Protective cap on Ribbon tweeter

[pypes]

I read that you should not hook up a ribbon tweeter without a protective filter in between it and the amplifier - even if it's simply a capacitor. If I am using a DCX as an active crossover with a ribbon tweeter in the loudspeaker, do I still need a capacitor in between the amp and the ribbon? If so, what size cap should I use? (ribbon will be crossed >3khz)

Thanks in advance.

[lhwidget]

I'd use one to catch the LF content of any turn-on/turn-off transients from the amp driving them, or the occaisional oops with an input cable. Just make it large enough that it doesnt affect the driver's response in its pass band.

I'd also measure the driver's response with the cap in place so you can dial in the dqx (if it includes phase correction).

[pypes]

Any idea what size that would be? When I read that a ribbon has been tested (for instance, Zaph's "Battle of the Non-Domes") it seems to be 20uf. Would this be sufficient? I'm not sure exactly how to figure it out. The reason I was going with active in the first place is because I don't know how to model a passive crossover. Another post I noted mentioned that 20uf may not be sufficient and suggested at least 40uf... even just for test measuring. Is this more the range I should be looking at?

Btw, I'm using the Fountek NeoCD3.5H ribbon, which can be seen here - https://www.madisound.com/store/prod...oducts_id=8818

[lhwidget]

Bigger won't hurt anything, just starts getting expensive. It is simply forming a first order filter (6 dB/octave) with an f3 way below the XO point. Using the nominal impedance of 7Ω, then

f3 = 1 / (2•pi•R•C) where R = 7Ω and C = 20 µF = 0.00002 F

then f3 = 1/ (2•pi•7•0.00002) = 1137 Hz

The filter is passing everything above 1137 Hz and attenuating everything below 1137 at a rate of 6 dB/octave, and it's 3 dB down at f3

The 40 µF cap will have an f3 of 568 Hz, in other words it will pass more lower frequency energy to the tweeter, providing less protection, but having less chance to mess with the XO response.

If the impedance stays close to the nominal value, I'd use the 20 µF. I haven't used a ribbon with a transformer yet, so I don't know how the impedance behaves, I think it stays pretty flat (in other words holler at Wolf or Jeff ).

[Jeff B.]

Bigger won't hurt anything, just starts getting expensive. It is simply forming a first order filter (6 dB/octave) with an f3 way below the XO point. Using the nominal impedance of 7Ω, then

f3 = 1 / (2•pi•R•C) where R = 7Ω and C = 20 µF = 0.00002 F

then f3 = 1/ (2•pi•7•0.00002) = 1137 Hz

The filter is passing everything above 1137 Hz and attenuating everything below 1137 at a rate of 6 dB/octave, and it's 3 dB down at f3

The 40 µF cap will have an f3 of 568 Hz, in other words it will pass more lower frequency energy to the tweeter, providing less protection, but having less chance to mess with the XO response.

If the impedance stays close to the nominal value, I'd use the 20 µF. I haven't used a ribbon with a transformer yet, so I don't know how the impedance behaves, I think it stays pretty flat (in other words holler at Wolf or Jeff ).

The reason it is important to use a cap with a ribbon tweeter is due to the fact that with true ribbons (the kind with impedance matching transformers, not the planar types) drop to a near dead short in impedance at low frequencies. Even if you are using active or digital filters, you do not want your amplifier to see this short. A cap of 30 or 40uf will avoid any problems for your amp.

Jeff

[lhwidget]

Thanks Jeff, I was afraid there was more to it.

[pypes]

Alright, thanks a lot for the information. I'll get a pair of 40uf caps before I finish sealing up the cabinets then.