I have a modded dac-ah (passive I/V) which has about +3.25VDC on the outputs. I don't have DC blocking coupling caps to stop this DC 'getting out'.
This connects to a grounded grid preamp which has no coupling caps on the input. There are 4.7uF caps on the output, so I don't have a DC offset on the connection to the power amp.
I always assumed that the presence of the preamp's output coupling caps meant that I didn't need to worry about a DC offset on the input, so haven't used coupling caps between the dac and the preamp.
Yesterday I tried a couple of coupling caps between the dac and the pre. I was expecting a loss of focus, but the result was the opposite. With the caps in place, the sound became more focussed and yet smoother and warmer. To my ears, better all round.
Do you think this result is due to the preamp not liking DC on it's inputs? (I'm sure the improvement is more than just the added flavor of the coupling caps....)
This connects to a grounded grid preamp which has no coupling caps on the input. There are 4.7uF caps on the output, so I don't have a DC offset on the connection to the power amp.
I always assumed that the presence of the preamp's output coupling caps meant that I didn't need to worry about a DC offset on the input, so haven't used coupling caps between the dac and the preamp.
Yesterday I tried a couple of coupling caps between the dac and the pre. I was expecting a loss of focus, but the result was the opposite. With the caps in place, the sound became more focussed and yet smoother and warmer. To my ears, better all round.
Do you think this result is due to the preamp not liking DC on it's inputs? (I'm sure the improvement is more than just the added flavor of the coupling caps....)
The DC on the input ought only to get through the first stage of whatever amplifier it's applied to - the first coupling cap should block it from hitting the rest of the amplifier.
That said, the DC you're talking about could very well be loading the input stage excessively, depending on a number of factors (the gain of the stage being primary). If my understanding is accurate, with DC on the input, the tube is not operating where you "think" it is operating; that's assuming you banged out all the load lines and calculations for the stage yourself.
In any case, the coupling cap may just be the less ugly of the two beasts. Caps are bad, DC's worse (unless the stage is intentionally direct-coupled and you planned for it)
That said, the DC you're talking about could very well be loading the input stage excessively, depending on a number of factors (the gain of the stage being primary). If my understanding is accurate, with DC on the input, the tube is not operating where you "think" it is operating; that's assuming you banged out all the load lines and calculations for the stage yourself.
In any case, the coupling cap may just be the less ugly of the two beasts. Caps are bad, DC's worse (unless the stage is intentionally direct-coupled and you planned for it)
I always try to make sure that none of my equipment has DC on the inputs or outputs,even if that means using a coupling cap.
As koolatron noted,DC can shift bias points,and do other unexpected/unwanted things.
And of course,shifting bias points and stuff will alter the sound,because things aren't running where they're designed to be.
Plus DC on volume pots and stuff can make them scratchy,and wear
faster.
So,IMO,a coupling cap is a small sacrifice to make,to avoid other more problematic issues..
As koolatron noted,DC can shift bias points,and do other unexpected/unwanted things.
And of course,shifting bias points and stuff will alter the sound,because things aren't running where they're designed to be.
Plus DC on volume pots and stuff can make them scratchy,and wear
faster.
So,IMO,a coupling cap is a small sacrifice to make,to avoid other more problematic issues..
Sharpi31,
As the other posters have indicated, you need the blockage of DC a capacitor provides. Size the cap. so the corner frequency of the pole it forms with the preamp's I/P resistor is between 15 and 19 Hz. The DC blocking cap. is not within a NFB loop and there is no point in passing any infrasonic garbage into the preamp.
Experiment with several different types of capacitor construction. In particular, try PIO and film and foil parts. You will discover what pleases you most.
As the other posters have indicated, you need the blockage of DC a capacitor provides. Size the cap. so the corner frequency of the pole it forms with the preamp's I/P resistor is between 15 and 19 Hz. The DC blocking cap. is not within a NFB loop and there is no point in passing any infrasonic garbage into the preamp.
Experiment with several different types of capacitor construction. In particular, try PIO and film and foil parts. You will discover what pleases you most.
"I have a modded dac-ah (passive I/V) which has about +3.25VDC on the outputs. I don't have DC blocking coupling caps to stop this DC 'getting out'."
This is why all my designs have input coupling capacitors, whether it's solid state or hollow state. I've seen too many sound cards and other devices with DC on the outputs. Putting +3.25Vdc on the input probably completely nullifies the grid bias, making for a positive Vgk. Most small signal VTs weren't designed for that, and it's a wonder you didn't poof the input VT. Even if it doesn't roll over and die, it sure won't be operating at its design Q-Point. No wonder they sound better when that DC is blocked off.
This is why all my designs have input coupling capacitors, whether it's solid state or hollow state. I've seen too many sound cards and other devices with DC on the outputs. Putting +3.25Vdc on the input probably completely nullifies the grid bias, making for a positive Vgk. Most small signal VTs weren't designed for that, and it's a wonder you didn't poof the input VT. Even if it doesn't roll over and die, it sure won't be operating at its design Q-Point. No wonder they sound better when that DC is blocked off.
I think I've learnt a valuable lesson 
Many thanks for the explanations. The caps I've used are Epcos MKV 3.3uF, which should give me a corner frequency of around 1Hz (amps input impedance is 33Kohm).
Eli: I tend to avoid caps with corner frequencies so close to 20Hz, given the phase shifts that will be apparent within the audio band. Obviously this means I'm passing subsonic frequencies into my amps. Do you think living with phase shifts is less of a problem than exposing kit to >20Hz content?
Many thanks for the explanations. The caps I've used are Epcos MKV 3.3uF, which should give me a corner frequency of around 1Hz (amps input impedance is 33Kohm).
Eli: I tend to avoid caps with corner frequencies so close to 20Hz, given the phase shifts that will be apparent within the audio band. Obviously this means I'm passing subsonic frequencies into my amps. Do you think living with phase shifts is less of a problem than exposing kit to >20Hz content?
Looks like I'm a dissenting voice here - I too have a DAC with DC on the output. I tried a cap and a transformer out of it but nothing was as pure as going straight into the grids of my 10Y line stage via a stepped attenuator in shunt between the two halves of the 10Y differential pair - my DAC is balanced out.
I allow for the change in bias when calculating the 10y stage. I just don't see the problem here - caps degrade the sound to my ears.
Andy
I allow for the change in bias when calculating the 10y stage. I just don't see the problem here - caps degrade the sound to my ears.
Andy
Eli Duttman said:
sharpi31 said:
I prefer the bass when the F-3dB of the input high pass filter (DC blocking cap) is set to below 2Hz. To make the following stages economic, I set the Power Amp to ~90mS and the preamp to 180mS. The 3u3F & 33k give 109mS. The DC block on the lower leg of the NFB loop must be set to an even lower frequency to ensure no significant audio signal voltage across the cap.
sharpi31 said:I think I've learnt a valuable lesson
Many thanks for the explanations. The caps I've used are Epcos MKV 3.3uF, which should give me a corner frequency of around 1Hz (amps input impedance is 33Kohm).
Eli: I tend to avoid caps with corner frequencies so close to 20Hz, given the phase shifts that will be apparent within the audio band. Obviously this means I'm passing subsonic frequencies into my amps. Do you think living with phase shifts is less of a problem than exposing kit to >20Hz content?
You make a point that always bears repeating. All designs contain compromises. My view is that eliminating the infrasonic crud is more important than the phase shift. Is this particular phase shift audible? I believe it's not. Deep bass is non-directional.
Inside a NFB loop, additional factors come into play. The very last thing you want is satisfying Barkhausen's oscillation criterion. So, inside a NFB loop, keeping the "corner" freq. of high pass poles at or below 5 Hz. makes considerable sense.
Eli,
I have some 0.47uF russian teflon caps that would give me a corner frequency of 10Hz. I'll give them a try and see how it sounds... Obviously these are different caps than the MKVs, but the comparison should still be interesting.
FYI, I use the coupling cap calculator on this page. It's very handy:
Coupling cap calculator
I have some 0.47uF russian teflon caps that would give me a corner frequency of 10Hz. I'll give them a try and see how it sounds... Obviously these are different caps than the MKVs, but the comparison should still be interesting.
FYI, I use the coupling cap calculator on this page. It's very handy:
Coupling cap calculator
I was to try some cap couple experiments & direct this mourning & coverted my DAC-AH for experiment mode, Then found a couple UTC LMI 5000 10k/10k trannies . Results are that it is 7 pm & it is still playing, nice.. I thought they were bandwidth limited & never used em, but I do not hear any limiting. Also can not find them in my 1963 UTC catalog, they are hermetic, & suspect very high quality, cute lil sum bit....They do sound good though. I was using Auricap & many others direct off I/V resistors. Into 26 pre/ 300B PP, Kenwood7070, yep...Kenwood I said....
J & G
J & G
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.