Deep Blue Delay / Sea Urchin problems. Strange Voltages

[slashandburn]

I've been fighting with this for months and recently gave up and started over on a fresh PCB.  Low and behold, this one seems to suffer from exactly the same problem. It passes signal both engaged and disengaged, though little to non-existent delay and its pretty obviously not working.  The voltages are obviously wrong, though they're exactly the same ammount of "wrong" as my first attempt. 

I built from this PCB image here http://effectslayouts.blogspot.co.uk/2015/02/mad-professor-deep-blue-delay.html and on both builds I made sure to jumper pins 3+4 of the 20399 as per the notes in the comments.   No parts substitutions worth mentioning, except the 78L05 instead of the 7805.  As far as I'm aware these are the same in all but orientation, so I wouldn't normally expect any issues there though I mention it mostly because I think something around this area is my problem. On both builds, this 78L05 isn't even seeing a full 9v.  For some reason (identical on both builds, both with fresh components) it's getting 6.5v on the input and putting out 3.2v to the PT2399, which I imagine is at least 80% of my problem here.

I've ruled out the 33ohm resistor, it's definitely intact and the correct value (on both builds).  I've been round the rest of board 1 with a fine tooth comb, multiple times over the course of serveral months!   I was massively naive to think I could just solve the problem by scrapping it and starting again. I shouldn't be all that surprised to find seemingly exactly the same problem. I'm not really. Just annoyed with myself that I didn't ask for help sooner.

I guess at this stage there's three possibilities. A. The PCB/Build Doc is wrong. B. My Pot's or 3PDT are at fault (the only re-used components between builds).   Or, C. Somehow I've made exactly the same mistake twice.   My own money would be on C but really I'm just begging for suggestions.  If anybody has any idea.....please!

Here's a complete list of my voltages. I'll keep this to the most recent build to save this getting more confusing than it needs to be.   Photo's to follow!

Battery:   9.26V   (Okay, just noted seriously fast battery drain, by the end of gathering these voltages this was down to 8.36V. Indicative of a short?)

78L05:
I: 6.04
G: 0
O: 3.02

PT2399
1.  3
2.  1.5
3.  0
4.  0
5.  1.73
6.  1.49
7.  2.74
8.  2.72
9.  1.19
10. 0.72
11. 0.72
12. 1.35
13. 1.55
14. 1.86
15. 1.63
16. 1.63

TL072
1. 2.78
2. 2.78
3. 2.77
4. 0
5. 2.75
6. 2.75
7. 2.74
8. 5.44

Cheers in anticipation!  (edit: photos as promised)

[patrick398]

Can't offer much practical help here but the pt2399 needs 4.5v minimum supply so your not getting any delay because the chip is most likely functioning due to lack of power. In my experience if you're getting 9v out of your supply and that drops at the input of your 7805 it could be a short somewhere draining more current that the supply can handle. Someone correct me if i'm wrong there.

[theehman]

 Pinouts on the regulator may be different.  One of the diagrams I found shows the 7805 with a reversed pinout of the 78L05

  https://goo.gl/images/nfjws8

[duck_arse]

get one of your regulators onto a breadboard. measure the in and out voltages, then power off and turn the reg the opposite direction, retest your voltages.

datasheet, datasheet, datasheet, datasheet, datasheet, datasheet ......

[vigilante397]

datasheet, datasheet, datasheet, datasheet, datasheet, datasheet ......

+1 to that. If your battery is at 9V and your regulator input is at 6V that's a huge red flag. Is there any particular reason you put the regulator in the opposite orientation of the layout? I see you have it socketed, have you tried it the other way around?

[slashandburn]

Thanks for the quick responses. I'm fairly certain I've got the regulator orientated correctly but it wouldn't be the first time I've been wrong about stuff like that.

Actually I suspect the problem might lie before the 78L05. The 33ohm resistor doesn't even see 9v. Seems to be getting around 7v and whittling down to the 6 point whatever the regulator was seeing. When I noticed this I tried snipping the LED out of the circuit along with its current limiting resistor. Nada, still nothing. I've even swapped dc Jack sockets and used different battery snaps.

>>   Is there any particular reason you put the regulator in the opposite orientation of the layout?

The datasheet told me to!  At least I think it did. I swear it did. Hang on, better check that.  Yup, the build doc calls for a 7805. I've subbed in a 78L05 which would appear to need orientated 180% from the layout.

Maybe I'm losing my mind. You know what they say about doing the same thing multiple times and expecting different results.

[Kipper4]

Einstein wasn't it ?

[vigilante397]

Have you checked for cold solder joints going into the board? Or a cold solder joint on the power connection?

Have you removed the battery and tried with a wall-wart? Have you pulled the board out and tested it outside of the box? I've found removing as many variables as possible helps me get to the bottom of problems.

[ElectricDruid]

If the current drawn is massive, you'd expect to see a noticeable drop even across 33R, so don't worry about that. The problem is where the current draw is coming from. "Regulator back to front" is definitely favourite, but "solder bridge or other short" is close second.


I've cross-checked the datasheet and your orientation does seem to be correct.

Shorting on the box like Vigilante said is a good thought, and getting the board out of the box is an interesting test. Could it be a short through the box from the off board wiring? (I once had a dead short like this on a pedal that was fine on the bench - non-isolated power jack was shorting through the enclosure to the metal jacks which were grounded).

I'd pull all the chips out and test the voltages at the sockets. Do you get full 9V across the op-pamp power pins? Full 5V on the PT2399 power pins? If so, the chip/s is/are fried. If not, there's a problem on the board.

One other thoughts: 79L05 is easily confused with 78L05 (don't ask me how I know this).

HTH,
Tom

[slashandburn]

Well I'll be damned, well I fished out a few different 9V wall warts, and was getting more into the territory at the input pin of the regulator over batteries.   I also fixed the polarity of my dc socket wiring in the picture, that happened after ripping the socket out in frustration and temporarily wiring direct to battery snap. It was only like that for the photo. HONESTLY!

So I'd unboxed it and still couldn't get 5V out of the regulator. I was just about to take Tom up on his suggestion to take all the chips out and measure the empty sockets.  Instead, I did what I should've tried A LONG TIME AGO!   Flipped the regulator 180% in the socket, plugged in the best of the wall warts and measured that magic number on the 78L05.    FIVE VOLTS!

I can't plug it in yet to double check, hopefully in the next hour or so.  But i'm cautiously optimistic.  And also completely bemused.  Have I (and Tom, sorry to drag you down with me man!) completely lost it here?   It's definitely an 78L05.  Orientated as per my photo should be correct, no?    If this thing fires up now that i've orientated it the opposite way from the datasheet........how do I get all those months back?   

Anyway. Fingers crossed.  I'll report back shortly.

[slashandburn]

So I have now have this 78L05 as pictured here:


Now orientated as pictured here, against the datasheet stated pinout:



And it works?  I can't even begin to...   I swear this is the kind of thing that makes me drink on a Monday.

Sorry Patrick, Ron, Stephen and Nathan. I should've just flipped it round. But, the Datasheet!

I'm still baffled. But it works. Sounds alright too!
Cheers!  Bloody datasheets!

[vigilante397]

Where did you get the datasheet? The orientation shown in the original layout you posted was correct for a 78L05. If he had put the "L" in there then the full name wouldn't have fit so nicely on the layout.

[ElectricDruid]

That's a useless diagram, since it doesn't specify whether that's looking from above or below.

I've come across variants of the 79L05 that use the reversed pinout, but I've never seen it for 78L05 and I'd thought they were immune from that. Basic human error is always more likely...especially around me.

T.

[vigilante397]

That's a useless diagram, since it doesn't specify whether that's looking from above or below.

A fair point. I assumed it was from above as the surface mount packages are shown from above, but obviously assuming things in datasheets isn't the best way to go

[duck_arse]

it is my understanding/reasoning/(proven faulty before) logic that the pins cannot be seen from the top of the package, as the package is in the way. SOME datasheets draw and state "from above" or like, but draw dotted-line circles for the pins, to indicate that they are obscured. I think the Temic jfet sheets do this, from memory.

so, DRAW THE 3D VERSION.

[ElectricDruid]

+1 agree, Duck. A "proper" 3D picture of the part and the pins is so much better.

[duck_arse]

Vishay do it as well as Temic, see here in the J201 docs. [or it's the same docs with diff co. logo.]
http://pdf.datasheetcatalog.com/datasheet/vishay/70233.pdf

[slashandburn]

I'll leave you all to discuss this amongst yourselves. I'm off to hide until I've lived down the shame.

[vigilante397]

I'm off to hide until I've lived down the shame. and play with my sweet new delay

Fixed it for you

[Vikt0r]

That's a useless diagram, since it doesn't specify whether that's looking from above or below.

That's a totally correct diagram, assuming that you have no supernatural powers and can't see through the enclosure,
therefore, if you can see the pins - it means you are looking at the device from below.