Rectifying the imbalance in a filament secondary would appear to be an easy hurdle, snip the centre tap and create a new virtual centre reference with a couple resistors matched to any arbitrary precison. When floated the filament balance should (that word again) be perfect to the limits of the transformer's parasitics. Still, according to any references I've seen none of this should matter with DC-biased filaments. Biasing supposedly repels the emission which generates hum.
I had an issue with bass response in my pre amp, caused by an 40V elevated heater supply.
The heaters receive 12,6Vac. My fix is here:
An externally hosted image should be here but it was not working when we last tested it.
Measuring the voltages over the 10K resistor DC= 0,0V but AC is 2,9V.
How safe is this fix w.r.t. reliability of parts used and the other circuitry?
Last edited:
i would put the top resistor 120k to the last cap as ripple is lowest there....
if you are using dc coupling that circuit is a must, resistor values can be higher
as you are not really draining current that much...
if you are using dc coupling that circuit is a must, resistor values can be higher
as you are not really draining current that much...
Sorry, perhaps some explication is necessary first.
When I connected my freshly build pre amplifier and compared reproduction to the original audio, some bass notes were louder than they should be, especially with high impulse tones in the frequency of circa 70Hz. Certain toms and the hammer on the base drum were affected on some recordings. So I dug out the tone generator and the AC millivolt meter and tested the amplifier for abnormal behavior but could not find more than an 1 mV discrepancy in the sub 100Hz region, down to 10Hz. As the level was 400mV this could not be of importance.
There had to be something resonating at the frequency of interest. While I possess a hard copy of RDH4 I have no stomack to study the principle of networks. So, this simple filter came up, consisting of three components.
When I connected my freshly build pre amplifier and compared reproduction to the original audio, some bass notes were louder than they should be, especially with high impulse tones in the frequency of circa 70Hz. Certain toms and the hammer on the base drum were affected on some recordings. So I dug out the tone generator and the AC millivolt meter and tested the amplifier for abnormal behavior but could not find more than an 1 mV discrepancy in the sub 100Hz region, down to 10Hz. As the level was 400mV this could not be of importance.
There had to be something resonating at the frequency of interest. While I possess a hard copy of RDH4 I have no stomack to study the principle of networks. So, this simple filter came up, consisting of three components.
Last edited:
how did you arrive at that? did you do a
frequency scan to see response curves?
or did you just play it by ear?
frequency scan to see response curves?
or did you just play it by ear?
easy for you to say....😉
once i replaced a 3 terminal reg with a resistor sized to give correct voltage, that took care of the reliability issue...
For those in need of some aid, this is what I came across on the web.
A capacitor offers internal resistance to AC current, called the Equivalent Series Resistance (ESR). At lower frequencies, this is mainly the resistance of the dielectric. Some dielectric resistivity is inversely proportional to temperature, that's why their ESR at high frequencies decreases as temperature increases.
The power dissipated in the capacitor results in an elevation of temperature. The allowable temperature rise of a capacitor due to power dissipation is determined by experience. This in turn limits the power the capacitor can dissipate. When AC current is applied to a capacitor, the resistance (ESR) that opposes the flow of current results in heat generation: P = I x I x ESR.
The power a capacitor can dissipate is also limited by the applied DC voltage. The operating voltage should not be allowed to rise above the rated voltage (nor should it drop below zero, since the capacitor is a polarized component). Assuming the capacitor is biased at half the rated voltage, which is the optimum use condition, the limiting value of the voltage is, for a sinusoidal waveform: Vrms = Vpp/2root2 = Rv/2,8 (RV Rated Voltage). So, a 16 volts rated cap should ideally not be used above 5,7Vrms.
To find the limiting current Irms, we divide Vrms by the impedance at the desired frequency: Irms = Vrms/Z and Z=1/(2pifC) Simplified for 100Hz ripple: Irms = 0.23 x CV where Irms is the maximum permissible rms current in milliamperes, C the capacitance minus the capacitance tolerance in microfarads and V the rated voltage in volts. All above calculations assume the capacitor is properly biased at half the rated voltage. If this is not the case, Vrms becomes Vp/root2 where Vp=Vrated-Vbias or Vbias, whichever is lower.
Wherever you got this "information," you can profitably avoid visiting again. Nearly everything there is incorrect.
any cap other than tantalum will do....
the object of your circuit below is to bias the heaters up
so as not to exceed tube heater to cathode voltage specs...
you can change your resistors to 1.2M instead of 120k,
and 200k instead of 20k. wrt to the cap, i used 100 volt types
if the bias is 60volts or so.....this is a safe thing to do...
On one of my first SS amp designs I had lots of hum.
Hum from the power supply was getting into the driver and LTP stages.
So I decoupled the supply with 100uf and 39R on each rail.
Worked a treat.
Wow, that's remarkable. It truly is incorrect. The equivalent circuit of a cap is closely approximated by a capacitance (reactance of 1/2pifC) in series with the ESR (a resistance). They add in quadrature to get the AC impedance. The ESR is a very low number, generally well under 1 ohm and represents the asymptote of the impedance vs frequency curve, that is, instead of trending to zero impedance with increased frequency, the impedance trends to the few milliohms of ESR.
The other comments are relevant to high ripple applications, which this very much isn't. For your application, the ESR is entirely negligible.
Wrong, the purpose of this heater bias is to eliminate disturbance of the amplifier frequency response due to interaction of HV ripple with the cathodes. That's why I shared this information with the group, to see if any have experienced the same.
Can you pls underbuild your answers with some reasoning for the sake of the quality of this forum? I'm not a big fan of your work, no offense intended.
i am not here to build any fan base, i only build amps....😉
i think you are in the wrong thread.....
i do not explain opinions, nor will i demand you to explain your opinions....
technical discussions, those require more explanations...
but definitely using 16volt tantalums on circuits that have 40 volts is unsafe...
guaranteed to let the magic smoke out...i wouldn't even try...
i think this belong to the snake oil thread...
technical discussions, those require more explanations...
but definitely using 16volt tantalums on circuits that have 40 volts is unsafe...
guaranteed to let the magic smoke out...i wouldn't even try...
i think this belong to the snake oil thread...
- Status
- Not open for further replies.