Few questions regarding ferrite toroids
Njubster, Fri Oct 31 2014, 12:13PMHi,
I'm trying to isolate the mosfet (IRFP460) gate from driver circuitry. However, the sources of ferrite toroids in my country are pretty limited, I wasn't able to find any in the electronic stores. I did however manage to scavenge a lot of ferrite from broken devices. Most of the SMPS transformers are gaped, so the best candidates were these few, at least considering their sizes:
I have no idea about any of these, I don't know what these color codes mean, and I have no oscilloscope.
The first one looks like plain ferrite, it's not colored, and I found a bunch of these inside an old printer. Those were used as some kind of filters, mostly power filters, coming from the printer PSU.
The other two were scavenged from an old ATX power supplies. The first one's color is olive green, and the second one is light green with one side painted bluish-green.
My question is, does anyone know whether any of these could work? Are the other two powdered iron or ferrite?
Also, I have the highest hopes in the first one. Would stacking a few of them increase the flux and driving power transferred to the mosfet gate?
Re: Few questions regarding ferrite toroids
Sulaiman, Fri Oct 31 2014, 01:53PM
The ferrite cores used for common mode input filters (two separate isolated windings) are suitable
like the core on the left in your photo's
To be sure, with a multimeter on ohms range there should be some measurable resistance which indicates MnZn .. the good for GDT type
no measurable resistance indicates NiZn which are low permeability .. not good for a gdt.
Sulaiman, Fri Oct 31 2014, 01:53PM
The ferrite cores used for common mode input filters (two separate isolated windings) are suitable
like the core on the left in your photo's
To be sure, with a multimeter on ohms range there should be some measurable resistance which indicates MnZn .. the good for GDT type
no measurable resistance indicates NiZn which are low permeability .. not good for a gdt.
Re: Few questions regarding ferrite toroids
Njubster, Fri Oct 31 2014, 02:21PM
Thank you for your reply.
When touching the multimeter leads the resistance is 0.2Ohm, when measuring the winding on 3 stacked rings it's 0.4Ohm, is that ok, or is it too little?
Njubster, Fri Oct 31 2014, 02:21PM
Thank you for your reply.
When touching the multimeter leads the resistance is 0.2Ohm, when measuring the winding on 3 stacked rings it's 0.4Ohm, is that ok, or is it too little?
Re: Few questions regarding ferrite toroids
dexter, Fri Oct 31 2014, 02:27PM
i used E, RM and toroidal ferrite cores for GDT's and all worked just fine
just make sure there is no gap if E or RM cores are used and also make sure the core permeability is high
dexter, Fri Oct 31 2014, 02:27PM
i used E, RM and toroidal ferrite cores for GDT's and all worked just fine
just make sure there is no gap if E or RM cores are used and also make sure the core permeability is high
Re: Few questions regarding ferrite toroids
Njubster, Fri Oct 31 2014, 04:44PM
OK, I'm going to expand this thread a bit.
First, is the transformer core on the picture below big enough for isolating IRFP460's gate? It's from an old input filter, as Sulaiman suggested, and it has no gap. Only one mosfet is isolated.
Next, could anyone suggest me the best way to amplify the drive signal, and discharge the mosfet gate fast? I drew three options that popped in my mind.
(sorry about the image, threw away the paper before realizing how bad the photo was, so I made it clearer in paint a bit)
The first one (a) and the third one (c) seems OK to me, but I'm probably wrong.
The second one (b) concern's me a bit, I did saw some people use it on GDTs, but shouldn't the stored energy be allowed some time to pass onto the secondary circuit during the OFF time? Or does that only applies to the gaped cores? Wouldn't the PNP transistor take a part of the energy, not giving enough to the mosfet?
Also, what should the transformer ratio be? Is 1:1 good enough?
I know I ask too much, but I'm trying to learn and understand while I'm playing, and any help is greatly appreciated :)
Njubster, Fri Oct 31 2014, 04:44PM
OK, I'm going to expand this thread a bit.
First, is the transformer core on the picture below big enough for isolating IRFP460's gate? It's from an old input filter, as Sulaiman suggested, and it has no gap. Only one mosfet is isolated.
Next, could anyone suggest me the best way to amplify the drive signal, and discharge the mosfet gate fast? I drew three options that popped in my mind.
(sorry about the image, threw away the paper before realizing how bad the photo was, so I made it clearer in paint a bit)
The first one (a) and the third one (c) seems OK to me, but I'm probably wrong.
The second one (b) concern's me a bit, I did saw some people use it on GDTs, but shouldn't the stored energy be allowed some time to pass onto the secondary circuit during the OFF time? Or does that only applies to the gaped cores? Wouldn't the PNP transistor take a part of the energy, not giving enough to the mosfet?
Also, what should the transformer ratio be? Is 1:1 good enough?
I know I ask too much, but I'm trying to learn and understand while I'm playing, and any help is greatly appreciated :)
Re: Few questions regarding ferrite toroids
Sulaiman, Fri Oct 31 2014, 06:56PM
A lot depends upon the duty cycle;
for ANY transformer (or inductor)
winding volt.second integral for positive signals
MUST be the same as for negative signals
or else direct current will build up in the winding and the core will saturate.
So, for example, in 'flyback' mode the ON voltage may be 12V for 90us and OFF for 10us
so the average OFF voltage would have to be 12V x (90/10) = 108V
which would destroy a 50V rated BC337.
Also, in general, it is best to have some resistance in both the primary circuit to limit current if the transistor is turned on for too long
and in the secondary to 'damp' oscillations between transformer inductance and transistor capacitance
and to limit the peak transistor drive current (I doubt a BC337 could over-drive an IRFP460)
in a circuit like your b) diagram you also need series capacitance between transistors and transformer to allow negative as well as positive voltage swings accross the transformer primary (integral of v.t = 0)
SO
what is the duty cycle range for the IRFP460 ?
the cores with low resistance are suitable for GDTs,
I consider ANY low resistance a good sign
(opposite for r.f. ferrites)
Sulaiman, Fri Oct 31 2014, 06:56PM
A lot depends upon the duty cycle;
for ANY transformer (or inductor)
winding volt.second integral for positive signals
MUST be the same as for negative signals
or else direct current will build up in the winding and the core will saturate.
So, for example, in 'flyback' mode the ON voltage may be 12V for 90us and OFF for 10us
so the average OFF voltage would have to be 12V x (90/10) = 108V
which would destroy a 50V rated BC337.
Also, in general, it is best to have some resistance in both the primary circuit to limit current if the transistor is turned on for too long
and in the secondary to 'damp' oscillations between transformer inductance and transistor capacitance
and to limit the peak transistor drive current (I doubt a BC337 could over-drive an IRFP460)
in a circuit like your b) diagram you also need series capacitance between transistors and transformer to allow negative as well as positive voltage swings accross the transformer primary (integral of v.t = 0)
SO
what is the duty cycle range for the IRFP460 ?
the cores with low resistance are suitable for GDTs,
I consider ANY low resistance a good sign
(opposite for r.f. ferrites)
Re: Few questions regarding ferrite toroids
Njubster, Fri Oct 31 2014, 08:22PM
108V? How come?
And how do I determine the capacitor size?
And I'm not sure yet about the duty cycle, still hadn't got that far with the theory part :)
Njubster, Fri Oct 31 2014, 08:22PM
108V? How come?
And how do I determine the capacitor size?
And I'm not sure yet about the duty cycle, still hadn't got that far with the theory part :)
Re: Few questions regarding ferrite toroids
Sulaiman, Sat Nov 01 2014, 10:05AM
To design a GDT you need to have a reasonable idea of the maximum volts.time required to drive the transistor.
If the volt.seconds applied is more than designed for then the core will saturate,
so you could make a GDT with a large number of turns to allow for long on times
but if you have too many turns
the leakage inductance will be too large to drive the transistor at higher frequencies.
So for any GDT there is a range of on times that will work well
but above and below that the GDT will not work well
so you need to have a fairly good idea of what your operating range will be.
Unfortunately it's not practical to make a 'universal' GDT
SO
if you have a specific application in mind, explain it and you will get lots of specific advice
if you want a general purpose any frequency/any duty cycle driver
then I suggest that you use direct drive to the gate, non-isolated, and use an isolated supply for whatever it is that the transistor is driving
(e.g. isolation transformer + rectifier(s) + capacitor etc.)
OR have a separate isolated gate drive supply (e.g. 12V dc) and use an opto-coupler.
Sulaiman, Sat Nov 01 2014, 10:05AM
To design a GDT you need to have a reasonable idea of the maximum volts.time required to drive the transistor.
If the volt.seconds applied is more than designed for then the core will saturate,
so you could make a GDT with a large number of turns to allow for long on times
but if you have too many turns
the leakage inductance will be too large to drive the transistor at higher frequencies.
So for any GDT there is a range of on times that will work well
but above and below that the GDT will not work well
so you need to have a fairly good idea of what your operating range will be.
Unfortunately it's not practical to make a 'universal' GDT
SO
if you have a specific application in mind, explain it and you will get lots of specific advice
if you want a general purpose any frequency/any duty cycle driver
then I suggest that you use direct drive to the gate, non-isolated, and use an isolated supply for whatever it is that the transistor is driving
(e.g. isolation transformer + rectifier(s) + capacitor etc.)
OR have a separate isolated gate drive supply (e.g. 12V dc) and use an opto-coupler.
Print this page