Old Roland AP-II almost working... Busted FET?

[Beros]

Hi..

So I got this Roland AP-II off ebay. The seller said that it was working, but it's not. Instead of sending it back I thought i'd try to fix it.

The problem is that the effect clicks on, and you han hear a change in the guitar tone, but no there's real phasing going on. When I turn the internal trimpot, the sound changes, but still it's kind of "stuck" and wont swirl.

Here's the schematic:
http://rkerkhof.ruhosting.nl/Posts/AP-2.gif

What would my best bet be? To change the FETs?

[duck_arse]

start with checking the rate pot and its connections. it may be the osc isn't. follow the osc output, and check all the parts leading to the fets and their bias setup. it sounds like the bias is still ok, if the trimm shifts the sound. the fets have no real reason to be bad.

[R.G.]

The problem is that the effect clicks on, and you han hear a change in the guitar tone, but no there's real phasing going on. When I turn the internal trimpot, the sound changes, but still it's kind of "stuck" and wont swirl.
...
What would my best bet be? To change the FETs?

The JFETs don't do movement, the low frequency oscillator does. Changing the FETs is a complete waste of time and has the potential to make things worse.

The problem is that the LFO is not oscillating. Why it's not oscillating is not clear. To figure that out, you'd need to read and follow "what to do when it doesn't work", the sticky that's at the top of the list of posts in the forum that you have to skip over to get to the other posts.

In this case, just getting pin voltages for the two opamp sections in the LFO might be enough.

[Beros]

Hey.. Thanks for your replies..

Ok, last time i measured the batteries they were at 17.5 v, now they're at 14v. And with that I get:

The left hand one ( i guess it's IC1 on the schematic):
P1 0
P2 0
P3 0
P4  5,23
P5 0
P6 0
P7 0
P8 0
P9 0
P10 0
P11   -8,79
P12 0
P13 0
P14 0


The other one:

P1 0
P2 0
P3 0
P4  5,32
P5 0
P6 0
P7 0
P8 4,05
P9 -1.42
P10 0
P11 -8,78
P12 0,37
P13 0
P14 -8,16
The connections to /from the rate pot seems fine..

Another weird detail is that when i hit the switch it kind of phases a little bit, then stops after half a second or something.

[R.G.]

Ok, last time i measured the batteries they were at 17.5 v, now they're at 14v. And with that I get:
The left hand one ( i guess it's IC1 on the schematic):
...

I don't have time to dig through these in detail at the moment. Probably tonight.

Another weird detail is that when i hit the switch it kind of phases a little bit, then stops after half a second or something.

That's another vote for it being the LFO. That's a common behavior for an LFO that is disturbed by something (the switch in this case) but doesn't have enough signal/feedback/something to keep it going.

[Mark Hammer]

Is the bias trimpot where it should be?

EDIT:  Think of the FETs as huge boulders.  If you have enough of an initial push to get it moving, an extra set of hands will push the boudler faster/farther.  That extra set of hands will do nothing much on its own, though.  The FETs have to be biased into the appropriate range to work properly and yield sweep.  The LFO is added to that bias voltage like the extra set of hands. 

Most of the time the bias trimpot is factory set and secured via a dab of something to set the trimpot in place.  BUt, given the age of the pedal, the trimpot has had plenty of time to be jarred loose from its optimal setting...plus, it gets kicked a lot. 

I am not saying the bias is necessarily the problem, but it is a contender for a possible reason.  Especially given that you report a sonic change in effect-mode, but no sweep.

If the trimpot is NOT glued in place, then I would recommend setting the unit for a fast sweep, with maximum depth (so you'll be able to hear when you've reached the sweet spot) and adjusting the trimpot in search of sweep effect.  If the trimpot IS glued in place, then don't move it.

[R.G.]

Got some time to look. Pin 4, the +V supply pin on both opamps, is measuring a bit over 5V, no the nominal 9V the schemo says it should.

Until that gets corrected, not much else can be counted on to work or not work correctly. If you're running on batteries, put in fresh ones, then try again with new voltage measurements.

[Beros]

I put fresh baetteries in now, and it measures 9.5 on pin 4...

...And it started to work!

Thanks and sorry for not checking the obvious before bothering you wizards.

[Mark Hammer]

Could have been something more embarrassing, like not inserting the plug all the way.  You got off easy! 

[Beros]

All right. So It's up and running now with a millenium bypass. Sounds great, but the output is a bit low.. Not much. Just a tad. What would be the best way to up the volume? Work with the Q2 values? Or could even decreasing the value of R13 make it a bit louder?

[Mark Hammer]

R11/R12 mix the dry and wet signals in equal proportion.  R13 forms a voltage divider in conjunction with R11 and R12 (i.e., the output is attenuated by the ersatz 110k "pot) formed by R11/R13 and R12/R13).  Assuming the level difference is relatively minor, you could consider raising R13 to something between 150k and 220k.

If the level difference is more than miniscule, then you may want to tinker with the gain of Q1, since it adjusts the level of both dry and wet paths.  If I knew enough to tell you how to do that, I would, but I'm not that clever, so I'll leave it to others.

Note that, while FETs are susceptible to distortion if pushed too hard, this circuit incorporates the RC networks around each FET, that yield increased immunity to distortion.  It's not perfect, but it should be able to safely handle a nudge in an upward direction.

[Beros]

R11/R12 mix the dry and wet signals in equal proportion.  R13 forms a voltage divider in conjunction with R11 and R12 (i.e., the output is attenuated by the ersatz 110k "pot) formed by R11/R13 and R12/R13).  Assuming the level difference is relatively minor, you could consider raising R13 to something between 150k and 220k.

If the level difference is more than miniscule, then you may want to tinker with the gain of Q1, since it adjusts the level of both dry and wet paths.  If I knew enough to tell you how to do that, I would, but I'm not that clever, so I'll leave it to others.

Note that, while FETs are susceptible to distortion if pushed too hard, this circuit incorporates the RC networks around each FET, that yield increased immunity to distortion.  It's not perfect, but it should be able to safely handle a nudge in an upward direction.

Cool.. Could lowering the value of both R11 and R12 make it a bit louder? What would happen if those resistors have too low values?

[Mark Hammer]

What would happen?  I don't know.  But I don't think great harm would be done by making them both, say, 4k7 to 6k8.