Soldering SMD doesn't have to be out of your reach. I will admit to having been pretty daunted by the idea, but with a bit of self-confidence and no prior experience I successfully assembled neutrality's relay attenuator boards (see elsewhere on this forum) which are 100% SMD except for cable connectors, and use some pretty tiny ICs. I hand soldered everything with a £20 soldering iron with basic temperature control and an £8 set of bits and it wasn't too bad, the main thing to get used to is using very fine solder wire and generally using a separate bottle of flux to pre-apply the flux, as the very fine solder wire is too fine to have the flux built in. And a cheap set of optivisors from eBay helped enormously.
If the issue is unsteady hands, there are DIY possibilities to convert tabletop electric ovens into reasonable reflow ovens: I haven't tried this myself as the hand soldering approach has worked for me so far, but I wouldn't say it was out of reach.
Because the last thing you want in your measurement setup is a nice big ground loop.
Try, say, taking a loopback measurement of an amplifier. Soundcard output ground runs into amplifier, amplifier output ground goes to soundcard input ground, soundcard input ground connects to soundcard output ground. If that's not a ground loop I don't know what is.
Or let's say you want to measure some device that uses a floating (IEC Class II) SMPS as a power supply, with a computer tied to protective earth. Any SMPS mains filter leakage current will be making its way to earth via your signal ground once connected, and unbalanced cable shield resistance tends to be non-zero, so there's a finite amount of mains + filtered-out crap remaining that will appear together with real signals.
Being tied to ground really limits the utility of measurement gear. Not sure why differential probes aren't any more common than they are, really. They tend to be $$$ when bought as scope accessories, but building one seems to be doable on a rather modest budget.
People still like the old EMU 0404 USB, which had excellent converters for its day and actually still beats the very latest Focusrite Scarlett 2i2 3rd gen (which has improved over previous efforts and now is around 110 dB of SNR in A/D) that would also be within budget. The drivers apparently require a decently-sized vocabulary of expletives though, as well as an operating system no newer than Windows 7 (with the latest beta drivers, otherwise perhaps even XP is recommended). Having wrestled with Asus Xonar drivers (hint: prefer no later than build 1700 for D1 and D2), I know how that feels.
If you just want to measure output noise floor, most anything with a half-decent microphone preamp would provide sufficient performance. On a Behringer Q1002USB mixer, I found an equivalent input voltage noise level of ca. -126 dBu (20 kHz BW), that's ~ 0.4 µV or ~2.7 nV/√(Hz). A Mackie 402 VLZ4 dropped that by another 5-5.5 dB, so ~1.5 nV/√(Hz), with excellent distortion figures right up to clipping even at maximum gain. It'll be hard to find a preamp output that isn't at least 10 dB noisier - and 20 dB more often than not.
The greatest challenge with such a setup may be calibrating the levels accurately. Measuring 0 dBFS output and making an accurate -40 to -60 dB balanced attenuator may be useful.
Last edited:
Hi thanks a lot for the very kind and interesting advice ... but i really do not want to complicate the all thing.
My next step is simulation of very basic circuit to learn something
Then a simple prototype and then a board ... this is the plan
So i am not ready to solder now. It will take some time.
But i will keep your advice in my mind.
Thanks again.
Hi ! ok i see.
yes ... but if the amp has unbalanced inputs ... how i can connect xlr from the sound card to the rca in the amp and viceversa at the output ?
My amps have all rca inputs ...
Moreover ... But am i wrong when i say that scope probes are unbalanced ? and are not they use also for noise measurement ?
Am i missing something ?
i really do not know why ... still unbalanced probes are used.
But as the final aim is to use a sound card i need to know how to connect the xlr ins and outs of the card to the unbalanced unit to measure.
I am using only unbalanced amps and preamps
Actually i am much more interested in preamps to be honest ...
yes i have read about the Emu that is not supported any more from the house. So it is out of question ...
yes ! just noise for now ... distortion later ... i understand that a very clean signal generator is required for measuring low levels of THD ... so it is a more challenging task
I see. I understand i have to study more the topic.
Thanks again for the very kind and helpful support. I am a little hard to understan things ... 😱
Kind regards, gino 🙂
Last edited: