How to test POLARITY/PHASE of your Speakers and Entire Sytem?

[ccclapp]

I am tuning my room and system and want to verify that everything is in correct Polarity/Phase. But don't know how to do this. By this I mean that that the positive terminals stay positive through my system and the negatives stay negative. I understand that beyond user connection errors, if is reasonably common for polarity to be crossed inside components, due to manufacturing error, etc. I understand an out of polarity component/driver etc can have a meaningful impact on sound quality and room tuning results.

 

[NOTE: I believe I am speaking of "polarity" (which shifts the "phase") but if polarity is incorrect terminology, please correct me]

 

I have read that speakers can be tested using a battery and watching to see if the driver pushes out (correct polarity) or pulls in (reversed polarity). See link:

 

 

Here is a description of testing other components:

 

http://www.goodsoundclub.com/Forums/ShowPost.aspx?PageIndex=1&postID=12349#12349

 

I understand one can check speaker polarity with a multi-meter, but haven't found a good description...something about seeing of the voltage is positive or negative but issues due to AC power from our outlets (vs DC).

 

I am somewhat reluctant to follow the 2nd link, as it may require opening the chassis of my components an is not very clear of where/what to test. Nor do I want to start attaching batteries to my DAC and AMP to use the battery method there (don't worry, I'm not)

 

Is there a way to simply run a test signal from a source and test the outputs individually down the signal path from source to speaker to see of polarity/voltage reverses? If not, going down the signal path, how SHOULD this be done?

 

I have a multi=mater and REW (Room EQ Wizard).

 

Thanks in advance for your help!!

 

 

 

 

 

 

 

 

 

 

[ccclapp]

In reading about the question above, I've found some useful info, but still not full answer to method to test polarity along whole signal path.

 

Here is a link to a website with three simple phase-testing signals (to do on line)...

 

http://www.audiocheck.net/audiotests_polaritycheck.php

 

 

Here is a link to other speaker sets, summarized below:

 

http://www.yorkville.com/default.asp?p1=6&p2=17&p_id=26

 

 

"To test for electrical phasing, touch the positive and negative terminals of a 9-volt battery to the + and - leads of a speaker cable plugged into the enclosure. This will cause the woofer cones to move slightly. If the cones move out, the cable and enclosure are in phase. If they move in, the cable or enclosure wiring is out of phase.

Another test method requires a microphone and a VU meter (the level meter on your mixer may work if it is fairly precise). Face your enclosures towards each other, around six feet apart. While someone plays a single, low note through the PA, perhaps on synthesizer, hold the mic between the woofers of the two cabinets and slowly move it back and forth while watching the meter. If the reading gets higher when the mic approaches the half-way point between cabinets, they are in phase. If it decreases, they're out of phase.""

 

[accwai]

Actually there are two types of polarity: relative and absolute. Correct relative polarity means all channels in the system have the same polarity relative to each other. Correct absolute polarity means the polarity of the sound wave coming out of the speakers is the same as when things are recording, as oppose to compression from the speaker while the original mic actually sensed rarefication.

 

Relative polarity is easy to check--play anything through a system with incorrect relative polarity will result in a strange hollow sound. If you insist on using a test signal, then pull up the generator panel in REW and setup a low frequency pure sine wave, say 250Hz. Flip the speaker wires on one side and you immediately hear the difference.

 

As for absolute polarity, there is nothing to check. Some do not believe that absolute polarity effect is audible. If so, you're done. For those that do, the whole chain up to the original mic must be taken into account. And since the front end is not controllable, you must have an absolute polarity switch somewhere in your system to handle the case of incorrect polarity in the recording chain. Then you set it by ear on the fly, on a per album or peraps per track basis.

 

As for your second link, they are taking about power line polarity, not speakers. Romy the Cat is a rather crazy site. Read it at your own risk... :-)

 

Andy

 

[ccclapp]

Thanks Andy

 

Yes, as you suggest I am not trying to determine absolute polarity.

 

What I am looking for is a little higher granularity:

 

I have a 2.2 system and each tower has 5 drivers (plus subs). I was hoping to double check each driver. In addition, I imagine there could be a double reversal, so the end result of a test tone would result in speakers being in correct polarity, but potentially a source and an amp/speaker could both be reversed, yielding a in-phase end point. On one hand chances probably are small and possibly wont matter as long as nothing changes... However, since I am tuning my entire system, this is a set of checks I want to run through.

 

Similarly for electrical polarity, I figure I would test this too...

 

 

 

 

[esldude]

With multi-way speakers you sometimes have crossover designs that have drivers in reverse polarity to others in order for the crossover to function properly. If such is the case with your speakers holding 5 drivers you wouldn't want to change that. Your crossovers would alter the basic frequency response enough to be much worse if you did.

 

[ccclapp]

Yes, Ive read that.

I guess what I would look for is if they vary from one speaker to the other...

 

 

 

 

 

[esldude]

I think you can test for this at least up to the speaker terminals.

 

Will give it some more thought.

 

I think you could construct an asymmetrical tone with audacity or similar, then use a diode and multi-meter at the speaker terminals. Used in one direction you would read a higher DC voltage with the multimeter than you will once you reverse the multimeter connection. This at the speaker terminals. The direction of the meter connection with the higher reading would tell you if the entire system was correct phased or reverse phased. This probably doesn't make sense perhaps. If I get time, I will try it here and let you know if it works.

 

[esldude]

Audacity has a neat plug in called 'Diode Processor'. I am on linux, sometimes there are little differences in the linux and other versions, but I think this one is in all of them.

 

So open Audacity, select generate, select tone and it will by default offer 30 seconds of sine wave at 440 hz. That should be fine. So generate that tone.

 

Okay, you will need to have downloaded the Audacity plug ins for this part. They are listed alphabetically under the Effect menu then inside Plug Ins. So find Diode processor. You will get a slider with values between 0 and 3. An explanation that 1 is half wave and 2 is full wave rectification. Put the slider at 1.00 (need not be exact, but get pretty close). Then let it process your 440 hz sine wave. This makes the wave look like it would had it passed through a single diode.

 

Now take your diode tone and play through your system. You need nothing more than your multimeter for this part. Set your meter to read DC volts. Put your red lead on the speaker red terminal, put your black lead on the speaker black terminal. Play the tone and you should read some small voltage and it should be positive. If you read negative DC voltage then your system has reversed polarity overall. If reversed simply flip your speaker leads. Your system now has correct polarity.

 

The voltage reading will depend on the particulars of your system and might be a half volt or so or it could be a few volts. Still it will read steadily and the main thing is whether your reading is positive or negative. To confirm you also can flip your multimeter leads and it should read negatively that way.

 

[Julf]

Wouldn't your system need to be DC coupled for that to work? Won't any condenser/capacitor in your signal path block the DC component, so that a DC meter will show 0?

 

[ccclapp]

This appears to be a great test!

 

I have a couple of quick clarifying questions:

 

-- I forget is Audacity PC, which I am?

-- This is testing signal polarity, not electrical, is that correct?

-- Can I do this at each point in the signal path, i.e. at analog source outputs, then preamp out puts, then AMP outputs (speaker in)??

 

Thanks very much for helping with this!!

 

 

[esldude]

I don't think it will matter. What matters is if the bulk of the signal is more in one direction than the other. You may have a point, and I did think about that. I have already tried it with my system before posting the test procedure. It worked, but I am using a system that is DC coupled. Give me a minute and I will stick a cap on the multimeter and see what happens.

 

[esldude]

Just connected the multimeter and put a cap in series with the leads. Worked just fine. I had a small value cap so the voltage reading was smaller, but it was a steady positive reading, and reversing leads it became a steady negative reading.

 

[Synfreak]

Hi Caleb!

Most problems which might occur - and therefor make it sometimes a PITA to achieve that "absolute polarity test thingy - where already named by others.

 

If you would like to know if your speaker(s) - at least the bass chassis - are in the right polarity, so that a "positive bump" will give a an acoustic "positive bump" too, I have attached a small test file, which actually is computer-made, and consists of two very short pulses on the positive side only.

 

Just play this file, and spot your speakers if they move "forward" (out of the box, so to speak).

 

BUT BE CAREFUL WITH THE VOLUME!

 

If played to loud it could easily damage your speakers ... !!

 

Cheers

Harald

 

[esldude]

Yes, Audacity is available for Mac, Linux and PC. I just attached the file here. Audacity is a handy tool though worth having to play with.

 

This is testing the audio signal polarity.

 

Yes, you could do it with an analog outputs anywhere in the system. But there is no point anywhere except at the speakers. I suppose if you just wanted to know if a component altered phase it might matter to you. But for listening it only matters at the speakers.

 

For instance a number of brands over the years would have pre-amps and amps that flipped phase. As long as you pair them up, you flipped it once with the pre-amp, and the amp flipped it back. Result was the same as if it were never flipped.

 

Also, flipping phase by itself really does no harm. If your total system flips, just reverse your speaker connections and everything is fine.

 

[ccclapp]

Thanks again @esldude I will try this. I installed audacity on another pc and was able to find the diode plugin and install it

 

I notice you have an attachment to your post above. Is this a recording of the diode processed test signal? If so, do I just play this via any player instead of Audacity with the diode effect?

 

I look forward to giving this a whirl!

 

[ccclapp]

Thanks Harald. This may be handy as well. Thanks for your effort/help!

 

[accwai]

[...] you could do it with an analog outputs anywhere in the system.

 

Yes, but only for anything downstream from the signal generator. Upstream components will have to treated case by case, and would be extremely difficult to test for certain cases.

 

But there is no point anywhere except at the speakers. I suppose if you just wanted to know if a component altered phase it might matter to you. But for listening it only matters at the speakers.

 

Exactly. In terms of audible results, all that really matters is whether or not all channels are in relative phase. And you really don't need test signals for that. Problem with flipped relative phase is not subtle at all. As for whether or not the components are wired incorrectly inside, not being a heavy DIYer myself, I would tend to trust the manufacturer. For example, my main power amp is entirely hand wired. If I can't trust the wiring of the speaker terminal, there are much more serious things that I need to worry about.

 

As for flipping the speaker terminal to set absolute polarity, two things to consider. First, for systems where effect of absolute polarity is audible, you need to account for absolute polarity all the way to the mic in the original recording session. Unfortunately, there is no standard for absolute polarity in the recording chain and it can change from album to album or even track to track. So you have no choice but to have a phase switch in your playback chain, and set it by ear on the fly. And for systems that do not readily show absolute phase effect, you really shouldn't care.

 

Second, there is a school of thought that says the red speaker lead has stronger control than the black one. That leads to things like crossed crossover cable configuration on biwirable speakers, with red directly on the woofer red and black directly on the tweeter black. If you believe in stuff like that, the speaker leads should not be flipped indiscriminately.

 

Andy

 

[esldude]

Well, just got back from a quick bike ride. Nice day for it here in the Southeast USA.

 

Yes, I went back and attached the file. All you have to do is play it in any music player. It will unzip into a .wav file.

 

[ccclapp]

Ok, so I tried the tests and here is what I found:

 

A) I am clearly an idiot, but that's not new news ;-)

 

B) Harald's Knack-Knack file: the drivers on my towers are too small to see movement direction 6" drivers (my towers are actually 6 passive mid/tweeters which sit on a powered bass). On my subs, they appeared to go out (vs in) with the clicks, but I couldn't be sure. When I changed polarity, I did not see a difference. Each "click" seems to me a series of 2-3 very fast clicks, so I could not see a single or 1st distinct movement.

 

C) @esldude's diode wav file: Using MM set to Volts DC:

 

i) Reading speaker wire out of amp with speaker disconnected: I got a reading of about 40. If reversed -40. Unfortunately, this did not change with volume nor did it matter if file was playing or not. In all cases when touching speaker wire from amp that was the + or - reading

 

ii) Reading connected speaker wire at speaker terminal: Reading from the speaker & wire connection point (where the banana plugs go into the speaker and trying to read the connection point) its reads about 45. If red to red on wires and speaker, red probe to red yields +. If reverse probe same but -. If I reverse speaker wire connection polarity follows speaker wire, not speaker.

 

iii) Reading speaker wires with amp in standby: 34 with correct polarity.

 

iv) Reading speaker wire with amp switched off at read power from sp wire: 9 with inverse polarity

 

Having done the above, I get the feeling I may have misunderstood, or I just don't understand what the MM is telling me and what I should be looking for.

 

NOTE: MM set to volts DC without touching anything, the reading can range from 25-60 seemingly arbitrarily. When touching the probes to each other, it reads 0.

 

If you haven't guessed, I'm new to MMs.

 

Sorry to trouble you guys with this and thanks for your help!!

 

[esldude]

Caleb are you sure that is 40 volts and not 40 mV (millivolts)? Don't know if your multimeter is auto ranging or not.

 

If you are reading 40 volts DC on your speakers then something seems wrong?

 

BTW, you will get better results with the speakers connected. Prevents stray voltage readings.

 

With speakers connected, red lead of meter should be on red (+) lead on speaker. The black lead of the meter should be on the black lead (-) of the speaker. Measuring at the banana plugs is a good place to do it. Also just to confirm, the red and black leads of the meter should be connected at the same time.

 

Sorry about the poor illustration, I am attaching a sketch of the connection. I left out the speaker cables, but they should simply be connected as they normally are when you are listening to music. Meter set to volts DC.

 

 

 

[ccclapp]

Hi elsidude

 

The mm dues do auto range. It's a top unit ( and thus more options and a bit confusing)

 

I'm pretty sure it's 40 before the decimal point and a fresh rechargeable 9v batt read 10. Tomorrow I can double check that nothing on the screen indicates a lower scale. Possibly it auto changes scale as well??

 

Assuming it does, do the rest of my readings make sence?

 

If I choose to test other components, should I test at the junction of the next upstream component, as you did w speakers at speaker end of connection from amp, or just test at device output, eg on amp at SP terminal, at pre at line out terminal etc?

 

If so, would device be connected to downstream device or not?

 

Thanks

 

[esldude]

If you were reading millivolts, then your readings might make sense. Lots of amps have some DC offset. Though reading that while playing the test signal seems unlikely. Did you have the test signal turned up enough to hear at something like normal listening volume? If you had the signal at low volume it may have been masked by the small DC offset of the power amp.

 

Multimeters have a very high input impedance like 10 megaohms. So very low stray voltage signals can be read. If you connect across a lower impedance load, the load usually gets rid of the stray voltage as it has no power behind it, it cannot create enough current to matter. You did say touching the leads together gave a zero reading which is how it should work.

 

Now reading something like a pre-amp output probably would work okay. The reading you get on DC like that will be small and you will need to have the scale reading in millivolts to see it. So you don't necessarily need to test at the next upstream component. I haven't used this signal to read a low voltage level like that, but it should work.

 

[sdolezalek]

Audio Tools by Studio Six is a relatively cheap iPhone, iPad app that includes a polarity tester (together with downloadable test signal). I have used it quite successfully both to discover that the drivers in three way systems can intentionally be plus, minus, plus, but also that a speaker was incorrectly wired by the manufacturer. I have no affiliation with Studio Six, but I am a satisfied customer.

 

[ccclapp]

@esldude

 

In just double checking...by testing at the speaker terminals while receiving a signal from the amp, am I testing signal polarity of signal or or speakers? I had understood it to be speakers, but now imaging it's the signal from the amp, as - vs + follows speaker wire not speaker terminal if I invert connection.

 

 

@sdolezalek

 

Yes I tried audio tools (free version). The readings are negative on all drivers on all speakers, even 2nd set in 2nd zone from lexicon and cheap powered deck speakers from office PC . I don't actually think I am out of polarity because it seems that the 9v battery test on subs is correct. Also listening tests suggest it's correct. Its because I've not fould a fully reliable test yet, that I keep working at this...

 

 

I have REW and nay trying measuring the IR of the polarity pulse tones to see if the measurement shows an initial up or down trace...

 

Thanks guys

 

 

 

[esldude]

You are testing the signal from the amp. You really have to assume the speakers are marked correctly. As I mentioned before with multi-way crossovers they may intentionally invert some drivers vs. others. You really shouldn't need to worry about that. So get the signal at the speaker input terminals correct, by getting the amp correct and you should have it.

 

If something upstream of the speakers inverts polarity, then flipping the speaker leads will correct for it whether it is an amp, pre-amp or source that is doing the flipping.

 

I just checked my methodology here again. I dug up two 5 uF caps just to see if it mattered. Put these on the leads of the meter, and read almost exactly the same voltage as connecting the meter directly. I also generated a pure sine wave and tested it. Just as a for instance in this case, AC setting pure 440 hz sine wave read 1.88 volts while switching to DC gave a reading that fluctuated between -2 millivolts and plus 3 millivolts which basically is nothing. Reversing leads gave identical measurements. Next going with the asymmetrical test tone I uploaded in this thread, on AC I read .945 volts and switching to DC I read .757 volts which was pretty steady. The DC read positive in one direction and negative if I reversed leads. All readings taken at the speaker terminals.

 

So you should be able to get similar readings. Put on the test tone and set volume so you get maybe 2 volts using your meter on AC. Switching to DC you should get a reading somewhere between 1 and 1.5 volts and the polarity should tell you what you want to know.