Matching capacitors in phasers?

[Rob Strand]

I must ask, what's more effort ?! Measuring the specs of four devices and setting three trim pots or measuring a hundred or two devices and then singling out a quad that still won't be bang-on anyway ??!??!

For you and I it's no big deal, we are used to doing that type of stuff.  A lot of beginners get lost in the details.

the meter I'm using (a Circuit Test DMR 6000) exhibits +/-0.5% error on DC Voltage readings and +/-0.8% error on Resistance readings ... since the errors get compounded in this case this gives a +/-1.3% error total

If the DC voltage is say around 2V, then 1.3% is 26mV.  Lesser meters might be 1.5% or 30mV.
(You could add the mechanical stability of the pot if you like.)

So we a making all those measurements but only ending up with little better than matched JFETs.    Sure we are starting from a set of totally useless JFETs from the point of view of a one trimpot circuit.  So the set-up as you have it is an enormous improvement.

not sure if that's what you meant by *small* criticism ?!

You gotta understand all these arguments figure in my mind before I decide on a final circuit design approach (pretty typical thing in bone-fide analogue circuit design no?!) ... why would I post a circuit solution that overlooks such things ?!

Exactly that - *small*.   As you have written it is totally useable.     (I know you know your stuff - no need to get defensive man!)

LOL ... I'm sure I would get more than you think ...
false presumptions can cause the mind to distort



please geek away if you think you can actually provide something tangible ... careful assuming you can beat the accumulated error my solution exhibits as I have a pretty good sense of possible competing approaches ... still I'm open minded, let's see what you got!

I said you *would* get it.      I'm tired typing proofs because noone gets it.   (The idea: take a voltage divider, in this case a pot.  measure voltage wiper to clockwise terminal, and voltage wiper to anti-clockwise terminal assuming the meter had a finite impedance and a voltage error.  Take ratio of the two measurements and see what happens to the estimate of the voltage divider ratio.  Watch the magic  )

If, on the other hand, Vgs(off) were to be measured using something like an Atlas tester, where Vgs(off) is extracted at the 5uA line, then the error levels rise dramatically; to probably something like 5~10% ... it's hard to say exactly in advance because we'd have to extract numbers both ways and do a bit of math to determine the exact accumulated error difference

Yes that thing come-up in a recent thread.  Really need to interpret those numbers correctly.

[diffeq]

If I were to measure anything related to voltage dividers with impedance higher than 150k, I'd use stable voltage reference and low offset CMOS opamp buffer. For example, buffered TL431 reference ->> divider ->> buffer ->> DMM. Just my two cents.

[Eb7+9]

If the DC voltage is say around 2V, then 1.3% is 26mV.  Lesser meters might be 1.5% or 30mV.
(You could add the mechanical stability of the pot if you like.)

30mV error is still vastly better than the poor matching technique everybody used previously ... but I get your point here, which suggests using lower valued trimpots to reduce error due to meter loading ... mechanical stability is not an issue with my trimpots, I don't think, but it's a good point to bring up here

I'm tired typing proofs because noone gets it.   (The idea: take a voltage divider, in this case a pot.  measure voltage wiper to clockwise terminal, and voltage wiper to anti-clockwise terminal assuming the meter had a finite impedance and a voltage error.  Take ratio of the two measurements and see what happens to the estimate of the voltage divider ratio.  Watch the magic

Ok now I see what you are getting at ... thank-you for explaining

Yes, doing it your way acts to cancel out a great part of meter loading in measuring trimpots positioning ... but with a high enough Zin on the meter all this becomes a moot point

Thanks Rob!

[ljudsystem]

This is (more or less) what's on my breadbord for the moment it works but it distorts and has low outupt level. So either I've made a wrong connection or there something wrong with "my" design.

It's something I'm working on by request of our guitarist. The ideas is to build two tremolo/vibrato bands, one high passed and one low passed, and send them to different amps. Sort of a mock Leslie thingy.

[tubegeek]

The ideas is to build two tremolo/vibrato bands, one high passed and one low passed, and send them to different amps. Sort of a mock Leslie thingy.

Hope you take requests: "How Soon Is Now" !

[Kipper4]

@ljudsystem

I'd look at a smaller input cap as some pre emphasis. It might help with the distotortion on the output.
That 10uf after the buffer might be better to be smaller too.

You'll need a cap between the collector and the ldr, so it have a filtering action otherwise it will act more like a tremolo. Go look at the magnavibe.

See the 100nf after collector before the ldr (changing the Lyman cap 100nf to a smaller value might help with the perceived distortion.)
Here

http://revolutiondeux.blogspot.com/2012/01/bigfoot-fx-magnavibe.html

I'm not sure the tremolo will work ever so well either. It has no resistance to work against.

-------/\/\/------------------
                   /
                  >
                  >
                  >
                   /
                  /
               Gnd

I guess it all depends on how much dip in volume you want.

The series R can be a pot but defo a low K ohms. The bigger it is the more it will contribute to thermal noise and distortion.



The other thing is you'll need the the trem to trem when the vibe ain't vibeing or your tremolo will dip the volume of the signal that wants vibeing.
I sounded right in my head. Hope it makes sense.

Does it sound ok other than the distortion?

Are you using 1 led to wobble the 5x ldrs.
It might sap the power supply some driving that many leds if not.


How are you going to filter the low pass and high  passed signals, passive or active filters?

You might also need some series R between the buffer and the passive mixer.
Or it might fight with the other output being mixed.

These things can take some work to get right.

I hope it goes well.

Edit.
PS have you got it on the breadboard?

[j_flanders]

This is (more or less) what's on my breadbord for the moment it works but it distorts

Maximum headroom is 1/2 Vcc, so 4,5V.
You need half Vcc for the collector output and half Vcc for the emitter output.
Correct bias is therefore critical. Ideal bias would be 3/4Vcc at collector and 1/4Vcc at the emitter.
Can you post the collector and emitter voltages of the 4 phase splitting transistors?
According to a sim, in your circuit they would be at 7.5V at the collector and 1.5 at the emitter, giving you 3V headroom for the collector output and 3V for the emitter output. Slightly less than ideal but should be good enough.

You'll need a cap between the collector and the ldr, so it have a filtering action otherwise it will act more like a tremolo. Go look at the magnavibe.

Magnavibe has the cap at the collector output and ldr at the emitter output.
ljudsystem has the cap at the emitter output and ldr at the collector output.

Both will work and the final outcome is the same. There's no need (or want) for an additional cap at the ldr.

[jonny.reckless]

This thread has reminded me of all the pain I went through trying to control the properties of JFETs and LDRs, and why I ended up using operational transconductance amps (OTA) for my phaser and vibe designs. They are cheap, consistent from unit to unit, and pretty easy to use with predictable properties for designing current controlled allpass stages. Also, the transconductance is well defined, which means that for a vibrato effect, you can generate a sinusoidal LFO, exponentiate it with a bipolar junction transistor, and get really nice natural sounding vibrato. This is what I did for the Reckless vibe.
https://www.diystompboxes.com/smfforum/index.php?topic=119211.msg1146780#msg1146780

https://postimg.cc/5HzsFfYy

I think it's very hard to get such a natural vibrato effect using either LDRs or JFETs.

In a phaser I use 5% tolerance polypropylene or polyester capacitors. Tolerance is less important in a vibe circuit since you are not trying to get the poles coincident.

[ljudsystem]

Johnny, that vibe sounds fantastic. Maby I should give up on optocouplers and get into OTA designs instead. They just look so scary with all those pins... 

Thanks for the tips Kipper, I'll try to update the design tonight. I use 4 leds since it was less messy on the breadboard plus I wanted to be able to adjust the depth individually for each stage.

I use a 9v battery, and it's not fresh out of the package either. Could that be the source of the clipping? I'll measure Vcc when I get home.

I havn't had much time to work on it the last few days though, my son (who is two) refuses to go to sleep at night. Last night he didn't fall asleep until 10 o' clock then wakes up an hour later and wants to watch TV 

[Kipper4]

I made this with 6 stages and a wet/dry switch. for ota phaser.

[jonny.reckless]

I made this with 6 stages and a wet/dry switch. for ota phaser.

Neat design! I'd love to hear how that sounds.

BTW, those 27k / 1k8 in the allpass filters around the OTA will make it a bit noisy. If you could scale them down to 2k7 and 180 ohms you'll find it's noticeably quieter, especially if like me, you like to put the vibe before the overdrive for classic Jimi / Trower tones. OTAs are very sensitive to resistive noise (i.e. hiss) on the base of the input transistor pair. I generally try to keep the base impedance to 100 ohms or less wherever practical.

[Kipper4]

Was sure there was a demo but cant find it in my tube.

However I changed my version of Ricks design to reflect the values you use.
Any reason this wont work ? Havent breadboarded it yet.

[jonny.reckless]

It looks OK to me 

[Eb7+9]

hey Rob, I want to thank you for your critique ... you're entirely correct as far as meter error swamping Vgs(off) measurement accuracy ... of course with a high quality meter (something with a Zin of 20 or 50megOhm) it then becomes a mute point ... my oversight for relying on a cheaper unit with a mere Zin of around 1meg

In the process of thinking about what you said I've come up with a jFET driving technique that relies on a common (mirrored) current instead of direct (scaled) voltages ... all that is needed are variable trimpots to convert current to Gate voltage for each device ... by using a unipolar current that goes from 0 to Iref a voltage can be produced that goes from 0 to Iref x Rtrim ... with later term being set, now, by ear instead of thru measurement ... thus, doing away with voltage measurement and accuracy altogether

This insight would not have occurred to me without me challenging you to your original comment, yeah I needed to know what what you meant exactly ... so, as a result I have a new jFET phasor architecture that will allow very wide and balanced range of control using unmatched devices, again with each trimmer adjusted to reflect varying Vgs(off) parameters ...

I thank you again for that ...

[Rob Strand]

In the process of thinking about what you said I've come up with a jFET driving technique that relies on a common (mirrored) current instead of direct (scaled) voltages ... all that is needed are variable trimpots to convert current to Gate voltage for each device ... by using a unipolar current that goes from 0 to Iref a voltage can be produced that goes from 0 to Iref x Rtrim ... with later term being set, now, by ear instead of thru measurement ... thus, doing away with voltage measurement and accuracy altogether

Hey, pretty cool.