This resistors don't affect gain. Tey affect measurements by DVM or oscilloscope.
Try 47K, or even 10K.
Try 47K, or even 10K.
Unless dropped too low these resistors don't set gain, but they do set the output LF point. Reducing R increases the output LF point, which reduces the amount of supply rail subsonics coming through to the output. In some cases, of course, the input resistance of whatever comes next will determine the LF point.
I think, hrmmm....floating the filaments with the 200K 47K resistors is causing the problem. when he lifts the input with the volume control....buzzzzzz. Try grounding the filaments more by jumpering the 47k resistor.
NO! Actually there is a 10uF cap at that node expressly to provide a low impedance to ground - if you short that resistor to ground the filament to cathode insulation in your 6SN7 may fail. (Check the cathode voltage on the upper tube and if it is near or above 100V this is ill-advised.) Further this is unlikely to have anything to do with your issue, and necessary practice with single envelope dual triode based SRPP and mu-followers.
Please take a look at your plate voltage and compare it to the wander you see on the output of your line stage. Reducing the size of the output coupling cap so that it is -3dB at roughly 10Hz ought to be adequate protection. Consider adding a Maida style voltage regulator for the B+ if you are that concerned - having verified that your B+ is wandering of course.
Heater negative with respect to cathode is 200v for 6SN7. Measure your cathode voltages. My guess is you will be more than fine.
True from a data sheet perspective, (at least RCA HB3, and Sylvania 6SN7GT data sheets) direct experience with Sylvania and GE 6SN7 over 20yrs indicates that problems can start with as low as a 100V - 150V difference. I'll admit I've not had a very high failure rate, just a few here and there, but enough have failed to leave me questioning that rating. Modern production tubes IMLE are much worse.
And tungsol, however I bow to your experience.
Whenever I see AC filaments.....its the AC filaments.
Whenever I see AC filaments.....its the AC filaments.
Actually often my experience too. It did not initially register that the filament supply is not DC.. Easy enough to fix. 😀
Actually, the filament circuit seems easy enough for me just to attach a 6v dc lead acid battery to the AC 6-0 terminal on the pcb! This could reduce the noise floor quite easily!
I also have a simple filament regulator:
http://www.ebay.com.au/itm/220951565373?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
But the 6.3vac winding on the transformer I am using for the 6SN7 SRPP obviously doesn't allow me to get enough volts out of the regulator circuit - is there any way around this?
I also have a simple filament regulator:
http://www.ebay.com.au/itm/220951565373?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
But the 6.3vac winding on the transformer I am using for the 6SN7 SRPP obviously doesn't allow me to get enough volts out of the regulator circuit - is there any way around this?
Last edited:
Yes. Low drop-off regulator, for example implemented with MOSFET as a pass element, and Shottky rectifier.
Look whats new from Linear Tec.
LT3022 - 1A, 0.9V to 10V, Very Low Dropout Linear Regulator - Linear Technology 0.145mV drop. <= Ultra low drop voltage regulator....for all your 6.3 volt problems.
It might not be that new as I have not looked for awhile. You should be able to easily get 6.3v regulated out of a 6.3v AC tap with this.
LT3022 - 1A, 0.9V to 10V, Very Low Dropout Linear Regulator - Linear Technology 0.145mV drop. <= Ultra low drop voltage regulator....for all your 6.3 volt problems.
It might not be that new as I have not looked for awhile. You should be able to easily get 6.3v regulated out of a 6.3v AC tap with this.
Good part, but using 1 MOSFET, 1 BJT, couple of resistors and one Zener I can get much more than 1A. Well, one more Zener is needed to protect gate, since I provide 1 mA current from B+ rectifier through resistor.
Update on this - I changed the coupling cap from 1uF to 0.1uF and the bypass caps from 220uF to 470uF, as suggested by DF96. I don't know why this has worked- but the background noise has completely disappeared! No discernable loss of low end, either!
Increasing the cathode bypass caps will have had the extra benefit of reducing the frequency at which the anode impedance rises so improving LF PSRR.
What happens if the preamp is connected to, say, a solid amp with a 10k input impedance?
This is how I have always tested tube pre-stages: with a solid state TA2024 amp and some terrible speakers from an old sound surround system. At least then it magnifies and hissing/humming noises and assists (in the absence of a scope) of testing what changes in sound arise from changes in the circuit.
I let the unit run for about three hours last night and can confirm the noise levels are now really low. There appears to be some junk on the tweeters - should I install some RF filtering or LF bypass caps? I recall when building my DAC I soldered some 6.8nF caps at the output using these calculations:
1/(2 * 3.14 * 1000 * 0.0000000068) = 23.4 KHz
So the value was 6.8nF (0.0068uF)
Is there a similar low pass method used for line stages?
I let the unit run for about three hours last night and can confirm the noise levels are now really low. There appears to be some junk on the tweeters - should I install some RF filtering or LF bypass caps? I recall when building my DAC I soldered some 6.8nF caps at the output using these calculations:
1/(2 * 3.14 * 1000 * 0.0000000068) = 23.4 KHz
So the value was 6.8nF (0.0068uF)
Is there a similar low pass method used for line stages?
If the 10k was maintained down to DC then the 4.7uF output cap would give an LF cutoff of 3.4Hz (ignoring the SRPP output impedance). OK, provided that the 4.7uF is not an electrolytic.directdriver said:
You could put an RF filter at the input if interference is a problem. Most people would regard 23kHz as a bit low for a low pass filter, except that a DAC has special requirements.lordearl said: