https://www.qorvo.com/products/p/QPD1016 The part could be very useful in wideband/pulse applications [1] but there are no DC curves, no capacitances, no Spice models, just the silly load pull cartoons. [1] except for the price of course.
another RF part rant
Started by ●February 6, 2023
Reply by ●February 6, 20232023-02-06
On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote:> https://www.qorvo.com/products/p/QPD1016 > > The part could be very useful in wideband/pulse applications [1] but > there are no DC curves, no capacitances, no Spice models, just the > silly load pull cartoons. > > [1] except for the price of course.When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. https://www.qsl.net/n9zia/metricom/metricom-900-1.jpg
Reply by ●February 6, 20232023-02-06
On 2023-02-06 12:47, John Larkin wrote:> > https://www.qorvo.com/products/p/QPD1016 > > The part could be very useful in wideband/pulse applications [1] but > there are no DC curves, no capacitances, no Spice models, just the > silly load pull cartoons. > > [1] except for the price of course. >The evil phrase there is "pre-matched'. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
Reply by ●February 6, 20232023-02-06
On Mon, 6 Feb 2023 09:59:15 -0800 (PST), Simon S Aysdie <gwhite@ti.com> wrote:>On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote: >> https://www.qorvo.com/products/p/QPD1016 >> >> The part could be very useful in wideband/pulse applications [1] but >> there are no DC curves, no capacitances, no Spice models, just the >> silly load pull cartoons. >> >> [1] except for the price of course. > >When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. > >https://www.qsl.net/n9zia/metricom/metricom-900-1.jpgWhat was the frequency range? I worked with Metricom for a while, when they still wanted to do electrical metering. My impression, which aligns with this rant, is that they were mostly RF engineers and didn't much care about low stuff like 60 Hz. Smart meters are a giant business now, and other things made their Ricochet idea obsolete.
Reply by ●February 6, 20232023-02-06
On Mon, 6 Feb 2023 13:01:54 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 2023-02-06 12:47, John Larkin wrote: >> >> https://www.qorvo.com/products/p/QPD1016 >> >> The part could be very useful in wideband/pulse applications [1] but >> there are no DC curves, no capacitances, no Spice models, just the >> silly load pull cartoons. >> >> [1] except for the price of course. >> > >The evil phrase there is "pre-matched'. > >Cheers > >Phil HobbsAnd, basically, fiddle with the bias until it works. "Pre-matched" usually means they have tweaked the wire bonds to resonate parasitics in some RF band, which wrecks time-domain behavior. This one claims "DC to 1.7 GHz" and "The device can support pulsed and linear operations" which imply time-domain uses. Maybe they still mean pulsed RF. Maybe I can cage a sample.
Reply by ●February 6, 20232023-02-06
On Mon, 6 Feb 2023 09:59:15 -0800 (PST), Simon S Aysdie <gwhite@ti.com> wrote:>On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote: >> https://www.qorvo.com/products/p/QPD1016 >> >> The part could be very useful in wideband/pulse applications [1] but >> there are no DC curves, no capacitances, no Spice models, just the >> silly load pull cartoons. >> >> [1] except for the price of course. > >When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. > >https://www.qsl.net/n9zia/metricom/metricom-900-1.jpgAre those coaxial ceramic resonators up top? They can be useful in time domain too.
Reply by ●February 6, 20232023-02-06
On Monday, February 6, 2023 at 10:49:15 AM UTC-8, John Larkin wrote:> On Mon, 6 Feb 2023 09:59:15 -0800 (PST), Simon S Aysdie > <gwh...@ti.com> wrote: > > >On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote: > >> https://www.qorvo.com/products/p/QPD1016 > >> > >> The part could be very useful in wideband/pulse applications [1] but > >> there are no DC curves, no capacitances, no Spice models, just the > >> silly load pull cartoons. > >> > >> [1] except for the price of course. > > > >When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. > > > >https://www.qsl.net/n9zia/metricom/metricom-900-1.jpg > What was the frequency range?John, it has been more than 20 years. lol. I mean, it had to cover the 915 ISM band (~3%). It was well wider than that, of course, but "wide" is definitional and I don't recall. There was no particular attempt to make it narrow. Regardless, the Metricom radios were narrowband in fundamental requirements, as are most communications requirements.> I worked with Metricom for a while, when they still wanted to do > electrical metering. My impression, which aligns with this rant, is > that they were mostly RF engineers and didn't much care about low > stuff like 60 Hz.Our power supplies were energized by 60 Hz. We loved 60 Hz. But seriously, the IFs were bandpass sampled at 6 MHz for the 900 radio and 8 MHz for the 2400 radio. No, low frequency performance was not a concern with this particular part. I counted my blessings on having a great part available, sparse part data notwithstanding. 20 dB of gain and 4 W of power at 900 MHz seemed good to me at the time. Low frequency performance, aside from stability questions, was not important to the task at hand. So, you're right about that.> Smart meters are a giant business now, and other things made their > Ricochet idea obsolete.IIRC, Schlumberger bought the metering part from them as they went whole hog into the Ricochet thing. I mean, Ricochet was not the metering part of the business.
Reply by ●February 6, 20232023-02-06
On Monday, February 6, 2023 at 11:16:26 AM UTC-8, John Larkin wrote:> On Mon, 6 Feb 2023 09:59:15 -0800 (PST), Simon S Aysdie > <gwh...@ti.com> wrote: > >On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote: > >> https://www.qorvo.com/products/p/QPD1016 > >> > >> The part could be very useful in wideband/pulse applications [1] but > >> there are no DC curves, no capacitances, no Spice models, just the > >> silly load pull cartoons. > >> > >> [1] except for the price of course. > > > >When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. > > > >https://www.qsl.net/n9zia/metricom/metricom-900-1.jpg > Are those coaxial ceramic resonators up top? They can be useful in > time domain too.Yes. They were Transtech and/or Toko as were both on the AVL, IIRC.
Reply by ●February 6, 20232023-02-06
On Monday, February 6, 2023 at 11:17:51 AM UTC-8, Simon S Aysdie wrote:> On Monday, February 6, 2023 at 11:16:26 AM UTC-8, John Larkin wrote: > > On Mon, 6 Feb 2023 09:59:15 -0800 (PST), Simon S Aysdie > > <gwh...@ti.com> wrote: > > >On Monday, February 6, 2023 at 9:50:01 AM UTC-8, John Larkin wrote: > > >> https://www.qorvo.com/products/p/QPD1016 > > >> > > >> The part could be very useful in wideband/pulse applications [1] but > > >> there are no DC curves, no capacitances, no Spice models, just the > > >> silly load pull cartoons. > > >> > > >> [1] except for the price of course. > > > > > >When I designed this PA I just did a load line match per Steve Cripps. The Celeritek data sheet was one page. A spice model was not really necessary. > > > > > >https://www.qsl.net/n9zia/metricom/metricom-900-1.jpg > > Are those coaxial ceramic resonators up top? They can be useful in > > time domain too. > Yes. They were Transtech and/or Toko as were both on the AVL, IIRC.I mean, I am referring to the little ones. The giant ones (12 mm) up top are in the ComNav preselector. ComNav has exceptional Q because of their plating process. For a long time they were th only vendor who could meet specs. Eventually IMC in San Diego managed to meet the requirements. So we had ComNav from Maine and IMC from San Diego.
Reply by ●February 7, 20232023-02-07
On a sunny day (Mon, 06 Feb 2023 09:47:57 -0800) it happened John Larkin <jlarkin@highlandSNIPMEtechnology.com> wrote in <7qe2uh9mv99o5pt91in5j1j9b4drqqirrg@4ax.com>:> >https://www.qorvo.com/products/p/QPD1016 > >The part could be very useful in wideband/pulse applications [1] but >there are no DC curves, no capacitances, no Spice models, just the >silly load pull cartoons. > >[1] except for the price of course. >Bit puzzled by the bias on - bias off sequence for the test board... page 20 pfd How critical is that? Needs some external circuits..
