PC74hct4046ap Replacement

[dj_holmes]

Hi,

first I'm way over my expertise when it comes to this. I've read a few posts from here and I knew the clever members have a much better understand that I ever will and I would appreciate some advice.

I have this Digital to analogue converter with this Philips PC74hct4046ap, everyone says its rubbish and even one member talked about modern equivelants. I'm looking for less jitter or something to tray to reduce it as its directly affecting the sound quality. I've attached a schematic of where it goes.

Thank you in advance

[dj_holmes]

More info

[Benta]

You're barking up the wrong tree here, I think.
Issues with the '4046 are generally related to the PFD.
In your case, the '4046 is only used as VCO, where it's quite good.
If you have jitter problems, they're more likely caused in the "Custom Chip".

[dj_holmes]

Thank you for the feedback. Its in a very delicate location. Even a 1% improvement could potentially result in 25% better sound. I am already using the best regulator and caps and resistors but I have not altere the design in anyway. Its over 30 years old. Is there nothing that can be tweaked or slot in improvement?

As the description says its 11.2XXMhz which gets spilt into half from the other chip and appears to feedback into the 4060.  My skills are not great but am willing to do some experiementing if theer are any suggestions. Thanks

[moffy]

The VCO is in a feedback/PLL loop with PC2out as the comparator (the custom chip must be generating some reference) and R303,R304,R334,C302 as the loop filter. If you have stability issues it is most likely due to the phase comparator and loop filter, unless you have switching noise directly affecting the VCO. Can't comment much further as we don't know what is going on in the custom chip.

[moffy]

Ran across this discussion: https://e2e.ti.com/support/logic-group/logic/f/logic-forum/997347/cd74hc4046a-jittered-free-running-vco-signal
It mentions that for similar timing cap values that the jitter is high. The suggestion was to replace the 74HC4046 with the 74LV4046a part, not sure how that would work in your application you might need to change some values.

[dj_holmes]

Thanks for the replies. There are no issues I'm essentially trying to see if there is a slot in replacement with better parameters or some design tweaks.

[dj_holmes]

SN74LV4046A This chip has the same pinout. I think its only SMD where I need the pins but I can try and find an adaptor and solder myself. Its its an improvement its definetly worth a try


ok I think I found the dip version SN74LV4046AN

[BrianHG]

SN74LV4046A This chip has the same pinout. I think its only SMD where I need the pins but I can try and find an adaptor and solder myself. Its its an improvement its definetly worth a try


ok I think I found the dip version SN74LV4046AN

Good old 74HC4046 is still readily available in DIP and SMD.
The only difference with the 74HCT4046 are that the 'digital' inputs low-to-high voltage threshold and the HC variant can be a bit less noisy on it feedback input & PC inputs.

[dj_holmes]

Hi, are you saing my current PC74hct4046ap is a little less noise than 74HC4046? If so is the SN74LV4046AN even less noisy?   

[BrianHG]

Hi, are you saing my current PC74hct4046ap is a little less noise than 74HC4046? If so is the SN74LV4046AN even less noisy?   

HC4046 should be a little less noisy.
LV4046 isn't designed for 5v supply, it's designed for 3.3v supply and below.  (Don't quote me here, double check the data sheet.)

[dj_holmes]

Oh wow so the normal 74HC4046 is better!?!

SN74LV4046A  Optimized Power-Supply Voltage Range From 3 V to 5.5 V

[moffy]

Oh wow so the normal 74HC4046 is better!?!

SN74LV4046A  Optimized Power-Supply Voltage Range From 3 V to 5.5 V

There shouldn't be much difference between the HC and HCT except logic threshold levels, the LV is a much newer technology, faster and might require you to tweak some of the timing components, I say this without having checked the datasheet. It would be fine at 5V.

[BrianHG]

Are you sure the source of your noise is the 74HCT4046?

What about the VCO feedback input from your 'Custom Chip'.  That loop filter with the series diode doesn't look like it can create ever guarantee a super-tight loop.  That 4.7meg as a pull-down is kinda like a weak open circuit.  I would be surprised if that circuit could even be made stable unless you have a fluke of a tuned circuit.  I would toss out that design and do things properly, then you will have no oscillator noise when using any 74HC/HCT4046 variants.

[dj_holmes]

Thanks for the input:

1, its not my design
2, the colock works fine already, but I'm trying to get "more" out of it as its in a very important section.
3, I have no idea how to redesign it.

I'm just trying to get the lowest noise/jitter out of this VCO setup. I can't change the other custom chip only the VCO chip which is PC74hct4046ap made by Philips.

I have a scope and I did see the output 11.2XX frequency and it was pretty badish. I'm new to this and just want to get a better frequency output!  Thanks

[BrianHG]

Thanks for the input:

1, its not my design
2, the colock works fine already, but I'm trying to get "more" out of it as its in a very important section.
3, I have no idea how to redesign it.

I'm just trying to get the lowest noise/jitter out of this VCO setup. I can't change the other custom chip only the VCO chip which is PC74hct4046ap made by Philips.

I have a scope and I did see the output 11.2XX frequency and it was pretty badish. I'm new to this and just want to get a better frequency output!  Thanks

instead of playing with the IC, try playing with R303, R304, R334 and C302.
Also adjust R302.

[dj_holmes]

I'm going to send an image of the current waveform across. Any suggestions on what to change would be great. The idea of a drop in replacement IC really appeals to me 

[Benta]

instead of playing with the IC, try playing with R303, R304, R334 and C302.
Also adjust R302.
[/quote]

I agree. Optimizing the loop filter, aka loop controller will bring much more bang for your bucks.
Ultimately, using an active loop controller will probably result in best performance. But this is not an easy exercise.

[dj_holmes]

active loop controller- the heck is that 

[Benta]

active loop controller- the heck is that 

Well, R303, R304 and C302 comprise the "loop filter" or "loop controller" setting the dynamic behaviour of your PLL as well as suppressing ripple from your "custom IC".

Very primitive, but in many cases enough for a "charge-pump-PFD", which your "custom IC" might be.

It may be improved a bit, but a major leap would be replacing it with an active filter.

PLLs are math heavy in design.

[dj_holmes]

I hate maths 

[BrianHG]

active loop controller- the heck is that 

Your scope shot looks perfectly clean.  There is no issue I can see.

As for the loop filter, it is your 'custom chip' feeding pin 9 of the 74HCT4046.  That signal is an analog input which controls the speed of the output of the 74HCT4046.  If there is noise going into that pin, then the output will have FM modulated noise.  When we say loop filter, we mean how the resistors and caps going into that pin 9 are configured.  If the values or the custom chip feeding that pin 9 have noise, then no matter which 74HCT4046 you use, it will do as it supposed to, converting that voltage AM noise into FM oscillation noise exactly as that IC was designed to do.

[dj_holmes]

Ok, I already did the resistors and the caps best I can. What about the diode? Can that be replaced with some low noise? Thanks for info

[PCB.Wiz]

Ok, I already did the resistors and the caps best I can. What about the diode? Can that be replaced with some low noise? Thanks for info

The generic 4148 is 75V 1uA 4ns 2pF so if you want to feel you have 'improved' there, trawl the vendors tables for something better.
In the fine print, not all companies spec the parts the same.

[dj_holmes]

Accordig to this post IR LED is even less noise!

https://www.diyaudio.com/community/threads/some-noise-measurements-for-leds-and-zener-diodes.35821/