PT2399 thru-hole vs SMT issues

[d-rainger]

Hi!
I'm David, from Rainger FX guitar pedals, making my first post here! I'm wondering if anyone can shed some light on my recent problem with the PT2399 chip.

I've used the thru-hole version successfully in the past, and have recently been using the SMT version - the PT2399s. I've got this to work well too in the past, but with the current PCB project I'm working on, as I turn up the rate pot to longer delays, an out-of-control feedback quickly builds up and disappears into just crackling noises.

Meanwhile my breadboard thru-hole version works well. So one pin at a time I disconnected the SMT chip, reconnecting to the breadboard version and testing as I went, until finally the PCB was entirely connected only to the thru-hole chip in the breadboard... Working totally fine!

So then - using an SMT 'conversion board' - I replaced the thru-hole PT2399 on the breadboard with the PT2399s... And the problem occurred. I repeated this test with several chips of both kinds, confirming the situation; the bigger thru-hole works well in this circuit, and the SMT version doesn't.

Does anyone have any idea how to fix this?
Looking online, I find nothing about 'different conditions needed for thru-hole vs SMT'.
As I mentioned, I have got the PT2399s to work fine in the past, but I just cannot find any differences between 'good' and 'bad' layouts, and man, have I been looking....!

On the 'bad' circuits I've noticed the DC on pin 14 (the feedback/output pin) to increase as I turn the rate pot - which doesn't happen on 'good' circuits.

Any help/suggestions are much appreciated!

[pickdropper]

Do you have a picture of the SMT chips you are using?  There are some fakes out there.

[Scruffie]

The internal resistor values have often changed over time, worth checking the integrator resistor values between 9 & 10  and 11 & 12 to make sure there's not excess gain on the 'bad' chips.

[gtr2]

You didn't say which package you are using.  Is it the 150mil of 300 mil SOP?
I've only used the 300 mil which is the pt2399-S.  I haven't had any problems with anyone I've done work for with that package.

I'd also try them from a reputable supplier if you haven't already.

That being said, there is some variability between chips and some changes over time which are not disclosed.  Some batches have more quirks than others, in my experience.

Josh

[d-rainger]

Do you have a picture of the SMT chips you are using?  There are some fakes out there.

Here it is...

[d-rainger]

You didn't say which package you are using.  Is it the 150mil of 300 mil SOP?
I've only used the 300 mil which is the pt2399-S.  I haven't had any problems with anyone I've done work for with that package.

I'd also try them from a reputable supplier if you haven't already.

That being said, there is some variability between chips and some changes over time which are not disclosed.  Some batches have more quirks than others, in my experience.

Josh

It's 150mil, and it's from Tayda.
The resistance between pins 9 & 10 and 11 & 12 is around 5k - not that far from '4.7k'
BTW as I turn the rate pot, it works fine up to about haif way, then as soon as you turn up from there it's into full on feedback meltdown into crackles.....

[Scruffie]

Have you scoped what's happening on the clock output?

But... they came from Tayda, I would suspect they're seconds. I know people have had a lot of problems with batches of their through hole versions so I wouldn't want to trust their SMT either.

[pickdropper]

Try the ones in the wider SO16W package.

http://smallbear-electronics.mybigcommerce.com/ic-pt2399-s/

[pickdropper]

But... they came from Tayda, I would suspect they're seconds. I know people have had a lot of problems with batches of their through hole versions so I wouldn't want to trust their SMT either.

This.  First step is to get parts from a reliable vendor.

[juansolo]

But... they came from Tayda, I would suspect they're seconds. I know people have had a lot of problems with batches of their through hole versions so I wouldn't want to trust their SMT either.

This.  First step is to get parts from a reliable vendor.

With knobs on. We stopped buying ICs from eBay and Tayda long ago for good reason. The single only thing we get from them now are pots. Ironically we stopped getting ICs from them when we had a whole batch of bad 2399 through holes.

EDIT, even found the video. 2nd IC is the Tayda one. Need to turn the vol up as it's a crappy camera, but it was REALLY noticeable in person.

[garfo]

Wow, that is some crazy oscillation. Is there a noticeable difference on how those chips look from one another, or do they both look exactly the same? I mean same printing, same brand...etc.

[juansolo]

They were binned long ago. But you couldn't really tell the difference visually from what I remember. We've been getting them from Das Musikding since and had no issues.

[gtr2]

Most of Tayda's ICs and transistors are out of spec, I bet if you move to another vendor you'll have no issue.

[garfo]

They were binned long ago. But you couldn't really tell the difference visually from what I remember. We've been getting them from Das Musikding since and had no issues.

I get mine, usually, from Musikding or Banzai as well and have never had any problems.