I'm currently working on a design where I use a multiplier chip with differential inputs and output. Now, in my application all external connections are single-ended, and some app notes (e.g. the one for the ADL5391 chip) talk about using baluns for single-ended to differential conversion (and vice versa). The data sheet for the AD834 also mentions using a balun as an option, albeit only on the output side.

Now, being all new to RF design (and EE not being my trade), I have a couple of questions:

  • What are the principal advantages of using a balun for this conversion instead of going the simple way of just tying one side to ground? (I have some vague ideas here, but a good reference for the basics would be very much appreciated.)

  • How relevant are these for my application in the low hundreds of megahertz?

  • Which side of the balun is to be preferred for the impedance matching/AC coupling stuff? The example circuits in the above data sheets differ in this respect.

Sorry for the broadness of the question, but I'm still struggling to wrap my head around the issues involved.

  • 1
    \$\begingroup\$ I think baluns will improve distortion figures and DC offset issues with the device. Usually, if the input signals are small, signal/noise ratios are improved. As for "which side of the balun", it's bidirectional as far as I know - unbal_in to bal out or vice versa. \$\endgroup\$
    – Andy aka
    Sep 21, 2013 at 0:22

1 Answer 1


I don't think not using a balun and just connecting one side to ground is an option. As your signal is defined differentially, V = (V+)-(V-) having a DC component at one side and 0V at the other side will utterly destroy your intended calculation. You do need a single ended to differential converter, and mainly the two methods of achieving that are using precision amplifiers for low frequency applications, and baluns for high frequency applications. Low hundreds of megahertz is quite a high frequency, for which you'll almost certainly need a balun to operate this multiplier.

Also, I'd strongly advise you to consult an experienced EE for this design. Designing anything in the hundreds MHz range requires experience in board layout and signaling, and while recent standards operating at the GHz range may make a 100~200MHz design seem trivial, it certainly is not.

  • \$\begingroup\$ Thanks for your quick answer. Yes, I'll certainly run into issues, and I am also going to consult an experienced engineer, but right now I'm trying to get as far as I can on my own and pick up parts of the needed theory on the way (ah, the benefits of youth). I know that 100 MHz is quite a high frequency, but if you look at the data sheet for the AD834, they actually seem to use single-ended inputs without further conversion in some of their examples. There would be no DC component on the input signal (AC coupled, referenced to ground). \$\endgroup\$
    – aule
    Sep 21, 2013 at 12:35

Not the answer you're looking for? Browse other questions tagged or ask your own question.