Well, there's two more interested DIYers 😀 Hi John, how's things? Still listening to the ZV7 variant? I'm back to the SE game with Ls. Tonight I'm going to try some THD tests...
What's new with your circuit?
What's new with your circuit?
davidallancole said:
This would be true. But at lower freqs where the L's Xl lowers, it would be even worse. For instance at 10Hz or so the load would be down to 4 ohms or something. Some experimenting with a higher load R might uncover some info? Hmmm
davidallancole said:
I was suprised at the good sound but, I was driving it with F.B and low gain into cheapy speakers and the temptation to turn it up to high, probably got me. Overall the bass appeared to need help! I don't have anything better than a 90db/W/m pair (B20FU Pioneer) and I haven't tried them yet...
I was fooling around in PSPICE and noticed the same thing as the drawing you attached for the waveform. It seems to be an interaction with the inductor and load. The larger the inductor, the less "distortion" there is to the square wave and the low frequency cut off decreases towards 0 Hz. The current gets drawn out of the inductor for a longer period of time at the lower frequencies which decrease the voltage across it, thus decreasing the voltage to the load. Or, the resistance of the inductor gets lower as the current is drawn through it, causing the gain to decrease because the parallel impedance of the coil and load gets lower. Which ever way you want to look at it.
So your best bet for increasing the bass in the low end is using a bigger inductor cause it will be able to store more energy in the low frequencies. This will also decrease the "distortion" in the square wave.
So your best bet for increasing the bass in the low end is using a bigger inductor cause it will be able to store more energy in the low frequencies. This will also decrease the "distortion" in the square wave.
flg said:I got myself thinking 😱 with that square wave, harmonic post up there. Actually describing this stuff is pretty difficult even if you do know what your talking about 😕 But, I think I stumbled in my own mess here...
Trying to use a load as low as 8 ohms on a common source amp like this reduces optimum gain. I have used bulbs and big Rs on this design several months ago and this is true. That's why I said I expected the gain to go up with frequency due to the inductors Xl going up with frequency. But it's real flat, gain wise, from 30Hz on up to a Meg except for the low Q peaking. Like I said, lower -3db @ 6-7Hz! But, the square wave is full of multiple frequencies. And down low, the fundamental and low order harmonics of the Sq Wave could be operating in lower gain while the higher order harmonics are operating in a higher gain... Hence sq wave distortionThe average peak level of the sq wave is the same as the sine wave would be. I'll show a pic... Does this mean there is some frequency dependant gain variation causing the distortion, due to the rising Xl???
Perhaps kind of the reverse, insufficient inductance to maintain the response down to very low frequencies.. If you have some more inductors try adding them in series and see what effect if any that has on the square wave response. I think the inductance might be low enough that you are effectively differentiating the leading edge of your square wave. Insufficient LF response creates the reverse situation from insufficient HF response and from what you describe that is what it sounds like.
As an aside normally you wouldn't want to run an inductor into saturation because at that moment you are left mostly with winding dcr so the AC case should apply. You'll get some pretty odd waveforms depending on how long it takes that inductor to recover.
IMO an aircore inductor ought to work really well here.
You also want all of the audio load current to flow in the load not the inductor so at the lowest frequency of interest you probably want an inductive reactance up to 10X higher than your load impedance to minimize loading.
I agree with Kevin. Add more inductance to lower the cut off, the square wave will look better and there should be more bass present from the amp cause it will be flatter before it ducks off.
I have something a bit embarressing to admit 
The inductor I use is not 80mH, but 2*80mH.....the reason I forgot this, is that I made it for a differential pair type amp...and in that application it's 80mH allright 😉
Now it all makes sense again....you need more inductance.
Magura 🙂
The inductor I use is not 80mH, but 2*80mH.....the reason I forgot this, is that I made it for a differential pair type amp...and in that application it's 80mH allright 😉
Now it all makes sense again....you need more inductance.
Magura 🙂
Well everyone, I want to thank all 10 of you 😀 Is it it 10 yet Magura? Thank you, for the advice, help and suggestions but, we're not done yet...
Yes, I agree, some feedback, more inductance and no core (air) rather than these laminated things would be the next promising step. But, it's just not that simple or affordable.
Feedback - I am not the first one to say, "I want a descent linear performance before the application of "a little" feedback". No, I can't afford the feedback. Without more gain and better linearity, thats just asking for the compilcated problems feedback causes in addition to the need for more than I want to add (gain I'm willing to loose). The very first goal of this project, other than Single ended operation, is to have enough gain to clip the output with 2-3 Volts peak input, while driving an 8 (read 6) ohm speaker to 3 amps or so. Yes, that's probably 24V out and at least 3A Iq. I could be flexible
I think headroom is extremely important though🙄 So, 25WRMS or so is the target. 24V/3V= a gain of 8, after, or including feedback So, all of you who say just add some feedback, yea right, were do I get it from? You might remeber my original schez had an expected gain of about 16. When you use 6db of feedback your gain will end up being 8 😀
More inductance - I'm not done investigating inductance but, N.P.s 1000' of magnet wire suggestion, in the ZV7 paper, is rolling around in my brain... I might be willing to conceed to the "to little" camp, with my measly 120mH, 5A, inductor but, I just haven't been convinced all the distortion I see with a square wave is valid up to 1kHz, due to to little L. The trouble really is, the ones I have, are not cheap! To get one bigger I will have to have it custom made at a few multiples of more $ I believe the ones I have are an air gapped unit. Therefore, possibly I could disassemble them and add some air and reassemble them at the expense of DC, and AC, saturation charactersitics. Basically lower the current and up the inductance. My experience tells me I wont get ther from here...
O.K., what about a mid trebble amp only you say... It's not doing to good at 1k now and to me the mid spectrum begins at about 25oHz
What's left Magura?
Oh yes, Air 
An air core gets you away from the saturation and non linearity of inappropriate core materials. I beleive that could be a benifit. However, a core of any type, silicon steel or whatever, allows the size and cost to be lower. Less wire and DCR, less overall size. My 120mH, 5A unit cost $60.00USD. I knew the risk when I did it but I had to find out for myself. I'm an electronic kinda guy, I'll use them somewhere else if I dont use them here. Hey, I have 50mH chokes on my BOSOZ Preamp PSU (they don't saturate at the 200mA it draws). But anyway, an air core capable of the job, at the cost of many many more turns, which translates to a 1000' or more of wire, is even more expediture. A 1000' of wire has more DCR, unless it's big wire. Big wire is hard to deal with and expensive... If I need more L, how much more L 😕 Some would suggest the typical signal vs load ratio of 1:10 or whatever. So, If I go for 500-1000mH of L How much wire, at what thickness, length and cost will it take 😕 😕 😕
O.K. Before anyone confusses me with Grey Rollins 😀 (Hi Grey) I'll play somore and await further discussion... Thanks, all 10 of you 😀
Yes, I agree, some feedback, more inductance and no core (air) rather than these laminated things would be the next promising step. But, it's just not that simple or affordable.
Feedback - I am not the first one to say, "I want a descent linear performance before the application of "a little" feedback". No, I can't afford the feedback. Without more gain and better linearity, thats just asking for the compilcated problems feedback causes in addition to the need for more than I want to add (gain I'm willing to loose). The very first goal of this project, other than Single ended operation, is to have enough gain to clip the output with 2-3 Volts peak input, while driving an 8 (read 6) ohm speaker to 3 amps or so. Yes, that's probably 24V out and at least 3A Iq. I could be flexible
I think headroom is extremely important though🙄 So, 25WRMS or so is the target. 24V/3V= a gain of 8, after, or including feedback So, all of you who say just add some feedback, yea right, were do I get it from? You might remeber my original schez had an expected gain of about 16. When you use 6db of feedback your gain will end up being 8 😀 More inductance - I'm not done investigating inductance but, N.P.s 1000' of magnet wire suggestion, in the ZV7 paper, is rolling around in my brain... I might be willing to conceed to the "to little" camp, with my measly 120mH, 5A, inductor but, I just haven't been convinced all the distortion I see with a square wave is valid up to 1kHz, due to to little L. The trouble really is, the ones I have, are not cheap! To get one bigger I will have to have it custom made at a few multiples of more $
I believe the ones I have are an air gapped unit. Therefore, possibly I could disassemble them and add some air and reassemble them at the expense of DC, and AC, saturation charactersitics. Basically lower the current and up the inductance. My experience tells me I wont get ther from here... O.K., what about a mid trebble amp only you say... It's not doing to good at 1k now and to me the mid spectrum begins at about 25oHz
What's left Magura?
Oh yes, Air An air core gets you away from the saturation and non linearity of inappropriate core materials. I beleive that could be a benifit. However, a core of any type, silicon steel or whatever, allows the size and cost to be lower. Less wire and DCR, less overall size. My 120mH, 5A unit cost $60.00USD. I knew the risk when I did it but I had to find out for myself. I'm an electronic kinda guy, I'll use them somewhere else if I dont use them here. Hey, I have 50mH chokes on my BOSOZ Preamp PSU (they don't saturate at the 200mA it draws). But anyway, an air core capable of the job, at the cost of many many more turns, which translates to a 1000' or more of wire, is even more expediture. A 1000' of wire has more DCR, unless it's big wire. Big wire is hard to deal with and expensive... If I need more L, how much more L 😕 Some would suggest the typical signal vs load ratio of 1:10 or whatever. So, If I go for 500-1000mH of L How much wire, at what thickness, length and cost will it take 😕 😕 😕 O.K. Before anyone confusses me with Grey Rollins 😀 (Hi Grey) I'll play somore and await further discussion... Thanks, all 10 of you 😀
Naah, 10 persons with interest in this was maybe an overestimation 😉 Though I still wonder where Fuling is hiding....this is right up his alley!
Ok, how about this:
Throw away 6db gain on negative feedback.
Next step is to go buy a roll of 1.9mm magnet wire. That's not gonna break the bank, and allow you to make a pretty hefty air core inductor....mine are 2*80mH and weighs in around 40lbs. That's just about the weight of such a roll of wire, I think the ones I usually buy are like 44lbs, so you will end up with hardly any waste.
You can see mine here, actually they're not all that hard to make.
http://www.briangt.com/gallery/magura-inductors
As the price of copper have gone back down a bit lately, it's sure affordable as well.
Magura 🙂
Ok, how about this:
Throw away 6db gain on negative feedback.
Next step is to go buy a roll of 1.9mm magnet wire. That's not gonna break the bank, and allow you to make a pretty hefty air core inductor....mine are 2*80mH and weighs in around 40lbs. That's just about the weight of such a roll of wire, I think the ones I usually buy are like 44lbs, so you will end up with hardly any waste.
You can see mine here, actually they're not all that hard to make.
http://www.briangt.com/gallery/magura-inductors
As the price of copper have gone back down a bit lately, it's sure affordable as well.
Magura 🙂
Magura said:I have something a bit embarressing to admit
The inductor I use is not 80mH, but 2*80mH.....the reason I forgot this, is that I made it for a differential pair type amp...and in that application it's 80mH allright 😉
Now it all makes sense again....you need more inductance.
Magura 🙂
Ohh, embarressing, smallishing. Do you hear anyone lagphing
Any of the other 8 of us??? I actually have a center tapped unit for a project like that. Possibly several. Thanks N.P. A Carpenter circuit, or a ZV7T??? I just want an S.E. unit first😉 Not a differential/push-pull type of topology...How about the square wave thing dude? do you have a sig gen that does square waves? What does a square wave do in yours?
To be honest, I've wasted most of my evening on typing, thinking, and drinking a few Martini's and I still have to go visit Mom... I wanted to crank up the True RTA and RightMark Audio anylizer and measure some THD tonight... We'll see... Thanks bud 😀
Well square wave looks....ahem....square on mine 😉 I will have to admit that it get's less square when approaching 20Hz, but nothing like what you describe.
I have to admit that I have "wasted" 8db og gain on negative feedback, but then again I had some more to begin with, as I run mine off a 40V supply.
Magura 🙂
I have to admit that I have "wasted" 8db og gain on negative feedback, but then again I had some more to begin with, as I run mine off a 40V supply.
Magura 🙂
Magura said:
How about circlotron or whomever too...
And the guy's from:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=24737&perpage=10&highlight=&pagenumber=1
And:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=10258&perpage=10&pagenumber=1
I've seen your stuff dude. Your a pioneer in this realm
😀
My father left me all the tools I did not already have, including his lathe...
By the time I finish typing, you'll have responded again... I gotta go... for now... Later
just in case - just idea for test....
you have stereo setup?
right?
then use both chokes in series , for mono test and draw your conclusions from that.........
sorry for off topic ........
you have stereo setup?
right?
then use both chokes in series , for mono test and draw your conclusions from that.........
sorry for off topic ........
Zen Mod, Ugly Idea 😀
Actually, I did. I noticed a very slight improvement with the 2 in series. It certainly looks like more inductance is the answer I'm going to have to agree with.
When testing with bulbs or resistors, I don't see any of this. And the gain was lower. But then, of coarse, I don't get the benifit of the higher efficiency of the inductor in the circuit either...
Actually, I did. I noticed a very slight improvement with the 2 in series. It certainly looks like more inductance is the answer I'm going to have to agree with.
When testing with bulbs or resistors, I don't see any of this. And the gain was lower. But then, of coarse, I don't get the benifit of the higher efficiency of the inductor in the circuit either...
Just a thought...
Biamping solves the too-small-inductor problem by allowing the amplifier to play the higher frequencies where you need less inductance.
edit: oops, I see that you mentioned in an earlier post that your square wave is off up to 1000hz. My idea isn't so good, then. Is it off at 1K hz with air-core?
Biamping solves the too-small-inductor problem by allowing the amplifier to play the higher frequencies where you need less inductance.
edit: oops, I see that you mentioned in an earlier post that your square wave is off up to 1000hz. My idea isn't so good, then. Is it off at 1K hz with air-core?
- Status
- Not open for further replies.