4 pt2399's in series / long clean delay..

[deadastronaut]

cheers sam, i'l try that resistor...

re: hetero-dining...straight men eating?.. ...just read up about that. gotcha

[John Lyons]

^ Yes, it seems that the Zombie apocalypse has sexual leanings.
Zombies have rights too you know.

[markusw]

cheers sam, i'l try that resistor...

re: hetero-dining...straight men eating?.. ...just read up about that. gotcha

Impressive work!  
Might be that the tolerances of the PT2399 are high "enough" that the difference in clock frequency between the 4 PT2399 is large enough to be out of the audible range.
Might explain why you can get away with using the same value for all 4 emitter resistors.
Do you have a chance to measure the clock frequencies of the 4 PT2399s at different delay times?

[deadastronaut]

^ yes you may be right on the tolerances..makes sense.

out of curiosity i did change each emmiters resistor to a different value..seems ok. (the same)


i have noticed that when i turn up my mixer level full, there is 'noise' a little 'whine' when at longer delays....hmmmm..

shorter delays are quiet though..

[markusw]

^ yes you may be right on the tolerances..makes sense.

out of curiosity i did change each emmiters resistor to a different value..seems ok. (the same)


i have noticed that when i turn up my mixer level full, there is 'noise' a little 'whine' when at longer delays....hmmmm..

shorter delays are quiet though..

To my understanding longer delays should more easily result in heterodyning within the audible range.
How large was the difference for the four resistors when you used different values?
Maybe it helps if you set the values 5-10% apart.
Might also help if you give each PT2399 it's own 7805. Just speculating...don't know how much clock noise bleeds into the power rail...

[deadastronaut]

there wasn't any noticable difference really from using 560r on all of them..

i now have 470/560/680/820r's..

but yeah i'll give different values a go.....i may need to up the values of the caps going across pins 13/14-15/16

see if thats any better..i noticed brian uses 2.2n's, whereas i'm using 1n's..

[markusw]

there wasn't any noticable difference really from using 560r on all of them..

i now have 470/560/680/820r's..

These values in theory already should prevent heterodyning even at long delays. These values probably are even better than if the 4 values are spread evenly.
Interesting that there isn't any difference to 4x 560R.
Maybe tolerances are really "large enough" and/or the the interstage filtering is already enough to prevent any clock bleeding into the subsequent PT2399.
Thanks for sharing your observations!

[deadastronaut]

well i tried the ST FB path etc...but it introduced a LOT of noise..

so am back to the original schemo..nice n quiet again. ...and 560r on all emitters too...like markusw said its probably ok due to pt tolerances.

anyway i'm back to tinkering with the feedback/bypass equalisation adjustments..

[deadastronaut]

i think i may have to give up on this...

now that its really quiet in my house and no traffic outside , i can hear a fair bit of hiss which gets louder the longer the delay is....even without playing anything.

its very noticable especially when compared to bypassed or short delays...hmmmm....bummer.

[Jdansti]

Is it possible to filter the wet path to knock out the freqs above guitar range?

[psychedelicfish]

Try larger capacitors from pins 7 and 8 to ground, I seem to recall larger capacitors there reduce noise by quite a lot.

Merlin has made some notes on the PT2399, which can be found here. You may find that the noise is partially due to the long leads on your breadboard. Low ESR capacitors might be good for filtering the +5V rail and Vref. You may also find that if you use something like his zener follower circuit, you could apply the LFO to the base of the transistor. If you did use the zener follower you'd have to use a reasonably large transistor to handle the current draw of 4 PT2399s.

[deadastronaut]

cheers Ed, yep i read that...i'll  mess with those caps,

the hiss was more prominent due to that 220n on pt4, which made the delay level way above input level.(which i liked)

but it was just too much so i put it back to 100n....see how it goes..

[Mustachio]

i think i may have to give up on this...

now that its really quiet in my house and no traffic outside , i can hear a fair bit of hiss which gets louder the longer the delay is....even without playing anything.

its very noticable especially when compared to bypassed or short delays...hmmmm....bummer.

Awww Man !




I was really looking forward to this. Hopefully you can get it settled down with the noise . If not even the 3 chips and earlier tests you had going sounded great!

[deadastronaut]

hi jim,

yeah its a pita for sure...tried all sorts last night , i may have to try just 3 and see how that goes. bummer eh!

i can raise the caps across 13/14/-15/16 to 2.2n  but it loses 'fidelity' ..becomes soft delay..nothing's easy is it.. 

[Jdansti]

Disclaimer: I have no F-ing idea what I'm talking about. 

Each chip is adding its own hiss on top of the previous chip's hiss.  In the case of higher caps, your magnifying the loss in fidelity. Kinda like making a photocopy of a photocopy many times over.  Have you tried graduating the caps on each chip instead of making them all the same?  Start low or start high?

[deadastronaut]

hi john, me too   

yes ive tried that...

1 - 4 starting low to high...i think its just inherent in these poxy @#$%ing @#$% @#$%er chips...

rant:
why isnt there a nicer common delay chip. with plenty of ram not just 44k..

i know these are just meant for karaoke machines and cheapo delay stuff, but  come on...just a little bit more eh.

[Govmnt_Lacky]

@Rob

That is probably why the Beltons stopped at 3 chips.

Sorry to hear about the project. Was hoping for a breakthrough 

[deadastronaut]

^ yep, you may be right on that...

[deadastronaut]

well i tried just 3, but the delay is too short imo...and not really any noise advantage...

the magic 4 are good, guess i'll just have to 'limit' the uber long stuff with a smaller delay pot..just to keep it in a 'useable, practical  range'

i'm doing my testing with headphones on with this...

the reason its too hissy is because i like the delay to be a ittle louder than initial input...so i'm just being a greedy bastard..

if its used as 'normal' everyday delay ( behind initial notes ) its fine really (acceptable) ....but still tinkering..

[armdnrdy]

You might try incorporating a compander.

Look how it was used in the PT80 for ideas:

http://1.bp.blogspot.com/-T4AAqQL-BVs/Tprmwi7gbUI/AAAAAAAAA6g/jDlRxiX90cE/s1600/pt80schematic.jpg

See Mark Hammer's reply:

http://www.diystompboxes.com/smfforum/index.php?topic=80062.0