[Alex Pogossov]

The best square wave you can get using a CMOS Schmitt trigger. Using 74HC14
you can get about 6ns rise and fall time. If this is not enough, use more
advanced 74LVC14 -- 3ns. The above assumes the load is under 100pF.

A limitation is that you have 5V rail for the former and 3.6V rail for the
later. So the amplitude can not exceed 2.5V or 1.8V respectively. But it is
usually enough.

Another issue -- 50% duty cycle. It is possible to achieve it by using an
op-amp based bias control servo loop, or by using a high speed comparator in
front of the Schimitt trigger, e.g. MAX941 or even something more exotic if
operation above 10MHz is required.

Making Schmitt triggers of discrete transistors nowadays is laughable.

Regards,
Alex



"patrick-turner" wrote in message
...

Phil resounded with a near zero content post of....


"patrick-turner"
( snip 7K of self indulgent reminiscing and
wanking from an aging, retarded brick layer )
** My ****ing god, the excruciating, long running
" Turneroid Concerto for Solo Trumpet "
is reaching *megalomaniacal* proportions.
When will it ever ****ing finish ??????


Unfortunately, Phil has a limited visioin of the world, and about any of its
peoples, but then all opinions are "limited" in veracity or plausuibility.

But I take issue with Phil's phrase " ** My ****ing god "

Unfortunately, despite millions of years passing since our species evolved
to consider the possibility of a spirit world and a god of some sort or
other, or many others, there has never been convincing demonstration by
anything supernatural ever having existed. THEREFORE, how on earth does
Phil's god copulate, ie, ****, root, pharque, poke etc ?

Being Sunday, and a time to consider religious matters if one wishes, then I
invite Phil to talk more to us about his copulating god, and discuss the
properties of the god, and of the copulation methods and details.

I am just longing to hear the PT's last post.
Pun massively intended ......

Weel keep longing, but the world moves on Phil......

Meanwhile, I have tried unsuccessfully to get a transistor based Schmitt
trigger square sine wave to square wave converter to work up to 2MHz, using
3 x PN100 transistors, and produce +/-10V peak, with 20V rise time of
0.05uS, and with good symmetry and without having a huge number of R&C
compensation networks and a lack of readiness to work properly above 300kHz.

Perhaps the PN100 ( said to be equalivalent to 2N2222 and many other types )
is something that is allergic to working much above 1MHz, and seems like its
gain rolls off and base current increases renderig it impossible to use in
the many Schmitt Trigger circuits displayed when Googling. NONE if the
displayed websites have a complete circuit showing details of operation with
all voltages and wave forms and the F range over which reliable operation
can be had. In other words, there is a huge anount of regurgitation of basic
info that is practically useless, a typical feature of how the Internet
displays mostly crap that IS CRAP because most essential details are never
included and test results never given.

What would Phil have us drink about Schmitt triggers? A bowl of sour and
bitter un-helpful bull**** perhaps?

Now stay calm Phil, and, in your own words, and, in your own time, post
something genuinely helpful, please......

Meanwhile, in the absense of anything really helpful online about trying to
make square waves up to 2MHz, I'll explore what a 4093 chip will do, and
maybe get a +/-3.3V pk swing which can be amplified to be +/-10Vpk, and thus
match the +/- 10Vpk swing of the sine wave oscillator I just made with
tubes.

The harmonic content of a square wave is well posted online and consists of
fundemental sine wave and odd harmonics of the fundemental frequency, Ff. If
Vf = 2MHz, then for a FAIR looking square wave we need the correct relative
levels and phase of 6MHz, 10MHz, 14MHz, 18MHz. The devices which produce or
amplifiy the square wave must therefore not attenuate the harmonics, and
should be able to handle 10 times the Ff.
I think a PN100 is dismally incapable of what I wish it to do.

If Phil or anyone else would not wish to use a 4093, or 74LS14, what device
would they use?

I do not wish to have to pioneer the use of say 2 x 6CM5 or 2 x 6CL6, EL86
etc, to make reasonably good looking square waves between 1Hz and 2MHz. The
little BJTs are OK for 1Hz to say 250kHz, but beyond that they really
struggle.

Patrick Turner.