Zero autoformers -- how to measure impedance?

Good point, Roger!  Thanks.

Best regards,
-- Al

Yes it does, but the number is so small I wouldn't worry about it. 22.6 mv into 8 ohms is 0.000064 watts!

This is an old thread but needs some correcting. The user who has -22 mV of offset on his amp needs a little clarification.

First, offset can be negative or positive. We would like it to be zero but 22 mv is quite acceptable. However when we connect a Zero to this,whose DC resistance is almost ZERO we can have significant current flowing which can saturate the core and make for some very bad bass and IM distortion. The impedance selected matters not.

The Zero is wound with very heavy wire and I do not know its DCR nor is that easy to measure with a DMM. Actually you measure low resistances by putting a known current throught the wire and seeing what the resultant voltage is. This is absoultely accurate and is not affected by probe resistances. Say the wire resistance was 0.2 ohms, which is quite likely then the current is 100 mA and I would expect quite a bit of core offset. 

Ryan, congratulations and thanks for the nice words!!

agree. thanks for nice words.

But it looks like something upstream is feeding DC into your amp that, when amplified, creates a -22.6mV at the NAD output. What do you think, Almarg??

Yes it does, but the number is so small I wouldn't worry about it. 22.6 mv into 8 ohms is 0.000064 watts!

Also, perhaps it is actually some small amount of low frequency noise, such as a ground loop might cause, which the meter is somehow partially rectifying and indicating as DC, even though it is AC at some frequency or combination of frequencies. Just speculating.

So I don't know why the NAD would have gone into protection when its speaker-level outputs were connected to the Zero, and through it to the speaker. I listed three possibilities in my first post in this thread, but none seem applicable at this point.

Ryan, congratulations and thanks for the nice words!!

Best regards,
-- Al

6-17-15: Angelgz2
Hi Al and Bombaywalla

Thanks for the inputs. I am glad that the zeros didn't fry. My bad, the -22.6 is in mV as indicated on the multimeter.

Whew!! glad that it was just missing the "mV" units....

6-17-15: Angelgz2
I did another measurement with everything disconnected. The unit has already been on for a while. The measurement read -2.0mV on the front channels and -1.9 on the surround channels. So I take this is normal since the voltage is practically zero?

looks like the NAD by itself is just fine. yes, those low mV numbers should be ignored like Almarg indicated in an earlier post.
But it looks like something upstream is feeding DC into your amp that, when amplified, creates a -22.6mV at the NAD output. What do you think, Almarg??

I did another measurement with everything disconnected. The unit has already been on for a while. The measurement read -2.0mV on the front channels and -1.9 on the surround channels. So I take this is normal since the voltage is practically zero?

I've ran the zeros with the Parasound amp using the NAD as a pre amp for over 5 hours last night with movies and music and neither unit gets hot or enter into protection mode. The speakers sounds much better than before. Unlike using a 3ohm in-line resistor, they no longer sound like someone speaking in a box. The 1.5 ohm resistor definitely helped preventing the 4 ohm woofer from sounding too loud.

I would definitely call this a success. Thanks to all for your help completing this project!! God bless you all

Thanks,
Ryan Huang

Hi Al and Bombaywalla

Thanks for the inputs. I am glad that the zeros didn't fry. My bad, the -22.6 is in mV as indicated on the multimeter. I have one pair of speakers connect to it but will measure again taking everything off. Thanks!

Hello Ryan & Bombaywalla,

The reason I suggested making the measurement "with the component driving the amp turned on but not providing a signal" is so that in the event that all of the electronics is DC coupled, and a large DC voltage is being generated due to a fault upstream of the amp, the measurement at the amp output would pick it up. (In this case, of course, it turns out that the NAD T757 that is being used is an A/V Receiver, rather than just a power amp, so what I said would apply to whatever source component is connected to it and selected).

Assuming that the -22.6 is volts, not millivolts (and I suspect it is volts given your mention of a mild shock sensation), that is definitely abnormal and indicative of a major fault, that could conceivably burn out any speaker or Zero that is connected. However given what has been said I suspect the Zero is ok.

If you haven't already done so, the next step would be to verify that the -22.6 volts is present at the output of the NAD when nothing is connected to it, except power. If anything close to that much voltage is present under that condition, the unit needs to be professionally repaired.

Good luck. Regards,
-- Al

other than the meter, and with the component driving the amp turned on but not providing a signal.

I re-read Almarg's post & he did say that the above in quotes. From my experience I've found ANYTHING connected to the amp while measuring the DC offset can influence this measurement adversely i.e. you are likely to not read the correct DC offset voltage. So, I like to ensure
(1) NOTHING is connected to the amp except a power cable. All inputs & outputs have no connections
(2) the amp is turned on & allowed to sit there for 25-30 minutes for the bias to stabilize before i measure the DC offset. When the amp 1st turns on the bias is very erratic & changing all the time as the components/devices run current, heat up & come to some sort of thermal equilibrium. No point measuring DC offset while all this happening - your reading will be all over the place & you might even fall off your chair if you actually believe that reading....

I hope I didn't fry the zeros did I? I mean, how can they be fried since they didn't get hot at all and I've only ran them on the NAD unit for no more than 2 hours.

you probably did not fry the Zero since the DC offset from the NAD must have been small.
But let me tell you - time has nothing to do with electronics breaking down or not. If it has to break, it'll break instantly - pretty much like an incandescent light bulb. Ever seen how those break when it's their time to fail? That's exactly how electronics breaks.
If the DC current was too high for the Zero, 2 hrs is a life-time to do damage to them. They would have been fried.

The fact that you can play music thru them thru the Parasound means they are functional.

I measured the DC output of the NAD T757 and found it to be -22.6. How can it be negative? Is that normal?

1stly, did you follow Almarg's instructions to the T? Nothing connected to the amp means nothing connected to the amp.
2ndly, are you missing a "mV" after the -22.6? I.E. did you read the measurement correctly numbers & its units (Volts or milli-volts)?

DC offset can be positive or negative so a minus is definitely possible.

Just FYI: these amps have a plus DC rail (+20V or +30V, etc) & a minus rail (-20V, -30V, etc). So, when there is no music playing the red & black/white binding posts should sit a 0V (ground potential) + any DC offset (plus or minus DC offset).

if it's sitting at -22.6V then you have something driving that amp. Again, did you follow Almarg's instructions to the T?
Thanks.

Hi Al,

I measured the DC output of the NAD T757 and found it to be -22.6. How can it be negative? Is that normal?

Thanks,
Ryan

Thanks Al and Bombaywalla. I figured this is a separate topic so I started a new thread. I hope I didn't fry the zeros. I have an amp and a receiver. The zeros were originally connected to a NAD T757 V2 and this unit kept going into protection. Yesterday, I rewired it to the Parasound new classic 5125 (connected to the front pre-outs of the T757) and no more protection mode and the amps are not hot at all.

While it was connected to the T757, I can feel a relatively high current on the binding posts where the zeros are connected to the receiver. When I put my fingers on these binding posts, the current "zaps" my fingers a bit but not that bad.

I hope I didn't fry the zeros did I? I mean, how can they be fried since they didn't get hot at all and I've only ran them on the NAD unit for no more than 2 hours. I'll get a reading on the DC output on the NAD unit today and hopefully we can come to a conclusion whether they are fried.

Oh, I see! Thanks for this info, Almarg. I didn't know the OP had another paralller thread running.

Bombaywalla, based on info the OP has provided in another recent thread he is driving a combination of four drivers in parallel. Three of them are 8 ohms, and the other is 4 ohms placed in series with a 1.5 ohm resistor. The purpose of the resistor is to cause the 4 ohm driver to receive approximately the same amount of power as the other drivers.

The overall impedance of that combination is approximately 1.8 ohms. He is using the Zero to step that up 4x, so that the amplifier will see a load of 7.2 ohms.

Best regards,
-- Al

Assume nothing so here goes -
* how have you connected the Zero xformers? I.E. did you connect them up in division mode? Or, did you connect them up in multiplication mode? If you connected them in multiplication mode then the side that has only 2 binding posts should be connected to your speaker. The side that has 4 binding posts in a row should be connected to your amp.
* if you truly have the Zero xformer connected in multiplication mode, then, how is your amp going into protection mode?
Actually, it could. I'm thinking out aloud here so some of this could be gibberish - pardon me in advance.
If the Zero xformer is in multiplication mode - say 2X - and 8 ohms becomes 16 ohms, the voltage developed across the output stage of the amp might begin to clip as it reaches its plus/minus or both power rail supplies. That might inform the amp to go into protection mode due to excessive voltage. The current into 16 Ohms would be well below the max current limit of the amp i.e. the amp is not current limiting.

If you have a SS amp that has DC offset or excess DC offset, you risk frying that side of the Zero that the amp is connected to (because the coils of that side of the xformer are basically an inductor. And, an inductor is a short to DC voltages. So, having a DC offset in the SS amp means having that same DC voltage across the Zero windings. Since the winding resistance is usually very small, the DC current can be (very) high that could fry the windings of the Zero which is connected to the amp. Hence the caution from Paul Spelz.
Still that should not put your amp into protection mode - it'll destroy the Zero tho').

Check your connections.
(actually I'm not sure which mode to use the Zero in - multiplication mode might yield a too high voltage. Division mode might yield a too high current. Both modes could set off the protection mode depending on what the amp is capable of).
See if you can eliminate or nearly eliminate the DC offset.

Why are you using a Zero with a SS amp? Are you driving a 1-Ohm Apogee or something similar that you need the amp to "see" a higher speaker load impedance?

Ryan, yes, most multimeters can only measure impedance at a frequency of 0 Hz, or in other words, DC resistance. And given that, the numbers you cited seem reasonable, regardless of whether or not a speaker is connected (although the DC resistance of a speaker, if one is conected, will have a slight effect on those readings).

But as you are aware the amplifier should see a considerably higher impedance at audible frequencies, essentially corresponding to the speaker impedance x the impedance transformation ratio the Zero is connected to provide. The impedance of the Zero by itself, at audible frequencies when it is not connected to a speaker and an amplifier, is pretty much meaningless.

It does seem conceivable that if the amplifier is putting out significant amounts of DC, perhaps due to a fault, the amplifier could sense the presence of a near short at 0 Hz, and enter protection mode. You should be able to measure the DC output of the amp with the multimeter. Do that with the outputs of the amp connected to nothing, other than the meter, and with the component driving the amp turned on but not providing a signal. If you measure just say a few millivolts, I wouldn't worry about it.

The other possibility that occurs to me is that if you are running the amp with the Zero connected but no speaker connected to the Zero's output, the amp could shut down, or conceivably even be damaged, as a result of what is called "inductive kickback." Especially if the amp is processing a signal when connected under that condition.

Finally, of course, as we discussed in your other recent thread connecting the Zero backwards, so that it reduces the load impedance seen by the amp rather than increasing it, could obviously cause the amp to enter protection mode. Although I presume based on our previous discussions that you are not doing that.

Not sure what else to suggest at this point. Good luck. Regards,
-- Al