This post will most likely stand the hair up on the back of EE necks because it’ll be riddled with inaccuracies I will use to make a point about why devices sound differently in audio circuits.
What I have to share hopefully will be interesting to both the uninformed as well as those knowledgable in everything electronic except how to make something sound good. There are many fine electrical engineers that can make excellent amplification devices but they won’t necessarily be high-end and they’ll not likely produce musically satisfying results.
Remember in our earlier discussions I mentioned there are two types of transistors: the standard Bipolar Junction transistor (BJT) and the Junction Field Effect Transistor (JFET). The former acts like everything we associate with a transistor and the latter more like a tube. It is to this difference that I want to address today’s post.
In a typical BJT used in 99.9% of all solid state amplification devices such as preamps, power amps and DACS, there is a physical connection between our three nodes – the input, output and power connection. A change in one will be reflected in the other three – which means that our phono cartridge we’re using as a source is not isolated from changes we make in our amplifying device – something not good for sonics. Again, just to keep the feathers from being ruffled too much – yes there are tricks and methods to reduce these effects – but natively and without these modifications what I have mentioned is true.
In a tube or JFET this is not the case. The input is basically immune from any changes to the other two nodes in this type of device and the physical connection does not exist – instead a field is generated that impedes the flow of voltage in the device. We refer to these types of devices as voltage amplifiers and the BJT as a current amplifier. Voltage amplifiers are always on and the input merely turns them off – referred to as “pinching” the voltage – where current amplifiers are the opposite and considered more like a pump.
In fact, a good way to view these two types of effects would use a water analogy. In a BJT current device picture a hand pump for water. If you want the water to flow, start moving the handle of the pump up and down and you force water to flow. In a voltage device, picture the water already flowing in a hose and then pinch or squeeze the hose to restrict that flow.
What’s the big deal between these two types of devices? Simply put when used as an input to a source like a phono cartridge, CD player, or even a power amplifier the current amplifiers don’t handle micro dynamics and low level details anywhere near as well as the voltage amplifying devices do – because they don’t naturally respond to these low level signals having a certain amount of current required to get them moving at all.
I know, I know, there are many fine examples of this not being true, ways of biasing that overcome these obstacles and tons of technical reasons to argue the point – but in case after case tube and transistor input circuits almost always outperform their current amplification counterparts in these two critical areas of music. I have numerous examples of our designs at PS over the years where we simply replaced the input bipolar transistor with a JFET and a whole new world of musicality opened up. The hard edge of the upper harmonics and low level detail associated for so long with transistors is gone and in its place a musicality from both the lowest level details to the highest is achieved that is rare amongst amplification devices.
What we can conclude from all this is that when it comes to tubes vs. solid state used as a connection between a source and the internals of the amplifier, designers are much better off to place a voltage amplifying device on the input (tube or JFET) rather than a current amplifying device. The lack of a physical connection seems rather important to achieving the musicality anyone reading these posts would most likely want to hear.
Tomorrow I was going to cover transconductance (how a voltage amp works) vs. Gain (how a bipolar transistor works) but some of you have written asking me to focus a bit more on the sonic differences between the two and focus more on how they are implemented.