I'm interested in building the Big MoFo and using it to drive a compression tweeter in an upcoming 2-way active loudspeaker project. I'm still a little bit unsure (or clueless if you prefer!) about the tradeoffs when using various inductors in the MoFo circuit when and if I only need to reproduce frequencies above about 1kHz, as well as the inductance that is necessary to support that frequency band. Can a lower inductance air core be used? Air core inductors are not typically available in large inductances like 50mH, so if I use another type (e.g. powder core, iron core) are there problems with saturation, etc. of the inductor under the steady DC current that is used? How might distortion be effected by inductor type? Unfortunately I really do not have any insight into these issues at all or how I could investigate them on my own. My primary motivation in exploring options is to try and make the amp lighter and cheaper to build for this particular application. If the best option remains one of the Hammond inductors then that is fine, but I wanted to throw this out there and hopefully learn something from the responses that I receive. Thanks.
The inductor is the load for the mosfet: the higher, the lower distortion. The speaker will be connected in parallel to this load, so you will never get a load higher than your speaker impedance.
To calculate the reactance of the inductor, formula is 2 pi f l. So for the 100mH, the reactance would be about 630 ohms : much higher than the speaker load. So you can go with mayber 10mH. The air core wont saturate, I dont know about the others.
Last week I read your tutorial on the DSP with the raspberry - I am very tempted to try it out and hopefully learn some linux along the way.
Cheers, Erik
To calculate the reactance of the inductor, formula is 2 pi f l. So for the 100mH, the reactance would be about 630 ohms : much higher than the speaker load. So you can go with mayber 10mH. The air core wont saturate, I dont know about the others.
Last week I read your tutorial on the DSP with the raspberry - I am very tempted to try it out and hopefully learn some linux along the way.
Cheers, Erik
Hi Charlie,
The cutoff frequency of the filter defined by the output of the amp can be approximated as:
.2/6.28*L
So, even a .5 or 1mH air core should be workable with your requirements.
If I get the chance this weekend I'll try one and get some THD figures for you. 🙂
The cutoff frequency of the filter defined by the output of the amp can be approximated as:
.2/6.28*L
So, even a .5 or 1mH air core should be workable with your requirements.
If I get the chance this weekend I'll try one and get some THD figures for you. 🙂
Awesome! I was hoping to pick your brain at Burning Amp about this, but that was not in the cards this year. Thanks for the info. I will look for a post about the THD implications in a few days. Thanks for the help, Mike!
Ah, that makes sense. I can see that from the schematic: L1, R5, and the load are all in parallel with C2 providing AC coupling above some cutoff frequency.
While I am looking at the MoFo schematic, what is the purpose of R5? Is this just a "drain" for any accumulated charge on the output cap (C2) or does it help linearize the load impedance seen by Q1?
Maybe for my application a powder core inductor like this would be a possibility?
Jantzen Audio 3.3mH 15 AWG P-Core Inductor Crossover Coil
Low DCR and can probably dissipate heat well. Not sure how these would perform when fed almost 2A of DC current, however...
Will the lower inductance reduce the amount of ripple rejection at the PS switching frequency? The low noise is one of the main reasons why this amp is so attractive to me when the load is a 105dB/W compression driver.
Jantzen Audio 3.3mH 15 AWG P-Core Inductor Crossover Coil
Low DCR and can probably dissipate heat well. Not sure how these would perform when fed almost 2A of DC current, however...
Will the lower inductance reduce the amount of ripple rejection at the PS switching frequency? The low noise is one of the main reasons why this amp is so attractive to me when the load is a 105dB/W compression driver.
Last edited:
Hi Charlie,
Do not worry much about noise with SMPS, my unshielded output transformer can pick up 50Hz interference from domestic appliances nearby, but the level is less than -90dB, at the far end of 20kHz, grass line is quite elevated about 5 dB above the -120dB threshold.
I used a trifilar transmission line air core transformer, straight out of a 100m drum of 3 core electrical cable, no more no less than two 3 way connectors attached at both ends.
The result is stunning, not easy to tell the difference between my SIT amp and my “Reflow MoFo”
(don’t trust me, due to my old age hearing and speakers quality,
)
Well, the coil resonated at 1650Hz with 3.3uF cap, “on line calculator” told me that I have a 2.8mH inductor, that is the way for one who knows only +-x/.
Cheers from Pascal
Do not worry much about noise with SMPS, my unshielded output transformer can pick up 50Hz interference from domestic appliances nearby, but the level is less than -90dB, at the far end of 20kHz, grass line is quite elevated about 5 dB above the -120dB threshold.
I used a trifilar transmission line air core transformer, straight out of a 100m drum of 3 core electrical cable, no more no less than two 3 way connectors attached at both ends.
The result is stunning, not easy to tell the difference between my SIT amp and my “Reflow MoFo”
(don’t trust me, due to my old age hearing and speakers quality,
)Well, the coil resonated at 1650Hz with 3.3uF cap, “on line calculator” told me that I have a 2.8mH inductor, that is the way for one who knows only +-x/.
Cheers from Pascal
Attachments
@MossFed: Wow, very creative there! I like that - using what 'ya got.
Thanks for posting about it.
Thanks for posting about it.
Air core versus Ferrite
It is raining, so I have plenty of time to do the winding.
Got a pot core FX2243 with 3H1 (A13) ferrite material with an AL 8830.
The size of my final transformer is D=45mm and H=30mm
I manage to wind in 30 turns of 50 Ohm coaxial câble.
Trifilar air core transformer // 3.3uF @ 1650Hz L=2.8mH
Coaxial FX2243 transfo // 3.3uF @ 1675Hz L=2.73mH
Pretty close for a fair comparison
1) Bias Iq for best THD+N
Air core 2.5A
Ferrite 1A
2) THD+N for 0.3Vrms input
Air core 0.054% @ -11.1dB output/8 Ohm dummy.
Ferrite 0.30% @ -11.2dB
Pascal
It is raining, so I have plenty of time to do the winding.
Got a pot core FX2243 with 3H1 (A13) ferrite material with an AL 8830.
The size of my final transformer is D=45mm and H=30mm
I manage to wind in 30 turns of 50 Ohm coaxial câble.
Trifilar air core transformer // 3.3uF @ 1650Hz L=2.8mH
Coaxial FX2243 transfo // 3.3uF @ 1675Hz L=2.73mH
Pretty close for a fair comparison
1) Bias Iq for best THD+N
Air core 2.5A
Ferrite 1A
2) THD+N for 0.3Vrms input
Air core 0.054% @ -11.1dB output/8 Ohm dummy.
Ferrite 0.30% @ -11.2dB
Pascal
Attachments
Last edited:
Versus Iron Core
...
A 62 mH dc choke is loaded to the Mosfet Source and coupled via a 20kuF to an 8 Ohm load.
Iq is optimised to 3A, fearless IRFP240 is under 90W of heat dissipation to a manageable level of 51deg C at copper heat spreader.
Close to 1 dB gain compared to the air core circuit.
0.01% of THD+N better than the air core.
Pascal
...
A 62 mH dc choke is loaded to the Mosfet Source and coupled via a 20kuF to an 8 Ohm load.
Iq is optimised to 3A, fearless IRFP240 is under 90W of heat dissipation to a manageable level of 51deg C at copper heat spreader.
Close to 1 dB gain compared to the air core circuit.
0.01% of THD+N better than the air core.
Pascal
Attachments
90W ........ damn ABFAB area
Mate!
Was it a SOA that you referred? ABFAB not 😕
I have been in this forum for a year, enough time to reckonise that your language skill is far superior to many yanks and poms, my hat off to you
90W is sissy per your language.
Down here many crocodile dundees “that is the knife” use kW in his back yard.
Best
Attachments
Current Amplifier
Hi Mr Rothacher,
Lot of useful info I gathered from you, many thanks.
I am on the same Charlie’s path, looking for ferrite core audio transformer.
Dealing with common drain amp topo, I believe that a current transformer is the way to go.
I gather some flat ribbon cables with a hope to build low impedance (4 ~ 12.5 Ohm) transmission lines then to assemble them in a current transformer with ferrite core for my Zeus, an amp small enough to fit inside a 10x5x20 cm case.
Warm regards
Pascal
Hi Mr Rothacher,
Lot of useful info I gathered from you, many thanks.
I am on the same Charlie’s path, looking for ferrite core audio transformer.
Dealing with common drain amp topo, I believe that a current transformer is the way to go.
I gather some flat ribbon cables with a hope to build low impedance (4 ~ 12.5 Ohm) transmission lines then to assemble them in a current transformer with ferrite core for my Zeus, an amp small enough to fit inside a 10x5x20 cm case.
Warm regards
Pascal
Attachments
who even thinks of that ?
that's sort of virtual reality I'm simply not capable to accept
I tried , endured almost entire first season ......... well ...... it seems I'm not smart enough to dig it
that's sort of virtual reality I'm simply not capable to accept
I tried , endured almost entire first season ......... well ...... it seems I'm not smart enough to dig it
- Status
- Not open for further replies.