I need to wind a choke for loading a mosfet. However my transformer winder do not have the suffice knowledge on how to go about winding a dc choke. However he is willing to wind it if the calculations are done by me. He recall the size of the lamination would be of the 96 or 111mm variant. He has H14 and Z11 laminations.
I am having trouble finding specifications for the h14 specs i.e. BH,gauss etc with respect to lamination size. I have never done magnetics before but i am willing to learn how. All the transformer winder needs is the turns ratio and air gapped
My design needs to utilize a 100mh at 5A chokes. Can anyone help me by pointing out formulas for turns ratios and air gaps please. Thanks
I am having trouble finding specifications for the h14 specs i.e. BH,gauss etc with respect to lamination size. I have never done magnetics before but i am willing to learn how. All the transformer winder needs is the turns ratio and air gapped
My design needs to utilize a 100mh at 5A chokes. Can anyone help me by pointing out formulas for turns ratios and air gaps please. Thanks
If specs are not available you may have to design the inductor empirically
1- Determine energy storage requirements
E= .5 * L * I^2
E= .5 * .1 * 5^2 = 1,25 Joules
2- Find a suitable combination of core/gap capable of storing that energy before saturation
You'll need a huge core, maybe 10-20Kg to be able to use a small gap and reduce stray magnetic fields that otherwise will induce voltages in signal paths and ruin your amplifier [as an alternative, you can place the inductor several meters away from any small signal path]
Energy storage has little dependence on number of tuns [only due to leakage inductance] so you can test this with the same sample-coil, without worrying to rewind the full 100mH coil each time
I suggest appliying a constant voltage at adjustable length pulses and repetition frequency to the coil while looking at current waveform for saturation knee, leting the coil 'reset' between pulses
Inductance can be seen [and measured] as the DC voltage that applied during 1 sec to a coil causes 1A current variation [assuming negligible coi resistance], so it can be calculated from the current/time waveform obtained before
Having inductance and saturation current for a given gap, you can calculate energy storage and adjust gap lenght [longer gap means higher storage]
3- Find the number of turns to get 100mH
This is very simple kowing that inductance is almost proportional to the number of turns sqared [the error is introduced again by leakage inductance] so knowning actual inductance you can easily calculate number of turns required for the desired inductance
1- Determine energy storage requirements
E= .5 * L * I^2
E= .5 * .1 * 5^2 = 1,25 Joules
2- Find a suitable combination of core/gap capable of storing that energy before saturation
You'll need a huge core, maybe 10-20Kg to be able to use a small gap and reduce stray magnetic fields that otherwise will induce voltages in signal paths and ruin your amplifier [as an alternative, you can place the inductor several meters away from any small signal path]
Energy storage has little dependence on number of tuns [only due to leakage inductance] so you can test this with the same sample-coil, without worrying to rewind the full 100mH coil each time
I suggest appliying a constant voltage at adjustable length pulses and repetition frequency to the coil while looking at current waveform for saturation knee, leting the coil 'reset' between pulses
Inductance can be seen [and measured] as the DC voltage that applied during 1 sec to a coil causes 1A current variation [assuming negligible coi resistance], so it can be calculated from the current/time waveform obtained before
Having inductance and saturation current for a given gap, you can calculate energy storage and adjust gap lenght [longer gap means higher storage]
3- Find the number of turns to get 100mH
This is very simple kowing that inductance is almost proportional to the number of turns sqared [the error is introduced again by leakage inductance] so knowning actual inductance you can easily calculate number of turns required for the desired inductance
NickC: Seems like you have plans fore some serious class A amp, am I right?
Is so, wlecome to the Choked Mosfet´s Club!😀
Is so, wlecome to the Choked Mosfet´s Club!😀
Seat-of-the-pants method. Works for me anyway.
Try the 96mm laminations. Use a 0.5mm gap. Wind on a known number of turns of thick wire and apply 5 volts AC and measure the voltage across it and the current through it. Raise it to 10 volts and measure again. A you go, graph the results on a spreadsheet of volts, amps and impedance (volts divided by amps).
The impedance will remain fairly constant and this will enable you to calculate the inductance seeing you already know the AC frequency. When the impedance on the graph starts to drop at high current then you know the inductor has reached saturation at that particular airgap. The DC saturation current will be 1.414 times this AC RMS saturation current
The inductance varies with the square of the number of turns, so if for example you measured 50mH with the winding you first put on, then for 100mH you would need 1.414 times as many turns.
Try the 96mm laminations. Use a 0.5mm gap. Wind on a known number of turns of thick wire and apply 5 volts AC and measure the voltage across it and the current through it. Raise it to 10 volts and measure again. A you go, graph the results on a spreadsheet of volts, amps and impedance (volts divided by amps).
The impedance will remain fairly constant and this will enable you to calculate the inductance seeing you already know the AC frequency. When the impedance on the graph starts to drop at high current then you know the inductor has reached saturation at that particular airgap. The DC saturation current will be 1.414 times this AC RMS saturation current
The inductance varies with the square of the number of turns, so if for example you measured 50mH with the winding you first put on, then for 100mH you would need 1.414 times as many turns.
THanks for the reply.
Will be my first venture into a mosfet amp. I have many class a amps. Tubes and solid state. Gave up on so low efficient class a. Hence the idea of choke class a, eff should be 40-50 percent which is great. Also this would be my first semi solid state design so do jump in designing
Eva
thanks for the detail/complex testing method. Will try to read up more though
Fuling,
this idea could be applied to a cathode follower with a huge choke loading like 6c33 or lots of 6as7. Just that would be a nice break from my gm70 pp to build a no heater and ugly tube (mosfet). Will probably call this the no heater tube output stage. Driver would be a el34 probably draw high current low B+. I hear that hexfets like the irf540( what i plan to use) have high Cin hence i need a high current driver. No plans for lateral mosfets got some hexfets and want to test them out. I have on hand some 4 pieces of 15mh which would be fine for testing. Will wire them in series plus they are to be used with a 4 ohm speaker which would be fine down to 40hz.
Circlotron
With the air gap of 0.5mm, how many turns to start off for 100mh? The testing is similar to your "first class a". But just to clear up some details, you are injecting ac into the choke? but are you loading the choke with some resistance? How do it? just leave it unloaded but inject AC via a variac and have a DMN inseries for current measuremnent. Will probably used a z11 core, will get the higher henries than standard H14 variant plus only adding about 10-20 percent more in cost only. Trying to avoid microwaves style though. The cost for winding this inductor would be around USD15-20 say 25 for Z11 which is fine. Building a boatanchor for my tube not doing another for a solid state. I think my total tube amp weight will out beat your ss amp (using 4kva of trans)
Will be my first venture into a mosfet amp. I have many class a amps. Tubes and solid state. Gave up on so low efficient class a. Hence the idea of choke class a, eff should be 40-50 percent which is great. Also this would be my first semi solid state design so do jump in designing
Eva
thanks for the detail/complex testing method. Will try to read up more though
Fuling,
this idea could be applied to a cathode follower with a huge choke loading like 6c33 or lots of 6as7. Just that would be a nice break from my gm70 pp to build a no heater and ugly tube (mosfet). Will probably call this the no heater tube output stage. Driver would be a el34 probably draw high current low B+. I hear that hexfets like the irf540( what i plan to use) have high Cin hence i need a high current driver. No plans for lateral mosfets got some hexfets and want to test them out. I have on hand some 4 pieces of 15mh which would be fine for testing. Will wire them in series plus they are to be used with a 4 ohm speaker which would be fine down to 40hz.
Circlotron
With the air gap of 0.5mm, how many turns to start off for 100mh? The testing is similar to your "first class a". But just to clear up some details, you are injecting ac into the choke? but are you loading the choke with some resistance? How do it? just leave it unloaded but inject AC via a variac and have a DMN inseries for current measuremnent. Will probably used a z11 core, will get the higher henries than standard H14 variant plus only adding about 10-20 percent more in cost only. Trying to avoid microwaves style though. The cost for winding this inductor would be around USD15-20 say 25 for Z11 which is fine. Building a boatanchor for my tube not doing another for a solid state. I think my total tube amp weight will out beat your ss amp (using 4kva of trans)
- Status
- Not open for further replies.