yup.. I had to add a screw to secure PC mount (split bobbin) power transformers back in my old kit business. Too much mass for the solder leads to hold. Not to mention shipping these days is probably a higher G event.
JR
JR
Help Support GroupDIY Audio Forum:
jdbakker
Well-known member
[quote author="mediatechnology"]Any recommendations on a good input common-mode RF choke?[/quote]
Digi-Key have several. Want SMD or thru-hole ?
[quote author="RogerFoote"]Yes, drums! We (when we get time, my son and I) record hand drums, gong etc for our own enjoyment, but like huge drum sounds.[/quote]
Like this guy, you mean ?
JD 'Drums! More drums!' B.
Digi-Key have several. Want SMD or thru-hole ?
[quote author="RogerFoote"]Yes, drums! We (when we get time, my son and I) record hand drums, gong etc for our own enjoyment, but like huge drum sounds.[/quote]
Like this guy, you mean ?
JD 'Drums! More drums!' B.
ulysses
Well-known member
[quote author="mediatechnology"]Any recommendations on a good input common-mode RF choke?[/quote]
I like these ones.
I like these ones.
jdbakker
Well-known member
[quote author="ulysses"][quote author="mediatechnology"]Any recommendations on a good input common-mode RF choke?[/quote]
I like these ones.[/quote]
Those are nice transformers, but useless as common-mode chokes (see this thread for more details on CM chokes). A suitable CM choke would be Digi-Key # 513-1035-1-ND (SMD), or you could just DIY a bifilar choke by winding a few turns of twisted pair wire on a ferrite ring.
Wayne, are you sure you want a CM choke on this thing ? As PRR says in that thread, they may make matters worse due to resonance.
JDB.
I like these ones.[/quote]
Those are nice transformers, but useless as common-mode chokes (see this thread for more details on CM chokes). A suitable CM choke would be Digi-Key # 513-1035-1-ND (SMD), or you could just DIY a bifilar choke by winding a few turns of twisted pair wire on a ferrite ring.
Wayne, are you sure you want a CM choke on this thing ? As PRR says in that thread, they may make matters worse due to resonance.
JDB.
ulysses
Well-known member
[quote author="mediatechnology"]Were you suggesting transformers Justin? I did find these: http://www.jensentransformers.com/as/as072.pdf[/quote]
I was being a wise-ass, because I think an input transformer will choke common-mode RFI quite effectively and offer some other advantages as well (thereby rendering this circuit discussion moot). I don't see how a common-mode choke is any less of an intrusion on your purist signal path than a transformer would be.
But then I saw that PDF you found on Jensen's website and it kind of confused me. At the top of the web page I linked in my last post, they go on and on about exactly this stuff- CMR, RFI rejection, etc and so on. So why do they think it's necessary to put ferrite beads, shunt capacitors, AND a common-mode RF choke in front of their own beautiful transformers? Well, they DO say it's for the "MOST HOSTILE RF environments". (And doesn't the loose tolerance of the ferrite beads and 10pF capacitors screw with the common-mode-itude hitting the choke?)
I have to admit there's a lot I don't know about RF filtering on the front (or back) of an audio circuit, so I'm looking forward to reading your references. I've always felt weird about slapping a thousand pF across an audio input, and I think the ferrite bead makes sense because it'll isolate the shunt cap from the source and allow for a smaller cap to be used. I think there are few environments where you'd need much more than that. I've been around radio stations, and I know some are worse than others but do we all have to wear their bandaids on our gear?
I was being a wise-ass, because I think an input transformer will choke common-mode RFI quite effectively and offer some other advantages as well (thereby rendering this circuit discussion moot). I don't see how a common-mode choke is any less of an intrusion on your purist signal path than a transformer would be.
But then I saw that PDF you found on Jensen's website and it kind of confused me. At the top of the web page I linked in my last post, they go on and on about exactly this stuff- CMR, RFI rejection, etc and so on. So why do they think it's necessary to put ferrite beads, shunt capacitors, AND a common-mode RF choke in front of their own beautiful transformers? Well, they DO say it's for the "MOST HOSTILE RF environments". (And doesn't the loose tolerance of the ferrite beads and 10pF capacitors screw with the common-mode-itude hitting the choke?)
I have to admit there's a lot I don't know about RF filtering on the front (or back) of an audio circuit, so I'm looking forward to reading your references. I've always felt weird about slapping a thousand pF across an audio input, and I think the ferrite bead makes sense because it'll isolate the shunt cap from the source and allow for a smaller cap to be used. I think there are few environments where you'd need much more than that. I've been around radio stations, and I know some are worse than others but do we all have to wear their bandaids on our gear?
[quote author="ulysses"]
I don't see how a common-mode choke is any less of an intrusion on your purist signal path than a transformer would be.
[/quote]
I see an important distinction between an input transformer which is affecting the entire audio bandpass, and an input filter inductance which is pretty much out of the picture at audio frequencies and only introducing significant reactance at much higher frequencies.
I don't know that RF is a huge issue for modern mic preamps, but AM radio can be a PIA because it's not a high enough frequency to easily filter out with simple RCs, but still high enough to be problematic for slower electronics.
I still recall (painfully) squaring away a console in a recording studio that was in the main lobe of a AM tower (960 kHz).
JR
I don't see how a common-mode choke is any less of an intrusion on your purist signal path than a transformer would be.
[/quote]
I see an important distinction between an input transformer which is affecting the entire audio bandpass, and an input filter inductance which is pretty much out of the picture at audio frequencies and only introducing significant reactance at much higher frequencies.
I don't know that RF is a huge issue for modern mic preamps, but AM radio can be a PIA because it's not a high enough frequency to easily filter out with simple RCs, but still high enough to be problematic for slower electronics.
I still recall (painfully) squaring away a console in a recording studio that was in the main lobe of a AM tower (960 kHz).
JR
mikep
Well-known member
interesting. I wonder what jensen liked about that particular coilcraft part. maybe should order a few? that kind of filtering in front of a tranformer is IMO overkill and not very common, but it is not unheard of. Neve did something similar in the 82 series mic amp. I think using a common mode choke is maybe a good idea for transformerless inputs. I have often wondered what the "ultimate" common mode choke is(for a mic inputs), and if it even matters. my brain says: the reactance is outside the audio band. my gut says: don't assume anthing when it comes to magnetics.
mike p
mike p
Indeed the GBW helps to a point, but if differential input voltage is significant it becomes a power bandwidth (or slew rate) issue.
I would be careful about throwing two unshielded GP power inductors in the inputs as they may not treat stray magnetic fields common mode like the dedicated RF filter designs should, so may remove one source of interference while introducing another.
Caveat- this is speculation on my part, I have never tried them this way, so I don't know for a fact they would do this and shielded chassis with remote PS may make this moot. I recall a difficult design (for install market) where I had to make mic input transformers coexist with 70-100V audio output transformers in the same chassis. :sad:
JR
I would be careful about throwing two unshielded GP power inductors in the inputs as they may not treat stray magnetic fields common mode like the dedicated RF filter designs should, so may remove one source of interference while introducing another.
Caveat- this is speculation on my part, I have never tried them this way, so I don't know for a fact they would do this and shielded chassis with remote PS may make this moot. I recall a difficult design (for install market) where I had to make mic input transformers coexist with 70-100V audio output transformers in the same chassis. :sad:
JR
Sweet, any external magnetic field will decode common mode. It's worth note that the Jensen app was for a line level input which suggests the audio current flowing will be modest ( perhaps 10k). IIRC Jensen was pretty big on properly terminating transformer windings.
I would be inclined to put this after the nominal 2K mic termination and even resistive biasing, so it doesn't have to handle any current drawn by those and will only be loaded by device input Z and filter caps. Winding DCR is nice and low so that won't be a noise issue.
It is perhaps worth note that the windings are only passing the audio signal current drawn by any load after it. Transformer action further improves CMR.
I look forward to seeing bench results. I would look for worst case at 20Hz, minimum gain (max input V), but I wouldn't be surprised if results are better than typical bench sources.
JR
I would be inclined to put this after the nominal 2K mic termination and even resistive biasing, so it doesn't have to handle any current drawn by those and will only be loaded by device input Z and filter caps. Winding DCR is nice and low so that won't be a noise issue.
It is perhaps worth note that the windings are only passing the audio signal current drawn by any load after it. Transformer action further improves CMR.
I look forward to seeing bench results. I would look for worst case at 20Hz, minimum gain (max input V), but I wouldn't be surprised if results are better than typical bench sources.
JR
ulysses
Well-known member
A quick Googling of Coilcraft K0065-A turned up this link to a competitor's cross-reference list. The T-13706 looks nice, and there are some other interesting-looking things on their site. Check out the L11600 and L392 ("Low cost encapsulated torroid common mode choke").
Also see the "overstock inventory" list which includes prices.
http://www.rhombus-ind.com/xref/coilcraft/coilc-03.html
http://www.rhombus-ind.com/magcat/hlp-cmc.pdf
http://www.rhombus-ind.com/datashts/l-392.pdf
Also see the "overstock inventory" list which includes prices.
http://www.rhombus-ind.com/xref/coilcraft/coilc-03.html
http://www.rhombus-ind.com/magcat/hlp-cmc.pdf
http://www.rhombus-ind.com/datashts/l-392.pdf
Samuel Groner
Well-known member
I'm planning to use Würth Elektronik WE-CMB S, 744822233, for the next project: www.wuerth-elektronik.de/website/emc/eisos/alg/kat_layout.php?id=109
Samuel
Samuel Groner
Well-known member
In Switzerland I can get them from www.distrelec.com, part no. 335276.
From www.farnell.com I found the following parts which look (without extensive research) suitable:
* 969-4242
* 926-5856
* 752-800
Samuel
That is a fine looking square wave, not that music contains anything remotely like that :grin: .
The tilt may be little effected by the servo and hopefully is dominated by the one pole passive RC HPF in the direct path.
I suspect the avoidance of a capacitor in the gain leg has more to do with this nice response than the input blocking caps which typically aren't working quite as hard.
It might be instructive to to try a few different caps in series with the gain leg to see how they act, should you want some "capacitor" sound. :?:
sweet.
JR
The tilt may be little effected by the servo and hopefully is dominated by the one pole passive RC HPF in the direct path.
I suspect the avoidance of a capacitor in the gain leg has more to do with this nice response than the input blocking caps which typically aren't working quite as hard.
It might be instructive to to try a few different caps in series with the gain leg to see how they act, should you want some "capacitor" sound. :?:
sweet.
JR
This is a fairly esoteric concern and not an obvious problem especially with modern well designed parts, and in light of typical signal levels. If concerned I suspect 3-6 dB of gain would pretty much guarantee keeping the input stage in it's happy space. Balancing source impedance for such a low gain app may not be worth the trouble as any errors are not multiplied by very large noise gain factors.
I would be torn between making the stage inverting and using one of many acceptable jelly bean opamps, or looking at a new uber-tech opamp, perhaps out of curiosity more than design judgment. The inverting app puts the high impedance components in serious for noise analysis so may exhibit more Johnson (thermal) noise than non-inverting, perhaps suggesting use of a larger value cap...( the tweaks never end) :grin:
JR
I would be torn between making the stage inverting and using one of many acceptable jelly bean opamps, or looking at a new uber-tech opamp, perhaps out of curiosity more than design judgment. The inverting app puts the high impedance components in serious for noise analysis so may exhibit more Johnson (thermal) noise than non-inverting, perhaps suggesting use of a larger value cap...( the tweaks never end) :grin:
JR
bcarso
Well-known member
Kind of amazing 6320452 issued. Oh well. They will have some difficulty enforcing it I expect.
bcarso
Well-known member
The situation is pretty dismal. I would like to make some money though that wasn't just so many dollars per hour worked, so I'm considering some patents. They aren't worth the fees and maintnenance though unless you can license them effectively, and it had better be a decent-sized company since some of them have a cynical go-ahead-and-sue-us attitude.
A lot of people say just publish the ideas, and then you will find work and income somehow.
A lot of people say just publish the ideas, and then you will find work and income somehow.
I'm enjoying the project, but I've been messing with pretty much the same concept for years. I just haven't touched an iron to solder on it's behalf so for me it's mostly mental masturbation. Hearing about that project is what brought me here.
I am conflicted by the current state of the patent system and recent developments. I don't know if you have been following the news but the supreme court has passed down a ruling changing a few things including the concept of "obviousness". I have always found obviousness very difficult and am not sure I appreciate their help. It used to be that the test for obviousness was finding some prior art or specific teaching of the subject invention in the literature. Now I guess they will just declare something obvious? Every patent I ever applied for was first rejected as obvious. I think there's a form letter examiners use.
My problem with the two patents you listed is not that they're obvious, but I find it hard to believe they are novel (i.e. they were the first to use that technique).
In my experience, some of the better inventions I've seen, after you understand how they work, appear obvious. I have several patents and some of them are like that, some not. I rather prefer the old "If it was so damn obvious, why hasn't it been done before?" test. Now of course some ideas may not be worth doing, but if they're not worth doing they're not worth patenting so it doesn't much matter.
I imagine if we made PRR a patent examiner he might just declare everything obvious and close up shop..
. I understand that the system is a little wobbly and needs fixing but I'm not sure they're helping. They're taking power away from patent holders to reduce the impact of patent trolls who are disrupting some businesses, but us little guys don't need our patents made weaker... Thanks "other" John Roberts...
JR
I am conflicted by the current state of the patent system and recent developments. I don't know if you have been following the news but the supreme court has passed down a ruling changing a few things including the concept of "obviousness". I have always found obviousness very difficult and am not sure I appreciate their help. It used to be that the test for obviousness was finding some prior art or specific teaching of the subject invention in the literature. Now I guess they will just declare something obvious? Every patent I ever applied for was first rejected as obvious. I think there's a form letter examiners use.
My problem with the two patents you listed is not that they're obvious, but I find it hard to believe they are novel (i.e. they were the first to use that technique).
In my experience, some of the better inventions I've seen, after you understand how they work, appear obvious. I have several patents and some of them are like that, some not. I rather prefer the old "If it was so damn obvious, why hasn't it been done before?" test. Now of course some ideas may not be worth doing, but if they're not worth doing they're not worth patenting so it doesn't much matter.
I imagine if we made PRR a patent examiner he might just declare everything obvious and close up shop..
. I understand that the system is a little wobbly and needs fixing but I'm not sure they're helping. They're taking power away from patent holders to reduce the impact of patent trolls who are disrupting some businesses, but us little guys don't need our patents made weaker... Thanks "other" John Roberts...JR
I'm just not sure there's a commercial product there, unless you use gold plated knobs and do the audio-fool marketing dance. Sure it's better but I used to have a theory about audio advancements.. if you can't hear the difference through a screen door, out in the yard, you're going to have to work awfully hard to sell it.
Perhaps the more appropriate product is DC coupled to an A/D convertor all floating up at phantom voltage with an optocoupled digital output. Maybe use both halfs of a stereo A/D to buy another bit and use some digital trickery on the two polarities to parse out source impedance imbalance separately from and in addition to DC trims. Then you need to go ahead and digitally control the gain too, so maybe use that TI chip unless you are willing to do some more heavy lifting in designing good low noise gain switches.. (maybe more digital trix).
Oops :roll: did I say this out loud on the internet? I guess there goes my patent for that
JR
PS: I'll try to work up some sympathy but it sounds like you were using somebody else's content. Am I reading this wrong?
It appears that royalty fees may kill internet radio with new higher rates. They were barely scraping by before. I don't think I'd want to be within a hundred miles of any of those businesses... We can't be very far from just accessing our own music stash from wherever we are using magic beams (not magic beans).
Perhaps the more appropriate product is DC coupled to an A/D convertor all floating up at phantom voltage with an optocoupled digital output. Maybe use both halfs of a stereo A/D to buy another bit and use some digital trickery on the two polarities to parse out source impedance imbalance separately from and in addition to DC trims. Then you need to go ahead and digitally control the gain too, so maybe use that TI chip unless you are willing to do some more heavy lifting in designing good low noise gain switches.. (maybe more digital trix).
Oops :roll: did I say this out loud on the internet? I guess there goes my patent for that
JR
PS: I'll try to work up some sympathy but it sounds like you were using somebody else's content. Am I reading this wrong?
It appears that royalty fees may kill internet radio with new higher rates. They were barely scraping by before. I don't think I'd want to be within a hundred miles of any of those businesses... We can't be very far from just accessing our own music stash from wherever we are using magic beams (not magic beans).
Yes, by all means continue to pursue.
IMO it (DC coupled) is a clear advancement in the SOTA. The large capacitors in the front end, and the gain leg, more so than in the output are doing some heavy lifting and can have measurable deviations from ideal. It also happens to be a very merchantable (sellable) hook, in that consumers have already been taught that different types or even old caps are bad bad bad... What could be easier to communicate in advertising than "no caps in audio path" ? (Note: You'd have to lose the HPF cap while we all know it's fine.)
However, I repeat don't forget the gold knobs... If I were you I'd finish the design (perhaps less publicly). I haven't published my approach for that, and other reasons (one big one is that my current approach is about three different variants that I can't parse through without bench time, that I'm not ready to spend.). My bench time gets consumed by a less crowded product category. Once you get it happy, try to hook up with one of the handful of prestige names that could capitalize on your baby, ask for a fair royalty (less than we'd all like) and maybe get some good gas/beer money out of the transaction. Plus the psychological pat on the butt of having your product in the world making good sounds.
I would even be tempted to patent (if it wasn't blasted all over the internet). My take on the patent system is that it is easier to get the patent than defend it, but a patent gives you something to sell to a bigger partner. Why would anybody pay you for a published DIY design? Of course now I probably can't patent mine either, but I tested the waters years ago and got the standard NIH response.
I have a short list of things I believe still remain to be improved in mic preamps, and /or features that could be put into a high end unit that might actually do something useful, but this is a mature technology and likely to be eclipsed by some LSI chip with mic level input and WIFI digital output... or some other possible future.
I'm not sure where I read it, maybe here. I think it was Einstein who said "Vision without execution is just hallucination" :wink: I have notebooks full of hallucinations. Life is short.
JR
IMO it (DC coupled) is a clear advancement in the SOTA. The large capacitors in the front end, and the gain leg, more so than in the output are doing some heavy lifting and can have measurable deviations from ideal. It also happens to be a very merchantable (sellable) hook, in that consumers have already been taught that different types or even old caps are bad bad bad... What could be easier to communicate in advertising than "no caps in audio path" ? (Note: You'd have to lose the HPF cap while we all know it's fine.)
However, I repeat don't forget the gold knobs... If I were you I'd finish the design (perhaps less publicly). I haven't published my approach for that, and other reasons (one big one is that my current approach is about three different variants that I can't parse through without bench time, that I'm not ready to spend.). My bench time gets consumed by a less crowded product category. Once you get it happy, try to hook up with one of the handful of prestige names that could capitalize on your baby, ask for a fair royalty (less than we'd all like) and maybe get some good gas/beer money out of the transaction. Plus the psychological pat on the butt of having your product in the world making good sounds.
I would even be tempted to patent (if it wasn't blasted all over the internet). My take on the patent system is that it is easier to get the patent than defend it, but a patent gives you something to sell to a bigger partner. Why would anybody pay you for a published DIY design? Of course now I probably can't patent mine either, but I tested the waters years ago and got the standard NIH response.
I have a short list of things I believe still remain to be improved in mic preamps, and /or features that could be put into a high end unit that might actually do something useful, but this is a mature technology and likely to be eclipsed by some LSI chip with mic level input and WIFI digital output... or some other possible future.
I'm not sure where I read it, maybe here. I think it was Einstein who said "Vision without execution is just hallucination" :wink: I have notebooks full of hallucinations. Life is short.
JR
I have gotten one patent on my my own dime and several on others nickel. It kind of reminds me of that old joke about about a chicken participating in scrambled eggs but being fully committed to fried chicken... :roll:
I am far from turning a profit on my current venture but I felt it was worth being "fully committed". I have worked on enough patents that I found a lawyer who let me do as much of the work myself as I could... I leaned heavily on him (at too much $/hour) to tweak the claims, and make sure I didn't do anything stupid (easy in patent law). One could also represent themselves in the patent process but my take is there's an unwritten dance that lawyers and examiners tango to. If you represent yourself you may get exactly what you paid for.
There is probably a real market with discerning customers for a kit or quasi kit. There is so much bedroom recording going on that there is surely a market, but I still would like to see an integrated digital back end.
FWIW another web community associated with Pro SR generated a public domain design for sub woofer. Much of the design calculus was done by a true industry heavyweight with participation from the community to define target parameters. Now this may not translate because there is so must sweat equity available in building a big speaker box, but there may be same interest here to fine tune feature set, and maybe get some digital expertise to come forward with a compatible optimized back end design.
Or not...
JR
I am far from turning a profit on my current venture but I felt it was worth being "fully committed". I have worked on enough patents that I found a lawyer who let me do as much of the work myself as I could... I leaned heavily on him (at too much $/hour) to tweak the claims, and make sure I didn't do anything stupid (easy in patent law). One could also represent themselves in the patent process but my take is there's an unwritten dance that lawyers and examiners tango to. If you represent yourself you may get exactly what you paid for.
There is probably a real market with discerning customers for a kit or quasi kit. There is so much bedroom recording going on that there is surely a market, but I still would like to see an integrated digital back end.
FWIW another web community associated with Pro SR generated a public domain design for sub woofer. Much of the design calculus was done by a true industry heavyweight with participation from the community to define target parameters. Now this may not translate because there is so must sweat equity available in building a big speaker box, but there may be same interest here to fine tune feature set, and maybe get some digital expertise to come forward with a compatible optimized back end design.
Or not...
JR
