Ne555 function generator?

[WoundUp]

https://www.petervis.com/Electronics_Kits/555-signal-generator-kit/555-signal-generator-kit-circuit-diagram.html


Has anyone built one of these? I'm curious because I purchased an XR2206 based function generator from Amazon that apparently has a counterfeit 2206 chip so it isn't a good as it could be. I've still got to assemble and solder it together so I figured I'd ask about the NE555 version as a possible replacement for this thing instead.




Meaning building whatever the 555 circuit is, not trying to replace only the 2206 with a 555 in the parts they sent me. Would I be better off building the 555 circuit with a real 555 chip or is the fake 2206 good enough? Has anyone built one of those kits they sell on eBay and Amazon?

https://www.ebay.com/itm/401185073102

Anything I'm missing here? Feel free to add. I don't have the 555 pcb that's in the link at the top of the page. It'll be on perfboard if I do it.


Thanks.

[Ben N]

I built one, and it sorta works, but it's noisy as heck. But isn't the 555 limited to square waves?

[GibsonGM]

Yes, the 555 outputs square waves. But you can find circuits to shape square waves into triangle/sawtooth etc.   And even sine waves.

But at that point, you might be better off just whipping up a "quick and dirty oscillator' a la  R.G. Keene..2 bjt's and you're done.

Otherwise, I'd just buy an el cheapo from Ebay that does many things. 

All I ever need is a 555-based pulse generator, and the quick n dirty oscillator, really.   YMMV tho.

[ElectricDruid]

I built one, and it sorta works, but it's noisy as heck. But isn't the 555 limited to square waves?

No, not really. The "raw" waveform is a "shark's fin" shaped thing that sometimes gets called a triangle which goes from 1/3rd of the supply to 2/3rds of the supply. That waveform is also available. The comparator output turns that into a square wave which is much louder (almost the full supply voltage, so basically three times as loud, even before you allow for squares being perceptually louder than triangles).

The XR2206 was actually *intended* as a function generator chip, so the outputs are much more what you might expect, and it even includes a low distortion sine wave. It gos much higher than audio, since it was intended as a function generator for many uses, not just audio testing.

To be honest, if I wanted a analogue audio function generator chip today, I'd look at one of the CEM3340 clones - probably the best musical audio oscillator chip (Ramp, Triangle, and Square/Pulse with variable width) there's ever been, and now available from multiple manufacturers and from multiple sources (Full disclosure: including me!). That said, for a dead-cheap, basic tone signal to test stuff the 555 still does the job admirably.

[Rob Strand]

Why not try it fix your XR2206 function generator? What's actually wrong with it?
The XR2206 has a number of trimming features which might be missing from your circuit.
See the datasheet and XR2206 based function generators from old magazines for tips
[The XR2206 Function Gen (Elektor UK Dec 1984, EU Dec 1984 and India Jan 1985) is an example of improving an XR2206 design.  They hunt down the cause of each flaw and come up with a solution.  The result is much cleaner sine and triangle waveforms.]

The NE555 function generator kit is intended for fixed a fixed frequency.   The quality of the sine, triangle or sawtooth waveforms will be low.     You will find other NE555 designs which produce better triangle and sawtooth.  The "mods" give variable frequency but as is the design doesn't lend itself for variable frequency sine, triangle or saw-tooth.    The output level will vary with frequency.  With an NE555 a variable frequency sine will cause the most trouble.
[Forgot to mention, the 100n (C4) across the 4k7 at the NE555 output perhaps should be 10n not 100n but the PCB shows 100n]

A basic audio oscillator is sine and square, 20Hz to 200kHz, output impedance of 600 ohms and a level control (perhaps) with decade ranges.  Fairly low sine distortion, say < 0.1%.

A basic function generator is somewhat different.   Sine, Square, Triangle, Sawtooth.  Perhaps  <1Hz to 1MHz, output impedance 50 ohms,  output level control.   DC offset controls often present.  You can also get pulses with a duty cycle control.   Sinewave distortion is about 1%  sometimes visibly distorted.    The low frequency end means the designs are normally DC coupled.   The high frequency end and low output impedance means a common opamp isn't going to cut it. 

The output voltage varies a bit in both cases but some audio oscillators might only put out 3Vrms or so.

You have to ask yourself what you want out of it.  An audio oscillator is great for audio but isn't going to work for an LFO.
The XR2206 and chips like the CE3340 ElectricDruid mentioned try to be a one stop shop.    As far as whipping up something at home a simple audio quality sine-wave oscillator combined with a square wave option and a separate box with a simple opamp based LFO box can go a long way.    That means you can test the audio while using the LFO separately.

Edit: Added some stuff in []

[Rob Strand]

I found this link,

https://www.electroschematics.com/function-generator/
and
https://www.uctronics.com/download/Amazon/U6035.pdf
and
FYI,
https://www.sparkfun.com/datasheets/Kits/XR2206_104_020808.pdf

No trims that I can see.

[PRR]

.... apparently has a counterfeit 2206 chip so it isn't a good as it could be....

How "bad" can a counterfeit 2206 be? Only goes to 4 times the top of the audio band? Has 2% THD instead of 1% on the Sine? Actually the originals were made on a process so VERY ancient today that new fakes may work better than originals.

There are other "audio" signal generators. MAX038 is similar to 2206. Digital is taking over with better control and lower THD.

https://www.amazon.com/SainSmart-UDB1002S-Signal-Generator-Function/dp/B00JTR66CG
https://www.amazon.com/KKmoon-Frequency-Generator-1Hz-65534Hz-Amplitude/dp/B07FNLR6LX

[Rob Strand]

VERY ancient today t

Some people on the group use app's on their mobile phone to generate audio test signals.

The downside is if you accidentally connect the output across 15V DC or something you can blow-up the phone.   Passing the output through some protection resistors and zener clamps might  fend off some abuse.   If you have an old phone I guess it's your call.

On the topic of frying stuff, the output on the XR2206 kit feeds directly to the output.   So if you do connect it up to something damaging then it could take out the XR2206 chip.  If there was an output buffer with a protection resistor of some sort then the damage is somewhat less.

You will make mistakes.   Everyone does.   That's why it's good to bullet-proof stuff.

[WoundUp]

The k you for all the replies. I'll take the time and read through them thoroughly in just a few.


I went ahead and assembled the 2206. Whadaya know? It worked the first time I tried it and actually doesn't seem so bad for what it is. Apparently the sine wave gets pretty noisy above 11.9v when viewed on a scope. I'll never be using a scope so I guess that won't matter. Nor will I be using over 12v. Mostly 9v.

Quick function tests after assembly showed that it should be fine for what I need...which is basically just something to mess around with and use to provide audio signal to pedals when I'm building/troubleshooting.

Could I use this thing to "break in" the speaker in my amp? Just send the tone through it for a couple of hours/day until it's broken in? Or does it take cycling of the speaker in & out to break it in, like when actual music plays?

[ElectricDruid]

Could I use this thing to "break in" the speaker in my amp? Just send the tone through it for a couple of hours/day until it's broken in? Or does it take cycling of the speaker in & out to break it in, like when actual music plays?

You could, but you'll need to be *very* nice to your neighbours after subjecting them to that for hours on end.

The test signal will make the cone go back and forth just like music, but only at a single frequency. Whether breaking in speakers is really necessary and whether it needs a full range signal are both open to question.

[Rob Strand]

I went ahead and assembled the 2206. Whadaya know? It worked the first time I tried it and actually doesn't seem so bad for what it is. Apparently the sine wave gets pretty noisy above 11.9v when viewed on a scope. I'll never be using a scope so I guess that won't matter. Nor will I be using over 12v. Mostly 9v.

Try a putting a 470pF cap from the output (pin 2) to circuit ground (use the ground that goes to the output socket) .   That removes some noise above 0.5MHz.  Oscilloscopes show noise way above the audio frequencies.

Noise can also be a result of poor bypassing of some of the XR2206 pins.

[Rob Strand]

FWIW, here's the output of Elektor's improved XR2206 design (Elektor UK Dec 1984) compared to standard design (maybe the older Elektor design Elektor 1978). 


They say the key is to keep the wiring on pin 13 and pin 14 as short as possible (sine wave selection and trim pins).   Elektor's improvement was to use a MOSFET switch to keep the wiring length as short as possible.

Given the issue is related to pins 13 and pins 14, I wonder if putting the sinewave-trim trimpot on pin 14 is better or worse than putting it on pin 13.    I've seen designs with the trim on pin 14 without the glitch but it could come down to specifics of the wiring.  The XR2206 datasheet has it on pin 14.   However some designs put it on pin 13, so did Elektor's Jan 1978 circuit.  If it makes a difference it's 3040 years too late  .

[PRR]

A speaker will "break in". The suspension softens. The resonance frequency drops. On a sealed-box this is partly masked by box stiffness. On open-back G-cabinet it isn't.

I am of two minds. One is that you should break in with the actual sounds the speaker will play. Rock at home, Bach at work. The other is that 50/60Hz filament transformer power is a lot cheaper than an audio system, and an un-baffled speaker won't even be very loud while excurding to the limits.

[Rob Strand]

A speaker will "break in". The suspension softens. The resonance frequency drops. On a sealed-box this is partly masked by box stiffness. On open-back G-cabinet it isn't.

The displacement/excursion of the cone is a function of frequency.   Test frequencies above resonance aren't going to exercise the cone suspension much but they will heat the voice coil.   And you don't really want to stress the voice coil while you are breaking in the suspension.

At frequencies below resonance the cone moves as much as possible with the least amount of drive and stress on the voice coil.  In fact you can exceed the excursion limit of the driver at drive levels much lower than the label rating of the loudspeaker.

You don't want to break the driver while you are breaking it in!

[anotherjim]

I find it almost impossible to tell when the speaker has broken in with music/guitar signal. A low sine tone will yield a soft burr, even at low powers. Go away and come back an hour later and it may have become a lot smoother. I don't believe you have to bang the cone from end to end to run it in.

A clean bass tone can also show up build problems - loose fittings, wires touching back of cone (can't see inside a sealed cab) etc.

[Nasse]

https://www.elektormagazine.com/magazine/elektor-198406/45241 Elektor disco drum with 2206, two MOSFET opa env generator for piezo of ext input