Parafeed Cap size

[triode]

Hi,

What is the lower range of Parafeed Capacitor for "acceptable" low end roll-off?
Any simple formula to calc relation between the uF capacitance and -3dB freq roll off?

I want to upgrade the cap but the physical size (and the price) get very big for "boutique" caps of 12uF

Apologies if this already discussed before.
Thank in advance

[Paul Joppa]

Normally I would recommend 5.6uF to 27uF as the normal range for experimentation. But MonAmour uses a very high plate choke inductance, which can allow a smaller capacitance to be used without losing all the bass. It won't hurt the amp to go lower in capacitance. Roughly speaking:

1uF is 3dB down at 50Hz,

2uF at 25Hz,

3uF at 16.7Hz,

4uF at 12.5Hz.

This is the simple RC formula f = 1/(2*pi*R*C) where R is the transformer primary, 3000 ohms. For these small caps, the plate choke impedance is high enough that it can be dropped from the analysis, leaving only the capacitor and the load.

Of course, there will be interactions with the speaker impedance variations. By using a smaller capacitor, those interactions move up along with the cutoff. Just be prepared for some odd bass changes. Precise predictions are not possible - experimentation and listening are the tools of choice.

[triode]

Thanks Paul,

Certainly my speaker won't produce 10Hz !!!

[Paul Birkeland]

Certainly my speaker won't produce 10Hz !!!

Another way of looking at this is asking yourself when you want your amp to start rolling off on the bottom end. 

[Ivan303]

Normally I would recommend 5.6uF to 27uF as the normal range for experimentation. But MonAmour uses a very high plate choke inductance, which can allow a smaller capacitance to be used without losing all the bass. It won't hurt the amp to go lower in capacitance. Roughly speaking:

1uF is 3dB down at 50Hz,

2uF at 25Hz,

3uF at 16.7Hz,

4uF at 12.5Hz.

This is the simple RC formula f = 1/(2*pi*R*C) where R is the transformer primary, 3000 ohms. For these small caps, the plate choke impedance is high enough that it can be dropped from the analysis, leaving only the capacitor and the load.

Of course, there will be interactions with the speaker impedance variations. By using a smaller capacitor, those interactions move up along with the cutoff. Just be prepared for some odd bass changes. Precise predictions are not possible - experimentation and listening are the tools of choice.

So I have a couple pair of 'ugly' Solen Caps in both 5.5uf and 1.5uf 650V, so perhaps I can substitute them in the parafeed cap position and get an idea if I'm likely to hear anything 'odd' in the bass frequencies (what little real bass that can be had from the Edgar folded mid-bass horn) before I spend real $$$ on a 'gourmet' cap?

Also, as I really have no use for frequencies lower than about 50-60 Hz, would it be safe to go with a lower coupling cap value than the current .1uf?  Say .03uf? 

Good to hear the primary is 3K as that should not be a bad match for a 2.5V 300b I am currently using.  That plus the EdgarHorns are a solid  8 Ohms with first order x-overs.

Thanks

[Paul Joppa]

Those are reasonable things to exeriment with.

Coupling capacitor changes will affect overload recovery time constants as well as frequency response-probably not relevant with horn systems, but I mention it for completeness.