John Byrns wrote:
In article ,
Ian Bell wrote:
Eeyore wrote:
Ian Bell wrote:
On several preamp PSUs I have built in that last couple of years I have
noticed the same thing. A very low frequency low level variation in the
output voltage. My latest PSU uses a series of five 100uF caps with 1K
resistors in between and it still exhibits this problem. The 'noise'
seems to be below 1Hz in frequency, fairly random and peaks typically
between +- 10mV with occasional excursions to +20 to 30mV.
Well what do you expect without regulation ?
Unhelpful replies like yours I guess.
His reply may have been unhelpful, but it was certainly to the point. I don't
get it, if you don't want to use regulation how can you expect to avoid low
frequency voltage fluctuations? You should probably count yourself as lucky
that you are seeing only "occasional excursions to +20 to 30mV", I would expect
several volts, or more, without regulation.
--
Regards,
John Byrns
Considerable self regulation occurs in mains supplies in populos cities
and towns because where you have thousands of people all using something
and averaging say 1kW mains input per household 24/7, then if the mains
voltage is 240V, the current average is 4 amps and the load must be 60
ohms. Say there are 3,000 housholds sharing your circuit locally and
that the wiring between them is low resistance
then the load connected to the circuit you are on is 60 / 3,000, ie, a
very low load indeed on average, but one that changes its value
enough to cause Ian's observed LF voltages on a B+ rail of +/- 30mV.
That's about what I see here.
After several R&C stages of filtering the rail noise movements are much
less than +/-30mV.
If you have a CCS type of load supplying dc to a triode anode, then the
low Ra of ther triode makes a divider with the CCS, or with a resistance
load for dc feed, and the anode LF noise voltage will be much lower than
at the rail.
But where you have a pentode or cascode pair of triodes then Ra is high,
and any noise on the rail appears at the anode. The other source of LF
noise in phono and line level stages is where you have a large grid bias
R to bias a typical gain triode.
Its grid may be cap coupled to a preceding triode gain stage with low
Ra. At say 1kHz, the low Ra and low XC shunt the high resistance of the
grid biasing R, say 470k or 1M. But as F becomes lower, the XC becomes a
high impedance, and effectively the grid is then coupled toa good noise
source, ie, the high resistance. Its most important in phono or
microphone first stages to have a relatively low grid biasing R,
and 47k is typical for MM inputs. The carts are usually direct coupled
to a grid, and the standard 47k is so the grid is biased when no cart in
connected and the 47 happens to be the most common recommended loading
for most MM carts. If you have C&R coupling from a cart to grid, then
the bias R will make LF noise which is not shunted by lower Z of the
cart because of the rising XC of the coupling C. I have had to use about
9uF from cart to 47k to avoid unwanted LF noise in phono amps. Phono
amps have 10 times more gain at LF than at 1kHz, so the slightest LF
noise gets through. Its even worse with MC carts which may have 26dB
less sensitivity than an MM cart. If there is R&C coupling
from an MC cart to j-fet input, the best LF noise performance occurs
when the biasing R for the j-fet gate is say 470 ohms to load the cart
correctly and to keep LF noise low. MC carts have tyical output
reistance of less than 30 ohms, hence their noise overall is very low,
so an SNR of over 60dB is achieved despite their low signal voltage
level, but with C&R input coupling, C and R values have to be chosen
carefully.
See samples of phono amp circuits which have good LF noise performance
at
http://www.turneraudio.com.au/preamp...hono-2005.html
http://www.turneraudio.com.au/preamp...ated-2006.html
Patrick Turner.