Negative output clock generator help...

[liquidair]

Hey guys, this is my first post and hopefully my problem has a very easy solution.

I'm trying to generate a complementary negative output clock to drive a BBD chip at high speed (up to 3MHz clock, 1.5MHz when divided). I plan to set up an LMC555 timer as a relaxation oscillator, then send that clock to a CD4013 to be split and divided down, then sent to a CD4049 to be buffered. That's fine and dandy when the clock output is +9V-0V, but I need 0V - -9V. So would this work:

Convert the output of the 555 to a negative clock using reverse series/shunt diodes on the output, then run the CD4013 and 4049 from a Vdd = Gnd and Vss = -9V??? What do you guys think? I really appreciate your help and time!

[rjenkins]

Can't you build the whole clock circuit as single-supply, but actually run it as it's V+ being the BBD 0V and it's 'ground' being the -9V supply?

No level conversion needed.

[liquidair]

Well, originally I was going to do that but I read that the 555 timer's trigger has to be a positive voltage, unless I'm mistaken.

[Bill_Marsden]

A CMOS TLC 555 is rated at 2Mhz, older 555s are 1Mhz or less. You'd be better going for either a transistor crystal oscillator, or make one with a gate.

What exactly do you mean by a negitive clock? Are you talking 2 square waves inverted from each other?

[liquidair]

Hey Bill, your 555 work was what inspired me to use the 555 timer!! I was planning on using the LMC555 which is rated at 3Mhz. I realize now the error in my reasoning about the trigger, supplied as RJenkins said the trigger would be a positive voltage (with respect to Vss = -9V).

Bill, with the 555 wired as a relaxation oscillator like you have written about, could you use an LFO on pin 5 (control) to slowly change the output frequency very very slightly so that you don't have a comb filter effect with a delay (bbd)?

[Bill_Marsden]

As much as I like the 555, there are better VCO's out there. I assume that is what you're after.

You're after a monostable? I still don't understand what you're after exactly. A simple transistor can invert the signal, or a inverting gate.

[liquidair]

No, I'd like an astable. What I need is a variable frequency square-wave oscillator that is sweepable from roughly 160kHz to 3MHz. I need complementary outputs (like you'd get from a CD4047) and I need them oscillate between 0V (high) and -9V (low (roughly). This is to drive a BBD delay chip.

[liquidair]

Sorry, I made an error in that response....I need it sweepable from 80kHz to 1.5Mhz. My above frequency range is necessary only if I have to use the CD4013 to get the compl outputs.

[Bill_Marsden]

OK, the output levels and complementary are no problem, the frequency is going to be a bit harder I suspect. I know there are some good VCO chips out there, but not off the top of my head. 80KHz to 1.5Mhz is a pretty wide range, you need the sweep to be continous (and linear)?

[liquidair]

Sorry Bill, I think my language isn't very clear. It's my lack of electronics knowledge that is causing me to be confusing I think. By "sweepable", I just meant variable with a pot.

I've looked at other chips but the only other thing that comes close to working (setting the frequency with a pot in that frequency range) is the CD4047, but I hear that it is pretty much useless above 100kHz or so.

Here's your 555 circuit I was going to use:
http://forum.allaboutcircuits.com/showthread.php?t=15342

Your above circuit however seems perfect. With the right values, C=17pf, a series 13.7k, and a 250k pot, the output frequencies are right. A CD4013 would then divide the clock into compl outputs, and a CD4049 would drive the 700pF load (using 3 sections per compl output).

My question really is how to get the output negative and if it's as simple as connecting Vdd to Gnd and Vss to a negative voltage (say -9/10 V)?