Very slight distortion in max settings (Deep Blue Delay build)

[pigiums]

Hi everyone,

I've recently started building guitar pedals and boy is it fun! I've since done Mabean's Cupcake, tagboard's marshall bluesbreaker and tagboard's deep blue delay.. All sound fantastic, especially the bluesbreaker. It just adds so much character to the sound of the guitar! Can't wait to assemble the Aion Refractor (in a standstill because i'm lacking components.. 390n, 390pf and 820pf, tayda why won't you just stock them! ) and run it with the bluesbreaker!

Anyway, while the DBD works perfectly fine (repeats, duration, volume etc), i get very slight distortion and noise with the repeats in the heaviest settings (i.e longest and loudest repeat settings). Mark from tagboard effects recommended me to use an audio probe and i'll be doing just that after i come back from holiday. But the thing that baffles me is that its actually pretty darn quiet in the lower settings. In the lower settings i get more or less crisp and clear repeats with not much audible distortion/whitenoise with the repeats.

With all that said, wouldn't it be mostly likely to be the PT2399 chip that is the problem here? All components i have got is from tayda. Mark mentioned ceramic caps wouldn't affect anything (i used two ceramic caps). Should i get a whole pile of PT2399s and try them all out for the perfect one? Or could it really be a point in the circuit where the solder isn't done properly?

[anchovie]

PT2399s get noisy at long delay times, they weren't designed for much more sophisticated than karaoke echo. For cleaner long delays, have a look for a DIY circuit that uses multiple chips in series or put the cash you just saved by not buying a pile of PT2399s towards a shop-bought delay with a proper DSP chip. 

[midwayfair]

Anything above 300mS of delay time starts to get noticeable harmonic distortion. It may interest you to know that the newer version of the DBD just clamp the delay time to 450mS to avoid the issue rather than find a solution.

You can get it up to about 600mS with fairly heavy low and high pass filtering, but you have to make the repeats very dark to get there.

Overdriving the chip is another issue entirely, but it doesn't sound like you've got that going on.

There are literally hundreds of posts about this issue, but my honest opinion is that if you need perfectly clean delay, you don't want a PT2399 delay. The idea that so many companies try to pass them off as "tape-like" is laughable (tape delay is comparatively super clean even when the tape is bad). They are more comparable to bucket brigade chips, which, incidentally, can't usually do more than 250mS of delay. :/

[deadastronaut]

The idea that so many companies try to pass them off as "tape-like" is laughable

yeah, very naughty too.

[pigiums]

thanks for the replies! bummer i've always thought the pt2399 was the holy grail of delays. Then how does mad professor achieve their crisp and clear repeats? Or do their pedals distort like my build does?

So theres really no point in trying different PT2399 chips? They'll all sound the same in higher settings? Still makes me wonder because the demos on youtube all sound like perfect repeats..

[pigiums]

sorry i've just did some deeper digging.. and turns out some people do have trouble with Tayda's pt2399 chips. Maybe i'll do some sourcing and try others out!

[midwayfair]

thanks for the replies! bummer i've always thought the pt2399 was the holy grail of delays. Then how does mad professor achieve their crisp and clear repeats?

They don't.

They limit the delay below the max of the chip so that it doesn't get as much distortion, filter the crap out of it, and hope that the user doesn't put too hot a signal into it because then you'll distort the PT2399, because the first stage of the delay is boosted enough to make humbuckers overdrive the PT2399. The repeats in a DBD aren't crisp" or "clear" at all .. the chip simply doesn't do that (the repeats are always softened even when they aren't filtered at all; almost like it starts sampling after the pick attack), and they're not using any special chips, and even if they were auditing out the worst chips, nothing they do prevents the chip from running on 5V, which is an absolute ceiling on the voltage headroom for the chip.

It's true that every once in a while you'll get a PT2399 that is a little nicer or a little worse than average, but it's just not that big of a range of variation.

There's nothing special about the DBD. It's a Rebote with a little more filtering, and almost directly from the datasheet. I'm telling you this not to bash the company for it's $350 delay but to tell you that it's about as basic a PT2399 delay as you can get, and you're not building yours wrong. It's just the limitations of the chip, and part of it (running on 5V) is a limitation of digital technology in general.

[pigiums]

 makes me wonder why it's so desirable, or at least it's sold for so much. My dbd clone is as good as a hand wired boutique,  maybe a nice enclosure and people will be impressed Hahah

But regardless.. It does sound really nice,  thanks for all the replies! Many more pedals to build!

[chemosis]

how would you add a tone control and mod for longer repeats??

[vigilante397]

how would you add a tone control and mod for longer repeats??

Depends, what layout are you using? More repeats is a piece of cake (I like to add an oscillation switch to mine to get infinite repeats to the point of self-oscillation), there's a resistor before the repeats pot (15k in the Rebote) that can be tweaked to get more (lower the resistance) or less (increase the resistance) repeats.

Tone control depends on what exactly you're trying to achieve with it.

[chemosis]

ive built 3 of these. 1 a while ago and 2 recently. they all suck tone out of my intiall attack or pick strike. its too dark.it sounds darker than my bypass signal. the repeats and everything elese souunds fine. im using pedalparts echo blue pcb. i checked all component and cant find the issue. what should i change. maybe the C2 22pf capacitor near the input. i need more high end on my intiall pick attack

[ElectricDruid]

ive built 3 of these. 1 a while ago and 2 recently. they all suck tone out of my intiall attack or pick strike. its too dark.it sounds darker than my bypass signal. the repeats and everything elese souunds fine. im using pedalparts echo blue pcb. i checked all component and cant find the issue. what should i change. maybe the C2 22pf capacitor near the input. i need more high end on my intiall pick attack

If the schematic is similar to this version, that might work, yes:

http://revolutiondeux.blogspot.pt/2012/01/mad-professor-deep-blue-delay.html

This version uses 47pF across a 360K resistor, which rolls the top end off over 9KHz. That might trim the upper-treble a bit. If yours has got 22pF and the other value is the same, it should be fine and you'd have to look elsewhere. The output mixer has another similar cap...


HTH,
Tom

[chemosis]

im using the fuzzdog pedalparts.co.uk  echo blue  pcb/schematic i would send the schematic but i cant find attachment option for this page? C2 = 47pf  &   C3 -100pf. there is No 22pf on schematic. those are the only 2 pf values on the schematic. yea im experincing a little tone suck. my highs get diminished on my intiall pick attack,  maybe my 47pf needs changed to 22pf?? or would i need to decrease the value of C1.

[Rob Strand]

One good thing about using a pre-emphasis and de-emphasis circuit is it often hides some type of distortion.

As far as that circuit goes I can't say I like the design.  An inverting stage at the front-end isn't the best choice as it loads the pickup and/or is more noisy.

[ElectricDruid]

im using the fuzzdog pedalparts.co.uk  echo blue  pcb/schematic i would send the schematic but i cant find attachment option for this page? C2 = 47pf  &   C3 -100pf. there is No 22pf on schematic. those are the only 2 pf values on the schematic. yea im experincing a little tone suck. my highs get diminished on my intiall pick attack,  maybe my 47pf needs changed to 22pf?? or would i need to decrease the value of C1.

Ok, found it:

http://pedalparts.co.uk/docs/EchoBlue-v3.pdf

C2 and C3 are limiting the top end. C2 is across a 360K resistor, which gives you roll-off above 9.4KHz. That's not *too* bad, but you might notice. The second one (C3) shouldn't be a problem. It's across a 22K resistor, so the roll-off is from 73KHz up (worst case - likely to be higher if you've turned the trim down a bit). You *definitely* won't notice that. In many ways I'm surprised this value is so small. Given the noisy delay, I'd have limited the high-end here rather than at the front.
Changing C1 would alter the bottom end, but that's not where you've got a problem, so I wouldn't bother.

HTH,
Tom

[Mark Hammer]

Scott Swartz's PT-80 project ( http://www.generalguitargadgets.com/effects-projects/modulationecho/pt-80-delay/ ) also uses a PT2399 for delay, but includes a compander chip for both noise reduction and headroom management.  I built one, and while it is not entirely immune to headroom issues, is more tolerant of hot input signals than a simple Rebote 2.5, which is more or less the same circuit without the companding.

Sadly, the PT-80 is a costlier circuit to build, and needs a bigger box, which may not be what you want in your context.

[ElectricDruid]

I find it slightly ironic that with a digital chip like the PT2399, we're still throwing every trick from the ancient dusty pages (well, ok -1970's!) of the analog book at it to try and get reasonable performance - pre/de-emphasis, companders, whatever we can.
About the only benefit it has over a 4096-stage BBD is a slightly longer delay. Quality-wise, there's not a lot to choose; they're both terrible!!

Tom