How much DC is tolerable on a tweeter?

[LionIT]

Ok, the title I think is clear.

In your opinion, how much DC is tolerable on a tweeter, both in continuous (DC offset) and in peak (turn on / off spike)?

I currently use a Neurochrome Modulus 86 Rev 2.4 with SMPS. The Modulus has a DC servo circuit with a dedicated opamp, which reaches less than 1mV of DC offset and a few tens of mV during shutdown with 8 ohm load (with the multimeter I can't see exactly the peak but this should be the order of magnitude).

The tweeters that I drive directly with this amp are those of the Kef R3 Meta Uni-Q (crossover bypassed), that I don't know exactly how much power they tolerate.

There are no DC protection circuits on the amp, but I think I'll put them in the future.
Apart from that, what do you think of the aforementioned DC values?

 

[thewas]

I doubt that few tens of mV for such short time would damage it, but I would always use a series capacitor as a lower frequency highpass for operation safety and my brain comfort in any self made active design.

 

[LionIT]

I doubt that few tens of mV for such short time would damage it, but I would always use a series capacitor as a lower frequency highpass for operation safety and my brain comfort in any self made active design.

Thanks for the advice.
I imagine non-polar capacitor is necessary, but how do I determine the required capacity?

 

[thewas]

You are welcome, since you will use it outside your planned passband simple online calculators suffice, for example if you plan to use it above 2 kHz I would take as -3 dB point of the 6 dB filter 1 kHz which would mean approximately 20 µF for 8 Ohm and 40 µF for 4 Ohm.
https://www.diyaudioandvideo.com/Calculator/SpeakerCrossover/

 

[LionIT]

Thanks again for the replies.
I cannot measure the impedance of the tweeter and online there is only the impedance graph of the overall speaker (therefore altered by the crossover).
at 20khz it results about 5.5ohm ... can it be indicative?

 

[SSS]

Tweeters have like other speaker chassis a certain impedance like 4, 8, 16 Ohm. In your case it seems to be 4 Ohm. If it is a planar type other values like 6 Ohm are possible. Using a DC protection is mandatory since amplifiers can become defective and then the full rail voltage will apply. In order to keep transients save zener-diodes can be applied prallel to the tweeter. In former speakers a light bulb was in series to the tweeter.

 

[Doodski]

Tweeters have like other speaker chassis a certain impedance like 4, 8, 16 Ohm. In your case it seems to be 4 Ohm. If it is a planar type other values like 6 Ohm are possible. Using a DC protection is mandatory since amplifiers can become defective and then the full rail voltage will apply. In order to keep transients save zener-diodes can be applied prallel to the tweeter. In former speakers a light bulb was in series to the tweeter.

I have not seen a Zener diode in parallel with the speaker driver (tweeter). Are there any special considerations like voltage and power rating of the Zener diode? How would one size up a properly rated Zener diode.

 

[KSTR]

In terms of dissipation DC offset is not a problem for tweeters, even 2V would only generate 1W in a 4 ohms tweeter.

But of course it would severly off-center the rest position of the tweeter, compromising performance.
I would think an offset of below 10% of the maximum excursion might be a tolerable upper limit. If the tweeter's BL and Cms are known, one can calculate the maximum DC voltage from that:
Say, BL = 1 N/A, Cms = 0.1mm/N, Xmax = +-0.1mm and Re = 3R
Xdc := 0.01mm --> Idc = 100mA --> Vdc = 300mV.

 

[SSS]

I have not seen a Zener diode in parallel with the speaker driver (tweeter). Are there any special considerations like voltage and power rating of the Zener diode? How would one size up a properly rated Zener diode.

Depends on the tweeter specs and what music transients will be in the music signal. Therefore I can not give a specific value. Power is at transients (short time) no issue. So 1 Watt types will survive. Zener backwards behave like a regular silicon diode. So in series a diode needs to be inserted. And one rail should have an antiparallel second rail. And I assumed that a capacitor is in series from amp output to the tweeter connection.

 

[LionIT]

Ok thank you all for the information.
Building a passive protection circuit does not seem much different to me from maintaining the original passive crossover.
An active circuit with MOSFET seems to me more congenial for protection.
At least to preserve the advantage of driving the driver directly with the amp.
Before arriving at that protection circuit, however, I wanted to understand if the tweeter can withstand the DC offset / peak currently delivered by my Modulus, so to understand if the use is safe, catastrophes aside.
From the calculation of @KSTR it seems to me that in spans we are on a few hundred mV, so it is safe.
Correct me if I'm wrong.

 

[JaapD]

No need to chase problems that don’t exist in the first place. Any DC offset from the amplifier will be converted to heat (or mechanical offset) in the tweeter. Since you’re mentioning a magnitude of only 1mV it would result in about a quarter of a microwatt of power (P=U^2/R) or less being transferred to heat (or mechanical offset) which is absolutely negligible.

Cheers!

 

[SSS]

Best is to connect the tweeter direct to the amp output. In this case a so called crowbar protection can be done. It works as a short circuit parallel to the tweeter when DC voltage is discovered for a certain time span. As far as I remember such a circuit was used in the Quad 450 amp. The other non audio application is in lab electronic power supplies to save a device or circuit from overvoltage when the power supply fails.

 

[KSTR]

From the calculation of @KSTR it seems to me that in spans we are on a few hundred mV, so it is safe.
Correct me if I'm wrong.

No, you are correct. 100mV should not be a problem except for some very exotic tweeters maybe.

If your tweeter amp has enough headroom, I'd suggest a simple series resistor additional to the mandatory DC protection series capacitor. This adds some more protection and more often than not it gives better distortion values, except maybe at or below resonance, basically outside the working range.

Zeners for tweeter protection are very rarely used because they are not straightforward to dimension. Too low a voltage and they will introduce clipping all the time, too high a value and they won't protect.

 

[Wolf11Man]

The modulus is a very good amp and the chip has all kinds of protection in place. The DC at output is ridiculous low. The amp is protected from on/ off noise as well.
There is no reason to risk sound quality or amplifier function by implementing any semi conductor or capacitor construction.
If you have tested the amp modules function (with high load, like a complete passive speaker!), just hook it up directly to the tweeter.

Ask Tom Christensen what he thinks and follow his advice. Of course, nothing will save a tweeter from some wrongfull installation or module build.
In your shoes, I would use a 15uF capacitor (bipolar or film type) in series with the tweeter until your insatallation is finished. As a last step remove or bridge the cap. The advantage of a fully active speaker is to have no parts between driver and amp. You may hear the difference or not, but no part is the best. Any resistor, fuse, polyfuse or funny diode construction no one ever heard of, are a clear no-no.

PS in a PA speaker, that will see any kind of abuse, stupid connecting and handling, feedback from microfones etc, a capacitor, fuse, polypuse, light bulb or protective electronic circuit may be a life insurance. For HIFI usually not needed. Also, put the cost and time spend with such protection in relation to the cost of a new tweeter or a repair dome! Most HIFI speaker live for decades without a tweeter failure.

 

[LionIT]

The KEF R3 Meta seems a very well sorted design to begin with. What is your motivation in setting aside the stock crossover and going down this path?

Anyways, you're not going to have any problems connecting directly to a well-designed amp like the Neurochrome.
The "solutions" I see in this thread are for a problem that doesn't exist.

These days, audiophiles are much more likely to damage drivers with pilot errors in testing scenarios vice in normal usage. Sweep testing....things like that.

Have fun.

Dave Reite

I'm just moving to multi-way amplification with digital crossovers, as a hobby let's say...
I like the idea of aligning the UNI-Q drivers in the time domain and I'm curious to hear any differences.
Many use this approach with DIY speakers, but I honestly don't think they can reach the same engineering level as professionals like KEF, especially without anechoic measurements. So I start from a well-engineered speaker to have fun with this hobby.

Anyway, thank you all for your valuable contributions.
In the first instance I will use the Modulus without additional protections to see what acoustic performance I can get.
If they are satisfactory, I will evaluate any basic protections as you suggested.

 

[mhardy6647]

No reason short of crossover failure that any DC should get to a tweeter -- at least for passive XO schemes, that is!
That said, I think the best answer is no DC. The voice coils of most tweeters are small and also, as they say, exceeding fine.
Constant DC will generate heat.
Heat is the enemy.

Transient, small DC spikes to a tweeter (from an ill-behaved active crossover system) might be OK, but certainly represent, from my perspective, unnecessary risk.

 

[KSTR]

Constant DC will generate heat.
Heat is the enemy.

No. As I have shown in previous post, even several volts of contiuous DC won't kill a tweeter thermally (like 4Vdc into a 4Ohms VC will dissipate 4 Watts), but it will likely damage it from maxed out displacement, stretching the suspension beyond elasticy limits, disrupting tinsel leads, bottoming out and deforming the VC former, stuff like that.

 

[Old_School_Brad]

I have not seen a Zener diode in parallel with the speaker driver (tweeter). Are there any special considerations like voltage and power rating of the Zener diode? How would one size up a properly rated Zener diode.

Infinity did it.

Spoiler: Kappa tweeter

 

[SSS]

s

Did not know this one. Zener was just my idea for limiting voltages at the tweeter. Of course back to back zener will work instead of a extra diode in series with the zener diode. Depends on the forward voltage of the zener diode. And right, for every tweeter and listening habits the protection value needs to be calculated.
One post mentioned NO silicon around the tweeter. Zener as well a crowbar has absolutely no effect on the signal prior the protection level.

 

[Devnull]

I have not seen a Zener diode in parallel with the speaker driver (tweeter). Are there any special considerations like voltage and power rating of the Zener diode? How would one size up a properly rated Zener diode.

No special considerations, just the max voltage wanted across the tweeter. I think someone over at diyaudio measured increased distortion as the voltage across the zeners approached breakdown voltage.

EV used a different diode protection scheme. They used a full wave bridge across the tweeter. The output of the bridge would open a relay between the bridge input and the tweeter. Once the signal dropped below the relay hold voltage the circuit would reset.