RMS output differences in all-tube amps

[Bill Whitehurst]

I think that although there is, or should be differences due to impedence matching,
plate voltage, distortion levels and tube life expectations,
a lot of the differences are due to just how realistic are the manufacturers claim to begin with.

 

[lfmn16]

That's not RMS, it's "tube watts".

Everybody knows that tube watts are wattier than solid state watts!

 

[primusfan1989]

Regarding ultralinear operation, I've been a Talkbass member since 2009, and based on various posts I have read over the years, the general sentiment regarding ultralinear operation seems to be that it doesn't sound as good as nonlinear operation. Any comments on that, anyone?

I suppose it depends what you're going for, I wouldn't say ones better than other. They're just different. My Fender Studio Bass is rated 200watt at 4ohm runs a matching sextet of 6l6s ultra linear. I don't get too much power amp distortion out of the thing, stays pretty clean and that's what I want out of it.

 

[Rockbobmel]

The 35w per 6550A tube rating is for class AB fixed bias ultralinear operation.

Normally the rating for a pair, push pull, fixed bias for class AB1 is 100W. But keep in mind, the data sheet specs are for a long service life. With the right plate and screen voltages, fixed bias, and cooling, the tube can be pushed beyond the published limits. Is it good to do that is another question.

With a more conservative design, the tubes will last longer. An important design consideration.

My Sunn 2000S (4x 6550) is rated at 120 Watts. (BIG Iron), It is seriously loud and solid sounding.
My Mesa is 225-300 watts (12 x 6L6GC) but not that linear. (Smaller Iron)

 

[beans-on-toast]

My Sunn 2000S (4x 6550) is rated at 120 Watts. (BIG Iron), It is seriously loud and solid sounding.
My Mesa is 225-300 watts (12 x 6L6GC) but not that linear. (Smaller Iron)

It's all in the design!

 

[Mr. Foxen]

Regarding ultralinear operation, I've been a Talkbass member since 2009, and based on various posts I have read over the years, the general sentiment regarding ultralinear operation seems to be that it doesn't sound as good as nonlinear operation. Any comments on that, anyone?

My experience with UL output sections, covering Matamp Slave 300 (4xKT88), various Simms Watts (4x EL34 or KT88), STAK 130RMS and 230RMS (4 EL34 and 4 KT88 respectively) and Sound City Bass 150 (4 KT88), is they all give a certain impression of weight and size to the sound, even when run turned down, like there is still a lot of potential there. With the possible exception of the Sound City, these are all super high build quality amps with really big iron, some with power supply chokes, the Sound City is what makes me think the being UL is something to do with it, since that does the thing, with not that big iron and kind of low midrange build. I am definitely comparing them with fairly peer group standard pentode amps, as big valve bass amps are something I have in abundance. I wound up using the Sound City 150 over my GT200 for recording, much less gnarly break up, but that bigness to it.

 

[TinIndian]

All I can say in this is my 1980 Music Man HD150 gets 150 watts from 4 6L6GC's. And it appears that they got that much by running a crazy plate voltage of 700 volts.

 

[SactoBass]

All I can say in this is my 1980 Music Man HD150 gets 150 watts from 4 6L6GC's. And it appears that they got that much by running a crazy plate voltage of 700 volts.

Thanks TinIndian for that post!

It seems that the differences in the amount of B+ plate voltage being fed to the power tubes is a main factor in the differences of RMS output of different amps with the same type and number of power tubes. That is what my original post was all about......asking why different amps with the same type and number of power tubes have such different rated RMS outputs.

 

[RED J]

All I can say in this is my 1980 Music Man HD150 gets 150 watts from 4 6L6GC's. And it appears that they got that much by running a crazy plate voltage of 700 volts.

Yes. I can't contribute much more, but wondered how my Fender 300PS wrung 75 watts out of each 6550 in it. I found the answer: High plate voltage.

 

[SactoBass]

Yes. I can't contribute much more, but wondered how my Fender 300PS wrung 75 watts out of each 6550 in it. I found the answer: High plate voltage.

I'm seeing a pattern here!

 

[agedhorse]

My Sunn 2000S (4x 6550) is rated at 120 Watts. (BIG Iron), It is seriously loud and solid sounding.
My Mesa is 225-300 watts (12 x 6L6GC) but not that linear. (Smaller Iron)

The "size of the iron" does not matter if the materials are not the same. There are different permeability core materials, some are capable of MUCH higher flux density before saturation which means less "iron" is necessary, which also means that less copper is necessary, and less copper means less IR losses and higher efficiency, lower temperature rise etc. SOME manufacturers might boast about how big their transformers are without mentioning that perhaps they are using power permeability cores which are generally less expensive (ie. cheaper) but need to be bigger for the same performance.

In very large power transformers (ie. power grid distribution), and industrial motors, this higher permeability, low loss cores have been standard for years in order to achieve the super high efficiencies and extended performance. These transformer can exceed 99% efficiency and motors in the 98% range (larger sizes)

Thanks TinIndian for that post!

It seems that the differences in the amount of B+ plate voltage being fed to the power tubes is a main factor in the differences of RMS output of different amps with the same type and number of power tubes. That is what my original post was all about......asking why different amps with the same type and number of power tubes have such different rated RMS outputs.

Plate voltage is only one of 5 or 6 important variables that are responsible for the output capabilities of an amp. It COULD be the main factor in some cases but it could be WAY down the list in other cases, The answer is "it depends".

 

[TinIndian]

I am quite certain that Aged Horse is correct and there are several different variables that come into play in this. It seems that one fairly common way to squeeze more out of a set of tubes is running higher than usual plate voltages. As in many things though, there is more than one way to skin a cat, as they say.

 

[agedhorse]

I am quite certain that Aged Horse is correct and there are several different variables that come into play in this. It seems that one fairly common way to squeeze more out of a set of tubes is running higher than usual plate voltages. As in many things though, there is more than one way to skin a cat, as they say.

Another way is to design your output transformer to your specific circuit to match the plate to primary impedance more closely. I have seen some really poor mis-matches, and many are pretty mediocre. Also, it's possible to improve the transformer efficiency with good choices of materials and winding patterns, which also improves power transfer.

 

[Aidenn7]

basically match everything up to make better amp. like matching tubes. etc. power supply voltage to the transformer. how do you match a transformer to the voltage? Or is that stupid question.

 

[beans-on-toast]

Designing a good sounding audio transformer is a lot more complicated than one would imagine. When it can't be done with just a computer, they say that it involves art as well as engineering. This certainly applies to audio transformer design.

You can design an audio transformer that looks good on paper but you never know how good it is going to sound over the sonic landscape. Tweaking the design with listening tests is part of the process. Fortunalety, or unfortunately, this is often done by committee. Second guessing after you commit to a design is also part of the process.

Having experience is so important as it helps you to cope with the design tradeoffs dictated by real estate, desired performance features, heat, shielding, and cost. There are different types of frequency dependent distortions the dictates how well a transformer copes, especially in the low end. A bass amp transformer often has different requirements over one used for a guitar amp.

Even the materials factor in. Some metal formulations that were used in the past are too costly to use because of environments laws and other factors. In studio equipment design, some people pay a premium to use vintage audio transformers because they considered them superior.

The building blocks and the designs are always evolving and presenting new challenges. The field is so complex and far from stale.

 

[timv]

Actually, it has a B+ of 700 volts which isn't super high, the fender 300ps and 400ps push the power tubes with a transformer driven by a 6V6 in the 300 and a 6L6 in the 400. It pushes the tubes very hard but I've owned both and have another 400ps right now and on the bench they will put out the rated power clean when setup correctly.

Yes. I can't contribute much more, but wondered how my Fender 300PS wrung 75 watts out of each 6550 in it. I found the answer: High plate voltage.

 

[BassmanPaul]

Output power coupled with distortion are one of the design goals for the amp designer. The tubes he/she selects to reach those goals are his/her choice. Some go for the maximum power from the minimum number of output tubes. Others select a more conservative approach.

 

[beans-on-toast]

Output power coupled with distortion are one of the design goals for the amp designer. The tubes he selects to reach those goals are his choice. Some go for the maximum power from the minimum number of output tubes. Others select a more conservative approach.

Tisk tisk. What about the women that design amps.

 

[timv]

I share what I've tested for output power on my bench of several amps all the time on here. A lot of times I've gotten into forum arguments with my results but they are what they are.

I have a nice old all tube frequency generator and do all my tests at 100hz.
I have a nice old variac with current and voltage gauges but also connect a multimeter and keep the line voltage at 120 throughout the test so I'm always comparing like to like, even if my ac kicks on.
I have a load box assembled with 50watt 1% power resistors giving be 3 different 4 ohm loads and two 8 ohm loads.
I have a multimeter set to ac volts and an oscilloscope across the load.
I usually will use the power amp in if available and bypass the preamp.
Turn up the level on frequency generator until the peaks just start to flatten while making sure to keep line at 120volts.

With that setup I have tested:
Mesa bass 400, d-180, m-190 max clean power about 180-190, only about 10 watts more with 6550's.
Mesa bass 400+, max clean about 240-250.
Traynor yba-3a super custom special, about 240-250 clean.
Ampeg SVT's. I am kind of obsessed with SVT's and try to learn and see all the differences throughout and record the data.
My 1970 with 6146's does about 350-360 clean.
I have a 1977 that I just resurrected from death that has 695 volts b+ that does 300 clean.
I have an svt 2 that I just got for $100 that survived shipping disasters that I can only get about 250 ish out of and the b+ is only 650.
Most other SVT's I've tested from a 72' blue line to mti's usually only do 270.
My peavey classic 400's are really picky about tube type as stated in the mega thread. The most power I've tested was with the Chinese tubes they originally came with which put over 400 clean out. The worst bench tested was a nice matched octet of ge 6550's that did a bit over 200 clean. And that is due to the output transformer and tube matching.

Just figured I'd share this again. No arguments necessary, I listed my test setup and it isn't a lab or anything but for my home workbench it is repeatable and accurate and years later I can add to the data and know I'm still using the same base setup.
Thanks.

 

[beans-on-toast]

For consistency, I find it good to first configure the amp with the tone controls set flat, any boosts off.