I love tinkering, but it’s good to have design goals, especially since the laws of physics are, um, well, LAWS. and unforgiving. I’m not sure what you are trying to make. Maybe I just didn’t read through everything well too.
Baxandall puzzle
- Thread starter ruffrecords
- Start date
Help Support GroupDIY Audio Forum:
Hi Guava,
I made the Baxandall tone stage some years ago ,worked well . I always had some residual hum so I often use chokes for smoothing the ht , I had a small screened choke I ended up mounting on the pre , connected into the lead to the anode resistor ,it padded out the impedence seen by the anode, reduced the ht hum and I got an improvement in opperation of the Bax on high frequencies , the 10k load resistor and 1hy choke in series increased the gain and the amount of feedback around the Bax stage ,giving me lower THD and more boost on the high control ,maybe 25 db or more boost ,which isnt needed ,but the extra feedback also lowers the output impedence of the stage even more meaning its better able to drive both the Bax network and whatever load is presented to the output .The way things are that helping hand the choke gives me is only effective above around 1khz ,by increasing the chokes value the gain boost starts to take effect at a lower frequency. Remember almost all of the gain of the Bax when set flat gets used up by feedback , if we increase the gain of our bax with a choke most of that extra gain goes around the feedback loop ,at least when the settings are flat anyway .
What I still dont fully grasp is how input impedence of my bax network is effected by increasing the gain is such a way with a choke ,does input impedence go up or down ? Likewise with triode ul and pentode mode the output impedence of the tube itself will change ,
Anyway if anyone has a bax circuit of their own I'd love to hear your impressions of the choke/resistor load combo ,literally any Lf smoothing choke at all will do for the purposes of a listening test and measurments. To my ears everything the on paper theory say the choke will do, its doing , theres definately much more headroom in the high range ,on nominal signal levels the full 25-30 db boost is available without any sign of harshness or clipping ,where pre choke it could start to get a little mushy at extreme boost , again not that you ever would want 30 db boost ,padding resistors could be inserted into the network in anycase to limit cut or boost to +/-15-20 db and that might provide some small boost in input impedence .
Sorry I did get a little carried away in my own little creative bubble in the last few posts , bottom line is stick a choke in the anode load of your Baxandall EQ and tell me your impressions .
I know many will say oh I can to it better(cheaper) with silicon ,by adding buffers, CCs's,gyrators etc thats a load of old Blarney if you ask me ,your using your maths skills and an economists outlook and not your ears to design sound equipment . Yeah a choke will cost me a lot compared to silicon that could do a similar job ,but I gain simplicity ,the choke does exactly what it say on the tin ,a single component beats dozens of resistors caps and Ic's anyday . Perhaps the one concession to sillicon I'll make is to include Merlins Screen bias supply , any ideas about hardwiring in an LR8 regulator ;D
Upshot of choke/resistance loading of the bax is now you have the extra voltage swing the choke affords the output on top of another generous dollop of NFb around the loop ,at least over the frequency range the inductance allows ,with the C3g in this configuration I'm easily delivering to 20 k ohms line level unbalanced in self biased Pentode mode , before I used the choke the top end would get wheezy when driven hard into 20kohm, now its sweet as can be and my mixer input distorts before the equaliser does .
I made the Baxandall tone stage some years ago ,worked well . I always had some residual hum so I often use chokes for smoothing the ht , I had a small screened choke I ended up mounting on the pre , connected into the lead to the anode resistor ,it padded out the impedence seen by the anode, reduced the ht hum and I got an improvement in opperation of the Bax on high frequencies , the 10k load resistor and 1hy choke in series increased the gain and the amount of feedback around the Bax stage ,giving me lower THD and more boost on the high control ,maybe 25 db or more boost ,which isnt needed ,but the extra feedback also lowers the output impedence of the stage even more meaning its better able to drive both the Bax network and whatever load is presented to the output .The way things are that helping hand the choke gives me is only effective above around 1khz ,by increasing the chokes value the gain boost starts to take effect at a lower frequency. Remember almost all of the gain of the Bax when set flat gets used up by feedback , if we increase the gain of our bax with a choke most of that extra gain goes around the feedback loop ,at least when the settings are flat anyway .
What I still dont fully grasp is how input impedence of my bax network is effected by increasing the gain is such a way with a choke ,does input impedence go up or down ? Likewise with triode ul and pentode mode the output impedence of the tube itself will change ,
Anyway if anyone has a bax circuit of their own I'd love to hear your impressions of the choke/resistor load combo ,literally any Lf smoothing choke at all will do for the purposes of a listening test and measurments. To my ears everything the on paper theory say the choke will do, its doing , theres definately much more headroom in the high range ,on nominal signal levels the full 25-30 db boost is available without any sign of harshness or clipping ,where pre choke it could start to get a little mushy at extreme boost , again not that you ever would want 30 db boost ,padding resistors could be inserted into the network in anycase to limit cut or boost to +/-15-20 db and that might provide some small boost in input impedence .
Sorry I did get a little carried away in my own little creative bubble in the last few posts , bottom line is stick a choke in the anode load of your Baxandall EQ and tell me your impressions .
I know many will say oh I can to it better(cheaper) with silicon ,by adding buffers, CCs's,gyrators etc thats a load of old Blarney if you ask me ,your using your maths skills and an economists outlook and not your ears to design sound equipment . Yeah a choke will cost me a lot compared to silicon that could do a similar job ,but I gain simplicity ,the choke does exactly what it say on the tin ,a single component beats dozens of resistors caps and Ic's anyday . Perhaps the one concession to sillicon I'll make is to include Merlins Screen bias supply , any ideas about hardwiring in an LR8 regulator ;D
Upshot of choke/resistance loading of the bax is now you have the extra voltage swing the choke affords the output on top of another generous dollop of NFb around the loop ,at least over the frequency range the inductance allows ,with the C3g in this configuration I'm easily delivering to 20 k ohms line level unbalanced in self biased Pentode mode , before I used the choke the top end would get wheezy when driven hard into 20kohm, now its sweet as can be and my mixer input distorts before the equaliser does .
I assume you mean the input impedance of the whole circuit; the Baxendall network itself doesn't change. The input impedance is directly inverse-related to the amount of boost. Since you get a tad more boost with the choke arrangement, the input impedance decreases slightly, but only at max boost.Tubetec said:
It doesn't matter, since the tube operates as a current generator in this current-feedback arrangement.
I guess that’s why exchange off ideas is cool and fun. I’ve just been playing in the hollow sandbox lately 
Seems a Plate choke is the most appropriate post-eq tube arrangement since one may want crazy amount of boost. And the choke will allow for higher swings. Then just get a big Lundahl that can handle 15-20 mA or ElectraPrint May be able to make one. Then a 10:1 output tranny may be in order to tame the output
BTW where’s your output? CF?
Seems a Plate choke is the most appropriate post-eq tube arrangement since one may want crazy amount of boost. And the choke will allow for higher swings. Then just get a big Lundahl that can handle 15-20 mA or ElectraPrint May be able to make one. Then a 10:1 output tranny may be in order to tame the output
BTW where’s your output? CF?
There's no taming to be done; the Baxendall circuit is unity gain. Inserting a step-down xfmr would imply adding a gain stage.guavatone said:
As indicated on the schemo, it's at the side of the Bax network that's opposite the input.
Kingston
Well-known member
I put a variant of a tube "basic bax" in Drive-1: https://groupdiy.com/index.php?topic=41509.0
Directly from Max Robinsons http://www.angelfire.com/electronic/funwithtubes/Amp-Tone.html
Predictable, unity gain, works great. Very utilitarian. And kind of pointless and overkill when done with a tube.
Directly from Max Robinsons http://www.angelfire.com/electronic/funwithtubes/Amp-Tone.html
Predictable, unity gain, works great. Very utilitarian. And kind of pointless and overkill when done with a tube.
Yeah something like Luhndahl 1660 series multi wound choke would be nice , I was thinking much closer ratio output transformer such as Sowter 8940 3:1, bandwidth would be well beyond 100 khz then . 20 kohms is probably the lowest Ill need to drive ,but having a small step down on the output could come in handy sometime for driving mic inputs or other things of lowish impedence , its also likely we could trade some output level for lower impedence without any trouble ,C3g can swing lots of volts with choke coupling .
Ok ,
I can see my load varies quite a bit from low to high frequencies ,at low frequencies its more or less just the 10k resistor the anode see's then as frequency rises the choke impedence kicks in , looks like the rising impedence at higher frequencies causes more feedback and so a droop in the response. Could I add some extra resistance to the high pot in such a way that the central position of the control yeilds a flat response ? or could I tame the high load at high frequencies a bit by bypassing the choke to ht line with a suitable resistance ?
I suppose plotting the frequency response would be a good first move , I do see some of you have sim'd the Bax ,Ive never learned to use spice ,but if the model already exists can the inductor be added ?
I can see my load varies quite a bit from low to high frequencies ,at low frequencies its more or less just the 10k resistor the anode see's then as frequency rises the choke impedence kicks in , looks like the rising impedence at higher frequencies causes more feedback and so a droop in the response. Could I add some extra resistance to the high pot in such a way that the central position of the control yeilds a flat response ? or could I tame the high load at high frequencies a bit by bypassing the choke to ht line with a suitable resistance ?
I suppose plotting the frequency response would be a good first move , I do see some of you have sim'd the Bax ,Ive never learned to use spice ,but if the model already exists can the inductor be added ?
I have to ask. Does it even make sense to do this and expect impedance not to vary without a buffer stage before the bax? The cheap way would be a mosfet buffer. Maybe a mu-follower triode stage
Merlin says in his book to try not to make the resistor bellow the pot larger than the pot . Also the input and output caps of the bax should be the same values
Oh. And then you need to have another output cap and leaver bax as feedback only
Merlin says in his book to try not to make the resistor bellow the pot larger than the pot . Also the input and output caps of the bax should be the same values
Oh. And then you need to have another output cap and leaver bax as feedback only
Sorry Guava I know you dont get the full picture with just the Eq unit, but thanks for the interest all the same .
I have several other valve based preamps transformer coupled ,and other discrete op amp based mic amps that can easily drive the bax , so I just kept it simple . Almost all of my gear is modular so as levels require different units can be used or not .I know something with variable input impedence like a bax would have a buffer in front in any kind of commercial design ,simply so it works with a wide variety of sources , Im the only person who has to use this and I know the limitations as far as input impedence ,so I can work without the buffers .
And yes I know maybe on paper a 10 cent op amp might be better than any valve at all ,both on cost and performance , but what the valve gives me is a more gracious overload ,op amps simply hit maximum and over the space of a handfull of db the sound turns to dirt , a valve does nice things when you hit it with a big signal, an op amp sounds like its ready to die .Typically triodes were spec'd to 5% thd in the old days ,and when its almost pure 2nd harmonic even larger amounts of it arent so fatigueing on the ears , you drive your op amp to 5% thd its a different story ,it will sound attrocious and it could well go up in a puff of smoke . For me and my designs overload characteristics are the prime design factor , too most designers what goes on above max spec'd output there not concerned with in the slightest.
I have several other valve based preamps transformer coupled ,and other discrete op amp based mic amps that can easily drive the bax , so I just kept it simple . Almost all of my gear is modular so as levels require different units can be used or not .I know something with variable input impedence like a bax would have a buffer in front in any kind of commercial design ,simply so it works with a wide variety of sources , Im the only person who has to use this and I know the limitations as far as input impedence ,so I can work without the buffers .
And yes I know maybe on paper a 10 cent op amp might be better than any valve at all ,both on cost and performance , but what the valve gives me is a more gracious overload ,op amps simply hit maximum and over the space of a handfull of db the sound turns to dirt , a valve does nice things when you hit it with a big signal, an op amp sounds like its ready to die .Typically triodes were spec'd to 5% thd in the old days ,and when its almost pure 2nd harmonic even larger amounts of it arent so fatigueing on the ears , you drive your op amp to 5% thd its a different story ,it will sound attrocious and it could well go up in a puff of smoke . For me and my designs overload characteristics are the prime design factor , too most designers what goes on above max spec'd output there not concerned with in the slightest.
That is not possible; a rise in open-loop response indeed increases NFB but not as much as the gain increase. You need to make measurements rather than rely on subjective impressions.Tubetec said:
You can do that; it's commonly done in commercial products.
That would defeat the action of the choke.
That is a necessity at this point.
Check answers #68 & 71 in this thread.
You should; LTspice is quite easy to grab.
Indeed, it can. Are you using an EF86? That's about the only pentode model I have.
OK, I've done the sim with an EF86. The most striking thing is that with the 10k+1H plate load, the open-loop gain is only 11 (21dB), so the LF boost/cut graphs are very assymetrical at +14/-21dB.Tubetec said:
The OL gain rises steadily, starting at 1.5kHz, as predicted, reaching about 38dB at 10kHz. As a result the HF B/C graphs are almost symmetrical at +16/-17 at 10 kHz.
No doubt a pentode with higher Gm would be necessary with the existing values.
Connected in triode mode, the LF OL gain is 18dB and the HF OL gain is 28dB.
The resulting graphs would not be very different, though. About 1dB difference. that's the consequence of operating under significant NFB.
It's late now, so I'll take a rain check for the sim in UL mode.
Thank you very much Abbey
Interesting results ,
The tube I'm using is a C3g , a German postal pentode, its fairly high powered , about 14 ma plate current and a few more on the screen ,low impedence too . Still just to see in the sim what the choke does on an ef86 is good , an extra 20 db margin due to the choke at HF ,thats a hell of a lot of extra punch. Ill be out most of the day tomorrow ,but I'll get a chance to run some signals in soon and report back my findings .I have a standalone sig gen ,maybe room eq wizard could display the frequency response for me. I see what you mean about the small difference in overall gain regardless of pentode ,Ul or Triode , it does make a larger difference to the overload margin ,could be handy for dialing in grit . I have a massive 75 hy 70ma choke , dual coils ,operable in series or parrallel , might be interesting to try ,but its way to big for the chassis the Eq sits on at the moment .
Thank you all for your suggestions and questions , the more the merrier as they say .For anyone who wants more info on choke coupled amplifiers find it here , http://turneraudio.com.au/audiofilterchokes-page3.html
Interesting results ,
The tube I'm using is a C3g , a German postal pentode, its fairly high powered , about 14 ma plate current and a few more on the screen ,low impedence too . Still just to see in the sim what the choke does on an ef86 is good , an extra 20 db margin due to the choke at HF ,thats a hell of a lot of extra punch. Ill be out most of the day tomorrow ,but I'll get a chance to run some signals in soon and report back my findings .I have a standalone sig gen ,maybe room eq wizard could display the frequency response for me. I see what you mean about the small difference in overall gain regardless of pentode ,Ul or Triode , it does make a larger difference to the overload margin ,could be handy for dialing in grit . I have a massive 75 hy 70ma choke , dual coils ,operable in series or parrallel , might be interesting to try ,but its way to big for the chassis the Eq sits on at the moment .
Thank you all for your suggestions and questions , the more the merrier as they say .For anyone who wants more info on choke coupled amplifiers find it here , http://turneraudio.com.au/audiofilterchokes-page3.html
OK, I don't think I can find a workable Spice model for it, so I'll use a workaround (putting several EF86's in parallels).Tubetec said:
I would suggest you use REW as both generator and analyzer; synchronizing REW with a hardware sig gen is a PITN.
BTW: The grid resistor skews the Bax's response; galvanic continuity should be established at the input of the network.
I've done the sim using 7 off EF86 in parallels. LF OL gain is about 38dB, rising up to 55dB at 20kHz.Tubetec said:
As a result the EQ graphs are pretty much as expected, with about 1dB assymetry at LF (+18.5/-19.7dB @ 20Hz)
Excellent idea to use paralelled EF86's to simulate the C3g .
55db overload margin , so over 30 db improvement over the non choke coupled version , that really is astonishing .I can see how it was struggling a bit at full boost in the resistance only loaded version .
Certainly looks like its worth investigating further with bigger chokes . Maybe just for fun try a 75 Hy choke in the sim . Will the benefits in terms of OL and distortion continue to accrue indefinately the bigger I make the choke ? Im just wondering now what kind of output transformer would be able to take advantage of the performance , even the sowter Ac coupled line outputs run out of headroom way to early and surely any subsequent equipment you try and feed a hot signal like that would blow up .
Im very interested in the views of the other heavy hitters in tube design on my bax mod too , free lunches in electronics are few and far between ,but plonking a choke on the anode of the bax seems give a tremendous increase in performance and Im not seeing any real downsides ,well apart from the cost of the mumetal,silicon steel and iron that is .
55db overload margin , so over 30 db improvement over the non choke coupled version , that really is astonishing .I can see how it was struggling a bit at full boost in the resistance only loaded version .
Certainly looks like its worth investigating further with bigger chokes . Maybe just for fun try a 75 Hy choke in the sim . Will the benefits in terms of OL and distortion continue to accrue indefinately the bigger I make the choke ? Im just wondering now what kind of output transformer would be able to take advantage of the performance , even the sowter Ac coupled line outputs run out of headroom way to early and surely any subsequent equipment you try and feed a hot signal like that would blow up .
Im very interested in the views of the other heavy hitters in tube design on my bax mod too , free lunches in electronics are few and far between ,but plonking a choke on the anode of the bax seems give a tremendous increase in performance and Im not seeing any real downsides ,well apart from the cost of the mumetal,silicon steel and iron that is .
Just did it. The choke action is felt much earlier (about 20Hz) and reach a plateau at 200Hz. What happens is that the Bax network loads the plate, thus putting a limit to the OLG at about 55dB.Tubetec said:
Not really; what happens is that the Bax network loads the plate, thus putting a limit to the OLG at about 55dB.
The EQ graphs are better with less than 1dB assymetry.
Now, the simulation is done with an ideal inductor; using a real-world choke with large stray capacitance and non-linearities would probably seriously impair the actual performance.
Thanks Abbey,
Thats greatly increased my understanding of choke coupling with heavy feedback .
I suppose really you want any drop off in overload margin to be well below the audio spectrum , so hundreds of henries is probably the ball park figure required . I understand too that to give good HF performance I need to sectionalise the choke windings . I see two or three equal sections side by side on the bobbins in tab/tele V series chokes , I read somewhere the more sections the more the self capicitance can be reduced. If nessesary frequency response/load can be extend to hundreds of khz .
It seems from reading Turner audio's page up you need to caclulate the wire diameter on the basis of the full fault current that could flow in the event of an interelectrode short but seeing as theres a 10k resistor in the anode that should effectively limit things and allow a finer wire and more more turns and more inductance . Large multi would anode chokes dont come cheap ,but sectionalising a bobbin and winding two or more coils is way less work than making an output transformer. There is a guy about 20 miles away with a winding shop , I did contact him before about getting stuff done and although audio wasnt really his thing ,he was very helpfull , I could get him to do it ,and maybe even give me a small tutorial on winding while he's doing it .I see some Lundahl anode chokes are bifilar wound and can be connected either series or parallel and either one coil section reversed or normal in relation to the other in such way as the currents cancel or add in the core . I cant figure out which would be the best coil arrangement ,but in any case if I have two symetrical bobbins with multiple coils on each and a means to adjust the gap on test to suit the current, very very good performance could be achieved and Id have options as far as series, parallel and reverse orientations of the individual coils available too . Maybe external mumetal screening on the choke could be minimised if everything is nice and symetrical .
I'm starting to see now why many dismiss choke loading , its as extravagant as you could possibly be ,and probably the most costly topology of all, their being both choke and output transformer required
I was out chugging beer and listening to one of my fav blues bands last night , the ledgendary ' Hot Guitars' of Cork, Ireland
Damn there nearly all in well into their 70's and still kicking monumental amounts of ass on stage everytime they play . Up close and personal bar room blues ,just how I like it .
I cant face the work bench and measurments and high voltage today Abbey , I'd only end up making an entry in the shock log ;D
Thats greatly increased my understanding of choke coupling with heavy feedback .
I suppose really you want any drop off in overload margin to be well below the audio spectrum , so hundreds of henries is probably the ball park figure required . I understand too that to give good HF performance I need to sectionalise the choke windings . I see two or three equal sections side by side on the bobbins in tab/tele V series chokes , I read somewhere the more sections the more the self capicitance can be reduced. If nessesary frequency response/load can be extend to hundreds of khz .
It seems from reading Turner audio's page up you need to caclulate the wire diameter on the basis of the full fault current that could flow in the event of an interelectrode short but seeing as theres a 10k resistor in the anode that should effectively limit things and allow a finer wire and more more turns and more inductance . Large multi would anode chokes dont come cheap ,but sectionalising a bobbin and winding two or more coils is way less work than making an output transformer. There is a guy about 20 miles away with a winding shop , I did contact him before about getting stuff done and although audio wasnt really his thing ,he was very helpfull , I could get him to do it ,and maybe even give me a small tutorial on winding while he's doing it .I see some Lundahl anode chokes are bifilar wound and can be connected either series or parallel and either one coil section reversed or normal in relation to the other in such way as the currents cancel or add in the core . I cant figure out which would be the best coil arrangement ,but in any case if I have two symetrical bobbins with multiple coils on each and a means to adjust the gap on test to suit the current, very very good performance could be achieved and Id have options as far as series, parallel and reverse orientations of the individual coils available too . Maybe external mumetal screening on the choke could be minimised if everything is nice and symetrical .
I'm starting to see now why many dismiss choke loading , its as extravagant as you could possibly be ,and probably the most costly topology of all, their being both choke and output transformer required
I was out chugging beer and listening to one of my fav blues bands last night , the ledgendary ' Hot Guitars' of Cork, Ireland
Damn there nearly all in well into their 70's and still kicking monumental amounts of ass on stage everytime they play . Up close and personal bar room blues ,just how I like it .
I cant face the work bench and measurments and high voltage today Abbey , I'd only end up making an entry in the shock log ;D
Assuming two coils, each of inductance L.Tubetec said:
Parallels in-phase results in same inductance L, half the DCR, reduced leakage inductance, increased stray capacitance, total DC current vs. saturation unchanged (the core does not care if the emf comes from one or several windings).
Series in-phase results in inductance 4L, twice the DCR, increased leakage inductance, increased stray capacitance, DC current vs. saturation halved.
All in-phase combinations result in very low inductance.
However the arrangement can be used as a common-mode filter, with one coil inserted in each leg of a balanced signal.
Hi Abbey and thanks for info on choke windings.
I found some Hitachi AMCC32 cores , they come as a set for building a large output transformer , 8 used C sections ,and mounting bracket/fixings etc. cost about 90 euros
Ive contacted the seller to see if he has a vertical mounts for pairs of amcc32 cores and suitable dual bobbins .
I see a very high end company in Japan going under the Finemet tradename builds their chokes the exact same way ,dual vertical bobbins on Hitachi metglass cores . On ebay several places are offering used cores of this type quite cheaply ,
there does appear to be somekind of degrading of properties of metglass over time and with use , I also dont know anything about the origin ,do they come out of some weird machine thats shoots sub atomic particles or Teslas of electromagnetic energy , I have no clue . They do seem to come in smaller sizes too which might be interesting for making signal transformers
It looks like an easy job to get two coil sections wound side by side in one go across the width of the bobbin giving a total of four identical sections. I see in a typical V72 anode choke three sections are used , I presume in phase/ in series 1-2-3 and same winding direction , although I must find one of CJ's old teardowns of anode load chokes from V72 to get the lowdown.
To get good cancellation and symmetry relative to electromagnetic and best balance of incidental capacitance between the multiple coils is there a best strategy to use ? the more sections I make ,potentially the better the capacitance cancels . Could reverse winding one coil relative to the other on each bobbin (which I assume would flip the phase)then series connection in phase as regards signal ,but hum bucking with regards induced hum ?
There does seem to be software available from Hitachi specifically for calculating core flux and winding parameters for metglass , I tried the software but wasnt fully sure what I was doing and I got errors ,
http://hitachimetals.metglas.com/
I dont know which software module above to choose to design my inductor , and I'm just generating garbage.
Maybe CJ might have some ideas on how to roll C cores to best effect . 8)
I found some Hitachi AMCC32 cores , they come as a set for building a large output transformer , 8 used C sections ,and mounting bracket/fixings etc. cost about 90 euros
Ive contacted the seller to see if he has a vertical mounts for pairs of amcc32 cores and suitable dual bobbins .
I see a very high end company in Japan going under the Finemet tradename builds their chokes the exact same way ,dual vertical bobbins on Hitachi metglass cores . On ebay several places are offering used cores of this type quite cheaply ,
there does appear to be somekind of degrading of properties of metglass over time and with use , I also dont know anything about the origin ,do they come out of some weird machine thats shoots sub atomic particles or Teslas of electromagnetic energy , I have no clue . They do seem to come in smaller sizes too which might be interesting for making signal transformers
It looks like an easy job to get two coil sections wound side by side in one go across the width of the bobbin giving a total of four identical sections. I see in a typical V72 anode choke three sections are used , I presume in phase/ in series 1-2-3 and same winding direction , although I must find one of CJ's old teardowns of anode load chokes from V72 to get the lowdown.
To get good cancellation and symmetry relative to electromagnetic and best balance of incidental capacitance between the multiple coils is there a best strategy to use ? the more sections I make ,potentially the better the capacitance cancels . Could reverse winding one coil relative to the other on each bobbin (which I assume would flip the phase)then series connection in phase as regards signal ,but hum bucking with regards induced hum ?
There does seem to be software available from Hitachi specifically for calculating core flux and winding parameters for metglass , I tried the software but wasnt fully sure what I was doing and I got errors ,
http://hitachimetals.metglas.com/
I dont know which software module above to choose to design my inductor , and I'm just generating garbage.
Maybe CJ might have some ideas on how to roll C cores to best effect . 8)
Similar threads
