Triennial sine-shaper discussion

What's the state of the art in broadband DC-to-daylight (well, a few MHz, anyway) practical analog triangle-to-sine shapers?

Looks like the last discussion here was in 2017. Any new developments?

Here's a circuit someone posted as of April 2020:

Here's an array of lateral PNP diff-pairs, looks more suitable for as it says implementation on chip than anything discrete but idk.

Anyone know what the gist of this now gone-for-obv-reasons circuit was?

Reply to
bitrex
Loading thread data ...

What's the budget? How 'bout a demo board for some AD DDS chip?

George H.

Reply to
George Herold

The AD9833 is nice and can do 0.1 - ~10MHz sines but $9 in small quantity is too rich for me.

If the triangle is already available can we get a broadband sine for a couple bucks?

Reply to
bitrex

it

?

how about this?

formatting link
EEL_C116589.html

Reply to
Lasse Langwadt Christensen

That's a nice price, I've ordered from them before. DHL express comes very quickly but I have to pick up from the service point nearest me 20 miles away in Pawtucket RI; it's a tiny warehouse the 'hood by the train tracks, it look like something from Bulgaria after the fall of communism. but appropriately sized for Rhode Island I guess.

Next time I order from them I'll try DHL eCommerce and see what the difference is with the postal service covering the last mile.

Reply to
bitrex

Got a package from Russia one time there they look at you skeptically for a moment then rummage around behind the counter and in back cursing for 5 min, "Oh it got mixed in with these boxes from the Dominican Republic that's weird!"

y'know that kind of place

Reply to
bitrex

IDK, if the transistor circuit involves matching transistors... that's spendy. You could look see how wavetek did it. In the distant past I used these IC function generators... there were some 'distortion' adjustment pots, but the result was never satisfying. How good do you need? I'm just thinking there's probably some resistor/ diode thing that can round off the peak. ... think about approximating the sine with a few straight lines. It'll work better with lots of voltage drive. The temperature dependence will stink...

George H.

Reply to
George Herold

What happened to the rest of Dr. Phil's paper?

Reply to
bitrex

  • This scheme appears to be the best; a wowser.

  • A good idea and start; chip layout and definitely processing need tweaking.

  • The operation timed out when attempting to contact
    formatting link
  • A different resource for same info would be greatly appreciated.

Thanks.

Reply to
Robert Baer

RRO opamps mostly clip clean and fast, so a dual or quad could be made into a segmented breakpoint thing. Here's a quad:

formatting link

That's the general idea. It could no doubt be simplified.

The simplest triangle shaper is probably one resistor and two diodes. It is sort of like the tanh thing.

--
John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 
 Click to see the full signature
Reply to
jlarkin

Ahh, that's nice. Counting on opamp clipping seems like something you'd have to test. Are the + and - rails symmetric? or more importantly repeatable?

Yeah I recall something like that. Years ago there was a product that used a bunch of the 'function generator' ICs ICL8038. Huh available on amazon now. (I checked I had the right part number.)

formatting link

George H.

Reply to
George Herold

A good RRO opamp will go all the way to the rails, and come off clean and fast. They can make nice comparators, too.

--
John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 
 Click to see the full signature
Reply to
jlarkin

That's it. It's in Mathcad 2000i, running under Wine, so there are some font peculiarities I haven't bothered fixing. If you run off one margin, the PDF version contains a whole other column of pages.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

You DO know the description is incorrect; the Vcb of the transistors is NOT zero. In fact, the Vcb varies as the waveform: +/- 0.167V triangle Vb(Q1), +/- 0.167V sine Vb(Q2). Mmm.. running a transistor between 167mV normal collector voltage,to fairly deep saturation of 167mV can give non-symmetrical drive distortion which can cause undesirable harmonics.

Changing R1, R6 to 95.3K will drastically improve that. Gain, that is up to the student.

Thanks.

Reply to
Robert Baer

Evaluated one long ago - spawn of the devil. Distortion depended varyingly on waveform, frequency AND the distortion pot settings. If huge distortion is ok then go for it.

NT

Reply to
tabbypurr

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.