Pearl PH-03 -How to Increase Output Volume (w\schematics)

[HeaD]

Hi! I have this Phaser:

http://img509.imageshack.us/img509/1517/pearlph039my.jpg

Unfortunately when is engaged, the output volume decreases. I have thought that I could add a minibooster after the phaser output, but I would like to know if there is any other way to increase volume.

Maybe some resistor value to modify? Thanks to all

[johngreene]

The only way I can see to 'gain some gain' would be to increase R22. Maybe try a 10K pot in series with the existing 56K and adjust until you get equal volume as when bypassed and then measure the pot and add a fixed value in it's place.

--john

[Bernardduur]

Hmmmm, weird. Mine has no issue

[HeaD]

thanks johngreene, I'll try!!!

Bernardduur, are you sure? It doesnt decrease either a little?

[MartyMart]

I have one, if anything, I'd say there's a slight "increase" in level when engaged 
Is everything else working/sounding right ? ..... battery at 9 volts ....

MM.

[Bernardduur]

Yes, it is either a slight increase than a decrease......

They tend to differ between each other. I once had one that was on a moderate high speed setting on it's lowest "knob-speed" setting and there was nothing wrong with it (he used different opamps, that was all)...

[Mark Hammer]

Sounds like another case of 5% tolerance resistors.

I'm with John when it comes to suggesting an increase in the value of R22.  The one caveat I would add, though, is that R22 and C21, as shown, already roll off the high end starting around 2.8khz.  That will make the pedal a nicely warm and hiss-free pedal.  Increasing the value of R22 will have the ffect of making that rolloff lower, however.  For instance, a 68k feedback resistor will drop that rolloff down to 2.3khz

So, the optimum strategy would be to increase R22, while at the same time decreasing C21, so as to maintain the same bandwidth and tone, while increasing output level.  If you use a value of 69k for R22, and replace C21 with an 820pf cap, you wil restore the 2.8khz bandwidth while giving a bit more output.  If the thought od what the phaser might sound like with a bit more treble sounds interesting, use a 680pf or 560pf cap. 

The general principle, though, is that as R22 gets bigger, C21 needs to get smaller to maintain constant bandwidth.

[R.G.]

Why wouldn't you just make R7 be 100K to double the signal level (+6db that is)? Or reduce R5+R6 since the input is buffered.

Both the dry and phase line are fed from that first opamp, so increasing its gain does it all.

IC4 looks suspiciously like a CD4007.

[johngreene]

The LED must flash when you engage the pedal too. It looks like they are using a Lambda Diode oscillator in the LED circuitry.

--john

[Bernardduur]

IC4 is MC14069 UB.

I've never seen my LED flash though

[HeaD]

Ok, I have increased the R22 value (from 56k to 100k). The volume is quite increased and tones/bandwidth are not changed! Thanks to all!!! 

[Mark Hammer]

Why wouldn't you just make R7 be 100K to double the signal level (+6db that is)? Or reduce R5+R6 since the input is buffered.

Both the dry and phase line are fed from that first opamp, so increasing its gain does it all.

IC4 looks suspiciously like a CD4007.

Well here's the thing.  We don't know what IC4 really is, but it is conceivable that it is some sort of FET or MosFET array.  Can those transistors handle increased input signal amplitude cleanly?  Don't know.  From that perspective, the safest strategy I would think is to do the boosting n the output stage.  Ideally, hotter input signals result in lower S/N ratios in the output signal than would boosting the output stage, so your logic is sound, but in this case there would seem to be at least as much concern about FET distortion as about S/N ratio.

Of course, if it WERE a 4007, and if the anticipated signals they were handling were well within an acceptable range, I would openly and willingly prostrate myself to the suggestion of turning the gain up at input.  That's not a particularly pretty sight, though! 

[GFR]

Well here's the thing.  We don't know what IC4 really is, but it is conceivable that it is some sort of FET or MosFET array.

It is a 4069 cmos hex inverter for sure, as Bernarddur pointed out. There are 4 inverters used as VCRs. A 5th inverter (pins 8-9) is used as a "reference" in a loop to keep the other inverters correctly biased (if they're all matched).

And IIRC the Led is supposed to flash to warn you about a low battery. The 10k trimpot sets the low battery "threshold".

Edit: speaking of Pearl phasers take a look at the PH-44, it's very unnusual.
http://filters.muziq.be/model/pearl/ph44
It uses CMOS inverters as OTAs.
At
http://www5b.biglobe.ne.jp/~houshu/synth/
There's an explanation on how it works.

[Mark Hammer]

If it's a 4069, then all the more reason to be cautious about input level from what I understand. Somewhere in the archive are Mike Irwin's comments about the ETI 6-stage phaser that uses a 4049 as the control elements.  The S/N ratio is not especially great because the input signal has to be attenuated to accommodate the 4049's needs.  There is gain recovery after the phase-shifted and clean signal are combined, but that results in a noisy output, especially if any amount of resonance is added.

The chart that Sam Hoshuyama has (and to which I often refer), depicting the different ways to trade off FET headroom against tracking, shows that you can increase FET headroom somewhat (though never to the point where it equals that of LDRs) by means of an RC network bridging drain and gate.  If memory serves, Mike tried this out with the ETI circuit (it can be implemented with a hex invertor though don't ask me how), and found that he could successfuly increase the input signal level going to the series of allpass stages without risking significant distortion.  Ultimately it assisted in getting a better S/N ratio though certainly not what one might get from LDRs.

I have to say that the relative imunity to distortion by the transistors in a CMOS hex invertor, compared to the distortion-immunity in the FETs used in the many many thousands of existing phase shifters that make many many musicins very happy, is something I know very little about.  It may be much less of a problem than I imagine it to be.  Still, the rather low rolloff that Pearl uses on that mixing stage (<3khz!) suggests that noise IS something that needs attention in that design.

[GFR]

The chart that Sam Hoshuyama has (and to which I often refer), depicting the different ways to trade off FET headroom against tracking, shows that you can increase FET headroom somewhat (though never to the point where it equals that of LDRs) by means of an RC network bridging drain and gate.  If memory serves, Mike tried this out with the ETI circuit (it can be implemented with a hex invertor though don't ask me how), and found that he could successfuly increase the input signal level going to the series of allpass stages without risking significant distortion.  Ultimately it assisted in getting a better S/N ratio though certainly not what one might get from LDRs.

I think that's what the 680k resistors at pins 1-2, 3-4, 10-11, 12-13 do (linearization for less distortion).

Taking the first phasing stage IC2, if you consider pin 3 is the gate, pin4 is the drain and the source is at pin 7 it looks more familiar. The LFO is applied to the source (pin 7) while the gate is held at a fixed voltage.

The 8-9 gate, IC5 and Q2 are there to help with the "don't ask me how" part