Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[theo mack]

Sorry if this is a stupid question:

Why is it that tantalum caps often make audio circuits sound better?

Is it possible to replace any audio path cap with a tantalum equiv and make it sound better?

What is it about the properties of tantalum that makes it different?

theo mack

[Geoff_T]

Hi

This link explains it well...

http://www.cabot-corp.com/cws/businesses.nsf/CWSID/cwsBUS022 72001074813AM8998?OpenDocument&bc=Products+%26+Markets/T antalum/Technical+Papers&bcn=23/4294966876/4294966867&am p;entry=product

I cringe when I see re-capped Neve modules with tants replaced with radial electrolytics... they won't sound the same.

[theo mack]

Wonderful link.  Thanks Geoff.

So, would there be any advantage in replacing all the caps in an audio circuit with tantalums? Speaking in terms of tone.

[Geoff_T]

theom wrote on Thu, 26 August 2004 07:18

Wonderful link.  Thanks Geoff. So, would there be any advantage in replacing all the caps in an audio circuit with tantalums? Speaking in terms of tone.

Hi

Input caps maybe, but output cap values may run expensive or not available if the value is too high. The Neve 283 card uses lots of tants.

If you use them for power supply decoupling they have to work with a series resistor (which in most cases is normally there).

[Jakob Erland]

..but tantalum caps as psu decoupling can be a real P.I.T.A. as they tend to short when failing, not go open-circuit like the electrolytics. And shorted psu lines is less than desireable in most cases. Whenever I find tantalums used as PSU decoupling, I change them to electrolytics. In parallel with a small polyester cap, if there's real HF stability issues..

Jakob E.

[Geoff_T]

Jakob Erland wrote on Thu, 26 August 2004 10:56

..but tantalum caps as psu decoupling can be a real P.I.T.A. as they tend to short when failing, not go open-circuit like the electrolytics. And shorted psu lines is less than desireable in most cases. Whenever I find tantalums used as PSU decoupling, I change them to electrolytics. In parallel with a small polyester cap, if there's real HF stability issues.. Jakob E.

Hi

True... and that's why I said that a current limit resistor must be present that'll blow like a fuse and protect the power rails.

Neve 54 (suitcase) consoles use tants in the power decoupling and they do go... after around 25-30 years... they are pretty reliable.

Geoff

[Brian Roth]

One thing to keep in mind is that tants do NOT work well as audio coupling caps in a circuit with a "bipolar" power supply (as in typical desks, etc with opamps).  In these devices, the tants are reverse polarized through half of the signal swing.

Try putting a tant between the signal out and measurement input of even a simple distortion analyzer and look at the numbers...not pretty!

However, they are viable in circuits with a single-ended power supply, like older Neves, Amoex recorders, etc.

Bri

[Jakob Erland]

I think that this is over-simplifying the topic a bit.

Tantalums will work fine with bipolar supplies as well - you just need to know exactly what you're doing.

Our Calrec UA8000 has quite a few tantalums in signal path, and NTP are stuffed full of them (yes, also the +/- supply versions) Both are working well to say the least..

Jakob E.

[Geoff_T]

Brian Roth wrote on Thu, 26 August 2004 23:50

One thing to keep in mind is that tants do NOT work well as audio coupling caps in a circuit with a "bipolar" power supply (as in typical desks, etc with opamps).  In these devices, the tants are reverse polarized through half of the signal swing. Try putting a tant between the signal out and measurement input of even a simple distortion analyzer and look at the numbers...not pretty! However, they are viable in circuits with a single-ended power supply, like older Neves, Amoex recorders, etc. Bri

Hi

On that premise and circuit format, any electrolytic capacitor would be reverse biased with the audio signal swing but, boy, you must be swinging some volts there!

[John Klett]

Generally I don't like tantalum caps.  In all the testing I have done with solid tantalum and wet aluminum electrolytic caps, solid electrolytic caps have consistently contributed more distortion than wet electrolytic caps.  If people like the artifacts that tantalum caps contribute and want to ascribe that to the "sound of Neve" then by all means keep them in.  My Neve modules have no tantalum caps in them and they sound like Neve to me.  Technically a good electrolytic cap that is designed for audio will outperform any tantalum cap in any audio application when used correctly (i.e. non-polarized caps when there is no bias voltage and either polarized or non-polarized when there is a bias.

http://www.capacitors.com/picking_capacitors/pickcap.htm#tan talum

[Geoff_T]

Hi John

So are you saying that the designers of the circuits at Neve that used tants, especially for the input coupling caps where the very low leakage was advantageous, were not the brightest of fellows?

[bobkatz]

Jakob Erland wrote on Fri, 27 August 2004 05:54

I think that this is over-simplifying the topic a bit. Tantalums will work fine with bipolar supplies as well - you just need to know exactly what you're doing. Our Calrec UA8000 has quite a few tantalums in signal path, and NTP are stuffed full of them (yes, also the +/- supply versions) Both are working well to say the least.. Jakob E.

Did you measure the distortion? I really don't understand this hook on using Tantalums as coupling caps...  At least if you believe the studies done by Jung, etc. way back when I was a pup. What I learned from that is an an electrolytic (or a tantalum) without bias is a distortion generator. That "dirty" tantalum sound is good for something, but not for transparency or purity of tone. And sonically I've always gone for no cap, or polystyrene/polypropylene if at all possible.

[Dan Kennedy]

No, tant's with bias are pretty clean, and very low leakage, which is why Neve used them. There are very few electolytics with as little leakage, even new, even now.

Damn good reason to use them.

Terrible choice for anything supply related though. Probably related to their completely unforgiving manner in dealing with reverse voltages. Maybe ripple current takes them out, I don't know, but many, many times they just turn into charcoal, along with the pcb...

[dcollins]

Geoff_T wrote on Fri, 27 August 2004 13:41

Hi John So are you saying that the designers of the circuits at Neve that used tants, especially for the input coupling caps where the very low leakage was advantageous, were not the brightest of fellows?

How about those hacks that designed the ATR-100's?

DC

[Brian Roth]

Geoff, you raise an excellent point.  I'm just going from recollections from tests I ran eons ago.  In that testbed, I tried both solid tants and decent-quality aluminum electrolytics with similar values, and noted the tants yielded significantly higher THD in the "bipolar" AC circuit.

My only hunch is that tants are much more "hateful" about reverse polarization than aluminums. That "theory" might also explain why tants quickly "go boom" if miswired as PS bypasses, while a typical aluminum might merely "complain" in some fashion versus doing a meltdown.

However, my experiments were done a LOOONG time ago, and today's parts may indeed behave differently.

Bri

PS, I seem to recall both "flavors" of caps were rated at 25 or 35 VDC