On Thu, 6 Jan 2022 12:33:48 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>Jan Panteltje wrote: >> On a sunny day (Thu, 6 Jan 2022 08:27:14 -0800) it happened Artist >> <sepflanze@sj.gmail.com> wrote in <sr7594$ro5$1@dont-email.me>: >> >>> My last post was not clear. Apologies >>> >>> The VCO I seek preferably produces a sine wave to reduce the part count >>> required to filter out the higher harmonics of a square wave. >>> >>> The sine wave need not be clean. There can be some raggedness to it. >>> >>> It must be capable of frequencies at least within the band between >>> 70kHz, and 80kHs. That's 70 kHz or below on the low end., and 80kHz or >>> above on the high end. >>> >>> Output voltage level, and current source limits, are not important >>> because its output will be boosted to what is needed by a following stage. >> >> You will have to be more specific, >> but a few thoughts >> Those transducrs are normally capacitive, several nF, at leas the big ones I use >> Big ones >> http://panteltje.com/pub/40_KHz_ultrasonic_transducers_IMG_5133.JPG >> >> But I use small ones too >> http://panteltje.com/pub/44kHz_Doppler_radar_Rx_transducer_and_coil_IMG_4097.JPG >> >> the driving method differs! >> >> For a sine wave, drive it with an inductor (you can, once you know C_transdcer and frequency >> calculate L. >> >> opamp - L - [transducer]--- ground >> >> Now L sets the resonance. >> >> You will get very high voltage on the transducer for even a low drive that way, >> You can reduce Q by adding a series resistor for example. >> For a one stage you can take feedback from the transducer to an input >> and it will self oscillate (at value set by L). >> One way to go about it... >> >> If all you want is a sinewave between 70 kHz and 80 kHz >> wind a coil and use this >> >> +12 >> coil | >> |--- d >> ---------->| JFET >> | | |--- s >> | { === | >> | ( | 1 n |--------0 approx sine out >> | ( L |--------| >> | ( | | >> | === [ ] 470 >> | | 2 n | >> /// /// /// >> >> >> calculate L, wind it on some core. >> Note the C value ratio, could be 10 nF 20 nF etc >> >> > >Won't work. The mechanical resonance dominates the response. > >Cheers > >Phil HobbsSeems to me that the op should characterize the device before speculating about things to drive it. -- I yam what I yam - Popeye
Crystal Controlled Oscillator
Started by ●January 3, 2022
Reply by ●January 6, 20222022-01-06
Reply by ●January 6, 20222022-01-06
On a sunny day (Thu, 6 Jan 2022 12:33:48 -0500) it happened Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote in <3dcd26aa-0a7c-269d-5597-b39ce9bc3239@electrooptical.net>:>Jan Panteltje wrote: >> On a sunny day (Thu, 6 Jan 2022 08:27:14 -0800) it happened Artist >> <sepflanze@sj.gmail.com> wrote in <sr7594$ro5$1@dont-email.me>: >> >>> My last post was not clear. Apologies >>> >>> The VCO I seek preferably produces a sine wave to reduce the part count >>> required to filter out the higher harmonics of a square wave. >>> >>> The sine wave need not be clean. There can be some raggedness to it. >>> >>> It must be capable of frequencies at least within the band between >>> 70kHz, and 80kHs. That's 70 kHz or below on the low end., and 80kHz or >>> above on the high end. >>> >>> Output voltage level, and current source limits, are not important >>> because its output will be boosted to what is needed by a following stage. >> >> You will have to be more specific, >> but a few thoughts >> Those transducrs are normally capacitive, several nF, at leas the big ones I use >> Big ones >> http://panteltje.com/pub/40_KHz_ultrasonic_transducers_IMG_5133.JPG >> >> But I use small ones too >> http://panteltje.com/pub/44kHz_Doppler_radar_Rx_transducer_and_coil_IMG_4097.JPG >> >> the driving method differs! >> >> For a sine wave, drive it with an inductor (you can, once you know C_transdcer and frequency >> calculate L. >> >> opamp - L - [transducer]--- ground >> >> Now L sets the resonance. >> >> You will get very high voltage on the transducer for even a low drive that way, >> You can reduce Q by adding a series resistor for example. >> For a one stage you can take feedback from the transducer to an input >> and it will self oscillate (at value set by L). >> One way to go about it... >> >> If all you want is a sinewave between 70 kHz and 80 kHz >> wind a coil and use this >> >> +12 >> coil | >> |--- d >> ---------->| JFET >> | | |--- s >> | { === | >> | ( | 1 n |--------0 approx sine out >> | ( L |--------| >> | ( | | >> | === [ ] 470 >> | | 2 n | >> /// /// /// >> >> >> calculate L, wind it on some core. >> Note the C value ratio, could be 10 nF 20 nF etc >> >> > >Won't work. The mechanical resonance dominates the response.He said he uses an opamp amplifier AFTER the sine wave generator This LC oscillator circuit has been working for 50 years or more
Reply by ●January 6, 20222022-01-06
Jan Panteltje wrote:> On a sunny day (Thu, 6 Jan 2022 12:33:48 -0500) it happened Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote in > <3dcd26aa-0a7c-269d-5597-b39ce9bc3239@electrooptical.net>: > >> Jan Panteltje wrote: >>> On a sunny day (Thu, 6 Jan 2022 08:27:14 -0800) it happened Artist >>> <sepflanze@sj.gmail.com> wrote in <sr7594$ro5$1@dont-email.me>: >>> >>>> My last post was not clear. Apologies >>>> >>>> The VCO I seek preferably produces a sine wave to reduce the part count >>>> required to filter out the higher harmonics of a square wave. >>>> >>>> The sine wave need not be clean. There can be some raggedness to it. >>>> >>>> It must be capable of frequencies at least within the band between >>>> 70kHz, and 80kHs. That's 70 kHz or below on the low end., and 80kHz or >>>> above on the high end. >>>> >>>> Output voltage level, and current source limits, are not important >>>> because its output will be boosted to what is needed by a following stage. >>> >>> You will have to be more specific, >>> but a few thoughts >>> Those transducrs are normally capacitive, several nF, at leas the big ones I use >>> Big ones >>> http://panteltje.com/pub/40_KHz_ultrasonic_transducers_IMG_5133.JPG >>> >>> But I use small ones too >>> http://panteltje.com/pub/44kHz_Doppler_radar_Rx_transducer_and_coil_IMG_4097.JPG >>> >>> the driving method differs! >>> >>> For a sine wave, drive it with an inductor (you can, once you know C_transdcer and frequency >>> calculate L. >>> >>> opamp - L - [transducer]--- ground >>> >>> Now L sets the resonance. >>> >>> You will get very high voltage on the transducer for even a low drive that way, >>> You can reduce Q by adding a series resistor for example. >>> For a one stage you can take feedback from the transducer to an input >>> and it will self oscillate (at value set by L). >>> One way to go about it... >>> >>> If all you want is a sinewave between 70 kHz and 80 kHz >>> wind a coil and use this >>> >>> +12 >>> coil | >>> |--- d >>> ---------->| JFET >>> | | |--- s >>> | { === | >>> | ( | 1 n |--------0 approx sine out >>> | ( L |--------| >>> | ( | | >>> | === [ ] 470 >>> | | 2 n | >>> /// /// /// >>> >>> >>> calculate L, wind it on some core. >>> Note the C value ratio, could be 10 nF 20 nF etc >>> >>> >> >> Won't work. The mechanical resonance dominates the response. > > He said he uses an opamp amplifier AFTER the sine wave generator > This LC oscillator circuit has been working for 50 years or more >Quite right--I read too fast. 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 ●January 7, 20222022-01-07
On a sunny day (Thu, 6 Jan 2022 14:50:25 -0500) it happened Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote in <8af24784-0ed7-5329-b3d7-637937920e50@electrooptical.net>:>Jan Panteltje wrote: >> On a sunny day (Thu, 6 Jan 2022 12:33:48 -0500) it happened Phil Hobbs >> <pcdhSpamMeSenseless@electrooptical.net> wrote in >> <3dcd26aa-0a7c-269d-5597-b39ce9bc3239@electrooptical.net>: >> >>> Jan Panteltje wrote: >>>> On a sunny day (Thu, 6 Jan 2022 08:27:14 -0800) it happened Artist >>>> <sepflanze@sj.gmail.com> wrote in <sr7594$ro5$1@dont-email.me>: >>>> >>>>> My last post was not clear. Apologies >>>>> >>>>> The VCO I seek preferably produces a sine wave to reduce the part count >>>>> required to filter out the higher harmonics of a square wave. >>>>> >>>>> The sine wave need not be clean. There can be some raggedness to it. >>>>> >>>>> It must be capable of frequencies at least within the band between >>>>> 70kHz, and 80kHs. That's 70 kHz or below on the low end., and 80kHz or >>>>> above on the high end. >>>>> >>>>> Output voltage level, and current source limits, are not important >>>>> because its output will be boosted to what is needed by a following stage. >>>> >>>> You will have to be more specific, >>>> but a few thoughts >>>> Those transducrs are normally capacitive, several nF, at leas the big ones I use >>>> Big ones >>>> http://panteltje.com/pub/40_KHz_ultrasonic_transducers_IMG_5133.JPG >>>> >>>> But I use small ones too >>>> http://panteltje.com/pub/44kHz_Doppler_radar_Rx_transducer_and_coil_IMG_4097.JPG >>>> >>>> the driving method differs! >>>> >>>> For a sine wave, drive it with an inductor (you can, once you know C_transdcer and frequency >>>> calculate L. >>>> >>>> opamp - L - [transducer]--- ground >>>> >>>> Now L sets the resonance. >>>> >>>> You will get very high voltage on the transducer for even a low drive that way, >>>> You can reduce Q by adding a series resistor for example. >>>> For a one stage you can take feedback from the transducer to an input >>>> and it will self oscillate (at value set by L). >>>> One way to go about it... >>>> >>>> If all you want is a sinewave between 70 kHz and 80 kHz >>>> wind a coil and use this >>>> >>>> +12 >>>> coil | >>>> |--- d >>>> ---------->| JFET >>>> | | |--- s >>>> | { === | >>>> | ( | 1 n |--------0 approx sine out >>>> | ( L |--------| >>>> | ( | | >>>> | === [ ] 470 >>>> | | 2 n | >>>> /// /// /// >>>> >>>> >>>> calculate L, wind it on some core. >>>> Note the C value ratio, could be 10 nF 20 nF etc >>>> >>>> >>> >>> Won't work. The mechanical resonance dominates the response. >> >> He said he uses an opamp amplifier AFTER the sine wave generator >> This LC oscillator circuit has been working for 50 years or more >> >Quite right--I read too fast. >Cheers > >Phil HobbsOk no problem, As to those small 44 kHz transducers, some here may remember the remote controls that used ultrasound (before the IR thing took over). I took one apart and had a good look a the circuit. It is just a tuned LC with the transducer capacitance part of the LC. As Phil Alison already pointed out, not much resonance in those transducers, just are big capacitors. Speakers have also resonance (usually quite low) but are driven over the full audio range or more. So I digged in my old files http://panteltje.com/pub/experiment_with_44kHz_doppler_from_philips_remote_control_IXIMG_0758.JPG on the left that transducer X1 parallel to a tunable coil (measured the inductance of it windings back then) and an other similar transducer X2 tuned with L1 as receiver into a dual gate MOSFET to make a Doppler radar. That thing was so sensitive you could detect if you blinked an eye. Here that original Philips coil and transducer from the remote: http://www.panteltje.com/pub/44kHz_piezo_transmitter_from_old_acoustic_remote_IMG_4075.GIF All about tuning out the transducer capacitance. Cannot seem to find the circuit diagram of the original remote, but what I used as oscillator was pretty much the same parts and circuit. It makes sense to, in stead of driving a high frequency into the transducer C, use L combined with transducer C to get electrical resonance as power consumption counts in a TV remote. There are many projects on the web that drive the transducer from say an opamp directly however.
Reply by ●January 7, 20222022-01-07
PS did not even have to go back that far, acoustic phase wind speed meter, note the sine wave LC oscillator: http://panteltje.com/pub/acoustic_wind_speed_phase_only_test_circuit_diagram_IMG_4890.JPG test setup. fan makes the wind: http://www.panteltje.com/pub/acoustic_wind_speed_phase_only_test_setup_IMG_4887.JPG original Philips coil used bottom right. Was all published here I am sure Did many experiments with those transducrs: http://panteltje.com/pub/44kHz_radar_time_of_flight_test_in_wind_tunnel_IMG_4105.JPG all use the same simple LC oscillator to drive the piezo at 44 kHz IIRC or there about.
Reply by ●January 7, 20222022-01-07
On 7/1/22 7:18 pm, Jan Panteltje wrote:> As to those small 44 kHz transducers, > some here may remember the remote controls that used ultrasound (before the IR thing took over). > I took one apart and had a good look a the circuit. > It is just a tuned LC with the transducer capacitance part of the LC. > ... > and an other similar transducer X2 tuned with L1 as receiver into a dual gate MOSFET to make a Doppler radar. > That thing was so sensitive you could detect if you blinked an eye.Jan, your creativity is always a source of inspiration! I'd never thought to change the inductor for a different resonant frequency.> There are many projects on the web that drive the transducer from say an opamp directly however.For those who remember Gerry Coe at Devantech, he designed the now ubiquitous-on-Ebay SRF04 and SRF08 ultrasound ranging devices. These used two outputs of a MAX232 driver in push-pull to get 20V p-p drive into these transducers from a 5V supply. His other innovation was using an exponential RC decay as the reference level for the LM311 comparator, which increased sensitivity with increasing range. I had a lot of fun with a stereo version I made, two receivers 5cm each side of one transmitter module - yielding not only range but also angular measurement to about 4 degrees. The main issue was the threshold detector, which could trigger +- one cycle, depending on whether the return impulse arrived in-phase or out-of-phase with the mechanical ring-down of the receiver. With the extra compute power that's now available, it would be possible to digitise the signal to model that ring-down, and detect the return wavefront by divergence from the model. That wasn't possible using an MC68HC11 in 1995. Clifford Heath.
Reply by ●January 7, 20222022-01-07
Why not a circuit similar to this self excited one used for the muRata piezoelectric microblower: https://www.mouser.com/pdfdocs/MurataMicroblowerDriverInfo.pdf ? The op amp I would use instead is the OPA552P. It is known this has the bandwidth, and the output power, to do the job. I would, therefore, not need the buffering transistors. To determine a compensation network I recognize I will need to characterize the piezo electric actuator either with a resistor, and a function generator, or a network analyzer if we have one. It may be that it won't need one due to there being no buffering transistors. -- To email me directly remove sj. from my email address's domain name. This is a spam jammer.
Reply by ●January 8, 20222022-01-08
On a sunny day (Sat, 8 Jan 2022 09:39:26 +1100) it happened Clifford Heath <no.spam@please.net> wrote in <16c81e1bd248678a$1$4060634$30dd3a6f@news.thecubenet.com>:>On 7/1/22 7:18 pm, Jan Panteltje wrote: >> As to those small 44 kHz transducers, >> some here may remember the remote controls that used ultrasound (before the IR thing took over). >> I took one apart and had a good look a the circuit. >> It is just a tuned LC with the transducer capacitance part of the LC. >> ... >> and an other similar transducer X2 tuned with L1 as receiver into a dual gate MOSFET to make a Doppler radar. >> That thing was so sensitive you could detect if you blinked an eye. > > >Jan, your creativity is always a source of inspiration! I'd never >thought to change the inductor for a different resonant frequency. > > >> There are many projects on the web that drive the transducer from say an opamp directly however. > >For those who remember Gerry Coe at Devantech, he designed the now >ubiquitous-on-Ebay SRF04 and SRF08 ultrasound ranging devices. These >used two outputs of a MAX232 driver in push-pull to get 20V p-p drive >into these transducers from a 5V supply.Right, that MAX232 trick is nice! In the Philips remote the step up in that oscillator is a factor 42 (something like sqrt(8.9E-3 / 4.9E-6) from the inductance I measured of the coil. Do not remember what batteries it had, but makes sense to get a good signal from 3 V (2 x AA) or so.>His other innovation was using >an exponential RC decay as the reference level for the LM311 comparator, >which increased sensitivity with increasing range. > >I had a lot of fun with a stereo version I made, two receivers 5cm each >side of one transmitter module - yielding not only range but also >angular measurement to about 4 degrees. The main issue was the threshold >detector, which could trigger +- one cycle, depending on whether the >return impulse arrived in-phase or out-of-phase with the mechanical >ring-down of the receiver. With the extra compute power that's now >available, it would be possible to digitise the signal to model that >ring-down, and detect the return wavefront by divergence from the model. >That wasn't possible using an MC68HC11 in 1995. > >Clifford Heath.I also have some ultrasonic distance meter modules from ebay, something like this: https://www.ebay.com/itm/322948018241 those have some special chip and a crystal. Cheap way to get transducers :-) Now I just received some about 3 GHz? (still have to measure it) doppler motion detectors that work through glass etc. from a local shop, also for about 2 Euro a piece, on ebay those are even cheaper: https://www.ebay.com/itm/311911747154 Is using that frequency even legal here? Plenty of stuff play with!
Reply by ●January 8, 20222022-01-08
On Saturday, January 8, 2022 at 1:05:05 PM UTC+11, Artist wrote:> Why not a circuit similar to this self excited one used for the muRata > piezoelectric microblower: > > https://www.mouser.com/pdfdocs/MurataMicroblowerDriverInfo.pdf > > ? > > The op amp I would use instead is the OPA552P. It is known this has the > bandwidth, and the output power, to do the job. I would, therefore, not > need the buffering transistors. > > To determine a compensation network I recognize I will need to > characterize the piezo electric actuator either with a resistor, and a > function generator, or a network analyzer if we have one. It may be that > it won't need one due to there being no buffering transistors.You clearly haven't characterised your actuator, and whoever developed the original circuit may not have bothered to. You need to known the impedance of your actuator as a function of frequency. You seem to think that it is resonant somewhere around 80KHz but you have suggested that this varies from actuator to actuator. You need to monitor the amplitude and phase of the current going into the actuator and of the voltage appearing across the actuator . It would be a good idea to keep track of the mechanical movement occurring at the same time - this is the effect you are trying to create. You are going to need a function generator to generate the drive frequency, and at least a two-channel oscillisocope to keep track of the voltage and current. A network analyser delivers all of this in one package (if you are lucky). You should find that amplitude of the mechanical movement will peak at the resonant frequency - which means that it doesn't change at all for small excursion in frequency. You should find that the phase of the mechanical movement will change quite rapidly with frequency around the peak, and if you keep track of that you can use it to servo the drive frequency to stay at the peak. With any luck you won't have to - the mechanical resonance may well be stable for any given actuator, though this isn't guaranteed. -- Bill Sloman, Sydney
Reply by ●January 8, 20222022-01-08
PS see https://github.com/jdesbonnet/RCWL-0516>Now I just received some about 3 GHz? (still have to measure it) doppler motion detectors that work through glass etc. >from a local shop, also for about 2 Euro a piece, on ebay those are even cheaper: > https://www.ebay.com/itm/311911747154 > Is using that frequency even legal here? >Plenty of stuff play with!
