Low noise BJT.
A quick goo.... and saw this!😀
https://www.beis.de/Elektronik/LNPreAmp/5xTransistorNoise.gif
A quick goo.... and saw this!😀
https://www.beis.de/Elektronik/LNPreAmp/5xTransistorNoise.gif
Same basic circuit as for the standard denoizator, but with the resistors value around the transistor amplifier adapted as a function of the (high) output voltage.
Some caveats though:
As the input is bootstrapped to the output, it will change the loop characteristics, which could be problematic: the denoizator is supposed to be used with plain-vanilla, non-augmented circuits.
You are free to try it of course, but it might not be possible to stabilize it.
Another problem is the dissipation: to be effective, the amplifier needs to run at a few mA, typically 5.
As you will calculate the amplifier to keep the Vce under 10V, the collector load will dissipate a non-negligible amount of power: >1W for 250V for example.
The capacitors will need to withstand ~all of the output voltage: imagine the size of a 220µF/250V cap.
In principle, the nonoiser can use smaller caps, but I am practically certain that it will be impossible to stabilize with a Maida
THANKS, Elvee!! I can't wait to try it! How much reduction in noise voltage would you expect? RC4194 datasheet quotes 250µVolts noise for 15 volts out.
If it actually works, you can expect 40db or more, because the impedances are more favorable than in a 317.
But don't get too enthustiastic for now: you might never be able to get it stable
But don't get too enthustiastic for now: you might never be able to get it stable
I am considering adding the DeNoiser add-on to my pre-amplifier power section.
It has multiple LM317/337 combo's with one high voltage with +/- 52 Vdc:
(from the service manual)
The BC337 is limited to 45 Vdc so what alternative higher voltage rating transistor would be advised and do I need to change some resistors R4/R5/R6 to other values too ?
Does the control benefit using a higer value reference resistor R1 (220 vs 115 ohm) which would set R2 as 9 kohm to get 52V.
It has multiple LM317/337 combo's with one high voltage with +/- 52 Vdc:
(from the service manual)
The BC337 is limited to 45 Vdc so what alternative higher voltage rating transistor would be advised and do I need to change some resistors R4/R5/R6 to other values too ?
Does the control benefit using a higer value reference resistor R1 (220 vs 115 ohm) which would set R2 as 9 kohm to get 52V.
On power up, the regulators see a 59V transient between Vin and ADJ.
I would not bet that they tolerate that forever.
regards, Gerhard
I would not bet that they tolerate that forever.
regards, Gerhard
It is a NAD C162 pre-amp ... working fine for 10+ years now ...
Between the rectifier and the LM3x7 is also a transitor in both lines for switchting off power in standby mode.
Never checked if they charge the regulaters slowly during power up ...
Between the rectifier and the LM3x7 is also a transitor in both lines for switchting off power in standby mode.
Never checked if they charge the regulaters slowly during power up ...
Last edited:
Enthusiasm effectively curbed. I think there are mistakes in your adaptation to the RC4194. The positive side seems all wrong to me---using a BC327 (same as the negative side) and feeding the + output signal to the emitter seems wrong---it's fed (via a 22uF DC blocker) to the base of a BC337 in the original LM317 version. What's happening here?
I think my corrected version is right: the 4194 has a very peculiar topology, with the + and - sides completely different
Looking a little further I realised the ZTX851 is rated for 60V.
I think I will install LTspice and start learning how to use it and see what 52V out will give me 😀
I think I will install LTspice and start learning how to use it and see what 52V out will give me 😀
The amplifier transistor doesn't need to be rated for the full input or output voltage: 15V is more than sufficient, if the resistor values are suited to the output voltage (keeping in mind all the caveats about using the denoiser with a HV supply, see above)
@Sadface
Big thank you for your work on the board layout, Gerbers, BoM, et al. I definitely have use for at least 3 PCBs for older RIAA headamps I want to get on the go again, plus a preamp. I'll have to wait to get them made though - lock-down in place.
Kevin
Big thank you for your work on the board layout, Gerbers, BoM, et al. I definitely have use for at least 3 PCBs for older RIAA headamps I want to get on the go again, plus a preamp. I'll have to wait to get them made though - lock-down in place.
Kevin
Keeping Q1 at 10Vceo and max 10 mA means R4 should be 42/0.01= 4200 ohm -> 4k7 (0,5 W)
R5 = 10-0,6/base current -> with hfe approx. 400 —> base current is 25 uA
9,4/25u= 376 k > 330 k ?
So far I understand transistors ... i think ;-)
??
R5 = 10-0,6/base current -> with hfe approx. 400 —> base current is 25 uA
9,4/25u= 376 k > 330 k ?
So far I understand transistors ... i think ;-)
??
G'day Guys,
If one was trying to maximise the efficacy of the CRC filter before the denoiser reg, then how low would one want to set the corner frequency?
I intend to use an LM317 denoiser for a tube preamp heater supply.
2x 6SN7 in series for 12vdc at max 650mA.
With this sort of current draw, using a 10R resistor for the RC filter results in unacceptably high voltage drops of something in the order of 6v which would not leave me with enough voltage drop left for sufficient regulation.
Thus, I am trying to balance capacitor size to resistor value to come up with something more like a 1v drop across the resistor. This seems to end up somewhere in the margin of 1.5R but now the capacitor size is going up which means cost and pcb real estate.
Values such as: 1.5R with 4700uF yield a corner of 22.5Hz. Is this low enough?
Do I need to target more like the 7.2Hz yielded by 10R with 2200uF?
If one was trying to maximise the efficacy of the CRC filter before the denoiser reg, then how low would one want to set the corner frequency?
I intend to use an LM317 denoiser for a tube preamp heater supply.
2x 6SN7 in series for 12vdc at max 650mA.
With this sort of current draw, using a 10R resistor for the RC filter results in unacceptably high voltage drops of something in the order of 6v which would not leave me with enough voltage drop left for sufficient regulation.
Thus, I am trying to balance capacitor size to resistor value to come up with something more like a 1v drop across the resistor. This seems to end up somewhere in the margin of 1.5R but now the capacitor size is going up which means cost and pcb real estate.
Values such as: 1.5R with 4700uF yield a corner of 22.5Hz. Is this low enough?
Do I need to target more like the 7.2Hz yielded by 10R with 2200uF?
With a filter having 6dB/octave, you need a corner frequency as low as practicable, meaning either a large cap, or a large resistive drop or a mix of both.
Not very attractive, and a waste IMHO for a heater, even in preamp applications
Not very attractive, and a waste IMHO for a heater, even in preamp applications
