7 Transistors Triangle Wave Generator

Thread Starter

iimagine

Joined Dec 20, 2010
511
After many days of messing around with LTSpice, I came up with this 7 Transistors Triangle Wave Generator circuit. It gives out a very nice wave form, highly linear. The only problem with it is that when I change the capacitor value down to 100pF or lower the output is no longer linear and I can not figure out why. Can someone help me out please? I have not build and tested this circuit yet, if anyone doubt it would work please explain as I am still learning
Thanks in advance.
 

Attachments

GopherT

Joined Nov 23, 2012
8,009
First, why do you want to? if you are looking for higher frequencies, make sure the transistor can handle those frequencies - check for specific brand and submodel DATASHEET.

If it should work, try making the r3/r4 voltage divider lower resistance. 100k or even less.

Is there a background for this circuit? Has anyone said it should handle smaller capacitance?
 

Thread Starter

iimagine

Joined Dec 20, 2010
511
Thanks all for responding to my post

To GopherT:
Higher frequencies is what I am looking for, yes. I will look up datasheet as you suggested. Thanks

Lowering r3/r4 voltage divider too much will mess up the symmetry as it will take longer for Q6 to discharge the capacitor, and less for Q5 to charge it. the output wave form still bent

This is my circuit, I came up with it, no one has seen this yet, so there is no back ground
 

ErnieM

Joined Apr 24, 2011
8,377
After many days of messing around with LTSpice, I came up with this 7 Transistors Triangle Wave Generator circuit. It gives out a very nice wave form, highly linear. The only problem with it is that when I change the capacitor value down to 100pF or lower the output is no longer linear and I can not figure out why. Can someone help me out please? I have not build and tested this circuit yet, if anyone doubt it would work please explain as I am still learning
Thanks in advance.
Hands down, this is an excellent and most incredible first post I have ever seen.

iimagine, you will certainly go far.
 

KCHARROIS

Joined Jun 29, 2012
311
Just out of curiosity, is it possible to achieve triangular wave generator in this format in the 100MHz range? Or what is the highest output range with this kind of layout?

Thanks
 

crutschow

Joined Mar 14, 2008
34,285
After many days of messing around with LTSpice, I came up with this 7 Transistors Triangle Wave Generator circuit. It gives out a very nice wave form, highly linear. The only problem with it is that when I change the capacitor value down to 100pF or lower the output is no longer linear and I can not figure out why. ..................
The transistor speed is a factor. If you substitute 2N2369's for the 2N3904's, the waveform looks better with 100pF. But it still has a somewhat rounded bottom. If you also change R3 and R4 to 300kΩ, then the sawtooth look fairly linear.
 

GopherT

Joined Nov 23, 2012
8,009
The transistor speed is a factor. If you substitute 2N2369's for the 2N3904's, the waveform looks better with 100pF. But it still has a somewhat rounded bottom. If you also change R3 and R4 to 300kΩ, then the sawtooth look fairly linear.
Crutschow,
Thanks for info on the R3/ R4. I wasn't sure how much they had t come down, I overestimated (above).
What frequency is possible with the 100p cap and faster transistor?
 

#12

Joined Nov 30, 2010
18,224
I have not build and tested this circuit yet, if anyone doubt it would work please explain as I am still learning.
Hello. I seem to be party pooping today. I just embarrassed a microprocessor by erasing more than half the parts on the schematic and rearranging what was left. Same job, less than half the parts count, and no transistors.

http://forum.allaboutcircuits.com/showthread.php?t=89503

Now I'm going to pick on you. It's a brilliant design. I can't fault you there, but...it depends on matched transistors. The only way that will work properly is as an I.C. In the analog world, current mirrors don't work like they do on an integrated chip. You will need to use degenerating resistors in the emitter circuits to get the mirrors to work properly. You might even have to hand-pick your matched pairs, and buy name brand transistors to avoid rejects re-marked as those very well known transistor numbers you used.

I would not have posted anything except you used the word, "build". I hope you build this because there is a world of education to be found in that activity. Personally, I am not educated enough to quantify the properties you will need to get this working. People like crutschow have that kind of education. Stay tuned in and I expect he will help you get a physical circuit working.

Number Twelve
 

Thread Starter

iimagine

Joined Dec 20, 2010
511
The transistor speed is a factor. If you substitute 2N2369's for the 2N3904's, the waveform looks better with 100pF. But it still has a somewhat rounded bottom. If you also change R3 and R4 to 300kΩ, then the sawtooth look fairly linear.
Thank you very much for pointing that out. The waveform does indeed looks much better!

What I was hoping to achieve is to design a triangle wave generator circuit that consumes current as low as possible and with high linearity waveform. Its time for me to search for a faster transistor and see how it goes. Do you or anyone have a recommendation?
 

Thread Starter

iimagine

Joined Dec 20, 2010
511
Thank you #12 for your suggestions and comments. I understand that I have to hand matched those transistors...but they are so cheap :D
I'm planning to build this circuit soon, hope you hang around to see how it goes :)
 

#12

Joined Nov 30, 2010
18,224
One more thing, if you're going above a megahertz, you can't use a protoboard to build it. Too much stray capacitance in those. Think about whether you want to work in dead bug style, use perfboard, or invest the work required to do a double sided board.
 

RichardO

Joined May 4, 2013
2,270
r3/r4 = 3k/3k
r1/r2 = 1k/100Ω
c1 = 100pF
Q5/Q6/Q7 = 2N2369

Frequency = 7200khz
If you use the 2n2369's then you should also use a better PNP for Q1 and Q2. 2N3640 or 2N5771 come to mind.

I doubt that you will get anywhere near 7 MHz in a real circuit and I am eager to see what you do get. As #12 says, this is not easy. But... that is what makes it fun.

-- Richard
 

#12

Joined Nov 30, 2010
18,224
I was thinking all you have to do is go to a vendor site and use their search engine to pick a transistor number with a high Ft when I was reminded that you can buy double transistors...matched pairs in a single package.

You can also get quads and hexes, but I don't know if they are matched. Theoretically, they are, just because they are on the same piece of silicon, but they aren't specified like the matched pairs are.
 
Top