Listening to Tube Amps

[gdmoore28]

Even though I'm 63 years old, it's been a number of years since I've been able to listen to tube amps through good quality speakers. Because my current speaker build is based upon Altec A7 components (416Z woofs; 511B horns; 802D drivers; custom network), I wanted to have likewise period-correct amplification. Through much trading, selling, and bargaining, I now have some gear that I could afford - two Dynakit Mark IV monobloc amps and a Dynaco PAT 4 preamp. All three components are going into the shop for the usual upgrades, but I just had to hear them once before turning them over to the tech.

The speakers used were my DIY pieces utilizing a PE 10" aluminum-cone woof, 5 1/4" aluminum-cone mid, and the "silkie" tweeter, crossed at 500/4000Hz. I used my Pioneer PL400 turntable and Shure M97x cartridge. This system is normally powered by my Pioneer SX980 receiver. (Yes, I live on the "poor man" end of the audio spectrum and love classic old gear :D)

As you tube guys might well have predicted, there was a huge difference in sound between the tube amps and my SX980. Not better or worse, just wildly different. What surprised me the most was the abundant bass response from the tube amps (all tone controls were set to flat). I'm not sure if I like it, as the bass seemed a bit on the loose side - not well-controlled as on the Pioneer. However, the mids were sweet and smooth and very impressive. The treble sounded smooth, as well, but not crystal-ly and sparkly as from the solid state. In all honesty, I would much prefer the Pioneer for use with these speakers.

Everything I've read indicates that will not be the case with the Altec speakers, though. Experienced users of the Altec/tubes mix say that the tube amps (along with proper horn compensation) remove the harshness frequently heard when the horns are driven by solid state amps. We'll see. I have no idea what to expect from the bass response.

Anyway, those are my impressions. I was able to get in about 1/2 hour of listening time before one of the Mark IVs blew it's fuse, so extended listening may change my mind. Now it's off to the tech with the amps and the preamp. I hope to have the speakers built by the time I see them again.

Just curious as to whether my experience with the tubers reflect your own.

GeeDeeEmm

[JRT]

Looking at some online pics, your Pioneer SX980 receiver should have a pair of jumpers across preamplifier output and power amplifier input ports. It might be interesting to try using the Pioneer SX980's preamplifier section in combination with the Dynakit Mark IV monoblock power amplifiers to see if that affects perceived response, and likewise a separate listening test of the Dynaco PAT-4 preamplifier in combination with the power amplifier sections of the Pioneer SX980. If nothing else, it should be fun to play with, and you might find some interesting results.

[gdmoore28]

Looking at some online pics, your Pioneer SX980 receiver should have a pair of jumpers across preamplifier output and power amplifier input ports. It might be interesting to try using the Pioneer SX980's preamplifier section in combination with the Dynakit Mark IV monoblock power amplifiers to see if that affects perceived response, and likewise a separate listening test of the Dynaco PAT-4 preamplifier in combination with the power amplifier sections of the Pioneer SX980. If nothing else, it should be fun to play with, and you might find some interesting results.

Yep! You're reading my mind. I'd set aside the entire day for doing that exact experiment, but when one of the Mk IVs blew it's fuse, that put an end to the experimentation. Drats.

GeeDeeEmm

[Psycoacoustics]

I'm a fan of tube preamps and solid state amps. Not quite the best of both worlds, but from where i sit I like the way the solid state amp controls the woofer better.

Just my .02. I'll be interested to read your thoughts after you do your experimenting. Mark

[gdmoore28]

I will. I picked up a Furman active crossover from Goodwill the other day, so I will be able to try bi-amping these as well. I've read numerous posts that recommend using the tubes for the horn and SS for the woofer. We'll see.

GeeDeeEmm

[JRT]

I picked up a Furman active crossover from Goodwill the other day, so I will be able to try bi-amping these as well. I've read numerous posts that recommend using the tubes for the horn and SS for the woofer. We'll see.

If you want to use the Dynakit Mark IV monobloc amplifers in that way, I would suggest adding a Zobel passive network to flatten the impedance of the load driven by the tube amplifier.

The output tubes see undesirable variation in load impedance at the amplifier output. Load impedance reflects back through the transformer such that the load impedance at the primary can be approximated as the product of load impedance at the secondary and square of the turns ratio.

[Æ]

If you want to use the Dynakit Mark IV monobloc amplifers in that way, I would suggest adding a Zobel passive network to flatten the impedance of the load driven by the tube amplifier.

Oh fooey, a Boucherot cell.

[Æ]

I'm a fan of tube preamps and solid state amps. Not quite the best of both worlds, but from where i sit I like the way the solid state amp controls the woofer better.

Just my .02. I'll be interested to read your thoughts after you do your experimenting. Mark

You'll certainly need a fan if you have them both running at the same time!

[gdmoore28]

I'd intended all along to do a Zobel network, so thanks for confirming my choice. Now, from what I understand, the Zobel network placed after the crossover network functions primarily to "normalize" the impedance seen by the crossover. I'd never heard the term "Boucherot cell" before, so had to look it up. The latter is intended to control the impedance seen by the amplifier itself, correct? We're wading in waaaay over my head here. Am I understanding it correctly?

GeeDeeEmm

[bwaslo]

Whatever name you use, I think what you want to do is get an impedance curve for your Altecs including crossover as seen by the amps (easy with a DATS or WT2 or WT3) and configure a network to put in parallel with each speaker to make it's impedance more flat like a resistor's. If you can get the measurement, I'd be happy to design a network to do it.

[Psycoacoustics]

You'll certainly need a fan if you have them both running at the same time!

Nah,,,,, my air conditioning works well! Actually the Conrad Johnson pre does run a little warm, but my B&K st140 doesn't unless I push it hard for 2 or 3 records.

Now my modified Hafler is a different story. Biased high, it is warm to the touch Always! It's in my basement though so no worries there. Mark

[Æ]

I'd intended all along to do a Zobel network, so thanks for confirming my choice. Now, from what I understand, the Zobel network placed after the crossover network functions primarily to "normalize" the impedance seen by the crossover. I'd never heard the term "Boucherot cell" before, so had to look it up. The latter is intended to control the impedance seen by the amplifier itself, correct? We're wading in waaaay over my head here. Am I understanding it correctly?

GeeDeeEmm

It is very basic stuff, not way over your head. I remember way back in the very first or second issue of Speaker Builder Magazine an article discussing amplifier output damper networks. I think it was written by Nelson Pass and it also mentioned Mathew Polk, but I don't remember all the details. It seems that many if not most solid state amplifiers and receivers incorporate a generic circuit consisting of a resistor, capacitor and even a small inductor at the output. For amplifier stability, to prevent oscillations with very reactive loads, especially capacitive loads. I usually leave the parallel cap and resistor in place, but will remove the series inductor.

[Pete Schumacher]

It is very basic stuff, not way over your head. I remember way back in the very first or second issue of Speaker Builder Magazine an article discussing amplifier output damper networks. I think it was written by Nelson Pass and it also mentioned Mathew Polk, but I don't remember all the details. It seems that many if not most solid state amplifiers and receivers incorporate a generic circuit consisting of a resistor, capacitor and even a small inductor at the output. For amplifier stability, to prevent oscillations with very reactive loads, especially capacitive loads. I usually leave the parallel cap and resistor in place, but will remove the series inductor.

That's something completely different than impedance compensation for a speaker. You're talking about negative feedback loop compensation to keep the phase shift less than 180 degrees while the gain is greater than unity.

Amplifiers with very low loop gain, like many tube amps, won't require such networks. But because of their low loop gain, they also exhibit poor load regulation and the use of conjugate networks to linearize the load impedance keeps the amplifier gain across the audio band far more linear as well.

[Æ]

That's something completely different than impedance compensation for a speaker. You're talking about negative feedback loop compensation to keep the phase shift less than 180 degrees while the gain is greater than unity.

Amplifiers with very low loop gain, like many tube amps, won't require such networks. But because of their low loop gain, they also exhibit poor load regulation and the use of conjugate networks to linearize the load impedance keeps the amplifier gain across the audio band far more linear as well.

Well, I didn't mean to imply that even though it may have seemed that way. Even if tube amps don't need it, apparently they benefit the most from it! I wish I had that copy of Speaker Builder Magazine so that I could have quoted it correctly in its entirety. I understand the concept, but next time I run into Nelson Pass I'll ask him about it in detail. He does show up every now and then at DIYs here in Northern California.