DIY Phase Meter?

[Guest]

Jakob
I have the meter but no PCB - and not quite sure of where to go to find the circuit (Phase Meter)
Of course I could buy it from Canford - at £200! (and I get another meter as well)
But the Canford pdfs with component layout (but no components) indicate a quite simpler circuit than the ones posted already
ie there is no way the SSL circuit or a similar one fits onto that board

 

[Guest]

Doing more and more research on this
I too am after a phase meter driver board taking a 2 track signal and displayign it on a centre is zero meter

The SSL schematic looks massively complex, franks is for LEDs, so I found this..

http://www.intersil.com/data/an/an9637.pdf

i quote..
4 ICs and a handful of components - provide 1% accurate phase measurements.. in this low cost circuit

I am a bit thick in these matters
So reference input could be Left,
and therefore Signal could be Right - but... there are 2 right signals

Help!
Or am I totally up the wrong tree here

Interesting though
They convert the signals to square waves then check the phase

 

[SSLtech]

[quote author="uk03878"]Interesting though
They convert the signals to square waves then check the phase[/quote]...from the 4th reply on page 1:

The most popular way to make things level independant is to amplify, clip, then compare the ratio of TIME that each channel is in the same pole compared to at opposing poles.

:green:

That looks like an RF type phase comparator. They don't use them for audio anywhere that I've seen...

I like the SSL approach rather better I have to say.

Keith

 

[clintrubber]

[quote author="uk03878"]So reference input could be Left,
and therefore Signal could be Right - but... there are 2 right signals

Help![/quote]
The signal input is balanced in that circuit.

Note the non-off-the-shelf parts, but for audio these could be easily substituted though. Each company says the components of their own application notes are easy to source and are low cost :thumb:

Bye,

Peter

 

[clintrubber]

That looks like an RF type phase comparator. They don't use them for audio anywhere that I've seen...

Edit: If I recall that circuit correctly it's in the end functionally equivalent to EXORing the signal (multiplying squarewaves), and that's not too rare for not-too-critical audio use.
Hey, let's not abandon it for audio because it happens to work up to 10 MHz  8)

I needed to ask for some other Intersil samples anyway, so while at it, I asked for a few of that application note as well. Looks like a handy circuit to have on the bench. And having two LEDbar-phase meters with somewhat differing indications, I'd like to know who's right 

 

[barclaycon]

I always liked the SSL one.
Did the job fine, and certainly alerted you to any out-of-phase stuff - if your ears hadn't already told you already!

Do you think this would make a good DIY project ?
I wonder if there would be reasonable interest.

 

[bradb]

This would be a majorly awesome DIY project for the forum.  I've ALWAYS wanted a phase meter that i can use to quickly check phase visually.

Hell, you may be working in a studio with NO control room and its impossible to hear out of phase drum sounds when you're in the same room.

 

[bradb]

Did anyone build any of these successfully?  Can you give us a feel for how it works in use?

 

[clintrubber]

To be honest (and/or as a sanity check), as we know there are various good plugins (also free ones) out there that tackle this job perfectly.
And from a plugin like this we can't say that it sounds better than the hardware 

But luckily ( ;D) there are still occasions away from any computer that justify a 'hardware circuit'.
The thing is, after getting a second Beh. Edison (which has a LED-bar phase correlation meter) I found that the both give slightly different indications. No big problem, and with some test-signals both boxes might be aligned (IIRC there's a pot in side). But if one would like then it could also be the justification for building a 'measurement-type of phase meter' like the Harris/Intersil circuit.

Bye,

  Peter

 

[sahib]

Attached image is a very simple circuit that displays phase ona DVM. It is from Building Valve Amplifiers by Morgan Jones page 255-256. It was apparently used at BBC Radio 3. In order not to infringe the copyright I have redrawn it, though excuse the botched discharge diode around the relay. I have also re-worded it.

Audio signals at the output of TL072s are applied to 311 comparators and converted to square waves. Both square waves are applied to the XOR gate. If both signals are in phase XOR outputs 0, if they are out of phase it ouputs 1. Because the change in phase from 0 to 180 degrees also changes the pulse width on the output of XOR, the mean level of pulse becomes directly proportional to phase. This mean level charges the 470nF poly cap on the output of XOR and scaled down to 180mV by 50K trimpot acting as the voltage divider. So the output has to be set to 180mV by the 50K trimpot. The D type determines if the inputs lead or lag each other. If Ch2 lags the ouput will be loaded with 1 which will turn on BC549 and activate the relay. The output of the relay will swap the polarity and activate DVM's +/- sign.

However, you can also connect a digital panelmeter on the output but in any case you have to make sure that either the DVM or the panelmeter has a resolution of 1mV so that from 0 to 180mV can be incrementally read.

Finally the usual de-coupling procedures have to be carried out on the power sonnections of all the ICs.

 

[clintrubber]

Attached image is a very simple circuit that displays phase ona DVM. It is from Building Valve Amplifiers by Morgan Jones page 255-256. It was apparently used at BBC Radio 3.

Hi,

Thanks for posting.

Most circuits we've seen here are in essence pretty alike.
From a distance they mainly look differently because of the method of displaying.
If we want to know a reasonably accurate phase-difference between two applied 'static' signals
then a DVM/DMM-type of display is more suited, but for a dancing mono/stereo indication
obviously a LED-bar is more suited (fast changing DMM-digits look cool, but provide little info).

Note that most circuits have a clear point where the 'calculation-of-info' goes
to the 'displaying-of-info', so adaptation from a meter to LED-bar (or vv) is easy.
(Connecting a mV-displaying meter to a cheap'n'simple calculating circuit might give
a false impression of accuracy though)

Regards,

  Peter

 

[sahib]

It was just another idea. It could also be adapted to visually more appealing display read-out. I was in the local civic amenity, dumping some plastic refuse today, and as usual I couldn't help practicing my midgy raking ritual where they dump discarded electronic equipment. I noticed this old Amstrad receiver with a centered tuning meter (as bcarso suggested). Took it home and dismantled in fifteen minutes. Image is attached. This could also be coupled to the output with further scaling of the output. The front glass is easily removable and replacing the facia with a new incremental scale layout would give a visual indication of the phase angle.

 

[Sleeper]

I wonder if anyone has a schematic for the Neumann u479.
Kubi has the pdf here  http://audio.kubarth.com/rundfunk/index.cgibut it doesn't have a schemo.

I've also tried a different implementation that's merely really distorted tl072's fed into a diode matrix.
I can't publish it because of copywrite issues...
It works OK, but it's a bit prone to strange behavior, enough so that I don't really trust it, and then what's the point.

Despite the complexity I'm pretty seriously thinking about drawing up a pcb for the SSL design.  
The component side of the neumann board is pictured in Kubi's PDF and it looks pretty simple.

Sleeper

 

[clintrubber]

[quote author= link=topic=9301.msg199094#msg199094 date=1153831593]

http://www.intersil.com/data/an/an9637.pdf

i quote..
4 ICs and a handful of components - provide 1% accurate phase measurements.. in this low cost circuit


[/quote]

As it turned out Ron Mancini also got a patent on it back then.
FWIW, but let's be complete:

http://www.pat2pdf.org/patents/pat5831423.pdf

Regards,

  Peter

 

[clintrubber]

And if you want to keep building to keep building then also make this,
a phase meter calibrator (don't ask for a calibrator calibrator  ):

http://www.audioxpress.com/magsdirx/ax/addenda/media/hansen2674.pdf

but you won't actually need it for calibrating for instance the Intersil/Harris/Mancini circuit
since once you get 0' & 180' right (the latter by means of a simple polarity-swapped signal),
the stuff in between is also correct.

 

[bradb]

Sleeper, i'd be down for one of those SSL PCBs,

talk about a useful project.

 

[Sleeper]

Hey Brad, 
The SSl that ssltech posted are mostly dead stock parts... I've seen them on quest etc. but I don't want to have to buy up an entire surplus lot.the newer 82005 schematic also has a few things that are hardish to find. like the 4583, available only as NTE at this point.  I'm certain I don't have the skillz to redesign test equipment. No thanks.

I think I'm going to revisit the phase meter that I have.
At least for the moment I'm going to recalibrate it against a software meter and see if what I considered to be wierdness is really just a setup problem. 
The design came from someone who makes a very Neve like product line.
I'd sure like to see the Neve schematic for comparison.

I'll post results here.
Kelly
 

 

[Script]

Free phase meter plug-ins?

Over the last couple of years I tested a few of these free plug-ins -- some are just too lame, others give readings that don't seem right... some wouldn't even recognize if there was an 'audible' problem. I also have that Edison thing: yet another (unreliable) reading  :-\  [--> Withered parts, leaking caps?]

So sometimes I end up cross-checking the different readings and also do the time-proven switch-to-mono compatibility check. It's surprising that some commercial electronic music CDs by independant labels actually have massive phase (L-R polarity) problems. But since this music doesn't get much airplay anyway (if at all), some labels/engineers don't seem to care and, if they are aware of the problem, seem to accept out-of-phase as yet another sonic flavour  . For CD-only releases this might be OK, to some degree at least -- and especially if switching to mono still yields satisfying results --, for vinyl however it's deadly.

Well, I am pretty sure I cannot spend $2000 plus for a precision hardware phase meter right now, but a DIY high-precision phase meter project and PCB would be great.

I will dig all the links mentioned above. Thanks to everyone so far for all the info.

script

 

[clintrubber]

Free phase meter plug-ins?

Over the last couple of years I tested a few of these free plug-ins -- some are just too lame, others give readings that don't seem right... some wouldn't even recognize if there was an 'audible' problem. I also have that Edison thing: yet another (unreliable) reading  :-\  [--> Withered parts, leaking caps?]

So sometimes I end up cross-checking the different readings and also do the time-proven switch-to-mono compatibility check. It's surprising that some commercial electronic music CDs by independant labels actually have massive phase (L-R polarity) problems. But since this music doesn't get much airplay anyway (if at all), some labels/engineers don't seem to care and, if they are aware of the problem, seem to accept out-of-phase as yet another sonic flavour  . For CD-only releases this might be OK, to some degree at least -- and especially if switching to mono still yields satisfying results --, for vinyl however it's deadly.

Interesting to hear about the Edison. Curious, how do you know its reading is inaccurate?
What are you using as your 'trusted reference' ?

The two Edisons I have show different readings, indeed does give me that the feeling that they're
then most likely both off, but IIRC there's an internal adjustment, maybe it's just a matter
of calibration.

And let's actually first start by saying what we demand of a phase meter.
A highly accurate response on steady state signals ? Or little more than
some idea 'how mono' or 'how stereo'or 'how over-stereo' our mix is ?

Well, I am pretty sure I cannot spend $2000 plus for a precision hardware phase meter right now, but a DIY high-precision phase meter project and PCB would be great.

I don't think there's a need to spend that much on a phase-meter.
How much accuracy do you need ? Risking to sound like an advertisement
for that Mancini/Harris/Intersil-circuit, it claims 1% acc.

Need more for audio ? I doubt it.
So then little reason to use a more elaborate circuit than this one. While not revolutionary, note that it's 'smart' by using matched comparators (the transistor-array). Low parts count then results, while still having good accuracy. 

While sure nice to have a PCB for it, that circuit is in essence just
three IC-packages, so putting it on veroboard (or the likes) is easy.

The only spot where a dedicated PCB might be nice is for the transistor-array,
which doesn't seem to come in 0.1" pitch, but SMD-only. I'm using an adapter-board
to regular pitch for it and the rest will be on vero.

The suggested ground plane in the Application Note might be ignored
when only using it for audio.

Bye,

  Peter

 

[bradb]

has anyone built and tested the intersil schematic?