MXR PHASE 90 SCHEMATICS...

[Gordonjcp]

No, look at the circuit diagram.

http://www.geofex.com/FX_images/p180.gif

That's not a great diagram, but it's better than nothing.  Now, look at the bottom right of the diagram.  The positive terminal of the battery goes through a 10k resistor to a Zener diode (there's also a normal diode to protect the opamps in case you wire the power up backwards).  Across that 250k trim pot you should see around 3V - this is used as the "zero" reference for the opamps.  To work properly, opamps need a positive and negative supply.  Since you've only got a single 9V supply from the battery, you can split that and pretend that it has a dual-rail supply.

Okay, so what have we got?  Well, you'll see that there's an opamp all by itself at the bottom.  That's the LFO.  The 500k pot sets the speed, along with the 10uF capacitor.  A proportion of the voltage across the cap is fed into the inverting input of the amp by the 150k resistor.  As the cap charges up, the output of the amp goes down and tries to discharge the cap through the 500k pot.  With the pot at its highest resistance it can only discharge slowly so the LFO oscillates slowly.  Reduce the value of the resistor, and it will discharge more quickly.  Once the cap has discharged the output of the amp starts going positive, and tries to charge it up again.  Once more, the resistance of the speed pot controls how quickly this can happen.

The practical upshot of all this charging and discharging is that you should have a rough squarewave at the output of the LFO opamp, and a rough triangle wave at the top of the 10uF capacitor.  This triangle is used to modulate the phaser.

So how does it phase?  Well, four of the five opamps in the top row are all-pass filters (the left-hand one is a buffer amplifier).  All-pass?  Isn't that just a bit of wire?  No.  An all-pass filter doesn't affect the frequency response but it *does* affect the phase response when it's adjusted - different frequencies are delayed by different amounts.  The 0.05uF capacitor and the 22k resistor and FET set the response.  The output of the LFO is fed to the FETs via that 3.3M resistor, along with a bias voltage from the trimpot (through the 1M resistor).  You can use really high value resistors like that - and indeed you *need* to - because the FETs have a very high input impedance, so they only need a tiny current flowing to make them work.  As you adjust the bias control, you set the "centre" point of the phaser.  If you've fiddled with the bias pot, set it about half-way to begin with.  The last piece of the puzzle is the 22k resistor providing feedback to give that deep swooshy phasing effect.

Finally, the last stage of the phaser is the output amplifier, which is the transistor labelled "2N4125", and the resistors and capacitors around it.  There's nothing particularly clever or interesting about that circuit, except maybe the 0.05uF cap on the output which seems a little small - you might get better bass response by fitting a larger capacitor there.

Right, so how can we track down crappy distorted sound?  Well, start off by inspecting both sides of the board for any damage, poor soldering or solder blobs from previous repair attempts or modifications.  If you see any, put them right first.  If you don't know how to solder, find something to practice on first - crappy soldering is the main reason why people fail to get kits working and end up making things they're repairing worse!  If you're not happy setting about your prized phaser with a soldering iron, find someone to help you.  You really don't want to make it worse.

Next, you want to set that bias pot about half-way if you suspect it's been fiddled with, and set the speed control about half-way up too.  Measure the voltage across the bias pot - it should be about 3v.  If it's much less than 3v, then something is shorting out the bias supply.  Either one of the diodes, the capacitor across them or the trimmer has failed.  If the bias voltage goes up and down as you turn the trimmer pot then something is wrong around the wiper of the pot or the 1M resistor.  Measure the voltage at the wiper of the bias pot - with it about half way you should get 1.5V or so.  Next, check that you've got about 9V across the supply pins of the 741s (pin 4 is -ve and pin 7 is +ve).  They should all be the same.  If they're not all the same, you've missed a bit of crappy soldering, or maybe a broken track.  Also, check that you've got about 2.5V-3V on the non-inverting input of the opamps - measure between pin 4 -ve and pin 3 non-inverting input.

If you've got this far and everything checks out, then measure the voltage between ground and pin 6 of the LFO opamp.  This should be going up and down, and vary as you change the speed pot.  If it doesn't, then temporarily wire a resistor of about 220k to 470k across where the speed pot should go - it might be that the pot is knackered.

Right, so the LFO is O-ng, the bias circuit is biasing and you've still got crap sound.  Feed in a signal at about instrument level (ie. don't drive it with a speaker level, because that *will* distort).  Slowly adjust the bias pot from all the way down to all the way up.  If you find a spot at which the phaser works properly, great - you've fixed it.  If you found a point where it was "better but not right" then leave it there, otherwise put it back to the middle.

Now you're going to need to trace the signal through the circuit.  I would use an oscilloscope for this at least to begin with, but often if you're tracking down distortion in a circuit you just want an audio probe.  You can make one with a small amplifier and speaker.  Something like an amplified PC speaker will work - make a lead to plug into it with a pair of clips on the other end.  Clip the ground connection to the ground in the phaser pedal, and clip a capacitor about 0.1uF or so (not important as long as it's more than about 0.01uF and less than about 1uF) in the one that goes to the input.  "Buzz" the input with your finger.  Bzz Bzz.  Yup, that's working.  TURN THE VOLUME OF THE AMP DOWN!  You don't want to launch the speaker cone across the room if it goes wrong.  You could wrap some sticky tape around it to stop your fingers buzzing the input.

Dab the free end of the capacitor you're using as a probe onto pin 6 of each of the "phaser" opamps in turn, starting from the input.  Turn the volume up until you hear the signal, remembering to turn it back down before you try the next stage.  We're listening to the output of each section, so if you find one that's quiet and distorted you've found the faulty stage.  The first stage is just a buffer amplifier, which has no voltage gain (doesn't make the output bigger or smaller) but does have current gain (makes the output "stronger", more able to drive a load.  Think of it as power steering for audio.) so the output should be much the same as the input.  As you go up each stage you should hear a more pronounced phasing effect.  If you've got lovely deep swooshy phasing at the end of the last opamp, but still crappy output then there's a good chance that output transistor has failed - replace it with any small-signal PNP tranny.  I'd use a BC558 because I've got a bag of about a thousand.

If you find a phaser stage is faulty, try and work out what's gone wrong.  The two most likely causes of problems would be the FET and the opamp.  Desolder the FET first, taking care to note which way round the legs go.  You might find the phaser starts working without it, in which case you know it's the FET.  If you find one has gone, you might have trouble getting a replacement.  There's nothing to stop you replacing all four with identical FETs - 2N3819 would be suitable and is a cheap and readily-available part.

It's not the FET?  Might be the opamp then.  Desolder the end of the capacitor that connects to pin 6 of the opamp you're suspicious of, and do the same for the opamp that precedes it.  Temporarily run a short wire from pin 6 of the preceding opamp to the dangling end of the capacitor that would go to pin 6 of the suspect chip, bypassing that stage.  Works now?  Change the chip.  Desoldering and replacing the chip is probably easier if you snip its legs away and remove the legs one at a time.  It's far better to use braid or a proper desoldering tool.  Exactly how you get it out is up to you - try not to ruin the board, the copper tracks will lift quite easily if you overheat them.

Right, if you've made it all the way to the bottom, congratulations.  I hope you get it working.

[jimmybjj]

@ Gordonjcp

I just registered to this forum just to say thanks for your post, very educational. I really enjoyed reading your post, especially to someone just getting started in electronics. I just built the tonepad version of the Phase 90 and haven't got it working yet, but I'm going to try and follow your informative post and see if i can get it going. I'll post again with my results either way, singing your praises or requesting help  thanks again.

[jimmybjj]

I got it fixed! thanks, I don't know if I would have ever got this working if you hadn't written that "walk thru" I ended have having a faulty trim pot. thanks again.

do you happen to have anymore post about guitar effects? I really learned alot fromyour previous post.

[Gordonjcp]

Post a link to a circuit, and I'll see if I can explain it.

[jimmybjj]

Post a link to a circuit, and I'll see if I can explain it.

really? awesome! how about the Ross compressor?
http://www.generalguitargadgets.com/pdf/ggg_dnr_ross_sc.pdf

[Gordonjcp]

Sorry, didn't notice this!  I'll reply in a new thread, since we're drifting off-topic.

[jimmybjj]

did you ever get around to describing the compressor? i enjoyed reading the mxr on and learned alot

[Gordonjcp]

I started to build it to work out exactly what was going on, and then Real Life intervened.  I will get back to it ;-)

[jimmybjj]

I started to build it to work out exactly what was going on, and then Real Life intervened.  I will get back to it ;-)

Great!!! Looking forward to it. Do you By chance have any other walk throughs? I find your explanations easy to follow.