GGG PT80 problem - varying pitch

[robottrouble]

Okay, so I finished my new PT80 and was really happy when it powered right up. only... one problem. the pitch of the repeats is shifted quite a bit. the first repeat is noticeably lower than the original note, with each repeat getting increasingly higher in pitch. I checked all of the components, and all of the voltages and everything seems to be correct. I also tried many different combinations of components (charge pump, no charge pump, no 12v regulator, etc), and a bunch of different power sources. no real change. I searched around quite a bit, but sorry if I missed this posted already. any ideas?

thanks,
adam

[anchovie]

Have you tried another PT2399?

[robottrouble]

Have you tried another PT2399?

not yet... ordering replacement parts today from smallbear.

[robottrouble]

ok, got the new parts today. swapped out the PT2399 and the 571, but I'm still getting the same result. I looked over all of the discrete components and everything seems to be correct. all voltages seem to be correct, except pin 10 on the 571 seems a little low. ~4.5V instead of 6V. any ideas?

[robottrouble]

I've worked on this the past couple nights and still haven't found anything useful. It seems like when you first hit a note is when it slows down, and then ramps back up to the correct frequency about a second later. if you hold a note, you can hear that the repeat starts slow, and then ramps up to the correct pitch. To me, it almost sounds like the initial hit increases the load on the circuit, which drops the current to the PT2399 and slows down the sample rate. That seems like it would be a problem with the SA571 compander, but I've replaced it and that didn't seem to fix anything. I also connected the circuit to a regulated power supply and noticed that the current draw does increase when I hit louder notes. I guess that could be a problem with the TL072 as well, but I'm really thinking it has something to do with the 571.

if anyone else has ideas about this circuit, or knowledge of these chips, I'd really appreciate any help.

thanks,
adam

[Govmnt_Lacky]

Just a heads up but, I ran into some problems with the PT80 build a while back and it turned out to be two problems:

1) If you look at the GGG layout, at the bottom of the layout you see there is the 5088 tranny and the 102 cap. Look closely at the LEFT leg of the cap and you will see there is a very easily missed possible solder bridge there. It would connect the left leg of the 102 cap to the right leg of the 333 cap.
2) I had a BAD run of 2399 chips. I had 4 bad total. The key to finding them was that ALL of the bad ones had only 2 lines of alphanumeric characters on them. One line being the chip number (PT2399) and the other was a batch number (I assume). I compared them with a WORKING chip that I had and the working chip had 3 lines of characters on it. It is unlikely that you have this since you ordered from Steve however, it is still worth a look. The bad chips were causing a host of issues including pitch and runaway repeat problems.

Hope this helps and good luck,

[cpm]

my thoughts go into that direction too, some slight starvation is causing the change in internal references and clocking.

does the PT80 us a charge pump? that would be suspect...

also, how are the power lines and gnd wired? may be a somewhat high resistive path on power or gnd may cause senough fraction of a volt variation when current increases

[cpm]

2) I had a BAD run of 2399 chips. I had 4 bad total. The key to finding them was that ALL of the bad ones had only 2 lines of alphanumeric characters on them. One line being the chip number (PT2399) and the other was a batch number (I assume). I compared them with a WORKING chip that I had and the working chip had 3 lines of characters on it. It is unlikely that you have this since you ordered from Steve however, it is still worth a look. The bad chips were causing a host of issues including pitch and runaway repeat problems.

i've never user one of those "single line code" chips... but looks like its one of those on this "fab echo" board:
http://www.diystompboxes.com/smfforum/index.php?topic=64519.msg609677#msg609677

i am really interested in clearing if its just bad luck you had or a proven risk of deceiptive chips

[Govmnt_Lacky]

The bad chips I ran into had a printed format like this:

PT2399
PTC 0327Z

The good chips (Directly from Princeton I might add) were in this format:

PT2399
PTC CSH13Y
       AU1T21

OR

PT2399
PTC ISD35Z
       YR0938

The Xs are just batch numbers and letters. It may be a coincidance however, I found it odd that the BAD chips had that single batch line format and the GOOD chips from Princeton had 2 lines of codes. 

[robottrouble]

my thoughts go into that direction too, some slight starvation is causing the change in internal references and clocking.

does the PT80 us a charge pump? that would be suspect...

also, how are the power lines and gnd wired? may be a somewhat high resistive path on power or gnd may cause senough fraction of a volt variation when current increases

Yeah, it has a charge pump, and I really want to think that's part of the problem, but I've tried the configurations listed in the manual (tho I don't think the instructions are exactly clear) and I've tried other solutions, running straight 12V and pulling the charge pump and the 12V regulator. also, the charge pump is definitely outputting the correct voltage, and I'm getting even 5V and 12V past the regulators.

It's nice to get a bit of reinforcement that I'm on the right track, tho. I'll try some other power configurations/supplies tonight. I dunno if this is common or not, but is there a way I could limit the amount of current that the SA571 can draw? I suppose I could remove it completely and just hard wire its inputs and outputs, but I'm not sure how valid of a test that would be.

Also, thanks for the info on the PTs, gvmt. The one from GGG does look different than the one from SBE, but I'm pretty sure I checked them against your description in the other poast, and they all (4) have the exact same behavior, so I'm just assuming they're all good. No real way to tell without a known good chip. I'll post some pictures of them tonight, just for reference.

thanks,
adam