Question on HoSo56

[jazbo8]

Ran some simulations on the Da Geezer's HoSo56 design, and I have some questions:

1) the smoothing caps in V1 has a huge effect on the frequency response - it is not subtle at all. So my question is, why is it such large value (1nF)? And a follow on question, if 1nF is used on V1, then what's the purpose of the 250pF smoothing cap on V2A? As its corner frequency is way above that of the V1, can the effect still be audible?

Here are the frequency responses for stage 1 (left) & 2 (right), sorry I did not mark the curves with switch positions, but they should be easy to figure out... Note how far apart the HF corner frequencies for stage 1 & 2 are from each other.

[jazbo8]

2) Output tube bias. Based on the voltages from the schematic, the EL84 appears to be operated quite a bit above its 12W dissipation limit (perhaps not as high as the plot indicates after factoring in the cathode voltage). Could someone please verify the readings on Geezer's schematic? Or perhaps the EL84s can handle a lot more than what the specs say? Here is a comparison between the datasheet (left) and the HoSo 56 (right):

[sluckey]

Hmm, using a simple calculator shows that the power tubes are idling at 25 watts. That's BOTH tubes and totally ignoring the screen current. Looks right on target to me.

[jazbo8]

Yes, that's what I did first, and everything looked ok, then I tried to plot the load line, which led to some discrepancies, and why I asked for a confirmation on the voltage readings. Plugging in Geezer's numbers, I get the plot on the right with Ip=57mA at idle, which not only looks high and implies a bias of ~-17V with Rk of 150 Ohm (the sim was fixed bias). The EL84 SPICE model should be ok since the left plot match up well with the datasheet's figures. May be I missed something else?!

[EDIT] Looked over the calculations and the sim results again, I think either the indicated bias voltage or the screen voltage is off. For example, lowering the bias to -14V could get the Pdiss close to spec. Still confused 

[jazbo8]

Since the amp is one the most popular designs at Hoffman, I thought many have built it... Could someone please confirm the readings on Geezer's schematic or your own? FYI, from the schematic - bias of -11.4V with Vg2=337V, Vp=340V and Ra-a=8k.

TIA,
Jaz

[sluckey]

Have you sent a pm to Geezer? He's always been accurate and thorough with his projects so I have high confidence in his numbers.

[jazbo8]

Have you sent a pm to Geezer? He's always been accurate and thorough with his projects so I have high confidence in his numbers.

I will try, didn't think of that...