Paul Badenhorst wrote:> On Friday, 19 August 2016 at 04:38:37 UTC+2, Kevin Glover wrote: >> I have benn trying to find a decent alternative to the LM566 VCO. It >> does not seem to be sold anymore at reputable sellers and the ones on >> EBay (from China) I don't trust to be authentic or unrecycled. Does >> anyone know of a 8 to 16 pin alternative (just square and triangle >> out) that won't break the bank. >> Thanks >> -Slackmeister > I see i have two unused NE566E's. >Re: Google Groups Welcome to Usenet (not Google Groups). This group is sci.electronics.design, which is actually older than Google. It's entirely unmoderated and unrestricted, except for the generally accepted informal rules, known as *netiquette*. Don't let Rob discourage you--c'mon and join the fun. Re: NE566 The 566 is a really crappy oscillator. It's perfectly okay for a junkbox project, of course--I've done lots worse things myself--but for a real design there are many, much better choices nowadays. On the simple end, the oscillator of a CD4046 will work as a very wide-range VCO with a pretty linear tuning characteristic over about 100:1 in frequency, though the center frequency isn't that accurately controlled by the RC time constant. The 4046 has only a square-wave output, so if that's a worry one could use one of many dual op amps. One section would be a Howland current source driving an integration capacitor with its other end grounded, and the other would be a Schmitt trigger controlling the polarity of the Howland. That will be better than the 556, but of course needs a lot of extra resistors. (On a PC board, one could make both the Howland and the Schmitt using one 8x resistor array, so it wouldn't be that many parts in real life.) 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
LM566 alternative.
Started by ●August 18, 2016
Reply by ●July 28, 20222022-07-28
Reply by ●July 31, 20222022-07-31
On 7/28/2022 6:27 AM, legg wrote:> On Wed, 27 Jul 2022 23:27:48 -0700 (PDT), Paul Badenhorst > <paulbpb@gmail.com> wrote: > >> On Friday, 19 August 2016 at 04:38:37 UTC+2, Kevin Glover wrote: >>> I have benn trying to find a decent alternative to the LM566 VCO. It >>> does not seem to be sold anymore at reputable sellers and the ones on >>> EBay (from China) I don't trust to be authentic or unrecycled. Does >>> anyone know of a 8 to 16 pin alternative (just square and triangle >>> out) that won't break the bank. >>> Thanks >>> -Slackmeister >> I see i have two unused NE566E's. > > Now all you need is a time machine to transport them back > to 2016.No, they're likely already there! The issue is getting the INFORMATION regarding their availability to the OP!
Reply by ●July 31, 20222022-07-31
On Friday, July 29, 2022 at 12:43:13 AM UTC+10, Phil Hobbs wrote:> Paul Badenhorst wrote: > > On Friday, 19 August 2016 at 04:38:37 UTC+2, Kevin Glover wrote:> Re: NE566 > > The 566 is a really crappy oscillator. It's perfectly okay for a > junkbox project, of course--I've done lots worse things myself--but for > a real design there are many, much better choices nowadays. > > On the simple end, the oscillator of a CD4046 will work as a very > wide-range VCO with a pretty linear tuning characteristic over about > 100:1 in frequency, though the center frequency isn't that accurately > controlled by the RC time constant. > > The 4046 has only a square-wave output, so if that's a worry one could > use one of many dual op amps. One section would be a Howland current > source driving an integration capacitor with its other end grounded, and > the other would be a Schmitt trigger controlling the polarity of the > Howland. That will be better than the 556, but of course needs a lot of > extra resistors. (On a PC board, one could make both the Howland and > the Schmitt using one 8x resistor array, so it wouldn't be that many > parts in real life.)Another option is to run the 4046 eight or 16 times faster than the sine wave frequency you need, divided it down by eight or sixteen, to get a square wave of the desired frequency, and clock the slower square wave through an eight or sixteen stage shift register at the faster clock rate. By tacking the right resistor values onto the taps on the shift register , and running all the resistors into a summing junction you can make finite impulse response low pass filter and get a pretty clean sine wave over quite a wide range of frequencies, You use up a lot of E96 value O.1% resistors, and you have to worry about putting a Hamming window on your sinc function resistor vales to suppress Gibbs oscillations, but it can make a nice sine wave over a wide range of frequencies. -- Bill Sloman, Sydney
