SIx Stage Phase B. Report

[petemoore]

I'd like to try that, but I'm getting lost with all the many caps values discussed for the particular 'enhanced depth' mod.
 I follow the basics of the texts, as applied to the phazer, just came back up from some retunings...
 If you state just the values you're suggesting for stages 1-6, I'd be glad to check stox and plug what I have that on or close to the values.
 Are you saying
 3 Stages 'in a row?' with 100n's...or [code '103'] or .01uf ?
 and the last 3 stages with 2n2's ?

Also after messing about with these caps...
 Is there a way to test how well each stage is doing ? Some seem more sensative or responsive to cap value changes.
 I set the input cap at .022uf.

[petemoore]

After the LDR posts...
That said, I re-Jfetted my Phaze 90, 4 stage shifter, now that I'm able to get Matched sets.
  I semi.vibed the capacitor value stggering on the phase shifts when I built it.
  This one is getting a great wobble, more of a pronounced phase swoop than the 6 stager is getting.
 It makes me think...Hmm...what's a fine tuned 6 stage phazer 'supposed to sound like in terms of sweep depth compared to a 4 stager.
 The Phaze 90 was built from the tonepad schematic. I'll see if I can get a link...
 It could well be something's just not quite right in the 180. Seems like it should be good, all the knobs work, OA's are biased...etc. Jfets are very closely matched. Maybe someone knows where to look for something to try.
 Anyway, my matching by ear, final chapter, is that Jfet matcher makes phazer work much 'more' than I was getting by ear....after trying some different ones, a Jfets matcher is like mandatory for Full Phazer Function...I'd wondered about that.

[petemoore]

http://www.tonepad.com/getFileInfo.asp?id=42
 I think that's it..

[stm]

I'd like to try that, but I'm getting lost with all the many caps values discussed for the particular 'enhanced depth' mod.
 I follow the basics of the texts, as applied to the phazer, just came back up from some retunings...
 If you state just the values you're suggesting for stages 1-6, I'd be glad to check stox and plug what I have that on or close to the values.
 Are you saying
 3 Stages 'in a row?' with 100n's...or [code '103'] or .01uf ?
 and the last 3 stages with 2n2's ?

Pete,

Your phaser has six PHASE stages, each having a capacitor that will define the overall phaser signature.

My suggestion is: on the first three phase stages, use a 100nF capacitor on each phaser stage (100nF = 0.1uF = 104).  On the following three phase stages use a 2n2 capacitor on each stage.  In other words, instead of having six different-valued capacitors, you will have two different groups with three identical values in each.

Will see if I can post some curves later.

Regards,

STM

[petemoore]

Excllent...makes sense and I will try those values.
 I'd tried a silimar setup, will try those exact values, because I have them there IN AVX box Metal Films, very socket friendly.
 I think the LFO is getting loaded or taxed by the extra stages. I say think as more like suggestion of possiblity.
 Yupp I re-Fet'ed my Phaze 90, and it is producing [thing does sound really great] and has a more profound sweep sound, more depth than the 6 stage withonly it's 4 stages.
 I don't think it's the cap values, because I tried a wide variety for values, the 90 I think followed a suggestion post of cap values for Vibing a 90.]
 So the 90 sounds just great, the 6 stage is weak in comparison at this time.
 What about a buffer for the LFO ? Just seems to ring a bell or something.

[stm]

Good.  Regarding the weakness and the buffer for the JFET's, I'm not really sure.  Maybe the range of the sweep is not large enough.  You should try increasing the range of the sweep until you find it is getting into a dead zone on each side of the sweep, then backoff a little.  You also need to have a "centering" adjustment on the sweep, so when you increase it until it gets to the dead spots it does it simultaneously on both sides.

An important aspect of phasers and flangers is to have a large range of variation produced by the sweep.  In other words, this is equivalent to varying the equivalent FET or LDR resistor in a range at least 10:1, and ideally over 30:1.

Regarding the two values I mentioned before (100n and 2.2n) I made additional simulations and have more faith in using 100n and 4.7n

This is a graph of the phase curves:



RED: six equal caps 22nF (0.022u)
BLUE: six different caps (NatSemi spacing)  100n, 47n, 33n, 15n, 10n, 4.7n
GREEN: two groups of three equal caps (recommended) 100n, 100n, 100n, 4.7n, 4.7n, 4.7n
PINK: two groups of three equal caps (perhaps too much curvature) 150n, 150n, 150n, 2.2n, 2.2n, 2.2n

Regards,

STM

[petemoore]

Good.  Regarding the weakness and the buffer for the JFET's, I'm not really sure.  Maybe the range of the sweep is not large enough.  You should try increasing the range of the sweep until you find it is getting into a dead zone on each side of the sweep, then backoff a little.  You also need to have a "centering" adjustment on the sweep, so when you increase it until it gets to the dead spots it does it simultanously on both sides.
 <<<I'm not sure if this refers to tweeking the knobs I have on it or doing something else. "Increasing the range of the sweep" sounds like a very good idea, I'm not sure how to go about it though.

An important aspect of phasers and flangers is to have a large range of variation produced by the sweep.  In other words, this is equivalent to varying the equivalent FET or LDR resistor in a range at least 10:1, and ideally over 30:1.
 >>>Is there a way to measure this, or how do I try to maximize the range?

Regarding the two values I mentioned to rey (100n and 2.2n) I made additional simulations and have more faith in 100n and 4.7n

This is a graph of the phase curves:



RED: six equal caps 22nF (0.022u)
BLUE: six different caps (NatSemi spacing)  100n, 47n, 33n, 15n, 10n, 4.7n
GREEN: two groups of three equal caps (recommended) 100n, 100n, 100n, 4.7n, 4.7n, 4.7n
PINK: two groups of three equal caps (perhaps too much curvature) 150n, 150n, 150n, 2.2n, 2.2n, 2.2n
 I've tried all .047uf's...for a 'basic phase' sound...
 I'm going to try the "GREEN" mod. today and will report on that.
 Thank you for all the help STM and everyone !!
 Any future help on 'increasing the range of the sweep' is greatly appreciated and thanked, 'in advance !!!  

Regards,

STM

[stm]

Mmmhhh...

Checking the sweep range with an oscilloscope is far more easy than any other means.  I will assume you don't have one though, and try to figure out a way of doing so with a DVM (digital voltmeter, multitester, or even a needle meter like a Simpson 260)

1) First, some tweaking of the LFO speed is necessary during the measurements.  You need to make it sweep real slow, so you have time to see the DVM readings (they update the reading about 2-3 times per second).  To do this, set the RATE or SPEED control to the minimum. If not slow enough, multiply the LFO capacitor 10 times.

2) Measure the drain-to-source Resistance of one FET with the (+) lead on the drain and (-) lead on the source.  Wait for the minimum resistance value to appear, then for the maximum resistance reading and write them down.  Repeat the above minimum/maximum readings now with the test leads with the opposite polarity.  When done, take the average of the two minimum and then the two maximum resistances (FETs are not exactly symmetrical!--this is why they distort on high input levels).  The ratio of the average minimum and average maximum resistances measured will give you an idea on how you are doing.

3) If using LDR instead JFET, polarity of the DVM is irrelevant, so you need to take just one minimum and maximum reading.

4) Now read the gate-to-source VOLTAGE (don't forget to set the instruments for Volts reading now) on one FET and write down minimum and maximum values.

5) Assuming you have no parallel resistor with the FETs, your goal could be to adjust the range and centering of the LFO control voltage so you get equivalent FET or LDR resistances varying between 2k and 50k (25:1 ratio) or even better like 1k to 100k.

6) If you matched FETs before, measuring on one device should give a reasonable indication.

7) The adjustments on the LFO are done usually on trimmers, not on pots accessible to the outside. This depends on the particular LFO circuit you are using.

Good luck!

STM

[petemoore]

Mmmhhh...

Checking the sweep range with an oscilloscope is far more easy than any other means.  I will assume you don't have one though, and try to figure out a way of doing so with a DVM (digital voltmeter, multitester, or even a needle meter like a Simpson 260)
>>>Yupp I have no Osci Scope...
>>>This is 'alot' about which I know 'a little'..
1) First, some tweaking of the LFO speed is necessary during the measurements.  You need to make it sweep real slow, so you have time to see the DVM readings (they update the reading about 2-3 times per second).  To do this, set the RATE or SPEED control to the minimum. If not slow enough, multiply the LFO capacitor 10 times.
 >>>It's able to go probably slow enough the way it's set up, with speed at min...I'll try it and mod if necessary.

2) Measure the drain-to-source Resistance of one FET with the (+) lead on the drain and (-) lead on the source.  Wait for the minimum resistance value to appear, then for the maximum resistance reading and write them down.  Repeat the above minimum/maximum readings now with the test leads with the opposite polarity.  When done, take the average of the two minimum and then the two maximum resistances (FETs are not exactly symmetrical!--this is why they distort on high input levels).  The ratio of the average minimum and average maximum resistances measured will give you an idea on how you are doing.
 ...I can do this...
3) If using LDR instead JFET, polarity of the DVM is irrelevant, so you need to take just one minimum and maximum reading.
 >>> Jfets so far...
4) Now read the gate-to-source VOLTAGE (don't forget to set the instruments for Volts reading now) on one FET and write down minimum and maximum values.
 >>>I can do this too !!!
5) Assuming you have no parallel resistor with the FETs, your goal could be to ">adjust the range and centering of the LFO control voltage<" so you get equivalent FET or LDR resistances varying between 2k and 50k (25:1 ratio) or even better like 1k to 100k.
 >>>This is done with the "Bias' Knob ?? I'd put the parallel resistor between the Jfet gates and the LFO...and that would____ ?
6) If you matched FETs before, measuring on one device should give a reasonable indication.
 >>>Very Closely matched set of 6.

7) The adjustments on the LFO are done usually on trimmers, not on pots accessible to the outside. This depends on the particular LFO circuit you are using.
 Which resistors would use trimpots on ?
Good luck!

STM

[gez]

Don't forget the National Semiconductor 6-stage phase shifter; it has a nice capacitor staggering.  This could be thought maybe as the sixtuple version of the Univibe

I'm a bit it late to this thread.  Sebastian, where does this schematic come from?  Is it from an application note or something, and if so is there a link to the rest of the document?  Just curious...

[stm]

Gez, the article belonged to a National Semiconductor Application Book.  As you may have noticed, the schem is on the firts page of this thread.  I don't have the original text, but it wasn't special, so you are not missing anything.  As I remember (read it around 1988 or so), just a basic description on the good matching of the JFETs, their rather large turn-off voltage (around 8V or so), the use of a 10K resistor in parallel with the JFETs to limit the maximum resistnace to 10k, and the fact the minimum resistance (with FETs full on or at 0V Vgs) of around 100 ohms, giving a total resistance variation of 100:1.  I found the schematic on the web some time ago.

Regards,

STM

[stm]

Pete:

The sweep range trimming and adjustment is done mainly on the LFO.   I would need to see the circuit you are using to be able to provide some pointers as to which you should tweak.  Usually there will be at least one pot to adjust.

Regarding the parallel resistor with the FETs, they are added just to limit the maximum resistance in case the FET goes fully OFF to avoid pops and other undesired behavior.  Notice that if you have matched FETs and properly tuned LFO you may not need them, or you can use a large value (over 100k) just in case.

Reducing its value will certainly reduce the Max-to-Min ratio of the total resistance, which is not what you want!

Regards,

STM

[petemoore]

Here's a link to the schem:
 http://www.geofex.com/
 I think that's it..
 GEO / Circuit Sweepings / Scroll down...
 PHaze 90 / 180 PLUS scheamtic.
 I'm sorry Im A bit new in this area and still a bit confuzed...
 I'll try taking some of those measurements, and maybe that'll bring it to light a little better for me.

[puretube]

ain`t it phase-cinating?

[petemoore]

ain`t it phase-cinating?

>>Sure IS...Keeping me on the edge of my bench for days !!!

[stm]

Pete:

The BIAS and DEPTH pots on the schem you mention are trimmers and set exactly what I mean.  BIAS sets the offset of the control voltage, while DEPTH sets its maximum variation.

The 22k resistors in parallel with the FETS set the maximum resistance.  You have to make sure at one extreme of the sweeping the FETs are getting almost OFF, so you effectively have the 22k equivalent resistance.

On the other side, when the FETs are ON (gate-to-source voltage around 0) you could expect perhaps something like 500 ohms resistance for the FETs, which should produce an overall resistance variation of 22000/500 = 44:1 (which is pretty good).

Make sure the above is actually happening by moving said trimmers so the effective gate-to-source voltage on the FETs ranges from 0 to -3 V or so.  Also,  measure the drain-to-source resistance on one of the FETs to see if you are getting the (theoretical) resistance variations I mentioned.

Regards,

STM

[stm]

ain`t it phase-cinating?

This is my favorite phrase of the Star Trek series.  Usually Mr. Spock says "Fascinating" when something very bad but at the same time most incredible happens, like when his brain was stolen to control a mainframe computer that took care of the life support system of a civilization that lived on caves...

Did I go too off the rails here?  :shock:   I think I'm working too much...  :|

[petemoore]

Ok....took some measurements..
 can't say I really understand what they mean other than understanding the basic G/S voltage variance controls D/S resistance...and I have both of those.
 I did however try the Green' cap mod involving .1uf's and .047uf's, and one 2n2. This is where it stays. definitely a more profound sweep with these values in this phaser of mine.
 Another thing I did was swap some OA's, ears decieve me or replacing the 1rst OA [input stage and first phase stage] from NE5532 to a 4558 made a slight gain boost and 'thicker sound.
 I put a TL072 in there just for kicks, seems like these OA's have a recognizable tone when used, slight as the differences may be.
 Then I tried the usual 'cheats'...
 Ross Compressor for input treatment makes a very nice phazer 'accessory', as did any clean boost, I think the compression allows me to keep the signal input high, but peaks not so high as to cause it to distort, hence greater percieved depth for general playing and nice thicker sounding 'usable' settings [boost helps gets the thickness out of it, but player must be more careful not to hit the strings so hard, at such boost levels].
 On the Big Amp, these Phazers are tons of fun...add FF and it's just heavenly !! Using a delicate hand and careful listening while playing a bit at different settings using the Mix Knob I think I wasn't actually doing enough of...I found setting the speed sort of high, then setting this where the wobble sounds, then a little left, a little right ... and playing a bit at each mix setting at a slower rate to hear what the differences sound like slower...tunes in a better depth of sweep...It figures !!!...I put a dot where the wobble sounded most intense with the rate kinda fast.  
 Oh I'm having tons of FUN, PLAYING with these STOMPBOXES !!!!!!  
 PHazer, Ge FF, Tycho, Echo, Tubes Speakers...how could it possibly be more fun than that