Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

MOSFET interchange help

Status
Not open for further replies.
M

mrbuell

Member level 1
Joined
Nov 22, 2010
Messages
32
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,514
MOSFET help

Hello, I have an inverter welder that blew up a few MOSFETS. I replaced the MOSFETS with some higher amperage ones, but they won't fire. My scope tells me that the gate is getting a signal,and they are rated the same gate threshold voltage. So why won't my MOSFETS fire?
 
Last edited:

Re: MOSFET help

Should the wave form be the same for the high side and low side gates? My scope is showing only 8 volts to the low side and over 20 volts to the high side gates. I don't see any reason for it though. All the components check out. (Ie diodes, resistors, and capacitors as part of the gate drive circuitry. I'm puzzled.
 

The mosfet bias in measured between the gate and source (Vgs),
if you are measuring the voltage with reference to gnd then you will get higher voltage for the high side mosfet.
Do you know the part number of the original mosets ,also what was the replacement mosfet part number.
8v should be enough to open any mosfet, the Rds depends on the mosfer and may be higher compared to the original mosfet but the output should still be working.

Alex
 
Hi and thank you for your help. The originals were Toshiba K2837 and the new ones STF30NM60. When I use a DMM and using ground reference, I get a 10V difference at the gates on each leg of the bridge. With the scope I get a significant difference between the high and low sides. Should I not be hooking up my scope probe to ground?
 

Because of the fact that all 4 mosfets are N type the driver for the upper side has to drive the gate with a voltage higher than the actual power rail so bootstap circuit is used, like this https://www.edaboard.com/threads/201855/#post849633
This is needed because when the high side mosfet is on the source (mosfet pin) will have a voltage almost equal to the power rail so the gate has to have a voltage about 10v above that so a circuit is used to generate that voltage that is higher than the power rail.

A multimeter will not show correct reading because the driving signal is a pulse wave and may have variable duty cycle too, only an oscilloscope can show it correctly and the actual bias is the difference between the source and gate, the lower side mosfet has the source connected to the gnd so the reading is already correct.
For the upper side if the oscilloscope ground is not floating then you can't connect it to the source, maybe use 2 probes, one to the gate and one to the source and see the difference.
If the voltage there is not high enough you probably have a problem with the bootsrap driver.

Alex
 
If you look at this schematic:
**broken link removed**
Is the bootstrap circuit the diode, resistor, capacitor combination at the gates?
Colin
 

Can you please help me interpret these measurements? With the oscope: A/C setting I get a nearly perfect square wave at 25 volts to all 4 gates. When I flip it to D/C, on the low side I get a square wave with huge spikes, 25 volts A/C at -150 VDC. On the high side I get the square wave 25VAC at 0VDC. (At the gates) The drain on the high side shows 150VDC. Does this still sound like a bootstrap problem? My bootstrap seemed to check out as OK.

Thank you for your help
Colin
 

I don't have experience is service of circuits like this but i will try to help if i can.
In your previous posts you said that the bias was 8v, if the bias is in fact 25v then you will have a problem because your original mosfets had a max Vgs voltage of 30v while the replacements that you have has a max Vgs breakdown voltage of 25v.
I think 25v seems too much for mosfet bias, i think usually about 15v is used but if this is the normal voltage used then STF30NM60 can't be uses in that circuit.
The datasheet of Toshiba K2837 has the Rdson rating for a bias of 10v so i can't imagine that a circuit would use 25v, is there any kind of adjustment for the bias voltage?

Alex

---------- Post added at 23:57 ---------- Previous post was at 23:16 ----------

Is you circuit like the one in the schematic you have provided?
Does it use a IR2110 driver?
If it does check the voltage at pin 3 because it is obviously getting 25v instead of 15 as the typical application voltage.
25v is the max Vcc that can be applied to IR2110, so it should be lower.

Alex
 
Last edited:
Hi, no the schematic is not like the one I'm fixing. On mine, the pulse is sent through a transformer to a circuit of diodes, caps, and resistors to the gates.(I think that this is the bootstrap) There is an identical circuit for each gate. On the low side, the gate signal is negatively biased -150 VDC but on the high side it is not biased at all. I would expect it to be positively biased 150 VDC, no? I was measuring the voltage peak to peak, I guess my gate voltage is actually 10-12 volts. My mistake.
Colin
 

The voltage applied to the gates is a square wave peak to peak, you don't measure the rms value in that case,
biasing is the peak value and if it goes beyond the gate breakdown voltage you will have problems.
Yes normally the offset would be on the high side mosfet,the negative voltage probably has to to with the gnd reference,
, you also have the transformers in the circuit so there in no need for the bootstrap circuit, i think your circuit is similar to this

GDT (Gate Drive Transformer).jpg

But the 25v is still high.

Take a look at **broken link removed**
at page 30 and below, you can see the description of Gate Drive Transformer

Alex
 

Good info, thanks. My oscope probe is a little goofy, so I don't know that I'm getting true readings. Should my low side voltage rail go to ground? There is no continuity to case ground.

Colin
 

Yes,to complete the circute the source of your low side mosfet has to be grounded,sometimes through a very low value of resistors.
 

Here's what I don't understand...the low side isn't actually ground, its -150 VDC. It is a 240 VAC welder, each leg rectified one positive one negative. So, should the low side still be grounded in this case?
 

I finally found a good power electronics book that explained a few things. I know now that I have a transformer coupled push pull gate drive. That would explain the 20 volts I was measuring at the gates. 10 pos, 10 neg. The pwm circuit is working 100%. But I still don't get why the inverter won't work. I have -150 VDC at the low side drain, and +150 VDC at the high side source. The low side gates are biased -150VDC, and the high side gates have no bias whatsoever. From what I read, I should have a +150VDC bias on the highside gates. Is this correct?
 

What type of H-Bridge driver are you using?

---------- Post added at 11:29 ---------- Previous post was at 11:18 ----------

also if you apply 150v at the gate of your mosfet i dont think they will survived,therefore try and sketch the driver section and post it we will find solution to it.
 

Hi. I have been able to put my o-scope on a working version of this inverter circuit. I get the same signal from the driver on both, so that's how I know that the PWM part is working. I have attached a sketch of the circuit for the gate drivers. There is one of these circuits for each mosfet, all coupled to the same transformer. The diodes are 1N4148 and the one I have marked as ZD is a 15 volt zener 1N477A. I rebuilt this circuit board with all the same as original parts except for the Zeners. The originals were marked as C15. I assumed that they were 15 volt Zeners. Could I be mistaken? https://obrazki.elektroda.pl/82_1299094156.jpg

I think I may have made an error identifying the original diodes. They are covered with shellac and very hard to read. They are definitely NOT 1n4148. I can make out ST (manufacturer), 15 and 1n I think. Are these also Zeners? They're glass with the anode marked with a black band.
 
Last edited:

yes if you see c15 at the body of the zener diode it means 15v zener diode and it serve as protection diode,it should work even without the diode but it is use as a regulator.
 

I still can't get this thing to work. Everything checks out fine though. Is it possible that the gates aren't getting enough amperage? The datasheets dont list the amp requirements.
 

Re: MOSFET help

Just a quick suggestion, when you put it all back together did you ensure that all the relevant ground returns were put back? It is possible that the gate drive is connected to the gate but the source return is missing, on a scope it looks like proper gate drive is there due to the capacitance of the mosfets, but in reality - not.

Hope this may help, Regards, Orson Cart
 
Last edited:
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top