Got bored, decided to revise my transistors triangle wave gen and here is the result:

For anyone thats bored too

 

One problem with this circuit is that, the charging and discharging time is not the same, close but not equal because of ~0.2V drop between CE of Q4. I cant think of anyway around this ...yet

 

450KHz

Got rid of current mirror, using long tail pair to get rid of ~0.2V diff between charge and discharge. There is still another ~0.2V error though.

Wave form still looks ok...

.

 

1MHz

Still looks decent...
Now, time to look for cheap, fast switching bjt and test it out

 

hi ii,
Have you tried your circuit driving a load.?
E

 

That's a pretty nice looking triangle wave. How fast can it go and still look that good? Just curious.

 

1MHz was the max for those 2N5769 BJTs

1.5MHz is the max for these MPSH10 BJTs.... unless there is away to bypass their internal capacitance without effecting the waveform

 

hi ii,
Have you tried your circuit driving a load.?
E

Hi eric,
Sorry for late reply, no I have not

 

hi ii,
I assumed you would be adding an output buffer, interested to see how it works out.
E

 

1 MHz makes it usable for lots of things.

 

3.18MHz

Have to go with current mirror to achieve this
Doesnt look too good though

 

Simple Discreet PWM
BJTs having a hard time keeping up at this speed, falling behind

 

Looking good.

Its nice to see a good discreet design.

 

Thanks DickCappels!

Does anyone know how to search for faster BJTs without reading thousands of datasheets?

 

Spikes at current mirror Q2ic ...why?