PT2399 tempo indicator needs demystified!

[MrStab]

hi guys,

i've been working on putting both a Little Angel chorus and Small Time Delay inside the same enclosure, because it's lazier and cheaper than buying two enclosures. now, i'm not sure i can even get this to fit without halving it, but i built this layout i found on the web just to test it out:



the only semi-useful thread i can find on this subject is this one: http://www.diystompboxes.com/smfforum/index.php?topic=50185.0 , which doesn't offer much insight given my IQ of 5.

first things first: would it even work with the Small Time Delay or (much less importantly) the Little Angel?

i'm still in the debugging stage, ie. my build of the above circuit (if the layout even works) might contain errors, but so far it doesn't seem to work. i should point out that only the PT2399 in the Little Angel works, and i need to wait a week to score a new IC for the Small Time (i burned out a couple by being a bit careless) - i have sockets installed but i fear the pins on my surviving chip are brittle.

all this seems to lack info badly online. i'm aware of the tap-tempo circuit for the PT2399, but i'm not too interested in that for this particular build. any n00b-friendly guidance would be really appreciated. i'm really, really stumped as to the removing ground thing with Magnus Modulus, despite it being explained on the thread i linked to - what's that all about and could it be why i'm not getting any results here?
also, i can't figure out from the layout which way the LED should go, but i've tried both ways to no avail.

cheers!

[MrStab]

update: pin 5 of the chorus flashes constantly with no variation in speed when hooked up. success? time will tell. i think i read the Little Angel's rate is controlled outwith the PT2399, which would make sense. it's a start at least!
will report back when i can, gonna try cutting the board in half.

in case i or anyone else needs it, could someone explain the ground thing? ie. how you'd do that without breaking the circuit? i think 90% of the appeal of this is its obscurity. i don't care what it does anymore. lol

[samhay]

I built one of these to work with either a rebote or small time delay (I can't remember which). It worked, but ticked quite badly. I think I tried using a seperate 5V regulator, and when this didn't tame the ticking, I shelved it - it will be in my 'to do' box somewhere. Does yours tick?
IIRC, it takes the timing from the PT2399's clock, so will work with delays where you vary the resisance from pin 6 to ground. The PT2399 chorus' don't work this way, so it may not work as expected - but then, what did you want it to report on? If you want a tempo indicator for the chorus' LFO, you would be better off just tapping the LFO directly. The delay time for the PT2399 in this case will be 10s ms, and the LED will be flashing like crazy.

[MrStab]

funnily enough, i did use a separate 78L05 as i figured it would just be more convenient. it was late at night so i couldn't test for ticking unfortunately. the ultimate "destination" for this will be the Small Time,  just thought i'd try to test it on the Little Angel first.  

i built an EA Trem with a rate flasher the other day - i'm guessing any such thing for the chorus would be closer-related to that, but i don't really think there'd be much point.

hmm... not the first report i've heard of at least similar circuits ticking. maybe there's some way to isolate the feed off pin 5 with a buffer or something? i read the 2395 chip had a similar clock output that required a buffer, but presumably that's a whole other system.

thanks for the input! sorry i cant offer more from my end right now. will test for ticking when i get home tomorrow, could maybe try a few suggestions as well.

PS: 10ms doesn't sound too far off what i saw, but i could be talking crap. lol

[MrStab]

had a free moment to test: no obvious ticking off pin 5 on the chorus, but as you mention that could be because the clock's redundant and/or permanently too fast in that circuit. it does seem to add some kind of mild interference, though. i don't doubt the Small Time will be more problematic.

[MrStab]

just tried with the Small Time, no audible clicks so far for what it's worth, but i haven't tried it at insane volumes with loads of gain yet. the flashing effect is kinda cool.

[samhay]

just tried with the Small Time, no audible clicks so far for what it's worth, but i haven't tried it at insane volumes with loads of gain yet. the flashing effect is kinda cool.

Excellent. Might have to dig mine out again then.

[R O Tiree]

As long as you:

(1) Keep the lead from pin 5 of the PT2399 short and

(2) Connect the GND wire from this circuit to the input jack where the other GND wire(s) to the circuit board(s) come from, then you shouldn't have too much problem. Even better, if you're not going to run this off battery power, connect it direct to the DC jack GND pin.

That's quite a big board - 2.9" x 1.2".  I can run up a smaller daughter-board for you, if you like - I've done a very similar circuit for an Echo Base that I built a while ago.  PM me if interested?

As to the ticking, it's not pin 5... the clock frequency out of that pin is insanely high - from a few to many MHz - so you'll never hear it. It's the LED and its transistor switching on and off that make that noise.  Give it its own 7805 power supply, properly smoothed, properly protected, with its own, separate GND connection back to the DC jack and you should be fine... it sez 'ere.

[cathexis]

I made that layout way back, but I have no recollection as to where I found the schematic. I certainly didn't invent it, and I had no clue how it worked. Not then, not now. IIRC the only way I could get it to work right was by cutting the ground connection completely... But it did work!
/Lars

[MrStab]

thanks for the clarification, Mike, and thanks for the layout, Lars!

i pretty much hooked it up as described - most of it seemed like common sense, but using a separate voltage regulator is one of the few times my chronic laziness has paid off. thanks for offering a smaller layout, Mike - i mentioned considering just cutting it in half, but in the end it didn't really matter as i used an old Marshall footswitch enclosure with plenty of room. maybe someone could benefit from it in future, tho.

i'm usually too self-conscious at my crap skills to post pics, but what the hell, this rat's nest is my rat's nest!



and the outside (missing a bypass LED & a bezel or two):



Edit: forgot to say: if a 100W amp through a 1x12" at max volume produces no ticking (i avoid ye olde halfe-stack for DIY testing), it has my thumbs-up lol.

[MrStab]

weird observation:

when i put the effect into the front of the amp, there's a weird whining which varies with delay time (and i heard some minor ticking at one point). when i put it through the FX loop, it's fine. why would that be? impedance-related?

could only find this post, by RG, which brings PT2399 clocks and whine together: http://www.diystompboxes.com/smfforum/index.php?topic=94484.msg815422#msg815422

not such an issue for me as i think that type of effect belongs in the loop anyway, but thought i'd mention it.

also, running at 12v sometimes seems to produce a tick but im not sure that's advisable with this circuit. can't repeat it "on demand" tho

[Ben N]

Just a guess--if you have line-level fx loop, maybe the signal is big enough to swamp the whine.

[MrStab]

interesting thought, Ben - i just tried switching between instrument & line level, but no whine either way. weird!

[R O Tiree]

If you're running it at 12V or more, then you're asking those 7805s to waste a lot of energy as heat.

The PT2399 operates at +5V and the max signal it can cope with is about +/- 2V. Anything more than that and you are asking for clipping (and not the good kind). So, operating the pedal at more than 9V is a waste of time and energy.  That energy has to go somewhere and it comes out as heat radiated from the 7805 chips.  If they get too hot they shut down and/or eventually degrade and fry themselves.

Bottom line, if it's not ticking perceptably at 9V, go with the flow.

The whining won't be the PT2399.  It's much more likely to be current-spikes as the 4040s are doing their division and ANDing the results to (ultimately) get the LED to flash at the delay rate.  Common practice would be to have a fairly large cap (about 47µF) from Vout to GND on the 7805, together with smaller caps (about 10nF) from V+ to GND as close to each digital chip as you can possibly get them.  If you're going to put a large cap in as suggested to smooth out the "bumps" in the current demand, you'll also need a 1N4001 diode connected across +9V and +5V so you don't fry the regulator at power-down - the end with the stripe needs to be connected to the +9V terminal.

[MrStab]

i only tried at 12v briefly, for the most part i've been testing with 9V batteries. the 78L05's lack of a heatsink makes me paranoid about pushing it too much - i spaced out and accidentally played it at 12V for a half an hour or so, was okay but that's not a reliable test by any means.
just out of curiosity's sake, wouldn't running at a higher voltage give the TL072 more headroom, or would the PT2399's AD conversion render that negligible?


are you talking about the tempo board, Mike? just so i'm clear. is it really an issue if it's fine in the loop (where i plan to use it) but not in front, or is there any risk to not correcting this? cheers for the info!

[R O Tiree]

At +9V, and assuming you have Vref for the TL072 centred at +4.5V, that is ample headroom for a +/-2V signal, which is about the maximum that the PT2399 can cope with without clipping. The PT2399 sets its own Vref internally at 2.5V, so you can see why anything much above +/-2V signal would get distorted. Therefore, there's no need to run the TL072 at any higher voltage.

You're right - I was talking about the tempo board at Reply #13. If it's not causing you any issues in the loop, then "if it ain't broke, don't fix it". If, however, you find that there is an issue, then you're going to have to make those modifications. You'll also need another big-a$$ cap (about 100µF) between Vin and GND on the 7805, as well as the stuff I already mentioned - the smaller cap (47µF) on the Vout side of the 7805, the protection diode and the smaller caps immediately adjacent to the 4040s between V+ and GND (pins 16 and 8 ).

[MrStab]

i'll take a wee note of what you advised and test it on another amp or 2 to confirm no ticking in the loop.
i think i get what you mean about voltages - it's good to have the random figures on datasheets put into a real-world explanation. cheers!

[MrStab]

random thought - please bear in mind i'm an electronics novice - if the voltage regulator is doing its job and can handle up to 35v and PT2399 voltages at 12v all check out at the usual 5/2.52v combo, shouldnt it be fine?

[R O Tiree]

As long as:

(1) The current draw is not high
(2) The 7805 has a heatsink (yours is the TO92 version which doesn't)

It's hard to say how much current each of those 4040s is passing but, the TO92 package can only cope with 200mA and a power dissipation of 500mW. Each 4040 has a quiescent current of 50mA and the LED about 25mA, making a total of 125mA*5V = 625mW out and 500mW wasted at a supply voltage of 9V - you're probably right on the limit. Feeding it 12V will take it over the limit. Unless I read the datasheet for the 4040s incorrectly...

If it was mine, I'd replace the TO92 regulator with the TO220 version (beware - the pinout is different) and pop those extra components in there that I mentioned before. Try running it for a while and touch the plastic casing of the 7805 with your finger. If it's hot you'll know that you're flirting with early failure of the regulator chip and possible damage to the 4040s depending on the failure mode. If it's cool, then you'll be OK.

[MrStab]

i didn't even consider the 4040's! thanks for clearing that up, makes sense. i tried a 7805 in my first attempt at a Little Angel - just by looking at it it's clear it can handle more juice.

going back to the original topic: through a coupla weeks' usage, i've noticed that the ticking from the Small Time's tempo indicator circuit seems variable. sometimes it'll tick, sometimes i'll cycle power and it won't. so definitely not sorted yet. the only parallel for such variability i can think of are the reports of intermittent PT2399s, but they work fine at the actual effect when ticking. weird.

may try the previously-mentioned decoupling at some point. i've built another Small Time for a friend but i omitted all that (would've been an unnecessary pain in spatial and financial terms), so my own build is open to experimentation.