On this site I often read about how bad 12A*7, and other 9 pin miniature preamp/driver tubes out there. (mostly due to their non-linear properties). And worshiping most things octal. Most hatred seems to come from the 12AU7. If these tube are so bad, why did so many audio people from the "golden age", like McIntosh and Heathkit, use them?
~John~
~John~
This is my opinion. Some will disagree.
I think it's not really that these are so bad, but that there are many times other valves that can do the job better. There were many regarded amplifiers of earlier days that used these valves and they sounded very nice, in spite of the ugly duckling valves, but upgrading them today can make them sound even better. Today, for new construction, it only makes sense to use what has been found to sound best in that particular function. And that usually is not a 12AX7 or 12AU7.
There are some nice sounding 9 pin small signal valves. Two that come to mind are 6FQ7/6CG7 and 12BH7. There are lots of mixed opinions on 6DJ8 and 12AT7.
Wade
I think it's not really that these are so bad, but that there are many times other valves that can do the job better. There were many regarded amplifiers of earlier days that used these valves and they sounded very nice, in spite of the ugly duckling valves, but upgrading them today can make them sound even better. Today, for new construction, it only makes sense to use what has been found to sound best in that particular function. And that usually is not a 12AX7 or 12AU7.
There are some nice sounding 9 pin small signal valves. Two that come to mind are 6FQ7/6CG7 and 12BH7. There are lots of mixed opinions on 6DJ8 and 12AT7.
Wade
amp or tube comparison ?
IMHO you should listen to the preamp without looking at the tube type. Compare completed amp circuits rather than just the tubes themselves.
Some designers use certain tubes based on what they like and if the end result sounds good , so be it .
I don't think it make sense just to pull out tubes and put in others of similar type unless they are all working in the region the designer intended to. That would mean that some/most resistor values around it might have to change.
Want to compare tubes......... try them in their own competently designed circuits . Easier said than done !
Cheers.
IMHO you should listen to the preamp without looking at the tube type. Compare completed amp circuits rather than just the tubes themselves.
Some designers use certain tubes based on what they like and if the end result sounds good , so be it .
I don't think it make sense just to pull out tubes and put in others of similar type unless they are all working in the region the designer intended to. That would mean that some/most resistor values around it might have to change.
Want to compare tubes......... try them in their own competently designed circuits . Easier said than done !
Cheers.
I have always stated that there are very few tubes that are "bad", only nonlinear sets of operating conditions. Granted there are tubes that were not designed to be linear, like those that were designed for "gain controlled" (AGC) amplifiers. Most other vacuum tubes can be biased for very linear operation. Often those conditions are very different than the "commonly used" conditions, and often they vary considerably from tube to tube.
The 12AX7, 12AT7, 12AV7, and 12AY7 and their industrial offspring all have operating points useful for audio amplifiers. The 12AU7 might have a set of useful conditions, but I haven't found them yet. I am not saying that the 12AU7 is useless for audio (yet), but there are better tubes to consider (6CG7, 6FQ7, 12BH7, 7044, 5687, 7119 etc) with similar gain specs.
The 12AX7, 12AT7, 12AV7, and 12AY7 and their industrial offspring all have operating points useful for audio amplifiers. The 12AU7 might have a set of useful conditions, but I haven't found them yet. I am not saying that the 12AU7 is useless for audio (yet), but there are better tubes to consider (6CG7, 6FQ7, 12BH7, 7044, 5687, 7119 etc) with similar gain specs.
What's good or bad?? I truly believe it's your personal choice.
Through A/B comparison, it would be easy for anyone to pick the one they like.
I am glad we are the DIYers that we could build some amps which we could spend as much as we can for the best result.
My friend built 9 Marantz 7C line amps with different components and architeture. He is going to build the last one that he call it the ultimate version M7.
Last night, my friend invited me to listen to a rather famous Japanese factory made line amp and compared to his clone Marantz 7C. Surely, I would say the DIY M7 is way superior.
12A*7's are bad??? I would say hamburgers are great!!!!
Cheers,
Johnny
Through A/B comparison, it would be easy for anyone to pick the one they like.
I am glad we are the DIYers that we could build some amps which we could spend as much as we can for the best result.
My friend built 9 Marantz 7C line amps with different components and architeture. He is going to build the last one that he call it the ultimate version M7.
Last night, my friend invited me to listen to a rather famous Japanese factory made line amp and compared to his clone Marantz 7C. Surely, I would say the DIY M7 is way superior.
12A*7's are bad??? I would say hamburgers are great!!!!
Cheers,
Johnny
This forum is great. I have learned to appreciate rare things like diyAudio.com while they're here-never it take for granted.
Back on this subject:
Is there somewhere perhaps you can refer me to for finding drop-in replacements for tubes? As an excellent example, perhaps for the 12AX7?
Thanks a lot.
Back on this subject:
Is there somewhere perhaps you can refer me to for finding drop-in replacements for tubes? As an excellent example, perhaps for the 12AX7?
Thanks a lot.
The words to use in a web search are "tube substitution." There have been a lot of tube substitution guides written, including nice ones printed in the RCA tube manual.
Why would you need to replace a 12AX7 with something else? It is one of the easiest-to-find tubes on Earth. If you're not happy with how a circuit works, fix the circuit.
Why would you need to replace a 12AX7 with something else? It is one of the easiest-to-find tubes on Earth. If you're not happy with how a circuit works, fix the circuit.
I'd like to try a 12BH7 for the long tail pair but I have one problem. I have all the preamp tubes wired in series and connected to a separate DC (unregulated) power supply. The heater string consists of three 12AX7's and 2 of the dreaded 12AU7's. Wired in the 12V mode, they use 150 mA. Swapping the 12AU7's for 12BH7 bring down the voltage and possibly strain the P/S because they use twice the current. (300 mA in 12V mode) So what can I do, besides redesign the whole thing?
TubeHead Johnny said:
put a suitable resistor in para to the 12ax7s to deliver the extra 150ma that you need. if the psu can deliver it of course, there are 3 12ax7s, at 12v @ 150ma, so that's approx [when heated up] 80 ohms, three of same, = 240 ohms, put a 240 ohm R in parallel to these, then join one end to the 12bh7s so that they can get the 300ma that they want. the resistor has to dissipate just under 6W so a 10 w component should do the trick.
just a thought
Your dilemma is only one of many reasons that you can't simply pull one valve out and plug another one in. The operating parameters of the entire circuit must be considered and to change to a different type may involve redesigning the whole thing to work optimally. With new construction you can simply design for the type you have determined is best suited to do the job you are expecting it to do.
I would like to convert my Heath W5Ms to use 6CG7s rather than 12AU7s, but I face similar problems in that the power supply already runs hot with KT66 output valves, and their heater current requirements. If I were to use only 6L6GCs it would probably work but I would rather keep the KT66 capability. So for now I am living with the 12AU7s.
Wade
I would like to convert my Heath W5Ms to use 6CG7s rather than 12AU7s, but I face similar problems in that the power supply already runs hot with KT66 output valves, and their heater current requirements. If I were to use only 6L6GCs it would probably work but I would rather keep the KT66 capability. So for now I am living with the 12AU7s.
Wade
TubeHead Johnny said:
Yes, but what does "bad" actually mean anyway? In the first place, most of the 12*7 family were designed as small signal RF types. So far as 12AU7s are concerned, it's an RF tube, and most of the circuits using it are: oscillator/buffers, push-pull frequency triplers, push-push doublers, push-pull drivers -- all operating in Class C. You frequently see it used as a mixer/detector where linearity is definitely not a good thing. The other big class of circuits is digital (logic gates, RS latches, flip-flops) or semi-digital apps like multivibrators, Schmidt triggers, and so forth. Audio was definitely an after thought. Does that mean you can't use it for audio? Not necessarily. You could probably use it for a cathode follower grid driver. If it produces mainly h2, then you could use it in a balanced topology such as an LTP, or SRPP, where h2 is largely cancelled. It just might work well in that situation.
My design philosophy is to "never say never", and I have used types like the 6BQ7 that don't even bother to make a pretense of being an audio tube, which don't necessarily have good loadlines, and which I wouldn't use in any conventional audio amp circuit. However, as an audio cascoded LTP, this type sounds just great. Use it in a conventional small signal stage, and it probably won't sound so swell.
As for other types, devices like the 12AX7, a high-u triode, these can produce more of the higher order harmonics, like a small signal pentode (large u-factor, and high r(p) ). The demand for this type arose during the later 1940s, early 1950s, when the Big Box manufacturers wanted to really horse up the gNFB in order to force impressive THD measurements to please the marketing depts. Throwing in 20+db(v) of gNFB requires an open loop gain that's considerably higher than that in order to preserve any sort of sane sensitivity. Many designs from previous times didn't have 20db(v) of open loop gain, and so the older octal small signal types were designed to be a good deal more linear (because they were used without gNFB at all, or only a small amount of gNFB to correct for the OPT and much smaller circuit nonlinearity) and to have lower u-factors. Types like the 6J5 and the 6SN7 have u= 20, and the 6SL7 has a u= 70. Compare to the 12AX7 and its u= 100. Doesn't mean that you can't use a 12AX7, but if you do, you also need to pay attention to the possibility of higher order harmonics, just as you would if using small signal pentodes.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.