LME49710 Stereo Preamplifier

This is a project that started many years ago as a headphone amplifier. It was designed around two NE5534 opamps (which were considered reasonable at the time) and as I remember, it never got much use as it was one of those projects that I thought I needed but really didn't. Anyway, I recently found it again (in my parents loft!) and decided to put it into service, this time as a preamplifier to go between my PC and power amplifier

                

I changed a couple of resistor values slightly but other than that it remains pretty much the same. Now though, the old 5534's have been replaced with more modern and superior LME49710 opamps which are pin for pin compatible. I came upon these after upgrading my PC with an ASUS sound card that had a plugable output op amp. I swapped the original NE5532 for an LM4562 dual device, and being impressed with it's quality, decided to search for the single version. The LME49710 was the result of that search

The preamplifier circuit is standard text book stuff and completely DC coupled from input to output. To quote Peter Walker of Quad, an amplifier should be a piece of wire with gain, and I reasoned anyway that most sources will have an output capacitor and most power amplifiers (including mine) will have an input capacitor, so no problem. Volume control potentiometers can produce noise when there is DC present on their wipers, but as there is no DC component at the input or output, this too is not an issue. So we're good to go with our amplifying wire!

In the diagram above I show the volume control as being a 10K linear pot rather than a logarithmic type which are more usually used for audio. The reason for this is that it was the only pot I had available during testing, but actually, I found it to behave exactly the way I would want a volume control to work, very smooth and not at all 'odd', so I left it in. Also, dual linear pots are supposed to track more accurately than logarithmic ones, so that could be an advantage. If you're horrified by the thought of a linear volume control then use a logarithmic pot. I probably could have used a better quality pot as well, but this would have cost more than the rest of the preamp put together!

               

The board was built using a method us radio hams used to call 'Scruffy Construction'. I think the photo is self explanatory. This does have some advantages though, one being that it's very easy to design the layout, with most of the board consisting of the earth plane. The finished result can look pretty good, but it does take skill with a soldering iron. Basically, with the copper side facing up, the circuit is drawn onto the board and 1mm holes are drilled (that's the smallest drill bit B&Q sold at the time!) where you want the component legs to go through. Where there is an earth connection, the leg can just be soldered directly to the copper. After drilling, the copper side can be spayed with clear lacquer to prevent it from tarnishing. Just scrape the lacquer away from around the holes when it's time to solder


When I started doing initial tests on this circuit with the NE5534 opamps installed, I found that it became unstable when being fed by a long input cable (5 metres) from my computer located in another room. This occurred when the volume control was advanced beyond half way, causing a weired popping sound. With the LME49710 op amps fitted, it's completely stable. I came across an article which may have the explanation for this, stating that some opamps are unstable at, or near to, unity gain, so my thinking is that as this preamp operates at a gain of just 4.7 (Rf/Ra) maybe this is low enough to cause problems under certain conditions. Checking the NE5534 data sheet confirms that it becomes unstable at gains of 3 or below, unless external compensation is applied. All this is of little importance here as the LME49710 is unity gain stable and behaves extremely well under all conditions in this circuit

The position of the mains transformer is important if located near to the electronics. At first I mounted the transformer in the normal upright position. This caused a slight hum to be induced into the audio and was not acceptable. After trial and error I found the best position was to mount it horizontally (see photo above). This completely eliminates any hum, even when listening on headphones. These small laminated transformers are not really the best choice here. As mentioned, they do radiate, but they can also get quite hot and depending on how they are mounted, can buzz mechanically. Input/output cabling must be kept away from the transformer too. I used this enclosure because I aquired it free (and nice enclosures have to have something built in them right?), but in hindsight I should have followed good hi-fi practice and used the more expensive option of a toroidal transformer mounted in a separate box


I started out by stating that this was originally a headphone amplifier project, and although it now performs the task of a preamp, the output is fed through a 1/4 inch jack socket via its switched contacts, so it still maintains its function as a high quality headphone amp. The LME49710 op amps seem to be well up to the job, driving my low impedance Sennheisers without any problem. Overall I am very happy with the sound quality, bearing in mind you can throw vast amounts of money at these simple designs to milk them for everything they've got. Know how much you wish to spend and stick to it!