Need an audio brainiac...Stumped...


I just scratch built a "dynaclone" ST35 6BQ5 power amp from the schematic and stayed true to the dyna design save i used a 5u4 tube HV rectifier vice the diode dyna design.

The amp sounds fine on 3 different sets of speakers (Infinitys, Magnavox/Jensen alnicos, Realistic Nova 10's) but motorboats/farts when hooked up to my newly aquired Altec alnico Santanas. 2 other tube amps and 2 other solid state amps work fine on the Sanatnas.

This happens only with that one amp and speaker combination (Dynaclone and Altec Santanas)
trichifi
You might want to ask this question over at www.diyaudio.com

There's lots of guys over there with the knowledge to give you some real, valid guesses backed by experience.
This Wikipedia writeup seems relevant.

My suspicion is that there is nothing wrong with the speakers, and that some marginal instability in the amplifier is being brought out by the particular impedance characteristics of the Santana.

The following statements in the writeup particularly grab my attention, especially given that you have modified the original design to use a tube high voltage rectifier:
Low frequency oscillations like motorboating indicate that some device or circuit with a large time constant is involved, such as ... the filter capacitors and supply transformer winding.

One common cause is feedback through the plate power supply circuit.[2][4] The power supply provides DC current to each tube's plate circuit, so the power supply wiring (power busses) can be an inadvertent feedback path between stages. The increasing impedance of the filter capacitors at low frequencies can mean that low frequency swings in the current drawn by output stages can cause voltage swings in the power supply voltage which feed back to earlier stages,[2][4] making the system a subaudio oscillator. This is caused by inadequate power supply filtering or decoupling.
Regards,
-- Al
Hey Guys, Its solved. Most of us were on the right track with the various inductive properties mismatch theory and it turns out, after studying crossover theory that the cure seems to be a "Zobel" series RC network in paralell with the amps output. Many crossover designs actually incorporate these in the LF driver portion of the crossover to make the amp "feel" a constant impeadance and thusly make it a happy camper with "bangy" high inductance, super reactive unstable feeling alnicos.

I started with a 14uf + 10ohm resistor which stopped the motorboating but seemed to take away some of the "umph". I reduced to 8uf + 6 ohms and it sounds smashing. Installed a switch and those components in my amp so I can add or delete the Zobel depending on my speaker choice with a flick of the switch. Looking forward to trying out on some of the other speakers that were non-problemaic to see how fidelity does in comparison with in or out of the Zobel.

Interesting to note; only the big alnico altecs seem to have the massive ammount of inductive kickback to cause the issue. Thanks to all of you for the input and and thanks to Otto Zobel up there in the big laboratory in the sky!
Found a weak HV filter cap (old salvage part) have new one on order. Will be intetresting to see if the Zobel is still needed to prevent the motorboat on the Altecs...