F5 power amplifier

[AndrewT]

add a 40dB wideband low noise amplifier between the test equipment and the noise source.

There are a couple of threads on this Forum that describe this technique and they link to web sites that describe the amps in detail.

[Melon Head]

Thanks Andrew

I found this via google. http://tangentsoft.net/elec/lnmp/
I hope it does the job

[AndrewT]

I bought some OPA228 as the second best choice for ICs for that amp.
Not built it up yet, so can't report on it's performance.

[sangram]

At 1m distances, the 40dB could possibly be audible, but the noise floor in my room is 38dB. At listening position, we're talking 24/34dB, well below noise floor except in the dead of the night. And that's at 0dB PSRR, even 10dB of PSRR will knock this well down below the noise threshold of most rooms.

I still think they're not a problem. The F5 notes that

raw supply with two 29,000 uF capacitor and 70 mV ripple will give about 100 uV noise (measured in the band from 20 to 20,000 Hertz).

We're talking 1/35th part of 70mV, so I assume it is not a problem at all, and much quieter than a linear supply.

[jackinnj]

Originally Posted by thanh1973

Does anyone know how to do measurements down to microvolts?
I have the equipment, I just don't know if there are any special techniques required to do ultra low noise measurenements

Linear Technologies has a number of application notes on low noise measurements -- you can search through the archives. Texas Instruments also has an easy to implement design on the product descriptor for the TL431 shunt regulator.

Noise measurement is most relevant when you specify the bandwidth over which it is measured, and the filter topology of the measurement. TI builds a filter in analog for you while Linear uses their switched cap filters. Myself, I prefer CCIR-486 with a 2kHz peak -- but that's beyond the scope of this discussion.

Dennis Colin has an easy to use amplifier using JFET's -- i believe it is archived on the AudioXpress website -- he uses "A" weighting.

[Melon Head]

Originally Posted by sangram

We're talking 1/35th part of 70mV, so I assume it is not a problem at all, and much quieter than a linear supply.

Well in reality it might even be 1/100th part of 70mV.
2mV is the noise of the equipment not of the supply. I need to amplify the noise of the supply to be able to measure it.
At this point in time, I only know it is better than 2mV pk-pk.

[Melon Head]

Originally Posted by sangram

Maybe SMPS would be a feasible option for some other builders (I have already purchased my supplies).

If someone knows how to mod the voltage of switch mode power supplies then I will start a new thread.

[AlanElsdon]

Originally Posted by thanh1973

If someone knows how to mod the voltage of switch mode power supplies then I will start a new thread.

If you search the web I think you will find a lot of people making SMPS of differing voltage / power and some are adjustable within quite wide ranges. It would be better (and in the end possibly cheaper) to buy one fitting your requirements.

[Melon Head]

If you can tell me where I can get a 850Watt supply for around $150 with +/-24V rails and with the performance of a computer supply let me know.

[sangram]

Not the right rails, but it's a start:

Connexelectronic

Hypex has a few of them as well. both the creators are active on DIYA, a chat would help figure out how the rails can be modded.