Forums

PLL recommendations?

Started by William Meyer January 22, 2004
I need a good PLL, and the frequency I need to generate is well beyond
the range of any 4046 type device. In particular, I have a need for
almost 80MHz. Any recommendations? I'm looking for inexpensive, small
parts count, and CMOS.

-- 
Bill
Posted with XanaNews Version 1.16.1.1
William Meyer wrote:

> I need a good PLL, and the frequency I need to generate is well beyond > the range of any 4046 type device. In particular, I have a need for > almost 80MHz. Any recommendations? I'm looking for inexpensive, small > parts count, and CMOS.
Do you need an oscillator or a PLL? If you just need an oscillator, use one. If you need a PLL to lock to an external clock, you're going to have to supply more information.
William Meyer wrote:
> I need a good PLL, and the frequency I need to generate is well beyond > the range of any 4046 type device. In particular, I have a need for > almost 80MHz. Any recommendations? I'm looking for inexpensive, small > parts count, and CMOS. >
Of great importance is the type of application. Do you need analog 80MHz with little phase noise, or digital 80MHz just as clock. What would be the input ? Another XCO ? Rene -- Ing.Buero R.Tschaggelar - http://www.ibrtses.com & commercial newsgroups - http://www.talkto.net
oN 22-Jan-04, Rene Tschaggelar said:

> Of great importance is the type of application. > Do you need analog 80MHz with little phase noise, or digital > 80MHz just as clock. What would be the input ? > Another XCO ?
I have a clean reference frequency, and need to generate another which is at an irrational relation to the first. It is desirable for the PLL output to be as stable as possible. Loop response can be very slow, as the reference is stable, and will not be switched. A secondary requirement, however, is that it may be necessary to alter the generated frequency, depending on operating mode of the system. This would still not require a rapid response, however. The primary concern will be long-term stability of frequency. -- Bill Posted with XanaNews Version 1.16.1.1
oN 21-Jan-04, Jim Stewart said:

> Do you need an oscillator or a PLL? If you just > need an oscillator, use one. If you need a PLL > to lock to an external clock, you're going to have > to supply more information.
I need a PLL. I have a reference frequency, and need to generate another frequency related to it. Except for the frequencies involved, it is straightforward. -- Bill Posted with XanaNews Version 1.16.1.1
William Meyer wrote:

> oN 21-Jan-04, Jim Stewart said: > > >>Do you need an oscillator or a PLL? If you just >>need an oscillator, use one. If you need a PLL >>to lock to an external clock, you're going to have >>to supply more information. > > > I need a PLL. I have a reference frequency, and need to generate another frequency related to it. Except for the frequencies involved, it is straightforward.
Ok, now we're getting somewhere. What's the exact input and output frequencies? Any idea as to phase noise requirements? What's your power and cost budgets?
> I need a good PLL, and the frequency I need to generate is well beyond > the range of any 4046 type device. In particular, I have a need for > almost 80MHz. Any recommendations? I'm looking for inexpensive, small > parts count, and CMOS. >
Take a look at this one: CY22393. It goes up to around 200MHz, programmable and available from DigiKey for example. -- Regards, Andras Tantos <http://andras.tantos.homedns.org>
William Meyer wrote:
> oN 22-Jan-04, Rene Tschaggelar said: > >>Of great importance is the type of application. >>Do you need analog 80MHz with little phase noise, or digital >>80MHz just as clock. What would be the input ? >>Another XCO ? > > > I have a clean reference frequency, and need to generate another
>which is at an irrational relation to the first. It is desirable >for the PLL output to be as stable as possible. Loop response >can be very slow, as the reference is stable, and will not >be switched. A secondary requirement, however, is that it may >be necessary to alter the generated frequency, depending on > operating mode of the system. This would still not require >a rapid response, however.
> > The primary concern will be long-term stability of frequency.
You have the longterm stability of the source oscillator anyway. If your source clock is as accurate as 1e-8, then your derived clock will also be 1e-8, as a PLL does not add cycles. The phase noise is a different matter altogether. There the phase noise of the VCO is the one that matters. You wouldn't want an external VCO, I guess. Have a look at the Cypress CY27EE16 clock generator, makes 200MHz. For lower phase noise, requiring an external VCO, the AD4001 from analog devices goes to 200MHz Rene -- Ing.Buero R.Tschaggelar - http://www.ibrtses.com & commercial newsgroups - http://www.talkto.net
oN 22-Jan-04, Rene Tschaggelar said:

> The phase noise is a different matter altogether. There > the phase noise of the VCO is the one that matters. > You wouldn't want an external VCO, I guess. > Have a look at the Cypress CY27EE16 clock generator, > makes 200MHz. > For lower phase noise, requiring an external VCO, > the AD4001 from analog devices goes to 200MHz
In this case, I can probably use one of the Cypress clocks, but I will = have a look at the AD4001 anyway. Thanks! -- Bill Posted with XanaNews Version 1.16.1.1
oN 22-Jan-04, Andras Tantos said:

> Take a look at this one: CY22393. It goes up to around 200MHz, =
programmable
> and available from DigiKey for example.
Thanks! That looks like a good fit. -- Bill Posted with XanaNews Version 1.16.1.2