Reply by Jim Thompson August 18, 20152015-08-18
On Wed, 19 Aug 2015 10:31:30 +0800, Kluge <nothing@none.com> wrote:

>On 18-Aug-15 3:07 PM, Martin Brown wrote: >> On 18/08/2015 03:09, Kluge wrote: >>> On 17-Aug-15 11:17 PM, John S wrote: >>>> On 8/17/2015 8:12 AM, Martin Brown wrote: >>>>> On 17/08/2015 11:53, Kluge wrote: >> >>>>>> A product has been using the HEF4060B for the last two or three >>>>>> decades >>>>>> without issue. But the last batch have issues with the oscillator >>>>>> starting reliably at the correct frequency (32767Hz xtal) >>>>>> >>>>>> >>>>>> The part originally used was manufactured by Philips. Then Philips >>>>>> became NXP around 2006. I'm not sure if it was taken over, renamed or >>>>>> whatever...... >>>>>> >>>>>> >>>>>> Parts branded NXP have been used in the past without issue however the >>>>>> last batch is causing problems. >>>>> >>>>> Does it start at a harmonic or a completely different frequency? >>>>> >>>>> Faster rise time or lower drive impedance parts can sometimes cause >>>>> curious effects and 32kHz xtals tend to react badly to being >>>>> overdriven. >>>> >>>> Excellent point, Martin. I wonder if the crystal manufacturer rather >>>> than the IC manufacturer has changed something. >>> >>> It starts up at roughly twice the fundamental. >> >> >> Less drive power and a low pass filter should fix it. >> >> As one of my dodgier mentors would say "slug it with a bit of extra R >> and C" (it was his fix for anything that misbehaved - amazingly it >> worked more often than not because he was very good at figuring out >> exactly where to put it). His fix for things mechanical was a bit of >> bent tin. He left a trail of things that worked but were ugly. >> >> I suspect the xtal loading capacitance is now too low and the output >> drive power too high so a bit more series resistance and some more >> capacitance ought to prevent the second harmonic from resonating. >> >> I'd bet the new part is faster with a lower output capacitance and >> smaller chip geometry. JT has explained why this happens. >> > > >Thanks Martin - more food for thought.
Yep. For the 'HCU04 replacement I had a helluva time matching the new gain-bandwidth-phase against the old... but I succeeded. Some "reworkers" may not be so dedicated. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by Kluge August 18, 20152015-08-18
On 18-Aug-15 3:07 PM, Martin Brown wrote:
> On 18/08/2015 03:09, Kluge wrote: >> On 17-Aug-15 11:17 PM, John S wrote: >>> On 8/17/2015 8:12 AM, Martin Brown wrote: >>>> On 17/08/2015 11:53, Kluge wrote: > >>>>> A product has been using the HEF4060B for the last two or three >>>>> decades >>>>> without issue. But the last batch have issues with the oscillator >>>>> starting reliably at the correct frequency (32767Hz xtal) >>>>> >>>>> >>>>> The part originally used was manufactured by Philips. Then Philips >>>>> became NXP around 2006. I'm not sure if it was taken over, renamed or >>>>> whatever...... >>>>> >>>>> >>>>> Parts branded NXP have been used in the past without issue however the >>>>> last batch is causing problems. >>>> >>>> Does it start at a harmonic or a completely different frequency? >>>> >>>> Faster rise time or lower drive impedance parts can sometimes cause >>>> curious effects and 32kHz xtals tend to react badly to being >>>> overdriven. >>> >>> Excellent point, Martin. I wonder if the crystal manufacturer rather >>> than the IC manufacturer has changed something. >> >> It starts up at roughly twice the fundamental. > > > Less drive power and a low pass filter should fix it. > > As one of my dodgier mentors would say "slug it with a bit of extra R > and C" (it was his fix for anything that misbehaved - amazingly it > worked more often than not because he was very good at figuring out > exactly where to put it). His fix for things mechanical was a bit of > bent tin. He left a trail of things that worked but were ugly. > > I suspect the xtal loading capacitance is now too low and the output > drive power too high so a bit more series resistance and some more > capacitance ought to prevent the second harmonic from resonating. > > I'd bet the new part is faster with a lower output capacitance and > smaller chip geometry. JT has explained why this happens. >
Thanks Martin - more food for thought.
Reply by Martin Brown August 18, 20152015-08-18
On 18/08/2015 03:09, Kluge wrote:
> On 17-Aug-15 11:17 PM, John S wrote: >> On 8/17/2015 8:12 AM, Martin Brown wrote: >>> On 17/08/2015 11:53, Kluge wrote:
>>>> A product has been using the HEF4060B for the last two or three decades >>>> without issue. But the last batch have issues with the oscillator >>>> starting reliably at the correct frequency (32767Hz xtal) >>>> >>>> >>>> The part originally used was manufactured by Philips. Then Philips >>>> became NXP around 2006. I'm not sure if it was taken over, renamed or >>>> whatever...... >>>> >>>> >>>> Parts branded NXP have been used in the past without issue however the >>>> last batch is causing problems. >>> >>> Does it start at a harmonic or a completely different frequency? >>> >>> Faster rise time or lower drive impedance parts can sometimes cause >>> curious effects and 32kHz xtals tend to react badly to being overdriven. >> >> Excellent point, Martin. I wonder if the crystal manufacturer rather >> than the IC manufacturer has changed something. > > It starts up at roughly twice the fundamental.
Less drive power and a low pass filter should fix it. As one of my dodgier mentors would say "slug it with a bit of extra R and C" (it was his fix for anything that misbehaved - amazingly it worked more often than not because he was very good at figuring out exactly where to put it). His fix for things mechanical was a bit of bent tin. He left a trail of things that worked but were ugly. I suspect the xtal loading capacitance is now too low and the output drive power too high so a bit more series resistance and some more capacitance ought to prevent the second harmonic from resonating. I'd bet the new part is faster with a lower output capacitance and smaller chip geometry. JT has explained why this happens. -- Regards, Martin Brown
Reply by Kluge August 17, 20152015-08-17
On 17-Aug-15 10:33 PM, Jim Thompson wrote:
> On Mon, 17 Aug 2015 18:53:59 +0800, Kluge <nothing@none.com> wrote: > >> A product has been using the HEF4060B for the last two or three decades >> without issue. But the last batch have issues with the oscillator >> starting reliably at the correct frequency (32767Hz xtal) >> >> >> The part originally used was manufactured by Philips. Then Philips >> became NXP around 2006. I'm not sure if it was taken over, renamed or >> whatever...... >> >> >> Parts branded NXP have been used in the past without issue however the >> last batch is causing problems. >> >> >> >> The older data sheets seem to be identical to the latest revision data >> sheets so we expect the performance to be the same. >> >> Is it acceptable to expect that a part will not subtly change in some >> way over this time period? What happens if they change to a different >> fabrication process or facility? I vaguely recall a discussion about >> this sort of thing in the last year or so. >> >> >> Thanks > > Are you still getting a part labeled HEF4060B or one labeled 74HC4060? > > (The 74HC4060 datasheet says, "... _pin_ compatible with the > HEF4060.") > > It's quite possible that the fab facility was changed. Some years ago > I was involved in moving all of ON-Semi's 'HC line to a different > process house. The hardest part to match new-to-old performance was > the 'HCU04 inverter... commonly used for crystal oscillators. > > You may need to adjust the values of the drive resistor and the "pi" > capacitors. > > ...Jim Thompson >
Its the same part number - HEF4060B. It's used in a battery supplied low power design. 4060's from other vendors were tested but the current consumption is too high with the alternatives.
Reply by Kluge August 17, 20152015-08-17
On 17-Aug-15 11:17 PM, John S wrote:
> On 8/17/2015 8:12 AM, Martin Brown wrote: >> On 17/08/2015 11:53, Kluge wrote: >>> A product has been using the HEF4060B for the last two or three decades >>> without issue. But the last batch have issues with the oscillator >>> starting reliably at the correct frequency (32767Hz xtal) >>> >>> >>> The part originally used was manufactured by Philips. Then Philips >>> became NXP around 2006. I'm not sure if it was taken over, renamed or >>> whatever...... >>> >>> >>> Parts branded NXP have been used in the past without issue however the >>> last batch is causing problems. >> >> Does it start at a harmonic or a completely different frequency? >> >> Faster rise time or lower drive impedance parts can sometimes cause >> curious effects and 32kHz xtals tend to react badly to being overdriven. >> > > Excellent point, Martin. I wonder if the crystal manufacturer rather > than the IC manufacturer has changed something.
It starts up at roughly twice the fundamental.
Reply by Kluge August 17, 20152015-08-17
On 17-Aug-15 9:01 PM, George Herold wrote:
> On Monday, August 17, 2015 at 6:54:03 AM UTC-4, Kluge wrote: >> A product has been using the HEF4060B for the last two or three decades >> without issue. But the last batch have issues with the oscillator >> starting reliably at the correct frequency (32767Hz xtal) >> >> >> The part originally used was manufactured by Philips. Then Philips >> became NXP around 2006. I'm not sure if it was taken over, renamed or >> whatever...... >> >> >> Parts branded NXP have been used in the past without issue however the >> last batch is causing problems. >> >> >> >> The older data sheets seem to be identical to the latest revision data >> sheets so we expect the performance to be the same. >> >> Is it acceptable to expect that a part will not subtly change in some >> way over this time period? What happens if they change to a different >> fabrication process or facility? I vaguely recall a discussion about >> this sort of thing in the last year or so. >> >> >> Thanks > > Hmm any other changes to the pcb? > New layout, different caps? > > George H. >
Nope, all EXACTLY the same. Some work ok some don't.
Reply by Jon Elson August 17, 20152015-08-17
Kluge wrote:

> A product has been using the HEF4060B for the last two or three decades > without issue. But the last batch have issues with the oscillator > starting reliably at the correct frequency (32767Hz xtal) > > > The part originally used was manufactured by Philips. Then Philips > became NXP around 2006. I'm not sure if it was taken over, renamed or > whatever...... > > > Parts branded NXP have been used in the past without issue however the > last batch is causing problems. > > > > The older data sheets seem to be identical to the latest revision data > sheets so we expect the performance to be the same. > > Is it acceptable to expect that a part will not subtly change in some > way over this time period?
Well, this happens all the time, unfortunately! One thing is, you may have to change the resonating caps on the crystal to get the crystal to start up un the right harmonic or fundamental. If the oscillator completely fails to start, a resistor across the crystal may be needed. if it starts at the wrong frequency, then the resonating caps are wrong. A few pF up or down may be all it needs to fix the problem. A small change to the chip layout or packaging technology could cause a few pF change in terminal capacitance. Jon
Reply by Tim Wescott August 17, 20152015-08-17
On Mon, 17 Aug 2015 18:53:59 +0800, Kluge wrote:

> A product has been using the HEF4060B for the last two or three decades > without issue. But the last batch have issues with the oscillator > starting reliably at the correct frequency (32767Hz xtal) > > > The part originally used was manufactured by Philips. Then Philips > became NXP around 2006. I'm not sure if it was taken over, renamed or > whatever...... > > > Parts branded NXP have been used in the past without issue however the > last batch is causing problems. > > > > The older data sheets seem to be identical to the latest revision data > sheets so we expect the performance to be the same. > > Is it acceptable to expect that a part will not subtly change in some > way over this time period? What happens if they change to a different > fabrication process or facility? I vaguely recall a discussion about > this sort of thing in the last year or so.
Parts change. Like JT and JL said, analog functions implemented on digital processes have the most trouble. Two or three decades without a problem is pretty good, actually. I was in charge of maintaining a board for a while that had an MC1594 on it: at least once a year the factory would change the process, which changed the offsets or gains, which required tweaking resistors on the board. So your life could be much worse. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com
Reply by John Larkin August 17, 20152015-08-17
On Mon, 17 Aug 2015 18:53:59 +0800, Kluge <nothing@none.com> wrote:

>A product has been using the HEF4060B for the last two or three decades >without issue. But the last batch have issues with the oscillator >starting reliably at the correct frequency (32767Hz xtal) > > >The part originally used was manufactured by Philips. Then Philips >became NXP around 2006. I'm not sure if it was taken over, renamed or >whatever...... > > >Parts branded NXP have been used in the past without issue however the >last batch is causing problems. > > > >The older data sheets seem to be identical to the latest revision data >sheets so we expect the performance to be the same. > >Is it acceptable to expect that a part will not subtly change in some >way over this time period? What happens if they change to a different >fabrication process or facility? I vaguely recall a discussion about >this sort of thing in the last year or so. > > >Thanks
Gate-based XOs are always tricky. The crystal parameters, the external caps, and the gate behavior all interact, so things need to be tweaked to work, and an apparently OK tweak could be marginal. The semi process could have changed, for instance the new gate could be faster than in the old process, have less phase shift, and not oscillate with your old parts. Play with the caps and any resistors in the circuit, to get you back to marginal operation. I always buy packaged oscillators, which always oscillate. -- John Larkin Highland Technology, Inc lunatic fringe electronics jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by John S August 17, 20152015-08-17
On 8/17/2015 8:12 AM, Martin Brown wrote:
> On 17/08/2015 11:53, Kluge wrote: >> A product has been using the HEF4060B for the last two or three decades >> without issue. But the last batch have issues with the oscillator >> starting reliably at the correct frequency (32767Hz xtal) >> >> >> The part originally used was manufactured by Philips. Then Philips >> became NXP around 2006. I'm not sure if it was taken over, renamed or >> whatever...... >> >> >> Parts branded NXP have been used in the past without issue however the >> last batch is causing problems. > > Does it start at a harmonic or a completely different frequency? > > Faster rise time or lower drive impedance parts can sometimes cause > curious effects and 32kHz xtals tend to react badly to being overdriven. >
Excellent point, Martin. I wonder if the crystal manufacturer rather than the IC manufacturer has changed something.