Keyboard Shortcuts
Likes
- Qex
- Messages
Search
|
"Systemizing RF Power Amplifier Design" by Roy Hejhall
Here's the app note referenced in the QEX article. It is much better written and gives a clear explanation of the design considerations. More importantly it describes how to measure the large signal parameters with a VNA using a very simple test fixture. Now that we have the nanoVNA, the technique becomes vastly more useful for the hobbyist.
So no need to find a transistor with a datasheet with specs or guess based on small signal parameters. "Measure twice and cut once." While the app note only treats class C, the measurement technique extrapolates to other amplifier classes very easily in an obvious fashion. By modularizing the matching networks separately from the transistor biasing network one can cover a broad range of use cases with very few fixture modules. The main take away is to use a broad tuning range test fixture and tune it for minimum S11 reflection using a directional bridge. Then remove the transistor and measure the BE and CE port complex impedances. The input and output impedances of the transistor are the complex conjugates of the measured values. From there it's pretty much pencil and paper work, though it also provides the input for a Spice program. Have Fun! Reg |
|
Re: RF amplifier design book recommendations wanted
Here's an Amazon list of candidates. These are not cheap, so I'd like to get things other people have read and like. NB I'm a paper book guy. Introduction to RF Power Amplifier Design and Simulation Abdullah Eroglu Advanced Techniques in RF Power Amplifier Design Steve C. Cripps Load-Pull Techniques with Applications to Power Amplifier Design Fadhel M. Ghannouchi & Mohammad S. Hashmi RF and Microwave Power Amplifier Design Andrei Grebennikov Linear and Switch-Mode RF Power Amplifiers: Design and Implementation Methods Abdullah Eroglu Modeling and Design Techniques for RF Power Amplifiers Arvind Raghavan The Load-pull Method of RF and Microwave Power Amplifier Design John F. Sevic Practical RF Amplifier Design and Performance Optimization with SPICE and Load- and Source-pull Techniques Amal Banerjee Advanced Design Techniques for RF Power Amplifiers Anna N. Rudiakova & Vladimir Krizhanovski Switchmode RF and Microwave Power Amplifiers Andrei Grebennikov, Nathan O. Sokal & Marc J. Franco Have Fun! Reg |
|
RF amplifier design book recommendations wanted
The QEX article has highlighted the age of my references on the topic. I'd like to get some more recent monographs.
Does anyone have books they think are particularly good? I am looking for things on a par with "Planar Microwave engineering" by Thomas H. Lee which I consider the best general treatment of RF I've ever read. Lee's enthusiasm for RF is quite infectious. Have Fun! Reg |
|
Re: July/August lead article in QEX
I agree that the editor should make certain the articles are well written. But that's a lot of work for one person. We really need to have a pool of reviewers. Not sure how to make that happen. I also would be happy to pay more and did until they dropped technical articles from QST. At that point I asked to get QEX instead of QST which was of no interest other than the ads. ARRL does not want to recognize the list or operate an alternative. Kai and other ARRL staff are aware of the list and both insisted on the disclaimer on the home page without ever looking at what I had written before I contacted them. Have Fun! Reg On Saturday, July 10, 2021, 12:29:02 PM CDT, Dave Daniel <kc0wjn@...> wrote: That explanation agrees with my thoughts, but editorship should insure that the articles are precise, concise and clearly written. I personally would pay more to have the ¡°full¡± articles appear in the magazine. However, my views apear to be unpopular. I recall making a similar argument to Kai when it was decided to issue the publication with thinner covers. I lost the argument. A notice about this group should probably appear in the magazine. I¡¯m not sure hiw to accomplish that. Perhaps an enail to the current editor would suffice. Hm.. I should do that once I get my computer back up and running. BTW, thank you for creating the group. DaveD |
|
Re: July/August lead article in QEX
¿ªÔÆÌåÓýThat explanation agrees with my thoughts, but editorship should insure that the articles are precise, concise and clearly written. I personally would pay more to have the ¡°full¡± articles appear in the magazine. However, my views apear to be unpopular. I recall making a similar argument to Kai when it was decided to issue the publication with thinner covers. I lost the argument. A notice about this group should probably appear in the magazine. I¡¯m not sure hiw to accomplish that. Perhaps an enail to the current editor would suffice. Hm.. I should do that once I get my computer back up and running. BTW, thank you for creating the group. DaveD On Jul 10, 2021, at 13:08, Reginald Beardsley via groups.io <pulaskite@...> wrote:
|
|
Re: July/August lead article in QEX
I think that's the page limit for an issue at work. The larger problem is a paucity of writers, reviewers and most importantly advertisers. After my last post I noticed the equation for the load is wrong showing a 3rd power of Vcc in the numerator instead of a 2nd power. And many of the figures use fonts that are difficult to read as printed. I had to blow up figure 2 by 300% to get something readable, but the grey background makes the result not very good. However, good technical writing is hard and choosing the boundaries for a short article quite treacherous. You have to say enough, but not too much. And correcting an erro quite a lot of work. I think the decline in DIY construction is in large part due to the intimidating complexity of modern radios and the high cost of T&M kit. My fond hope is that the nanoVNA and tinySA will lead to a resurgence in original designs for basic radios with competitive performance. I can't imagine anyone with a significant interest in RF not buying both the nanoVNA and tinySA from R&L Electronics or other vendors known to not be selling fakes. I have an 8560A, 8593A and 8566B. I bought both in part to see what it was like and in part to support the designers. I plan to buy revised versions of the tinyVNA by Erik as they appear. While the tinySA doesn't have the capability of my HP SAs, it's entirely adequate for HF and VHF design. BTW I very much wish other people would start threads on this list. I'm rather uncomfortable doing it all by myself. I did *not* start this list to hear myself talk. Have Fun! Reg On Saturday, July 10, 2021, 11:35:51 AM CDT, Dave Daniel <kc0wjn@...> wrote: Many QEX articles seem to be ¡ discontinuous ¡ in their presentation. Often there is a note that directs the reader to the ¡°full¡± article online. I¡¯m not sure what to make of this. DaveD On Jul 10, 2021, at 12:27, Reginald Beardsley via groups.io <pulaskite@...> wrote:
|
|
Re: July/August lead article in QEX
¿ªÔÆÌåÓýMany QEX articles seem to be ¡ discontinuous ¡ in their presentation. Often there is a note that directs the reader to the ¡°full¡± article online. I¡¯m not sure what to make of this. DaveD On Jul 10, 2021, at 12:27, Reginald Beardsley via groups.io <pulaskite@...> wrote:
|
|
Re: July/August lead article in QEX
I think that must be the case. The 2N3904 does not appear to be biased for class C operation. I'm going to construct the circuit and make some measurements. The biasing shown in figure 8 is also shown in the 1960's TI book on transistor design followed by the statement it should not be used in a practical design. After reviewing the subject in several monographs I have and reading the article a few more times, I think the issue is poor presentation. The X axes in 9A & 9B don't match nor in 13B & 13C or 15A & 15B. And I've never seen the oval and arrow before nor found any mention of what they signify. Packages are stated in various places as SO-8, SO, SOT, TO-39 and TO-32. And if we're talking about a 2N3866, why the digression with a 2N3904? With the appearance of the nanoVNA, I think a short, well written treatment of measuring the small and large signal parameters of a transistor and using those to design an amplifier would be useful to a lot more people than was the case a few years ago. Now if I can just document the work properly on the first pass ;-) Have Fun! Reg On Saturday, July 10, 2021, 10:36:41 AM CDT, John Marshall KU4AF <johnmars@...> wrote: Perhaps "bias at 20 mA" refers to the device collector current. This seems to be a common way to describe setting an amplifier's the bias point. KU4AF |
|
Re: July/August lead article in QEX
On Fri, 9 Jul 2021 at 15:43, Reginald Beardsley via <pulaskite=[email protected]> wrote:
Reg, I can¡¯t comment on the article as I don¡¯t subscribe to QEX. I used to, and have had a few things published in there, but I don¡¯t any longer. I might reconsider subscribing again.? However, given that you are the admin for the group, you have the option to turn off editing of posts. That¡¯s one of the first things I ever do on any groups I manage. When I set up the HP/Agilent/Keysight list, someone attached a number of large photographs to a post, then edited the post multiple times, so we all received multiple copies of his large photographs. We probably got 100 MB of his emails. It taught me that turning off editing was essential.? On the Keysight forums, I get an email when someone posts, with the contents of the post in the email. But if the author edits the post, the edited version is not emailed to me.? I think we need to accept that ¡°gatherings¡± of people with joint interests is moving away from email and more towards forums. Personally I use email 99% of the time to read or write anything to , but I think in the longer term, web based forums will overtake email. When I look at the HP/Agilent/Keysight list, most members are using email to post. However, I suspect that will change, as older people happier with emails die, and younger members join. Many technical technical forums are now starting up as ? Facebook groups. Dave Dr. David Kirkby, Kirkby Microwave Ltd, drkirkby@... Telephone 01621-680100./ +44 1621 680100 Registered in England & Wales, company number 08914892. Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom |
|
Re: July/August lead article in QEX
I decided I should construct the circuit in figure 8. So I made an enlarged copy to study more closely.
At the top left it states "bias at 20 mA" next to R1 which is 180k. Unless I have completely lost my mind, that requires a 3600 V drop across R1! That's rather difficult to get from a 12 V supply and in any case would blow a 2N3904 to bits in a microsecond. Long ago I had a TO-220 in a repair job go into thermal runaway and blow up. It was quite impressive. It made me decide that protective glasses were needed when working on such things. I'm going to build the circuit as shown and measure it. But GIGO comes to mind when I look at figure 9. Reg |
|
Re: July/August lead article in QEX
I'd like to clarify that I am not looking for help. My question is about the quality of the article. I'm a retired oil industry research scientist/programmer. I've written a couple hundred thousand lines of numerical software and fixed bugs in a couple million lines of other people's code. I can draw Fourier transform pairs on a bar napkin. The figures do not look right and I can find no explanation in the text that might shed light on the matter. From time to time I can be incredibly dense and completely miss what someone is saying. By the same token I've seen many instances of programs producing completely incorrect results which the user did not recognize was wrong. Posting to this list seemed the best way to find out if I was just being especially dumb when I read the paper. Reg |
|
Re: July/August lead article in QEX
One really needs to read the article to understand what I'm getting at. If I didn't know how it worked I'd be very intimidated. I couldn't make any sense of it. It's available on line for ARRL members if you don't get it in print. I was able to persuade ARRL to allow QEX as my print option as QST had become completely uninteresting. I find the appliance era in ham radio very distressing. I don't object to the appliances. What bothers me is the gradual dumbing down of the of the ham community. We have a local club which meets in a different restaurant each month. I offered to provide radios, T&M kit and first class commercial office space gratis for a club station if members were interested in staffing it on weekends to play Elmer to kids and adults. I never had anyone express any interest at all. After 5-6 months I quit going because there was *no* discussion of radio or electronics at the meetings. Just random BS. Have Fun! Reg On Friday, July 9, 2021, 01:06:41 PM CDT, Dave Benson <davek1swl@...> wrote: I don't have the QEX article, so feel free to take this as you will. The small-signal model indeed shows a sinusoidal input to the B-E junction with the 0.6V battery that Jim mentioned.? At one diode-drop forward, however, the junction begins conducting and the current is non-zero.? This is hardly a sinusoidal waveform anymore, and is in fact necessary for the high drive level expected for a class-C output.? The voltage at the device Base is sinusoidal when the B-E is not conducting, and limited to 0.6V positive or somewhat more during conduction.?? There's a second potential source of distortion in that the negative-going peaks may also conduct at the Base-Emitter reverse breakdown voltage.? If I understood Reg's original comment correctly, he surmised that the input waveform couldn't be sinusoidal. In a practical class-C amplifier, it isn't- whether immaterial to the class-C definition or not. .. back in my hole- Dave, K1SWL On Fri, Jul 9, 2021 at 10:43 AM Reginald Beardsley via <pulaskite=[email protected]> wrote:
|
|
Re: July/August lead article in QEX
I don't have the QEX article, so feel free to take this as you will. The small-signal model indeed shows a sinusoidal input to the B-E junction with the 0.6V battery that Jim mentioned.? At one diode-drop forward, however, the junction begins conducting and the current is non-zero.? This is hardly a sinusoidal waveform anymore, and is in fact necessary for the high drive level expected for a class-C output.? The voltage at the device Base is sinusoidal when the B-E is not conducting, and limited to 0.6V positive or somewhat more during conduction.?? There's a second potential source of distortion in that the negative-going peaks may also conduct at the Base-Emitter reverse breakdown voltage.? If I understood Reg's original comment correctly, he surmised that the input waveform couldn't be sinusoidal. In a practical class-C amplifier, it isn't- whether immaterial to the class-C definition or not. .. back in my hole- Dave, K1SWL On Fri, Jul 9, 2021 at 10:43 AM Reginald Beardsley via <pulaskite=[email protected]> wrote:
|
|
Re: July/August lead article in QEX
Hi, Jim, My confusion arose from the article showing the "input waveform" which I interpreted as the BE current and voltage, not the input to the amplifier. Once Vbe drops below ~0.6 V, Ibe drops to nearly zero. As one can feed a sinusoid to any class amplifier, it's difficult for me to grasp the relevance of the input waveform. The distinction among class A, B & C is simply the quiescent bias point. Indeed, Thomas H. Lee in "Planar Microwave Engineering" does not even show inputs, just the collector voltages and currents for the various classes of operation. I still do not understand what the FFTs are supposed to show. BTW This is a mailing list. If you edit a post, everyone gets a 2nd copy. Have Fun! Reg On Friday, July 9, 2021, 07:44:55 AM CDT, James Amos <jimamos@...> wrote: [Edited Message Follows] Hi Reg,To understand that simulation, try looking at some of the transistor models that get used.? ? ? As perhaps you surmised, class C is about they output and not about the input.? ?But, you confusion is with the input, so let's look at the base to emitter circuit for a simple? transistor model.? ?One simple transistor model would have a resistance and a 0.6V battery between the base and emitter pins without any feedback from the collector to emitter.? ?So, a sinusoidal input would have a sinusoidal current waveform, but not have any feedback from the output. More complex transistor models such as those used by spice do have feedback from the output.? ? I would say that diagrams in 9A, if generated with Spice, probably aren't perfect sine waves.? You just can't see the distortion because it is low.? ?The reflection coefficient should give you an idea of how big this feedback is.? If you're looking at Bowick, look at pages 98-102 where he talks about the transistor model, and especially the section on input impedance.? ?? 73's Jim N8CAH |
|
Re: July/August lead article in QEX
Hi Reg,
To understand that simulation, try looking at some of the transistor models that get used.? ? ? As perhaps you surmised, class C is about they output and not about the input.? ?But, you confusion is with the input, so let's look at the base to emitter circuit for a simple? transistor model.? ?One simple transistor model would have a resistance and a 0.6V battery between the base and emitter pins without any feedback from the collector to emitter.? ?So, a sinusoidal input would have a sinusoidal current waveform, but not have any feedback from the output. More complex transistor models such as those used by spice do have feedback from the output.? ? I would say that diagrams in 9A, if generated with Spice, probably aren't perfect sine waves.? You just can't see the distortion because it is low.? ?The reflection coefficient should give you an idea of how big this feedback is.? If you're looking at Bowick, look at pages 98-102 where he talks about the transistor model, and especially the section on input impedance.? ?? 73's Jim N8CAH |
|
Re: July/August lead article in QEX
Sigh... After posting that I realized my conception of class C is dominated by the output, not the input waveforms. Doesn't change the fact I can't figure out what he is trying to say. Reg On Wednesday, July 7, 2021, 11:56:11 AM CDT, Reginald Beardsley via groups.io <pulaskite@...> wrote: I'm curious what others think of this article.? I have read it a few times and find it completely incomprehensible to the point I question the correctness.? In particular, figures 9 and 13 are very hard to square with the central notion of a class C amplifier that it is in cutoff for more than 180 degrees.? If the BE junction is in cutoff, then I'd expect the input current to be a periodic impulse, not the visually perfect sinusoids of figure 9A.? Most troubling of all is figure 13A which purports to show the input current swinging between -58 mA and 89 mA.? Where's the cutoff? I'm a retired geophysicist, not an EE.? I started playing with RF over 50 years ago, but couldn't afford the T&M kit to do serious work.? I'm playing RF hardball for the first time using "RF Circuit Design" by Chris Bowick among numerous other books.? Bowick has the clearest design procedures I've found.? So far I've designed and built CC and CE small signal amplifier examples with CB next in the queue before moving on to power amplifiers.? The Spice implementations I've tried have resisted my efforts.? Thus I have relied on the methods Bowik describes and measurements.? But my goal is to compare design specification, simulation and measured data for each amplifier example. Have Fun! Reg |
|
July/August lead article in QEX
I'm curious what others think of this article. I have read it a few times and find it completely incomprehensible to the point I question the correctness. In particular, figures 9 and 13 are very hard to square with the central notion of a class C amplifier that it is in cutoff for more than 180 degrees. If the BE junction is in cutoff, then I'd expect the input current to be a periodic impulse, not the visually perfect sinusoids of figure 9A. Most troubling of all is figure 13A which purports to show the input current swinging between -58 mA and 89 mA. Where's the cutoff?
I'm a retired geophysicist, not an EE. I started playing with RF over 50 years ago, but couldn't afford the T&M kit to do serious work. I'm playing RF hardball for the first time using "RF Circuit Design" by Chris Bowick among numerous other books. Bowick has the clearest design procedures I've found. So far I've designed and built CC and CE small signal amplifier examples with CB next in the queue before moving on to power amplifiers. The Spice implementations I've tried have resisted my efforts. Thus I have relied on the methods Bowik describes and measurements. But my goal is to compare design specification, simulation and measured data for each amplifier example. Have Fun! Reg |
|
Re: Optimal Crystal Filter Design
It's quite different from Diophantine equations which impose the constraint that only integer values are allowed. That would make no sense in the context of filter design. The advantage is optimal use of the parts you have as opposed to requiring optimal parts which you do not have. I've not yet built an 8-12 pole xtal filter, but everything I've read suggests they get "interesting" when you have that many poles. I bought 100 40 MHz 3rd overtone xtals for $10 long ago with the intent of using them for a high order xtal filter. Have Fun! Reg On Sunday, June 13, 2021, 11:54:20 AM CDT, Mike Feher <n4fs@...> wrote: I believe I had that in a course in Number Theory and as I recall it was called Diophantine equations. Regardless, crystal filters have been around long before linear programing or even ready to use computers for that matter. Yet, they were able to achieve excellent results. Even making filters for 455 KHz using military surplus crystals did a great job. While I appreciate the math and the use of it to optimize, is it necessary, or even needed? 73 ¨C Mike ? Mike B. Feher, N4FS 89 Arnold Blvd. Howell NJ 07731 848-245-9115 ? From: [email protected] <[email protected]> On Behalf Of ebrucehunter via groups.io
Sent: Sunday, June 13, 2021 12:39 PM To: [email protected] Subject: Re: [qex] Optimal Crystal Filter Design ? Reg, ? Your certainly right about the value of Donoho's introduction.? I can now think of Linear Programming as a method of dealing with problems where you have fewer equations than unknowns. ? Thanks again, ? Bruce ? ? ? -----Original Message----- Yes, I am talking about linear programming as developed by Dantzig in the late 40's for solving logistical problems for USAF. Lots of linear algebra, but quite a lot more. It is Traveling Salesman and family territory. On Saturday, June 12, 2021, 02:38:17 PM CDT, ebrucehunter via groups.io <brucekareen@...> wrote: ? ? Reg, ? I apologize for this being off topic.? ? I am afraid I cannot be of any help with the design of crystal filters as my only experience with them has been to use a resistance-loaded 100 MHz crystal to clean up sub-harmonics from a multiplier.? However, when you mentioned "linear programming," a long standing question came to mind.? By linear programming are you referring to the methods utilized for business computations such as finding approximations to the Traveling Salesman problem?? Or you talking about linear algebra. ? I was an engineer in an earlier life and later went back to school for a business degree.? There we learned to use linear programming methods -? first computational and graphical, then a computer program, to solve various business problems and to perform optimizations.? I never really understood the underlying theory behind these business methods as they did not involve matrices, yet the school I attended required competence in linear algebra and calculus for entry. ? Regards, ? Bruce, KG6OJI? ? ? ? -----Original Message----- There is quite a lot of literature and software for designing filters using idealized xtals, but so far as I can find, no one has presented selecting which N? xtals from a set of P xtals will best match a specified N pole response. All the treatments I've found use the N xtals which most closely match the idealized design values. ? A VNA makes measuring the phase and amplitude response of xtals very simple.? And the nanoVNA has made it cheap.? This suggests simply multiplying the transfer functions of each possible? section and using linear programming to solve for the set of N of P xtals which most closely give the desired response.? The general term for this is "basis pursuit". ? Though NP hard on the surface, in fact, the optimal (L0) choice can be computed in L1 time using linear programming.? David Donoho of Stanford proved this in 2004. L0 is combinatorially intractable. L1 is more easily handled,? even very large problems which would be combinatorially impossible. ? Does anyone know of prior work on selecting which of a set of measured xtals will most closely match the required amplitude and phase response? ? Have Fun! Reg ? ? ? ? ? |
|
Re: Optimal Crystal Filter Design
¿ªÔÆÌåÓýI believe I had that in a course in Number Theory and as I recall it was called Diophantine equations. Regardless, crystal filters have been around long before linear programing or even ready to use computers for that matter. Yet, they were able to achieve excellent results. Even making filters for 455 KHz using military surplus crystals did a great job. While I appreciate the math and the use of it to optimize, is it necessary, or even needed? 73 ¨C Mike ? Mike B. Feher, N4FS 89 Arnold Blvd. Howell NJ 07731 848-245-9115 ? From: [email protected] <[email protected]> On Behalf Of ebrucehunter via groups.io
Sent: Sunday, June 13, 2021 12:39 PM To: [email protected] Subject: Re: [qex] Optimal Crystal Filter Design ? Reg, ? Your certainly right about the value of Donoho's introduction.? I can now think of Linear Programming as a method of dealing with problems where you have fewer equations than unknowns. ? Thanks again, ? Bruce ? ? ? -----Original Message----- Yes, I am talking about linear programming as developed by Dantzig in the late 40's for solving logistical problems for USAF. Lots of linear algebra, but quite a lot more. It is Traveling Salesman and family territory. On Saturday, June 12, 2021, 02:38:17 PM CDT, ebrucehunter via groups.io <brucekareen@...> wrote: ? ? Reg, ? I apologize for this being off topic.? ? I am afraid I cannot be of any help with the design of crystal filters as my only experience with them has been to use a resistance-loaded 100 MHz crystal to clean up sub-harmonics from a multiplier.? However, when you mentioned "linear programming," a long standing question came to mind.? By linear programming are you referring to the methods utilized for business computations such as finding approximations to the Traveling Salesman problem?? Or you talking about linear algebra. ? I was an engineer in an earlier life and later went back to school for a business degree.? There we learned to use linear programming methods -? first computational and graphical, then a computer program, to solve various business problems and to perform optimizations.? I never really understood the underlying theory behind these business methods as they did not involve matrices, yet the school I attended required competence in linear algebra and calculus for entry. ? Regards, ? Bruce, KG6OJI? ? ? ? -----Original Message----- There is quite a lot of literature and software for designing filters using idealized xtals, but so far as I can find, no one has presented selecting which N? xtals from a set of P xtals will best match a specified N pole response. All the treatments I've found use the N xtals which most closely match the idealized design values. ? A VNA makes measuring the phase and amplitude response of xtals very simple.? And the nanoVNA has made it cheap.? This suggests simply multiplying the transfer functions of each possible? section and using linear programming to solve for the set of N of P xtals which most closely give the desired response.? The general term for this is "basis pursuit". ? Though NP hard on the surface, in fact, the optimal (L0) choice can be computed in L1 time using linear programming.? David Donoho of Stanford proved this in 2004. L0 is combinatorially intractable. L1 is more easily handled,? even very large problems which would be combinatorially impossible. ? Does anyone know of prior work on selecting which of a set of measured xtals will most closely match the required amplitude and phase response? ? Have Fun! Reg ? ? ? ? ? |