?
What sort of driver are you using? It's very difficult to help without even knowing what circuit you're using.
Wolfram, Sat Jan 31 2015, 01:29AM
How do you measure the voltage on the upper gate with power on the half bridge?
What sort of driver are you using? It's very difficult to help without even knowing what circuit you're using.
Re: Voltage rise on gate
furnace, Sat Jan 31 2015, 05:35AM
This is my circuit. I tried increasing power on my driver but it does not make a difference.
furnace, Sat Jan 31 2015, 05:35AM
This is my circuit. I tried increasing power on my driver but it does not make a difference.
Re: Voltage rise on gate
Antonio, Sat Jan 31 2015, 04:29PM
Where are you connecting the oscilloscope probes? With a grounded oscilloscope you can directly measure the gate voltage of the lower transistor. To measure the gate-source voltage of the upper transistor you have to subtract the signals at the gate and at the source. (I don't recomment lifting the oscilloscope from the ground and measuring the gate-source voltage directly.)
Antonio, Sat Jan 31 2015, 04:29PM
Where are you connecting the oscilloscope probes? With a grounded oscilloscope you can directly measure the gate voltage of the lower transistor. To measure the gate-source voltage of the upper transistor you have to subtract the signals at the gate and at the source. (I don't recomment lifting the oscilloscope from the ground and measuring the gate-source voltage directly.)
Re: Voltage rise on gate
furnace, Sun Feb 01 2015, 04:21AM
The scope is grounded through a cap and the probes are directly on the gate of the mosfets.
But anyway this is not to be considered as this is not the source of my problem. The problem is that the voltage of my high side mosfet is rising on the gate or it seems to build up a charge.
furnace, Sun Feb 01 2015, 04:21AM
The scope is grounded through a cap and the probes are directly on the gate of the mosfets.
But anyway this is not to be considered as this is not the source of my problem. The problem is that the voltage of my high side mosfet is rising on the gate or it seems to build up a charge.
Re: Voltage rise on gate
mister_rf, Sun Feb 01 2015, 04:47PM
Question: are the 5V1 zener diodes connected as in the schematics? (D5/D6) Maybe one of the diode was reverse direction mounted.
In my opinion I’ll use an extra 10k resistor connected between the gate and the source of each MOSFET.
mister_rf, Sun Feb 01 2015, 04:47PM
Question: are the 5V1 zener diodes connected as in the schematics? (D5/D6) Maybe one of the diode was reverse direction mounted.
In my opinion I’ll use an extra 10k resistor connected between the gate and the source of each MOSFET.
Re: Voltage rise on gate
furnace, Mon Feb 02 2015, 09:55AM
The zener was connected wrong, yes! I corrected it but now the mosfets does not want to turn on?
furnace, Mon Feb 02 2015, 09:55AM
The zener was connected wrong, yes! I corrected it but now the mosfets does not want to turn on?
Re: Voltage rise on gate
furnace, Sat Feb 07 2015, 05:11AM
Can someone please assist me with this?
furnace, Sat Feb 07 2015, 05:11AM
Can someone please assist me with this?
Re: Voltage rise on gate
Sigurthr, Sat Feb 07 2015, 06:39AM
The gate could have been blown open circuit, in which case the fet will never turn on. It's rare, as they usually fail closed circuit, but not impossible.
Sigurthr, Sat Feb 07 2015, 06:39AM
The gate could have been blown open circuit, in which case the fet will never turn on. It's rare, as they usually fail closed circuit, but not impossible.
Re: Voltage rise on gate
furnace, Sun Feb 08 2015, 04:10AM
The mosfets are ok, because when I remove the zeners it turns back on again but than I get the high voltage on the gate. The high voltage reading is from the dc blocking cap being charged.
furnace, Sun Feb 08 2015, 04:10AM
The mosfets are ok, because when I remove the zeners it turns back on again but than I get the high voltage on the gate. The high voltage reading is from the dc blocking cap being charged.
Re: Voltage rise on gate
furnace, Wed Feb 11 2015, 12:24AM
I take it that this is normal for the half bridge to have a rise on gate of the high side?
furnace, Wed Feb 11 2015, 12:24AM
I take it that this is normal for the half bridge to have a rise on gate of the high side?
Re: Voltage rise on gate
Antonio, Wed Feb 11 2015, 12:57AM
Antonio, Wed Feb 11 2015, 12:57AM
furnace wrote ...
The scope is grounded through a cap and the probes are directly on the gate of the mosfets.
But anyway this is not to be considered as this is not the source of my problem. The problem is that the voltage of my high side mosfet is rising on the gate or it seems to build up a charge.
This would short-circuit the output of the bridge to the negative rail through the ground probes of the oscilloscope, if both are connected at the same time. The risk of extensive destruction is high... If you connect the ground of the oscilloscope to the output of the bridge, there are high-frequency signals everywhere, with all kinds of parasitic effects. If a capacitor is grounding the oscilloscope, it is across the output of the bridge. As a security measure with these measurements, connect the ground probe of the oscilloscope through a small 100 ohms resistor. If something is wrong the resistor burns, and not something more expensive.The scope is grounded through a cap and the probes are directly on the gate of the mosfets.
But anyway this is not to be considered as this is not the source of my problem. The problem is that the voltage of my high side mosfet is rising on the gate or it seems to build up a charge.
Re: Voltage rise on gate
furnace, Wed Feb 11 2015, 10:56AM
Thanks Antonio, this didn't solve the problem but my low side mosfet is behaving better, it seemed to short on the gate of the lower mosfet.
Any advise on the voltage rise of my high side mosfet?
furnace, Wed Feb 11 2015, 10:56AM
Thanks Antonio, this didn't solve the problem but my low side mosfet is behaving better, it seemed to short on the gate of the lower mosfet.
Any advise on the voltage rise of my high side mosfet?
Re: Voltage rise on gate
Antonio, Wed Feb 11 2015, 12:47PM
D5 and D6 should not exist, or at least should have small resistors in parallel, and be not zeners, but fast diodes. Their function would be is to accelerate the turnoff of the mosfets, but the effect is small. With the diagram as shown, the mosfets would not turn on. If the zeners are reversed they would not turn off.
Anyway, you can't measure directly the vgs voltage of the high-side mosfet without causing many loading problems, if not a straight short circuit. Try to observe the gate and the source with grounded oscilloscope probes and subtract the signals. The ground terminals of the probes can only be connected to ground.
Antonio, Wed Feb 11 2015, 12:47PM
D5 and D6 should not exist, or at least should have small resistors in parallel, and be not zeners, but fast diodes. Their function would be is to accelerate the turnoff of the mosfets, but the effect is small. With the diagram as shown, the mosfets would not turn on. If the zeners are reversed they would not turn off.
Anyway, you can't measure directly the vgs voltage of the high-side mosfet without causing many loading problems, if not a straight short circuit. Try to observe the gate and the source with grounded oscilloscope probes and subtract the signals. The ground terminals of the probes can only be connected to ground.
Re: Voltage rise on gate
furnace, Wed Feb 11 2015, 01:52PM
Yep! I figured out the zeners. I only have 15ohm resistors in series with gate and the signal is perfect. Can you please explain more in detail what you mean that I should observe the gate and source With grounded scope probes and subtrcact The signals?
At the moment my probe is directly on the high side gate and ground of the scope is on the main ground with 200ohm resistor.
furnace, Wed Feb 11 2015, 01:52PM
Yep! I figured out the zeners. I only have 15ohm resistors in series with gate and the signal is perfect. Can you please explain more in detail what you mean that I should observe the gate and source With grounded scope probes and subtrcact The signals?
At the moment my probe is directly on the high side gate and ground of the scope is on the main ground with 200ohm resistor.
Re: Voltage rise on gate
Antonio, Wed Feb 11 2015, 02:18PM
Many oscilloscopes have a function that subtracts the signals on both probes, or invert one and add. If yours don't have it, observe both signals, gate and source, with both probes grounded and subtract visually, what is probably better.
Antonio, Wed Feb 11 2015, 02:18PM
Many oscilloscopes have a function that subtracts the signals on both probes, or invert one and add. If yours don't have it, observe both signals, gate and source, with both probes grounded and subtract visually, what is probably better.
Re: Voltage rise on gate
furnace, Wed Feb 11 2015, 02:45PM
I have an inverted switch on my scope. Everything looks ok, it's just that my low mosfet is getting hotter than high one and the high signal is still rising when as I increase power on the variac.
furnace, Wed Feb 11 2015, 02:45PM
I have an inverted switch on my scope. Everything looks ok, it's just that my low mosfet is getting hotter than high one and the high signal is still rising when as I increase power on the variac.
Re: Voltage rise on gate
Wolfram, Wed Feb 11 2015, 08:52PM
If you're measuring without the top MOSFET source grounded, then you're measuring the sum of the lower MOSFET drain voltage, and the upper MOSFET gate voltage. As the drain voltage increases when you apply power with the variac, the gate voltage will appear to rise when it isn't
If you are measuring with the top MOSFET source grounded, then you're either shorting out your bridge or putting live mains on the scope chassis (only possible if the scope is not grounded and you're not measuring the bottom MOSFET gate voltage at the same time).
Wolfram, Wed Feb 11 2015, 08:52PM
If you're measuring without the top MOSFET source grounded, then you're measuring the sum of the lower MOSFET drain voltage, and the upper MOSFET gate voltage. As the drain voltage increases when you apply power with the variac, the gate voltage will appear to rise when it isn't
If you are measuring with the top MOSFET source grounded, then you're either shorting out your bridge or putting live mains on the scope chassis (only possible if the scope is not grounded and you're not measuring the bottom MOSFET gate voltage at the same time).
Re: Voltage rise on gate
furnace, Tue Feb 17 2015, 06:18AM
I'm blowing my switches, I've even had different mosfets on the bridge and tried igbts, the igbts get very hot when I use them at lower power levels and all the switches seem to burn at 450 v, I don't know what to do, I've seen on uzzors design that his switches does not have internal reverse diodes, is it possible that the reverse diodes could implement a short circuit on the bridge? When the lower mosfet turns on the current runs back to the high side mosfet through the reverse diode?
furnace, Tue Feb 17 2015, 06:18AM
I'm blowing my switches, I've even had different mosfets on the bridge and tried igbts, the igbts get very hot when I use them at lower power levels and all the switches seem to burn at 450 v, I don't know what to do, I've seen on uzzors design that his switches does not have internal reverse diodes, is it possible that the reverse diodes could implement a short circuit on the bridge? When the lower mosfet turns on the current runs back to the high side mosfet through the reverse diode?
Re: Voltage rise on gate
mister_rf, Tue Feb 17 2015, 09:43AM
Up to 450V?
Maybe that’s too much for such circuit. There occurs an overlapping conduction of both transistors because there’s no dead time between the upper and lower transistor.
mister_rf, Tue Feb 17 2015, 09:43AM
Up to 450V?
Maybe that’s too much for such circuit. There occurs an overlapping conduction of both transistors because there’s no dead time between the upper and lower transistor.
Re: Voltage rise on gate
furnace, Tue Feb 17 2015, 10:09AM
Cross conduction, that's what I thought. What to do? Do I need a new drive circuit and if so what would be a very efficient driver for high voltageses? Because I have 240v and mains.
furnace, Tue Feb 17 2015, 10:09AM
Cross conduction, that's what I thought. What to do? Do I need a new drive circuit and if so what would be a very efficient driver for high voltageses? Because I have 240v and mains.
Re: Voltage rise on gate
furnace, Fri Feb 20 2015, 07:16AM
So I took out the tc drivers and added a ir2113
mosfet driver which is suppose to be dv/dt immune as they say.
Stil I'm burning switches.
As can be seen on the picture I stil have little cross conducting on the mosfets.
I already have 15ohm resistors in series.
What should I do to get a pure square wave?
furnace, Fri Feb 20 2015, 07:16AM
So I took out the tc drivers and added a ir2113
mosfet driver which is suppose to be dv/dt immune as they say.
Stil I'm burning switches.
As can be seen on the picture I stil have little cross conducting on the mosfets.
I already have 15ohm resistors in series.
What should I do to get a pure square wave?
Re: Voltage rise on gate
furnace, Wed Apr 01 2015, 02:36AM
Can someone please help? When I connect my scope to ground it heats up the resistor and trips the mains power on my bridge.
furnace, Wed Apr 01 2015, 02:36AM
Can someone please help? When I connect my scope to ground it heats up the resistor and trips the mains power on my bridge.
Re: Voltage rise on gate
mister_rf, Wed Apr 01 2015, 07:49AM
You need to implement some sort of dead-time in that circuit.
I think originally the zener diodes were used there for this job. Also see the discussion here:
mister_rf, Wed Apr 01 2015, 07:49AM
You need to implement some sort of dead-time in that circuit.
I think originally the zener diodes were used there for this job. Also see the discussion here:
Re: Voltage rise on gate
furnace, Wed Apr 01 2015, 01:08PM
I do have the zeners on and the dead time is ok and I also have a very clean signal. My problem is the short that my probe is having. When I try to ground my scopes probe it starts heating up my resistor and on the probe until the main power trips. I loosened everything and checked all the connections. It seems that my probe ground makes a dead short to my variac even if the Mosfets are not turned on and also with the mosfets disconnected.
furnace, Wed Apr 01 2015, 01:08PM
I do have the zeners on and the dead time is ok and I also have a very clean signal. My problem is the short that my probe is having. When I try to ground my scopes probe it starts heating up my resistor and on the probe until the main power trips. I loosened everything and checked all the connections. It seems that my probe ground makes a dead short to my variac even if the Mosfets are not turned on and also with the mosfets disconnected.
Re: Voltage rise on gate
Bjørn, Wed Apr 01 2015, 02:02PM
You have not given any schematics that explain what you do so not much to say except you need to learn the simple things before you try to do build something complicated and dangerous.
Bjørn, Wed Apr 01 2015, 02:02PM
furnace wrote ...
My problem is the short that my probe is having. When I try to ground my scopes probe it starts heating up my resistor and on the probe until the main power trips. I loosened everything and checked all the connections. It seems that my probe ground makes a dead short to my variac even if the Mosfets are not turned on and also with the mosfets disconnected.
This is what is supposed to happen when you ground something that is not ground. The fuse is supposed to blow for safety reasons.My problem is the short that my probe is having. When I try to ground my scopes probe it starts heating up my resistor and on the probe until the main power trips. I loosened everything and checked all the connections. It seems that my probe ground makes a dead short to my variac even if the Mosfets are not turned on and also with the mosfets disconnected.
You have not given any schematics that explain what you do so not much to say except you need to learn the simple things before you try to do build something complicated and dangerous.
Re: Voltage rise on gate
furnace, Wed Apr 01 2015, 02:39PM
Oops! So sorry, but this is my schematic.
furnace, Wed Apr 01 2015, 02:39PM
Oops! So sorry, but this is my schematic.
Re: Voltage rise on gate
Bjørn, Wed Apr 01 2015, 05:04PM
Bjørn, Wed Apr 01 2015, 05:04PM
furnace wrote ...
Oops! So sorry, but this is my schematic.
And where exactly do you connect the oscilloscope when it goes wrong?Oops! So sorry, but this is my schematic.
Re: Voltage rise on gate
furnace, Wed Apr 01 2015, 10:02PM
My probe is directly on the gate and the ground of my probe is connected to earth ground through a 1k resistor.
furnace, Wed Apr 01 2015, 10:02PM
My probe is directly on the gate and the ground of my probe is connected to earth ground through a 1k resistor.
Re: Voltage rise on gate
GrantX, Thu Apr 02 2015, 05:25AM
Using an isolating step-down transformer to power the DC bus at a low voltage can help with troubleshooting, as there is no risk of shorting out mains power through your scope.
GrantX, Thu Apr 02 2015, 05:25AM
furnace wrote ...
My probe is directly on the gate and the ground of my probe is connected to earth ground through a 1k resistor.
I'm not sure if I'm missing something (the circuit diagram you posted was a linear DC supply), but you cannot scope the high side FET with the probe ground clip connected (well, not without extra effort). The ground clip is connected to mains earth and with rectified 230V on the bus the high side gate will be bouncing back and forth between the 325VDC rails (slightly more with negative drive from the GDT). Most scopes don't appreciate more than 40V or so, unless you're using 10x probes. Grounding either of the DC rails will instantly trip your breaker/RCD due to the massive fault current. Sorry if this part is already obvious, or has already been said.My probe is directly on the gate and the ground of my probe is connected to earth ground through a 1k resistor.
Using an isolating step-down transformer to power the DC bus at a low voltage can help with troubleshooting, as there is no risk of shorting out mains power through your scope.
Re: Voltage rise on gate
furnace, Thu Apr 02 2015, 09:47AM
Thanks Grantx! I'm only scoping low side mosfet. I just fixed the problem by removing the earth ground and connected the variac only on live and neutral. Now the scope signal is clear.
Another question!
Can anyone tell me how to choose the right mosfet or Igbt according to their volt and current raiting, because I keep on blowing irfp460 when I reach 400v at 10 amps. I'd like to use mains full power which is 230v.
furnace, Thu Apr 02 2015, 09:47AM
Thanks Grantx! I'm only scoping low side mosfet. I just fixed the problem by removing the earth ground and connected the variac only on live and neutral. Now the scope signal is clear.
Another question!
Can anyone tell me how to choose the right mosfet or Igbt according to their volt and current raiting, because I keep on blowing irfp460 when I reach 400v at 10 amps. I'd like to use mains full power which is 230v.
Re: Voltage rise on gate
furnace, Wed Apr 08 2015, 05:42PM
Can someone please tel me why my low side mosfet is getting hotter than my high side mosfet on this signal. Are they even suppose to get hot because I figured that the coupling transformer should anticipate all the current and voltage?
furnace, Wed Apr 08 2015, 05:42PM
Can someone please tel me why my low side mosfet is getting hotter than my high side mosfet on this signal. Are they even suppose to get hot because I figured that the coupling transformer should anticipate all the current and voltage?
Re: Voltage rise on gate
furnace, Sat Apr 11 2015, 06:35AM
Can someone please help?
furnace, Sat Apr 11 2015, 06:35AM
Can someone please help?
Print this page