# In LTspice can I convert the units of an FFT which by default seem to be shown in dBV to dBµV?

In LTspice can I convert the units of an FFT which by default seem to be shown in dBV to dBµV?

Cause of reason: The reason for this is that I've a modeled a LISN in LTspice where I want to simulate what is seen on the output, hence in real life what the Spectrum analyzer would see.

My first approach was to use FFT of waveform, then look at it in Linear mode so I could read out the voltage at one specific emission top then apply 20*log10(V)+120 then i got 86.6 dBµV, when applying 20*log10(Trace in LTspice)+120 i got 87.2 dBµV at the same top.

It can also be seen that I get a lot more noise where there shouldn't be any when applying LTspice math. My LISN where MP1 and MP2 is the measuring points • It seems a simple case of a multiplication in the dB scale: x10=20dB, x100=40dB, etc, so all you need is to find the number that converts the uV to V, 1e6. What you did is to actually add the value of 120 to the mV valued signal. The traces are limited to a certain number of points, so performing math will affect precision. The heavier the math, the more noise. That's why there's no need to work in the linear scale and apply logarithms, when there already is a dB scale, which implies simpler math. – a concerned citizen Oct 3 '18 at 16:27
• For LISN CE noise you want dBuV into a 50 Ohm load ,.to compare with spec. Logs are exponents – Tony Stewart EE75 Oct 4 '18 at 16:24
• @TonyEErocketscientist See my added picture of the LISN model and where I probe at MP1 and MP2. If I now have not missed something. – Anton Ingemarson Oct 5 '18 at 6:39
• FFT noise looks normal without a pre-amp and 30kHz noise with narrow dutycycle, I presume is your PWM SMPS? Do you need more filtering? – Tony Stewart EE75 Oct 5 '18 at 6:55
• Look at the following link from Wurth Elektronik engineers https://www.youtube.com/watch?v=u-WXTROc3lE Hope this helps – Majid Fard Mar 5 at 19:32