Leok

[Deang]

Leo,

I've concatenated some posts out of another thread that you may not have read yet. The other thread is slow to load because of all the pictures in it.

I originally asked the question if there was a way to tell how much power was going across the speaker terminal inputs from an amp. I've been dealing with a bit of a paradox -- and wanted to see what you thought.

Dean,

Without a scope you really can't . But I found one thing out as pertains to you using the 4 ohm taps. With tube amps they produce there rated wattage at all taps unlike solid state. The only way the rated output changes is if you hookup a 8ohm load to the 4 ohm tap and create a indepadance mismatch. So if your RF-7 are indeed a 4 ohm speakers hooked to 4 ohm taps on the Apollos or the Scott your actually getting the rated output. Also if it seems like the amp drops over all output using the 4 ohm taps then your RF-7s are not 4 Ohm speakers !

Craig

Craig,

The bottom line is this: Since the RF-7 has those low spots, it is best to accomodate them by using the 4 ohm taps. The overall output of the amp is halved, however -- it will deliver the necessary current when the impedance drops without attenuating the frequencies at those positions. At least, that is my understanding at this point.

Running the RF-7's on the 8 ohm taps will generate more current overall, but the frequency response will follow the impedance curve at those instances where the impedance drops.

My question asking if there was a way to tell what was going on at the terminal inputs of the speakers -- was for the purpose of validating the math.

Now, for the paradox. I've been running the Heresies on the Apollos since last night. The Apollos are wired in 4 ohm internally. I can't change them on the fly. I would have to remove parts and then resolder. I get clean output on the Heresies up to 105db. In fact, it sounds very good. There is no sense of strain or evidence of clipping. The sound is dynamic and crystalline. To be honest, compared to the RF-7's, the sound comes forward with less effort, and as far as my ears can tell -- sounds better at 105db than the RF-7s at 100db. At 100db, the RF-7's begin to harden up, and have me grinding my teeth.

So, why do the Heresies generate more output with no strain, then the RF-7's? The really odd thing, is that the Heresies are 98db/w, and the RF-7's are 102db/w. The Heresies, an 8 ohm speaker wired to my 4 ohm taps -- are taking the fullness the Apollos have to offer completely in stride (and according to the information in your post, should be an impedance mismatch, thus choking off the Apollos). The RF-7's OTOH, sound really great until they go over that 95db - 96db barrier, and then they start to sound, well...like $hit. The only way I could get the RF-7's to match the output of Heresies and Apollos, and have them maintain a semblence of integrity in their signature -- was with the Aragon. The Apollos definitely sound better at 95db and under -- but above that, the Aragon delivers a dynamite sound.

I'm just trying to sort some things out here, and will also be emailing Leo about this.

I'm beginning to wonder if some of this might have something to do with sheer mass of drivers in the RF-7, and possibly the additional parts and complexity of the crossover as compared to the Heresy. I don't know. It's odd.

At any rate, I'm looking forward to putting the SET amps on the Cornwalls. It may be that this where they belong. I would then of course put the Scott on the RF-7's -- where I think there might be less problems. Don't know, don't know.

[NOSValves]

Well my question is !! How does the Apollo accomplish its independance ? Does it actually accomplish it with seperate taps from the output transformers or are they changing resistor values ?? If this is so its basically hockus pockus and the Apollos only have one tap for independance and there just blowing smoke by adjusting values to make up for its lack of taps ! The Scott puts out 17 to 20 watts on all taps with the proper indepandance speakers hooked to them accomplished inside the tranformer rather than on the outside. The output is not halved even with a mismatch while it maybe lower its not that drastic. I bet you need to change your apollos to 8 ohms ! Just because the RF-7's have mild peaks that reach down near 4 ohms where does it spend most of its time ?? If its closer to 8 ohms than this is where the amp should be setup at so that there matched at the biggest percentage of the time.

Of coarse this is all me speculating here

Craig

[leok]

There are several items here. I may be able to help some.

Craig, on the output transformers: I don't think the various output impedances are actually "taps" as much as alternate ways to connect sets of output windings. The Hammonds I have in my pp amp are the same, and so are the ones in my Moondog kits. Connecting the windings in various series and parallel configs. produces the correct i/o winding relationships. That's why Dean says the setting is wired internally.

As far as how to best load the amp .. that's hard. One important part of the equation is how the output tube(s) is loaded in a nominal setting (4 Ohm resistor with the xfmr wired 4 Ohms, or 8 Ohm resistor with the xfmr wired 8 Ohms). In my case, with nominal loading, the KT66s see 8K Ohms, which is their min. distortion loading (the transformer was intended originally for EL84s). I don't want the KT66s to see more than that so with the RF-7s connected to 4 Ohm xfmr config. I attach an 8 Ohm resistor in parallel. If I don't, the highs get screechy. With the 8 Ohm resistors, no screech. This is great for the RF-7 also, because looking into the amp/resistor it sees better damping. When the RF-7s hit an impedance low point, the KT66s see just over 5K Ohms which is their max power loading. It's a pretty good compromise and sounds nice.

Dean, as far as I can tell, the 2A3s in the Moondog are loaded lighter than max power .. probably for lower distortion, similar to the way I loaded the KT66s. I'm guessing I can do the same trick (4 Ohm xfmr config with 8 Ohm parallel resistor). That's why I suggested you try it. I may be wrong, and have to remove the resistor, or go to 8 Ohms. Have you tried a parallel 8 Ohm resistor? Using the 4 Ohm xfmr config, the output tube on your Apollos may be insufficiently loaded.

Finally, or where you started .. high volume settings. I attended a quintet concert (string quartet + piano), the other day. The hall was quite small. At times, the violins were too loud for my ears and there was distortion .. guess where, and it wasn't the violins. I've found that I can very easily get too much high freq. power from the RF-7s or Fortes. One problem was the RF-7 voicing, which I fixed with that 10 Ohm resistor patch. But another part is my hearing .. I just can't take that much hf power. I find that the RF-7s are more likely to produce excessively loud hf, because they are so dynamic. Chances are, the speakers are getting it right, I just can't take it. Forte sound may have more compressed highs which means the rest of the spectrum will get louder before I overload .. so it seems the Fortes get louder before they distort. But really the RF-7s maintain a more accurate low-freq/high-freq power relationship.

I suggest you keep the volume below distortion for each speaker you try, and compare that way. The volumes may not sound or be the same, but at least you can hear what they are doing. Other suggestions, of course, the 8 Ohm parallel and 10 Ohm patch in the crossover.

leok

[mobile homeless]

Leok, I really recommend building the Moondogs stock and running them for the 8ohm tap first to see how it goes. Starting off with unusual configurations is not the best first running to get a handle on the way this amp behaves with your RF-7. Also, Ron DOES run the 2A3 at about 325v or so which is a bit higher than the 250v recommended in the RCA manual. Still, from what I have heard, most 2A3 does well run at the upper end of the voltage scale and will do fine. Tubes like the Sovtek will handle it find. Vintage RCA will not quite be as rugged but should do ok. You can always change the voltage although the last few things I have read put 300-325 in a good light.

kh

[leok]

Kelly,

That's a good idea about trying the 8 Ohm config. first. Since I believe the design can tolerate impedance dips that's a good start.

The tube loading I was refering to wasn't voltage, but transformer impedance as seen by the output tube with nominal loading on the output. I am under the impression that the transformer presents between 3K and 4K Ohms under those conditions. Any 2A3 data sheets I've seen suggest tube loading at around 2K Ohms, I suspect, for max power. Now comes the voltage, which maybe you were refering to. In order to get the 3.5 Watt output at the higher impedance loading, Ron would have to use a larger voltage swing.

Actually, with the RF-7s, I am more concerned with the high impedance for frequencies above that 4 Ohm dip. At those higher frequencies the output circuit is way under-loaded and I think this may contribute to hf distortion. That's why I lean towards the 4 ohm out with parallel resistor loading. Still, I should try it Ron's way first.

Leo

[Deang]

I realize I am a mere child here, but I have been living off of the archives at the SET Asylum over the last couple of days.

I just read an interesting thread last night, and it was being said was that using the 4 ohm taps on the transformer to accomodate the low impedance points of a speakers impedance curve is almost always better than using the 8 ohm taps. The result is supposedly lower distortion levels because the tubes aren't being driven to their maximum output the majority of the time -- and only rarely are they driven to their limits or slightly beyond. Another advantage from this is extended tube life.

OTOH, using the 8 ohm taps, though giving some elevated power output, actually increases distortion levels dramatically, as the tubes are being overdriven whenever they are forced to deliver current beyond their operating parameters to accomodate the low impedance.

Something I thought of last night (completely on my own:), was that the question really shouldn't be, "How many points on the curve does the speaker swing low", but rather -- "How often does the music take it there." I mean, we see an impedance curve with 3 or 4 low spots on it and think, "Well, that's not so bad." But what if the music has content that takes it to those 3 or 4 frequencies 90% of the time?

[NOSValves]

Dean,

That is indeed a great point so where in the curve are these low points ?

Craig

[mobile homeless]

You learn something new every day. I used to hang out in the SET Asylum for a few years and still frequent it every now and then, and I never heard this advice. I dont remember reading anything of the kind anywhere actually.

Personally, it doesnt sound correct to me. Alas, I would REALLY try to drop any preconceived notions and actually LISTEN to the two and compare. My understanding of the relationship seems to be opposite of some of the points in your post.

Could you please point me to the posts mentioning this? Also, who was it?

kh

[michael hurd]

Deang: Take a true rms dvm and set it to ac volt scale - measure the voltage developed by the amplifier into the load ( driving speakers ). Take the rms ac voltage, square it and divide by the speaker impedance. For example, 20 volts rms squared is 400 divided by 8 ohms is 50 watts. Now we all no that speakers are not just a resistive load, they have impedance peaks and nulls. So depending on frequency, your amplifier may put out less or more power than it's rating. Assuming that current is constant, halving the load, doubles the wattage ( 400 divided by 4 ohms is 100 watts ). But due to things like internal resistance in the amp, you may end up with something like 85 watts. But then again, the difference between 85 watts and 100 watts is not even discernable.

[NOSValves]

Dean,

Your listening experience you mentioned in the other thread stating that the RF7's sound like sh!t above 100db would make me think you need to try the Apollos on the 8 Ohm tap ! I wonder how often these independance peaks come into play on the music you listen too. But what really freaks me is why the Heresy's can bellow out 105db and sound better if indeed the apollos are being choked off with this halved output . Something weird is going on here. I wonder what the curve is on the heresy's.

Craig

[Deang]

http://db.audioasylum.com/cgi/m.pl?forum=set&n=21459&highlight=4+ohm&r=&session=

http://db.audioasylum.com/cgi/m.pl?forum=set&n=1866&highlight=4+ohm&r=&session=

http://db.audioasylum.com/cgi/m.pl?forum=set&n=21471&highlight=4+ohm&r=&session=

http://db.audioasylum.com/cgi/m.pl?forum=welborne&n=877&highlight=dip&r=&session=

http://db.audioasylum.com/cgi/m.pl?forum=welborne&n=551&highlight=4+ohms&r=&session=

I did find this hellish thread, and didn't understand a word of it.

http://db.audioasylum.com/cgi/m.pl?forum=set&n=15417&highlight=impedance&r=&session=

[Deang]

Craig,

I'll give you a call later tonight.

Something Leo touched on, that hasn't really been discussed, is the peaks. How bout going from say 3.9 ohms to 30 ohms? How about them K-horns on SET amps -- they seem to do O.K.

Think I'll just stay on those 4 ohm transformer taps. Going to the 8 ohm taps just brings in a different set of problems. As it is, the sound is fabulous, and is pretty much where I want it -- dynamic, articulate, and oh so wide open. I've learned my lesson -- I'm leaving it the hell alone.