|
On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson"
wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. |
On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson"
wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. |
Tony wrote in message . ..
This is a good learning experience for lots of us out here. Any chance of scanning the printed material and posting , say on a.b.s.e? Nope, sorry. I don't have access to the binaries groups, and in addition, it's copyrighted material. I'm willing to make a very limited number of copies, but not to post it. The articles I have are old HP Journal articles, available in "Inventions of Opportunity," copyright Hewlett-Packard Company, 1983, ISBN 0-9612030-0-5. You may be able to find that in your library. The articles in that book were chosen to represent seminal ideas and products that opened new horizons, so the SRD items are as much for historical interest as for technical info. You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. Cheers, Tom |
Tony wrote in message . ..
This is a good learning experience for lots of us out here. Any chance of scanning the printed material and posting , say on a.b.s.e? Nope, sorry. I don't have access to the binaries groups, and in addition, it's copyrighted material. I'm willing to make a very limited number of copies, but not to post it. The articles I have are old HP Journal articles, available in "Inventions of Opportunity," copyright Hewlett-Packard Company, 1983, ISBN 0-9612030-0-5. You may be able to find that in your library. The articles in that book were chosen to represent seminal ideas and products that opened new horizons, so the SRD items are as much for historical interest as for technical info. You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. Cheers, Tom |
You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. T'aint fair Tom! That URL is FB and I have it added to my library, but when I tried to trick it and deleted the AN-1054.PDF, and retyped AN-983.PDF, it can't find that one. Suppose I could write a little program to check AN-1 and step 1 then find out what if anything I had. Is there some secret code as to how this all works? Regards W4ZCB (I suppoe it wouldn't be a secret anymore if you told me.) |
You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. T'aint fair Tom! That URL is FB and I have it added to my library, but when I tried to trick it and deleted the AN-1054.PDF, and retyped AN-983.PDF, it can't find that one. Suppose I could write a little program to check AN-1 and step 1 then find out what if anything I had. Is there some secret code as to how this all works? Regards W4ZCB (I suppoe it wouldn't be a secret anymore if you told me.) |
"Harold E. Johnson" wrote in message news:YKjfc.38031$_K3.164147@attbi_s53...
You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. T'aint fair Tom! Ja, life's like dat. That URL is FB and I have it added to my library, but when I tried to trick it and deleted the AN-1054.PDF, and retyped AN-983.PDF, it can't find that one. Suppose I could write a little program to check AN-1 and step 1 then find out what if anything I had. Is there some secret code as to how this all works? Well, I just typed a search string about SRDs into Google to find the AN1054 reference. I don't have any magical ideas about finding ALL of them on the web, but when I moved out library a few months ago, I made a point of NOT tossing out the ap notes. There's a chance we have it at work. Now, can I remember to look? Luckily, I didn't find enough stamps to mail the other stuff off to you and John yet, so if I can find it, I can stick that one in too. Best magic I can think of for it at the moment. There are some other SRD-related ones listed in my Communications Components catalog from 1993: AN918, AN928, (AN983), AN984, AN989. It does list HP part numbers for each one. "To order literature, ... 1-800-537-7715" but that's from a long time ago... Cheers, Tom |
"Harold E. Johnson" wrote in message news:YKjfc.38031$_K3.164147@attbi_s53...
You'll find a little more modern technical info in HP AN-1054, available at http://rf.rfglobalnet.com/library/Ap...s/1/An1054.pdf. T'aint fair Tom! Ja, life's like dat. That URL is FB and I have it added to my library, but when I tried to trick it and deleted the AN-1054.PDF, and retyped AN-983.PDF, it can't find that one. Suppose I could write a little program to check AN-1 and step 1 then find out what if anything I had. Is there some secret code as to how this all works? Well, I just typed a search string about SRDs into Google to find the AN1054 reference. I don't have any magical ideas about finding ALL of them on the web, but when I moved out library a few months ago, I made a point of NOT tossing out the ap notes. There's a chance we have it at work. Now, can I remember to look? Luckily, I didn't find enough stamps to mail the other stuff off to you and John yet, so if I can find it, I can stick that one in too. Best magic I can think of for it at the moment. There are some other SRD-related ones listed in my Communications Components catalog from 1993: AN918, AN928, (AN983), AN984, AN989. It does list HP part numbers for each one. "To order literature, ... 1-800-537-7715" but that's from a long time ago... Cheers, Tom |
On Mon, 12 Apr 2004 12:23:09 -0700, John Larkin
wrote: The only distributor-stock SRDs I know of are the M/Acom MA44767, MA44768, MA44769 parts, all SOT-23 and dirt cheap. I think Penstock carries them. The '68 or '69 should be good for multiplication to 2 GHz. For high ratios, an SRD will beat a plain diode by a huge amount. There are lots of appnotes around about using them as multipliers. I have a bunch in stock and can send a few to anybody who wants to play. MPulse microwave used to be pretty good with samples, I have some MP4065 SRDs that work well. Tel # used to be 408 432 1480 Barry lennox |
On Mon, 12 Apr 2004 12:23:09 -0700, John Larkin
wrote: The only distributor-stock SRDs I know of are the M/Acom MA44767, MA44768, MA44769 parts, all SOT-23 and dirt cheap. I think Penstock carries them. The '68 or '69 should be good for multiplication to 2 GHz. For high ratios, an SRD will beat a plain diode by a huge amount. There are lots of appnotes around about using them as multipliers. I have a bunch in stock and can send a few to anybody who wants to play. MPulse microwave used to be pretty good with samples, I have some MP4065 SRDs that work well. Tel # used to be 408 432 1480 Barry lennox |
Well, I just typed a search string about SRDs into Google to find the AN1054 reference. I don't have any magical ideas about finding ALL of them on the web, but when I moved out library a few months ago, I made a point of NOT tossing out the ap notes. There's a chance we have it at work. Now, can I remember to look? DOn't worry about finding AN983, I have a dog-eared copy here. If YOU want a copy, I can do the reverse lend-lease thing. W4ZCB |
Well, I just typed a search string about SRDs into Google to find the AN1054 reference. I don't have any magical ideas about finding ALL of them on the web, but when I moved out library a few months ago, I made a point of NOT tossing out the ap notes. There's a chance we have it at work. Now, can I remember to look? DOn't worry about finding AN983, I have a dog-eared copy here. If YOU want a copy, I can do the reverse lend-lease thing. W4ZCB |
On Thu, 15 Apr 2004 20:12:40 +1200, Barry Lennox
wrote: On Mon, 12 Apr 2004 12:23:09 -0700, John Larkin wrote: The only distributor-stock SRDs I know of are the M/Acom MA44767, MA44768, MA44769 parts, all SOT-23 and dirt cheap. I think Penstock carries them. The '68 or '69 should be good for multiplication to 2 GHz. For high ratios, an SRD will beat a plain diode by a huge amount. There are lots of appnotes around about using them as multipliers. I have a bunch in stock and can send a few to anybody who wants to play. MPulse microwave used to be pretty good with samples, I have some MP4065 SRDs that work well. Tel # used to be 408 432 1480 Barry lennox M-Pulse is good guys; we use some of their faster parts, and they are very helpful. But they apparently don't stock anything, and wire-bond - even samples - to order. So if you only want a few pieces for yourself, and don't intend to place a production order, you'll have to lie to them to get samples. Metelics is similar; they seem to make the fastest SRD (35 ps) you can buy on the open market. Not as friendly as M-Pulse, though. John |
On Thu, 15 Apr 2004 20:12:40 +1200, Barry Lennox
wrote: On Mon, 12 Apr 2004 12:23:09 -0700, John Larkin wrote: The only distributor-stock SRDs I know of are the M/Acom MA44767, MA44768, MA44769 parts, all SOT-23 and dirt cheap. I think Penstock carries them. The '68 or '69 should be good for multiplication to 2 GHz. For high ratios, an SRD will beat a plain diode by a huge amount. There are lots of appnotes around about using them as multipliers. I have a bunch in stock and can send a few to anybody who wants to play. MPulse microwave used to be pretty good with samples, I have some MP4065 SRDs that work well. Tel # used to be 408 432 1480 Barry lennox M-Pulse is good guys; we use some of their faster parts, and they are very helpful. But they apparently don't stock anything, and wire-bond - even samples - to order. So if you only want a few pieces for yourself, and don't intend to place a production order, you'll have to lie to them to get samples. Metelics is similar; they seem to make the fastest SRD (35 ps) you can buy on the open market. Not as friendly as M-Pulse, though. John |
Paul Burridge wrote:
On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". One takes an inductor and places it in the center of a metal cylinder; one end of the inductor attaches to the inner wall (makes electrical connection and acts as support). The capacitance to the walls (and added ends) is the other half. Move an end for fine tuning. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". |
Paul Burridge wrote:
On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". One takes an inductor and places it in the center of a metal cylinder; one end of the inductor attaches to the inner wall (makes electrical connection and acts as support). The capacitance to the walls (and added ends) is the other half. Move an end for fine tuning. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". |
It may be of interest -
Doubling the length and diameter of a solenoid and reducing the number of turns of thicker wire to maintain the same inductance, doubles the Q until radiation loss resistance begins to predominate. And it's a big coil for radiation resistance to predominate. ---- Reg, G4FGQ |
It may be of interest -
Doubling the length and diameter of a solenoid and reducing the number of turns of thicker wire to maintain the same inductance, doubles the Q until radiation loss resistance begins to predominate. And it's a big coil for radiation resistance to predominate. ---- Reg, G4FGQ |
On Fri, 16 Apr 2004 08:19:10 GMT, Robert Baer
wrote: Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". One takes an inductor and places it in the center of a metal cylinder; one end of the inductor attaches to the inner wall (makes electrical connection and acts as support). The capacitance to the walls (and added ends) is the other half. Move an end for fine tuning. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. John |
On Fri, 16 Apr 2004 08:19:10 GMT, Robert Baer
wrote: Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( and efficiency drops pretty fast (Think like a rock) as the multiplication factor goes up. Have you ever actually defined what it is you're trying to do? Some control thing in your 70 MHz band? Or real power for some application? Hard to hit a moving target. Or is that the idea? Yeah, moving target's good. Keep the discussion generalised and it might help others as well. I'm not sure where the 70Mhz figure comes from, but it's a good enough guess by whoever made it. However, the final desired frequency in my particular case is in the region of 40Mhz which will be achieved by mixing down with the output from another oscillator and filtering. A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". One takes an inductor and places it in the center of a metal cylinder; one end of the inductor attaches to the inner wall (makes electrical connection and acts as support). The capacitance to the walls (and added ends) is the other half. Move an end for fine tuning. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. John |
Coaxial ceramic resonators are cool... they are small, extremely
stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? |
Coaxial ceramic resonators are cool... they are small, extremely
stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? |
Robert Baer wrote in message ...
.... A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. Based on an earlier P.Burridge thread, I'd say that's NOT small for him. Of course, you didn't mention the frequency (I'd guess around 10MHz), but in the earlier thread, I was suggesting that he use a coil at 18MHz or so with a Qu around 100, and he didn't seem to like even the rather small size that one could make such a coil. I did it on, um either a .68" OD or .80" OD powdered iron toroid, and that was apparently too big. I also suggested a multi-pole filter which could give the same effective filtering, and could use three small SMT inductors. I gathered even that was too big. And I suppose coaxial ceramic resonators for one-off projects at 18MHz aren't very practical... On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". There seems to be a popular misconception that a helical resonator gives better Q than an unshielded coil and capacitor. One of the key nice things about helical resonators is that they are well shielded...there's extremely little external field. That lets you stack several of them side-by-side, with appropriately chosen coupling apertures between the cavities, to make a nice, compact multi-pole filter. But let's not assign a quality that isn't the the same coil WITHOUT the shield will have a higher Qu, so long as it's not so huge that radiation is a significant loss mechanism, and as Reg suggests, that's BIG for most of the tanks we think about. In the older editions of "Reference Data for Radio Engineers," e.g. the fifth edition, there are some design nomographs for helical resonators in the Transmission Lines chapter. They will give you the Qu. If you find the Qu of the coil in air (see the same book, Fundamentals of Networks chapter, or use Reg's coil program or WAIRCOIL), you'll see that the coil's Qu is higher. And if you look also in the Fund. of Networks chapter, you'll find a graph for the decrease of inductance of a coil when shielded, and you'll find that that almost exactly accounts for the Q lowering: same effective series resistance, but lower inductance, gives lower Q. Is it significant? Well, I think for a typical helical resonator, it's a 15% to 25% lowering. Mainly I want to dispell the notion that a helical resonator is something magic that _raises_ the Q of a given coil, because it's not. It does have some very nice properties, but that just isn't one of them. Early helical resonator reference: W. W. Macalpine and R. O. Schildknecht, "Coaxial Resonators with Helical Inner Conductor," Proc. of the IRE, Dec. 1959 -- almost 45 years ago now. Cheers, Tom |
Robert Baer wrote in message ...
.... A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. Based on an earlier P.Burridge thread, I'd say that's NOT small for him. Of course, you didn't mention the frequency (I'd guess around 10MHz), but in the earlier thread, I was suggesting that he use a coil at 18MHz or so with a Qu around 100, and he didn't seem to like even the rather small size that one could make such a coil. I did it on, um either a .68" OD or .80" OD powdered iron toroid, and that was apparently too big. I also suggested a multi-pole filter which could give the same effective filtering, and could use three small SMT inductors. I gathered even that was too big. And I suppose coaxial ceramic resonators for one-off projects at 18MHz aren't very practical... On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". There seems to be a popular misconception that a helical resonator gives better Q than an unshielded coil and capacitor. One of the key nice things about helical resonators is that they are well shielded...there's extremely little external field. That lets you stack several of them side-by-side, with appropriately chosen coupling apertures between the cavities, to make a nice, compact multi-pole filter. But let's not assign a quality that isn't the the same coil WITHOUT the shield will have a higher Qu, so long as it's not so huge that radiation is a significant loss mechanism, and as Reg suggests, that's BIG for most of the tanks we think about. In the older editions of "Reference Data for Radio Engineers," e.g. the fifth edition, there are some design nomographs for helical resonators in the Transmission Lines chapter. They will give you the Qu. If you find the Qu of the coil in air (see the same book, Fundamentals of Networks chapter, or use Reg's coil program or WAIRCOIL), you'll see that the coil's Qu is higher. And if you look also in the Fund. of Networks chapter, you'll find a graph for the decrease of inductance of a coil when shielded, and you'll find that that almost exactly accounts for the Q lowering: same effective series resistance, but lower inductance, gives lower Q. Is it significant? Well, I think for a typical helical resonator, it's a 15% to 25% lowering. Mainly I want to dispell the notion that a helical resonator is something magic that _raises_ the Q of a given coil, because it's not. It does have some very nice properties, but that just isn't one of them. Early helical resonator reference: W. W. Macalpine and R. O. Schildknecht, "Coaxial Resonators with Helical Inner Conductor," Proc. of the IRE, Dec. 1959 -- almost 45 years ago now. Cheers, Tom |
On Fri, 16 Apr 2004 19:10:18 +0000 (UTC), "Reg Edwards"
wrote: Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John |
On Fri, 16 Apr 2004 19:10:18 +0000 (UTC), "Reg Edwards"
wrote: Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John |
Robert Baer wrote:
Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. That doesn't sound very small. On one extreme, one uses standard LC parts and get fair Qs in small size. You can also use positive feedback (negative resistance) to sharpen Q, if you are somewhat careful or don't mind tweaking. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". The ARRL Handbook has/had a design table for them too. -- Scott ********************************** DIY Piezo-Gyro, PCB Drill Bot & More Soon! http://home.comcast.net/~scottxs/ ********************************** |
Robert Baer wrote:
Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. That doesn't sound very small. On one extreme, one uses standard LC parts and get fair Qs in small size. You can also use positive feedback (negative resistance) to sharpen Q, if you are somewhat careful or don't mind tweaking. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". The ARRL Handbook has/had a design table for them too. -- Scott ********************************** DIY Piezo-Gyro, PCB Drill Bot & More Soon! http://home.comcast.net/~scottxs/ ********************************** |
Reg Edwards wrote:
It may be of interest - Doubling the length and diameter of a solenoid and reducing the number of turns of thicker wire to maintain the same inductance, doubles the Q until radiation loss resistance begins to predominate. And it's a big coil for radiation resistance to predominate. ---- Reg, G4FGQ ....and like i mentioned, put a (cylindrical) shield around it and you still have a high Q and a lower frequency *due to the higher distributed capacitance. Alternately, use it as a slow wave structure in a (resonant) cavity)... (same difference) |
Reg Edwards wrote:
It may be of interest - Doubling the length and diameter of a solenoid and reducing the number of turns of thicker wire to maintain the same inductance, doubles the Q until radiation loss resistance begins to predominate. And it's a big coil for radiation resistance to predominate. ---- Reg, G4FGQ ....and like i mentioned, put a (cylindrical) shield around it and you still have a high Q and a lower frequency *due to the higher distributed capacitance. Alternately, use it as a slow wave structure in a (resonant) cavity)... (same difference) |
Reg Edwards wrote:
Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? The three dog-bone method, perhaps? |
Reg Edwards wrote:
Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? The three dog-bone method, perhaps? |
John Larkin wrote:
On Fri, 16 Apr 2004 19:10:18 +0000 (UTC), "Reg Edwards" wrote: Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John Ringdown is the easist way when Qs are extremely high. |
John Larkin wrote:
On Fri, 16 Apr 2004 19:10:18 +0000 (UTC), "Reg Edwards" wrote: Coaxial ceramic resonators are cool... they are small, extremely stable, and have Qs in the thousands. How do you measure the Qs of resonators in the thousands? Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John Ringdown is the easist way when Qs are extremely high. |
Tom Bruhns wrote:
Robert Baer wrote in message ... ... A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. Based on an earlier P.Burridge thread, I'd say that's NOT small for him. Of course, you didn't mention the frequency (I'd guess around 10MHz), but in the earlier thread, I was suggesting that he use a coil at 18MHz or so with a Qu around 100, and he didn't seem to like even the rather small size that one could make such a coil. I did it on, um either a .68" OD or .80" OD powdered iron toroid, and that was apparently too big. I also suggested a multi-pole filter which could give the same effective filtering, and could use three small SMT inductors. I gathered even that was too big. And I suppose coaxial ceramic resonators for one-off projects at 18MHz aren't very practical... On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". There seems to be a popular misconception that a helical resonator gives better Q than an unshielded coil and capacitor. One of the key nice things about helical resonators is that they are well shielded...there's extremely little external field. That lets you stack several of them side-by-side, with appropriately chosen coupling apertures between the cavities, to make a nice, compact multi-pole filter. But let's not assign a quality that isn't the the same coil WITHOUT the shield will have a higher Qu, so long as it's not so huge that radiation is a significant loss mechanism, and as Reg suggests, that's BIG for most of the tanks we think about. In the older editions of "Reference Data for Radio Engineers," e.g. the fifth edition, there are some design nomographs for helical resonators in the Transmission Lines chapter. They will give you the Qu. If you find the Qu of the coil in air (see the same book, Fundamentals of Networks chapter, or use Reg's coil program or WAIRCOIL), you'll see that the coil's Qu is higher. And if you look also in the Fund. of Networks chapter, you'll find a graph for the decrease of inductance of a coil when shielded, and you'll find that that almost exactly accounts for the Q lowering: same effective series resistance, but lower inductance, gives lower Q. Is it significant? Well, I think for a typical helical resonator, it's a 15% to 25% lowering. Mainly I want to dispell the notion that a helical resonator is something magic that _raises_ the Q of a given coil, because it's not. It does have some very nice properties, but that just isn't one of them. Early helical resonator reference: W. W. Macalpine and R. O. Schildknecht, "Coaxial Resonators with Helical Inner Conductor," Proc. of the IRE, Dec. 1959 -- almost 45 years ago now. Cheers, Tom Yes, but the emphasis was on small size, and a helical resonator allows a goodly shrinkage of volume wihout a corresponding loss large of Q. But if the frequency is low enough, the ferite core method, if properly wound, then becomes a "preferred" solution for small size and high Q. Maybe his requirements are not too realistic? |
Tom Bruhns wrote:
Robert Baer wrote in message ... ... A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. Based on an earlier P.Burridge thread, I'd say that's NOT small for him. Of course, you didn't mention the frequency (I'd guess around 10MHz), but in the earlier thread, I was suggesting that he use a coil at 18MHz or so with a Qu around 100, and he didn't seem to like even the rather small size that one could make such a coil. I did it on, um either a .68" OD or .80" OD powdered iron toroid, and that was apparently too big. I also suggested a multi-pole filter which could give the same effective filtering, and could use three small SMT inductors. I gathered even that was too big. And I suppose coaxial ceramic resonators for one-off projects at 18MHz aren't very practical... On one extreme, one uses standard LC parts and get fair Qs in small size. On the other extreme, one makes a ersonant cavity to get very high Qs at the expense of size. In between there is something that can be called either a "shielded inductor" or a "resonant cavity with slow wave structure". There seems to be a popular misconception that a helical resonator gives better Q than an unshielded coil and capacitor. One of the key nice things about helical resonators is that they are well shielded...there's extremely little external field. That lets you stack several of them side-by-side, with appropriately chosen coupling apertures between the cavities, to make a nice, compact multi-pole filter. But let's not assign a quality that isn't the the same coil WITHOUT the shield will have a higher Qu, so long as it's not so huge that radiation is a significant loss mechanism, and as Reg suggests, that's BIG for most of the tanks we think about. In the older editions of "Reference Data for Radio Engineers," e.g. the fifth edition, there are some design nomographs for helical resonators in the Transmission Lines chapter. They will give you the Qu. If you find the Qu of the coil in air (see the same book, Fundamentals of Networks chapter, or use Reg's coil program or WAIRCOIL), you'll see that the coil's Qu is higher. And if you look also in the Fund. of Networks chapter, you'll find a graph for the decrease of inductance of a coil when shielded, and you'll find that that almost exactly accounts for the Q lowering: same effective series resistance, but lower inductance, gives lower Q. Is it significant? Well, I think for a typical helical resonator, it's a 15% to 25% lowering. Mainly I want to dispell the notion that a helical resonator is something magic that _raises_ the Q of a given coil, because it's not. It does have some very nice properties, but that just isn't one of them. Early helical resonator reference: W. W. Macalpine and R. O. Schildknecht, "Coaxial Resonators with Helical Inner Conductor," Proc. of the IRE, Dec. 1959 -- almost 45 years ago now. Cheers, Tom Yes, but the emphasis was on small size, and a helical resonator allows a goodly shrinkage of volume wihout a corresponding loss large of Q. But if the frequency is low enough, the ferite core method, if properly wound, then becomes a "preferred" solution for small size and high Q. Maybe his requirements are not too realistic? |
Scott Stephens wrote:
Robert Baer wrote: Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. That doesn't sound very small. On one extreme, one uses standard LC parts and get fair Qs in small size. You can also use positive feedback (negative resistance) to sharpen Q, if you are somewhat careful or don't mind tweaking. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". The ARRL Handbook has/had a design table for them too. -- Scott ********************************** DIY Piezo-Gyro, PCB Drill Bot & More Soon! http://home.comcast.net/~scottxs/ ********************************** A resonant cavity for the FM band would be roughly the size of a garbage can (which i stated); 6 inches is slightly smaller, i think. |
Scott Stephens wrote:
Robert Baer wrote: Paul Burridge wrote: On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson" wrote: An 8 MHz filter doesn't have to be physically large Paul, Hi Q coils in that frequency range in compact sizes... they don't seem to go together. :-( A high Q resonant circuit can be rather small. For example, i made a tunable LC with a Q approaching 1000, and it was not the size of a garbage can (resonant cavity); it was about 5 inches tall and about 3 inches in diameter. That doesn't sound very small. On one extreme, one uses standard LC parts and get fair Qs in small size. You can also use positive feedback (negative resistance) to sharpen Q, if you are somewhat careful or don't mind tweaking. Rather ingenious; ther was an IEE paper 20 years(??) ago covering the desigh equations. The terminology used was "Helical resonator". The ARRL Handbook has/had a design table for them too. -- Scott ********************************** DIY Piezo-Gyro, PCB Drill Bot & More Soon! http://home.comcast.net/~scottxs/ ********************************** A resonant cavity for the FM band would be roughly the size of a garbage can (which i stated); 6 inches is slightly smaller, i think. |
On Sat, 17 Apr 2004 07:43:49 GMT, Robert Baer posted
this: John Larkin wrote: Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John Ringdown is the easist way when Qs are extremely high. You must still account for the energy you extract from the circuit in order to measure the ringdown. Even the energy needed to drive a high impedance probe is significant when the Q gets high. IOW, the Q without the probe will be higher than the Q when you insert the probe to measure the Q. Jim |
On Sat, 17 Apr 2004 07:43:49 GMT, Robert Baer posted
this: John Larkin wrote: Well, all the usual methods: resonance width, phase shift, ringdown, stuff like that. I work with gadgets with Qs over 1e9, and people measure them without difficulty. John Ringdown is the easist way when Qs are extremely high. You must still account for the energy you extract from the circuit in order to measure the ringdown. Even the energy needed to drive a high impedance probe is significant when the Q gets high. IOW, the Q without the probe will be higher than the Q when you insert the probe to measure the Q. Jim |
| All times are GMT +1. The time now is 07:11 AM. |
Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
RadioBanter.com