Hello,

Anyone knows of an IC similar to the LTC6908 but with lower frequency output?
Looking for a dual output oscillator with complimentary outputs from 0 to 10khz with fast rise time.
That LTC6908 IC would be great but frequency starts at 50khz

Thanks

Ken

 

How close to 0Hz do you need?
What risetime is "fast"?

 

Are you making a squarewave inverter?

 

How close to 0Hz do you need?
What risetime is "fast"?

3 Hz lowest and 2ns or 3ns (the LTC6908 has 1.1ns)

 

Use a 555 configured as an astable oscillator, and connect it to a couple of high-speed inverting logic gates, such as the VHC family, to get fast inverted and non-inverted outputs.

 

Does it need to be a squarewave?
If so, use the 4047 as @Audioguru again says, because it has a divide by two which ensure that the outputs are really square.
Then buffer with a high-speed logic such as 74AHC04 as @crutschow said if you want 5V, or an MCP1401 if you need higher voltage outputs.
Or if "square" isn't so important, a 555 followed by a inverting/non-inverting driver such as TC4428.

Sorry for the pedantry, but it's "complimentary tea and coffee" but "complementary outputs"!

 

Use a 555 configured as an astable oscillator, and connect it to a couple of high-speed inverting logic gates, such as the VHC family, to get fast inverted and non-inverted outputs.

The rise time of a 555 is about 100ns, which is slow compared to other options.

 

The rise time of a 555 is about 100ns, which is slow compared to other options.

That's why you'd connect it to a couple of high-speed inverting logic gates, such as the VHC family, to get fast inverted and non-inverted outputs.

 

thanks, I found the CD74ACT109 which has complementary output. that should do it

 

Why the concern about risetime?

 

thanks, I found the CD74ACT109 which has complementary output. that should do it

but that's not going to oscillate.
and if you just want a buffer to speed up the risetimes there are plenty of simpler choices than a JK flipflop, such as a 74AC14 schmitt trigger.

 

Do you need the spread-spectrum modulation?

AND - what kind of frequency accuracy / stability / repeatability do you need. Depending on the answers, the whole circuit (variable oscillator plus two-phase outputs) can be done with one 74AC quad gate.

ak

 

If you look at the 109 internal schematic, you will see that there is some delay between the state changes of the two outputs (skew). What is your tolerance for this? That is, how many nanoseconds of delay between the Q output changing state and the -Q outut changing state can your system tolerate?

Also, for whatever that minimum skew requirement is, is it the same for both edge directions?

ak

 

If you look at the 109 internal schematic, you will see that there is some delay between the state changes of the two outputs (skew). What is your tolerance for this? That is, how many nanoseconds of delay between the Q output changing state and the -Q outut changing state can your system tolerate?

Also, for whatever that minimum skew requirement is, is it the same for both edge directions?

ak

I was going to use the LTC6991 for oscillating, then feed its output to the CD74ACT to get complimentary outputs, then feed that into separate MOSFET drivers.

 

TC4428 will give you complementary MOSFET drive outputs.

 

TC4428 will give you complementary MOSFET drive outputs.

Yes, but the rated current is less and rise time is higher then the ISL55110 MOSFET driver

 

Are you planning on putting a resistor in series with the MOSFET gate?
If so, the product of the resistor and the MOSFET gate-source capacitance will limit the rise-time.

 

Are you planning on putting a resistor in series with the MOSFET gate?
If so, the product of the resistor and the MOSFET gate-source capacitance will limit the rise-time.

Yes I know that, plan not to , I still have a few options on MOSFET, I need to get that gate-source capacitance filled up the fastest, thus thats why I need the current

 

I was going to use the LTC6991 for oscillating, then feed its output to the CD74ACT to get complimentary outputs, then feed that into separate MOSFET drivers.

Three simple inverters in a hex inverter package will do the same thing with less propagation delay. If the 6991 output signal is good enough to drive one MOSFET driver, then you need only one inverter to produce the other phase.

ak

 

Yes, but the rated current is less and rise time is higher then the ISL55110 MOSFET driver

The ISL55111 (same datasheet as the 55110) has an inverted B input. With this part, you no longer need a two-phase drive signal.

A - Oscillator - anything that meets your freq and stability specs

B - Buffer - a single inverter or non-inverting buffer, anything that meets your edge specs

C - MOSFET drivers - ISL55111 with both inputs connected to the buffer output.

Again, if your frequency stability requirement is not stringent, a single CMOS gate chip such as a 74AC14 will do parts A and B. If you use a gate part such as a 74AC132, you can gate the oscillator on and off if needed.

https://www.onsemi.com/pub/Collateral/MC74AC132-D.PDF

ak