questions about the 12AX7

what are some of the advantages and disadvantages of using the 12AX7 tube as the first gain stage in a mic preamp? Also, through my research of tube preamp schematics I noticed that the German designs used pentodes a lot and the American designs orften used triodes (on average, especially documented in Radiotron Designers Handbook). Does anyone know why the Germans favored pentodes over triode design? Can pentodes opperate with comparitively lower noise than triodes in preamp applications?

[quote author="kevinnyc"]what are some of the advantages and disadvantages of using the 12AX7 tube as the first gain stage in a mic preamp? Also, through my research of tube preamp schematics I noticed that the German designs used pentodes a lot and the American designs orften used triodes (on average, especially documented in Radiotron Designers Handbook). Does anyone know why the Germans favored pentodes over triode design? Can pentodes opperate with comparitively lower noise than triodes in preamp applications?[/quote]

As a general rule, no, since pentodes have an additional noise source called "partition noise". It's still possible to make quiet pentode-based preamps, though, especially with high-ratio transformers -- which have their own set of issues.

It's also possible to use pentodes in triode mode with good results. The EF86 and 5879 often get used this way.

As to 12AX7s...under certain circumstances they can be very quiet, although there are other tubes that, if properly operated, are even quieter. The distortion produced by 12AX7s depends very much on how it's biased and what its load is, but that's really true of any tube.

Finally, some folks strongly dislike the "sound" of high-mu triodes like the 12AX7, although I've never read a description of what exactly they don't like. (I put "sound" in quotes because a triode's sound can vary all over the place, as described above.) In general, medium- and low-mu triodes can be set up for lower and more benign distortion than high-mu triodes, but again there are plenty of exceptions.

Peace,
Paul

Companies got taxed by how many tubes were in thier products in Germany, so in order to reduce this surcharge, they invented a tube with twice the gain as the typical triodes.

Sometimes you will see a pentode wired like a triode in a mic pre.
I do not know what this does to noise.


The advantage of the 12AX7a is that there are so many brands and variations, that you have a lot of sonic choices to choose from, where as with say, a EF804S, there are very few choices from which to choose from.

There is a chart arounfd here with distortiion figures for varoius 12AX7's but I do not remember which thread.

Radiotron Designer's Handbook says that pentodes operated in triode have a noise advantage over regular triodes. This is one significant reason you see 6J7's in almost every pro preamp made before the common use of 9 pin types.

I think the Radiotron D. H. also talks about the 'unsuitability' of high-mu tubes for pro preamp use. This was the thinking at the time, at least.

The oddball American preamp is the RCA BA-1A from 1945-49 with two pentode connected 6J7's. Also the first RCA to use overall loop negative feedback.

You get wider use of 9 pin types, and 5879's and EF86's in pentode, once designers started using 20 db or more of loop negative feedback to counter the noise/distortion penalties of the average hi-mu, pentode, and 9-pin.

People today love their 12AX7's, thanks to Fender and Marshall, etc. I find a high gain tube to be a problem in a tube mic pre when recording loud sources, since you usually already have over 20 db of transformer gain. The higher mu just lowers the headroom of a stage that is usually being pushed with plenty of juice already.

I guess a fair comparison might be using two opamps at medium gain rather than one at a higher gain for better flexibility and gain range.


April fool's, CJ!

Use a chip, go to jail.
Thats the law.

Here is an interesting article that compares a number of 12a?7 triodes for distortion in two operating modes. The results are presented in figure 1 (a link at the bottom of the page).

http://www.tangible-technology.com/tubes/amplification/dual_triodes.htm

The big winners were the Mullard and Telefunken 12ax7s with partially bypassed cathode resistors (some negative feedback) and the 12ay7/6072as with somewhat less gain (avg. 3-6 db less) either bypassed or unbypassed.

This doesn't address noise performance but in two other articles

http://www.tubecad.com/articles_2002/RIAA_Preamps_Part_2/page2.html

and

http://members.aol.com/sbench102/dax7.html

the concept of using multiple parallel triodes is discussed with relation to noise and effects on odd-harmonic distortion (the bad kind) levels.

> why the Germans favored pentodes over triode design?

US studios used a lot of low-output ribbons. Germans used a lot of high-output condensers, in which the "first stage" is effectively inside the mike, and the board's "mic in" is more like a hot line input, much less gain than a US mic input.

I don't think triode-strapped pentodes offer any noise advantage over a good triode. But triode development stalled when pentodes appeared. The best tube designers and tube-material gurus were doing it to new types of pentodes, well into the 1950s. Only as VHF TV and UHF microwave links became common did they really wank on good triodes again. The TV tuner triodes are hard to beat for low voltage noise. But in conventional audio design, voltage noise is a non-issue because you can wind-up the input transformer to get high noise figure and still meet a 15KCPS frequency response spec. And a high noise resistance goes with low plate current, lessening the problems of supply filtering which can be severe.

The advantage of 12AX7 is that you can get one any place any time.

The "disadvantage" is that a high-gain tube with a reasonable supply voltage has a low input overload voltage, and probably a moderately high noise voltage due to low cathode current. Say 12AX7 with 250V supply, peak clean swing is maybe 50Vpk, 35Vrms. Max input voltage is around 0.5Vrms. If we use a 1:10 transformer, maximum mike input is 50mV. Some big condensers will make 50mV if you cough, and 200-1,000mV in musical settings.

What tube you use is NOT as important as how cleverly you use it. Tube types were usually dictated by "cost and profit": large operations were much happier if the whole studio used only 3 or 2 (or even just 1) tube type for everything. Shops like RCA had warehouses full of overstock tubes they needed to unload. The BBC remote console was entirely 6AU6, because you could get a replacement 6AU6 on a Sunday in Brighton or any town that had FM radio.

Well, dang it! I went looking for the reference I mentioned, and haven't found it yet. More specifically, I recall it had to do with the extra plate real estate (the other two tied grids) giving more surface area to capture stray electrons, and this contributing to lower long term noise.

RCA would certainly have had reason to push the 1620 and the 5879, since (I think) they came up with them both.

I did years of work in this area and had a lot of overload problems with the high mu type tube. I finally came up with a transformer input all tube pre that really is quiet and has good overload capability, but I'm sure the gurus would disapprove of such a radical departure from the "set" standard.

analag

[quote author="analag"]I did years of work in this area and had a lot of overload problems with the high mu type tube. I finally came up with a transformer input all tube pre that really is quiet and has good overload capability, but I'm sure the gurus would disapprove of such a radical departure from the "set" standard.

analag[/quote]

who pissed in your morning coffee, then?

The old warhorses of years gone by never used high mu tubes. Until they started using shiteloads of loop feedback. But I said that already....