12AX7 Tube Question

[duanestorey]

I'm considering building my own Hi-Fi tube amplifier. I'm actually an electrical engineer, but sadly enough most of my education was in the realm of digital, which is the norm nowadays. So my little pet project for the next 6 months is to attempt to learn all about tubes and Hi-Fi audio all while building an amp.

I've been wading through load lines and various tube datasheets. I have a really simple question regarding the use of 12AX7s as the pre-amp stage. I've seen a few designs where there are two 12AX7s in series (one stage after another), which ultimately drive a power tube. Looking at the datasheets for the 12AX7 though, I don't really understand why you'd want so much gain prior to the power stages. In fact, most of the 12AX7s seem to only work with a grid bias of around -1.5V or so, so I'm not even sure how you'd handle any significant voltage swing on the second stage if the first one amplified the signal very much.

Can anyone shed some light for me on why you'd want to do this? Thanks.

[blondamps]

Often it's a matter of gain V impedence matching. If you want to run the first stage to maximise input impedance and supply a good low inpedence source for the next stage, a LOT of feedback aroung the first (or first two) stages will achieve this. Valve power stages are actually quite capacitive and if you're setting out to deliver as much current to those grids as they need, the driving impedance (read plate resistor) can't be too large, otherwise the driven grid won't go UP as it should when the (12AX7) approaches cut-off. I hope this helps! Don't be afraid to experiment.... unlike junctions, valves are VERY tough! Good Luck and Have Fun

[kavermei]

It depends on the circuit... Sometimes the 2nd stage is there to recover the insertion loss of, for example, a tone stack. Further, in almost all cases some distortion will be desired in the preamp, and one gain stage alone does not create enough.

Kenneth

[kevinkr]

Originally Posted by kavermei

<snip> Further, in almost all cases some distortion will be desired in the preamp, and one gain stage alone does not create enough.

Kenneth

Usually the case with a guitar amplifier, not a hifi amp.

I recommend purchasing Morgan Jone's "Valve Amplifiers 3rd Edition" as it provides a lot of insight into amplifier design considerations.

[jrenkin]

I have been looking at the same issue. The advice I pulled together out of all the helpful comments and assistance through several threads is that all that gain is primarily to allow the use of gnfb, which improves disortion at the expense of gain to be really simplistic.
If you don't want gnfb, you need to be more compulsive in your amp design and component choices, but then only need to make sure that you have adequate gain to fully drive your output stage a little beyond clipping as you would rather the otput clip before the input stages.
I am making a PP EL34 with a single 12AT7 LPT and no gnfb. I get just enough gain in the LPT by using a CCS and so don't need to add a driving tube for it. If I did, I would need to add gnfb to bring it into line I suppose.
Haven't built it yet, so no idea if it is good or not, but the smart folks on this forum have not suggested that I am off track or anything.

Any one out there who knows better, correct me if needed.

[Miles Prower]

Originally Posted by duanestorey

Can anyone shed some light for me on why you'd want to do this? Thanks.

There are a couple of possibilities: you're making a preamp for a magnetic cartridge. These typically output not much more than 10mVp, and so require a lot of amplification to bring that up to a useful level.

The other possibility is that of an independent gain block. Here, you'd basically be designing a glass op-amp. You would need a lot of open loop gain in order to get that distortion down when you add the NFB to set the gain. It's the same idea as that of the solid state op-amp. (A lot of audiophoolz are going to reject that idea since they believe NFB is the Spawn of Satan. )

Originally Posted by jrenkin

I am making a PP EL34 with a single 12AT7 LPT and no gnfb. I get just enough gain in the LPT by using a CCS and so don't need to add a driving tube for it. If I did, I would need to add gnfb to bring it into line I suppose.
Haven't built it yet, so no idea if it is good or not, but the smart folks on this forum have not suggested that I am off track or anything.

I'd suggest ditching the 12AT7 and replacing it with two 6GK5s (singleton UHF triode). The gain will improve somewhat, and the 6GK5 is more linear than a 12AT7. This will be especially important since you said you're going to be running open loop. Being that this is a UHF triode, the internal capacitances are also a good deal lower, and that makes for a lower Cmiller.

[Eli Duttman]

Originally Posted by jrenkin

I have been looking at the same issue. The advice I pulled together out of all the helpful comments and assistance through several threads is that all that gain is primarily to allow the use of gnfb, which improves disortion at the expense of gain to be really simplistic.
If you don't want gnfb, you need to be more compulsive in your amp design and component choices, but then only need to make sure that you have adequate gain to fully drive your output stage a little beyond clipping as you would rather the otput clip before the input stages.
I am making a PP EL34 with a single 12AT7 LPT and no gnfb. I get just enough gain in the LPT by using a CCS and so don't need to add a driving tube for it. If I did, I would need to add gnfb to bring it into line I suppose.
Haven't built it yet, so no idea if it is good or not, but the smart folks on this forum have not suggested that I am off track or anything.

Any one out there who knows better, correct me if needed.

Miles' points are spot on. Another consideration is damping factor. You may be OK, if the EL34s are triode wired. Loop NFB of some kind is necessary, when UL and pentode mode are employed. Consider building a Poindexter style "Musical Machine". Poinz has already scoped out a 6GK5 LTP and triode wired EL34 setup. Going with a well proven design makes considerable sense to me.

[blondamps]

YAY to all above; it's worth mentioning too that although the gain remains primarily a function of the ratio between the Plate and Cathode loads, the 12AT7 is a relatively low gain device in voltage terms but CAN deliver (dissapates) higher (plate) current so it's more of a driver rather than a front end device. If you're not politically disinclined towards pentodes (in the pursuit of gain) there are rich pickings there too