Biphase clock

I read in sci.electronics.design that Niall 84 wrote (in ) about 'Biphase clock', on Wed, 23 Mar 2005:

Well, what 'trouble' are you having?

Not without knowing what your problem is, in some detail.

You'll get much better advice here than on most web sites, unless it's one recommended from here.

You may also get a free training course in Gratuitous Abuse 101, but, hey, it's all part of Life's Rich Pageant.

filters

meant

might

two D

outputs

two

state

D1.

I think he means non overlapping, eg ÷2,and the clock with Q, /Q

Use a crystal oscillator running at 4X speed and divide the output in two D flip-flops connected to generate quadrature signals. That is, the Q outputs of the two flops should be 00, 10, 11, 01 and 00, etc. This will be two square waves in quadrature which is what a biphase clock is. Use a state diagram to figure how to connect the flops. Hint, Q1 to D2, Qbar2 to D1. Bob

filters

meant

might

two D

outputs

two

state

D1.

I think he means non overlapping, eg ÷2,and the clock with Q, /Q

Yes, well I thought of that but figured he couldn't have trouble with a simple 180 deg wave train could he? If that's the case, he can use a single D flop with a 2X clock. D to Qbar. Bob

Why so much stuff? Why not just a Johnson counter?

.--- C1 | ---- | ---- .-------|D Q|---+-------|D Q|------ C2 | C0 --|> | C0 ---|> | | | /Q| | /Q|---. | ---- ---- | `---------------------------------'

Cheers! Rich

Does not work. Just try to analyze what happens when the flipflops starts from 00. Two in time seperated clocks will look like this: __ __ __ Cl1 __| |________| |________| |________ __ __ __ Cl2 ________| |________| |________| |__

With the Johnson counter one clock is shifted 90 degrees relative to the other. (Which is ok if that is what the OP needs. "Bi-phase clock" is a little ambigouos.)

petrus bitbyter

"Niall 84" schreef in bericht news: snipped-for-privacy@g14g2000cwa.googlegroups.com...

Niall,

If your targets are edge triggered, you only need a D-type flipflop as a divide by two circuit. If your clocks needs to be separated in time, you can use the circuit below.

Cl1 Cl2 | | __ | | +------| | .---. | .---. .---. | | +----|& |-------|D Q|--+------|D Q|---------|D Q|--+ | | +--|__| Cl--|> | Cl--|> | Cl--|> | | | | | /Q|o-+ | /Q|o-+ | /Q|o-+ | | | 74LS11 '---' | '---' | '---' | | | | | | | | | +--------------------+ | | | +------------------------------------+ | +----------------------------------------------------+ created by Andy´s ASCII-Circuit v1.24.140803 Beta

The flopflips can be 1.5 ordinary, old fashioned 74LS74. You can reduce chipcount by using one 74LS164 shiftregister instead.

petrus bitbyter

Maybe this will help - the C0 signal is the same clock, to both:

| | C0 ------+---------------'

legend: --)-- = no connect

Yes it does, I have, and no they don't. And there are only two 'e' in separated. ;-)

It does this: _ _ _ _ _ _ C0 _| |_| |_| |_| |_| |_| |_ _______ _______ C1 _| |_______| |__ _______ ______ C2 _____| |_______|

Thanks, Rich

The trouble i'm having is with the biphase clock, Basically every part of it, design implementation etc. I'm using a CD4066 chip, input and output to either side of where i'm switching,(The capacitor) and then i've to get a biphase clock to drive the control side of things.