Question about inductive spiking

Thread Starter

uraniumhexoflorite

Joined Oct 23, 2016
216
I'm working on building an electric scooter and yesterday the speed controller had a MOSFET failure that I suspect was caused by inductive spiking. This was very surprising to me because I had a schottky diode in parallel with the motor to prevent this type of failure. I took the speed controller apart and tested each of the MOSFETs as well as the diode. The diode was fine and one of the MOSFETs had a short between it's drain and gate. Some of the others appeared to be damaged since their VGD was lower than it was when I installed them. Any ideas on what could have caused this? I thought that the diode would prevent this, but clearly I was wrong. The diode I used is a 1N5402. The MOSFETs are P75NF75s. The speed controller contains 6 MOSFETs and runs at 21 kHz. The motor is a universal and I run it at a maximum of about 60 volts (4 12 volt SLAs in series). The motor has a resistance of about 2 ohms and I measured the inductance to be about 5.6 mH with my LCR meter. I'm pretty stumped so any advice is greatly appreciated. Thanks in advance for helping.
 

crutschow

Joined Mar 14, 2008
34,283
If there's a significant length of wire between the controller and the motor, the wire inductance may be sufficient to induce damaging voltage spikes at high motor currents.
To minimize that you can put a diode directly from the MOSFET drain (anode) to the supply voltage.
 

shortbus

Joined Sep 30, 2009
10,045
I don't understand why people continue to use mosfets like the one you chose in multiple units, instead of a single Isotop style that is made to do the job. The also choosing a mosfet of 75V to switch a 60V circuit.
Image of a proper isotop mosfet


1608564464130.png
 

Deleted member 440916

Joined Dec 31, 1969
0
I'm working on building an electric scooter and yesterday the speed controller had a MOSFET failure that I suspect was caused by inductive spiking. This was very surprising to me because I had a schottky diode in parallel with the motor to prevent this type of failure. I took the speed controller apart and tested each of the MOSFETs as well as the diode. The diode was fine and one of the MOSFETs had a short between it's drain and gate. Some of the others appeared to be damaged since their VGD was lower than it was when I installed them. Any ideas on what could have caused this? I thought that the diode would prevent this, but clearly I was wrong. The diode I used is a 1N5402. The MOSFETs are P75NF75s. The speed controller contains 6 MOSFETs and runs at 21 kHz. The motor is a universal and I run it at a maximum of about 60 volts (4 12 volt SLAs in series). The motor has a resistance of about 2 ohms and I measured the inductance to be about 5.6 mH with my LCR meter. I'm pretty stumped so any advice is greatly appreciated. Thanks in advance for helping.
Would it be possible to post a schematic ?
An 1N5402 is not a schottky and indeed rather slow so if it is being used as a flyback diode in conjunction with the mosfet and the latter is turning on fast it's just possible that the diode reverse recovery current is upsetting the mosfet, please take this with a pinch of salt as without a schematic its crystal ball gazing at best :)
 

Thread Starter

uraniumhexoflorite

Joined Oct 23, 2016
216
If there's a significant length of wire between the controller and the motor, the wire inductance may be sufficient to induce damaging voltage spikes at high motor currents.
To minimize that you can put a diode directly from the MOSFET drain (anode) to the supply voltage.
I'll give that a try. There is 2 or 3 feet of wire between the controller and the motor. Do you think the diode on it's own will be sufficient or will additional circuitry be necessary to protect the circuit?
 
Last edited:

Thread Starter

uraniumhexoflorite

Joined Oct 23, 2016
216
Would it be possible to post a schematic ?
An 1N5402 is not a schottky and indeed rather slow so if it is being used as a flyback diode in conjunction with the mosfet and the latter is turning on fast it's just possible that the diode reverse recovery current is upsetting the mosfet, please take this with a pinch of salt as without a schematic its crystal ball gazing at best :)
I'll see if I can draw up a schematic of the speed controller
 

crutschow

Joined Mar 14, 2008
34,283
Do you think the diode on it's own will be sufficient or will additional circuitry be necessary to protect the circuit?
Hard to say for sure, but a Schottky diode directly from the MOSFET drain to the plus supply should better protect against inductive spikes, if that's what's zapping the MOSFET.
 

Thread Starter

uraniumhexoflorite

Joined Oct 23, 2016
216
Here is a basic schematic of the speed controller. It looks like it uses a 555 to turn six MOSFETs on and off. The duty cycle is controlled by a potentiometer that also has a switch built in, which provides power to the 555 itself.
Notes_201222_142221.jpg
 

MisterBill2

Joined Jan 23, 2018
18,176
In addition to a diode, there may also be some value in having a capacitor across the motor circuit, closer to the controller end. Something like a 0.1 Mfd and a 0.47 mfd, rated a couple hundred volts, at least. The concept being to catch that very fast rise time and slow it a bit.
 

michael8

Joined Jan 11, 2015
410
MOSFETs are P75NF75, 6 in parallel, 21 KHz
motor universal, 2 ohms resistance, inductance 5.6 mH (measured)
power about 60 volts (4 * 12 volt SLA batteries)

speed controller powered from battery (60V) via lm317 regulator
555 generates/drives mosfet gates each with R in series to gate
and R from gate to ground (battery negative).

MOSFET datasheet
https://www.st.com/resource/en/datasheet/stb75nf75.pdf

LM555 datasheet:
https://www.ti.com/lit/ds/symlink/lm555.pdf

Maximum input voltage 18 volts.

LM317 datasheet:
https://www.ti.com/lit/ds/symlink/lm317.pdf

The maximum input voltage for the LM317 is only 40 volts,
and the LM317HV version 60 volts.

1N5402 datasheet: (Standard Recovery Rectifier)
https://www.onsemi.com/pub/Collateral/1N5400-D.PDF

Questions:

What is the power supply voltage for the 555?

What are the resistor values on the gates of the MOSFETS?
 

MisterBill2

Joined Jan 23, 2018
18,176
With devices in parallel there is usually some scheme to assure that they share the load equally. Not all devices are identical and so one or more may pass much more current than the others, leading to a failure. That is one very real possible cause, unless the actual system has some means of equalizing that is not shown. In addition, since there are two resistors in each gate drive connection, it is also possible that the mosfets were not getting equal gate drive. That, also, can lead to a device failure because the ratings of one device are exceeded.
You can use one pull-down resistor on the drive line to drain the gate charge and assure shut-off of all the devices.
A low value resistor in each source connection can help to assure some current sharing, but it will reduce efficiency a bit. There is information available about using mosfets in parallel and even some application notes that would be very useful.
 

Deleted member 440916

Joined Dec 31, 1969
0
Here is a basic schematic of the speed controller. It looks like it uses a 555 to turn six MOSFETs on and off. The duty cycle is controlled by a potentiometer that also has a switch built in, which provides power to the 555 itself.
View attachment 225662
OK great diagram thankyou, a number of problems
1/ You didn't show where the 1N5401 is but I assume you have connected it with the cathode to +out and anode to -out, am I right ?
2/ Could you tell me the resistor values in the gate circuit between the 555 and each mosfet ?

Assuming that is the correct circuit the mosfets will be switching EXTREMELY SLOWLY and are very likely to suffer SOA failures.
 

Thread Starter

uraniumhexoflorite

Joined Oct 23, 2016
216
It looks like the LM17 will supply about 8 volts (5600 and 30k ohms). The resistor that connects to the 555's pin 3 is 300 ohms and the resistor that connects to the gate and source of the MOSFET is 10k ohms. The resistors are the same for each MOSFET.
 
Last edited:

michael8

Joined Jan 11, 2015
410
Could we have a link to the speed controller on amazon?

> the LM17 will supply about 8 volts (5600 and 30k ohms).

I can't find a datasheet for a LM17, on the diagram it appears to say
it's a LM317 voltage regulator. If it's a LM317 then there needs to
be more than the output voltage setting resistors as the LM317 maximum
input voltage is 40 volts (60 volts if LM317HV).

5.6K and 30K on the output of a LM317 does sound like 8 volts output.
I'm assuming this is the power for the LM555 and thus the MOSFET gate
drive.

The LM555 output has a very approxmate typical voltage difference between
it's supply and the output of 2 volts. So the MOSFET gate drive is really
6 volts (assuming 100 mA to 200 ma total gate current for all 6 MOSFETS).

The MOSFETs are specified at a Vgs 10 volts. So they will have higher
Rds with only at most 6 volts on the gates.

Each MOSFET has about 3700 pF Ciss (input capacitance) which along
with the 300 ohm gate resistors will slow their switching speed
as will the lower voltage gate driving voltage.

> The diode was fine and one of the MOSFETs had a short between it's
> drain and gate.

So 60 volts from the drain of the shorted MOSFET could feed via
the 300 resisters to the gates of the other MOSFETs exceeding
their +/- 20 volts Vds maximum?
 

MisterBill2

Joined Jan 23, 2018
18,176
Some applications of multiple mosfets do add series diodes to the gate bias scheme to pprevent a failed device from sending an ecess voltage to the other devices. It is important to consider the voltage drop of that diode when designing the base drive and bias circuit capabilities.
It is entirely possible for a short circuit to the gate to damage other devices connected.
 
Top