UTC o-9 ouncers for headphone amp?

[mctavish]

Hello, got a pair of these UTC O-9 transformers and I was contemplating a one tube headphone amp. They are 15k primary with 50/250/600 ohm secondaries. The are rated for 4ma. of D.C. and thus their low-end frequency response is a bit compromised. The datasheet says they are 4db down @ 100hz but I was wondering if I connect them in a parafeed config will this improve? Or is the response down due to the airgap lowering the primary impedance? Anyone have any suggestions as to what to do with these? Is a headphone amp out of the question? Thanks in advance for the input, -Mac

[Palustris]

As a headphone amp you could try a 5687/7119/7044 with 100V on the plate and a cheap red LED as bias. The tube should run just over 20mA at quiescent so parafeed is mandatory. You can use a KBG 2uF - 4uF PIO as the parafeed cap.

[mctavish]

That sounds like what I was thinking. Could I use a CCS for the tube load instead of an inductor? Also, do you think the low end freq. response would be improved with parafeed operation despite the airgap in the trafo?

[kavermei]

Originally Posted by mctavish

do you think the low end freq. response would be improved with parafeed operation despite the airgap in the trafo?

The inductance of a gapped inductor does indeed decrease a little with increasing DC current. So at 0A DC I would expect that you have a little more bandwidth than at 4mA DC current. However, this effect is certainly not of such magnitude that an LF point of 100Hz @ 4mA suddenly becomes 20Hz @ 0mA.

My advice to you would be to use a parafeed circuit in any case. The limit of 4mA DC is so low that it rules out a lot of tubes anyway. Then measure/listen and see if you are satisfied with the LF response.

... So how can you improve the LF response in case you are not satisfied?

I see two possibilities. You could add a little feedback, or you could use a shelving filter at the input of the amp to compensate for the LF rolloff of the transformer.

Both solutions will result in a larger amplitude of LF signals being sent through the primary winding, to compensate for the reduced LF response of the transformer.
Luckily, in the absence of DC current, you can have more signal swing before reaching saturation. So if you go with parafeed, by eliminating the DC magnetization in the core, the transformer will (hopefully) be able to accomodate these larger LF signals.

So to summarize you have two reasons to use parafeed:
- 4mA DC limit is uselessly low for most tubes
- Absence of DC core magnetization helps accomodate larger LF signals if you use feedback or prefiltering.

Kenneth

[Palustris]

You could use a CCS or a resistor as the plate load for the tube in parafeed instead of the inductor, but you may have to try them to see what you like.

[jcx]

lower frequency from xmfr:

autotransformer connection?

also deliberate impedance mismatch, driving the headphone "current mode" by connecting them to higher impedance tap - makes them sensitive to headphone impedance curve, hurts power too so headphone sensitivity can be an issue