eVITAERC said:
No, if the preamp has an output cap, the extra one in the power amp doesn't do anything as far as offset from the preamp is concerned. Same with DC servo: assuming you mean a servo in the pre, that means the pre has 0 V DC, and you can connect that to the power amp with or without a power amp input cap (but why pay twice?).
The wire doesn't pick up DC out of thin air. However, but that's another story, it can (if not propoerly screened) pick up magnetic and/or electric fields (acting as antenna) and in that way end up with AC on the input of the poweramp that doesn't come from the preamp.
Jan Didden
eVITAERC said:
Actually a bad thing may happen, two caps in series give smaller capacitance, so those two may attenuate or phase shitf some bass.
The only reason for a cap being here and there too is that both amp and pre should be idiot audio chain proof. If all the power amps in the world have had a cap on input, preamps wouldn't have one on output.
cheers
P.S. Problems with cAPSLOCK ?
P.S. 2. There are no stupid questions
Hi eVITAERC,
Some units come without. I highly recommend the amplifier have the cap. I did get a pair where the amp cooked the speakers (Krell) because the servo in the preamp (Krell) failed and the amp passed the DC to the ex-speaker.
Generally the LF cutoff from both caps in series will not hurt the sound in most systems. Heck, even the room cutoff is higher. The speakers may be higher still. 0.2 Hz, 5 Hz, 18 Hz. For most, this is completely inaudible except the 18 Hz cutoff may sound cleaner.
-Chris
Some units come without. I highly recommend the amplifier have the cap. I did get a pair where the amp cooked the speakers (Krell) because the servo in the preamp (Krell) failed and the amp passed the DC to the ex-speaker.
Generally the LF cutoff from both caps in series will not hurt the sound in most systems. Heck, even the room cutoff is higher. The speakers may be higher still. 0.2 Hz, 5 Hz, 18 Hz. For most, this is completely inaudible except the 18 Hz cutoff may sound cleaner.
-Chris
Hi,
input caps on power amp and output cap on preamp will, as suggested earlier, add together.
If both were the same value then the bass roll off frequency would move up one octave. Decide if this is a problem. You can alleviate this by ensuring one of the caps is at least twice the size of the other, choose the poorest one to replace if you need to go that route. One of my preamps has a pair 10uF PP, that will never need replacing.
If you intend to run multiple parallel power amps from the same preamp source then all the power amps except one MUST have a dc blocking cap on the input. This prevents interaction of the input offset voltage between the paralleled input stages.
Both sides of the input and output caps should be connected to signal ground. You can add 1M0 to 2M2 at the RCA connectors to achieve this since it is very often omitted from amps and preamps. Croft are one company that are now recommending removal of these extra resistors from their preamps to improve sound quality (by removing excess loading from their direct coupled high impedance valve output stage).
input caps on power amp and output cap on preamp will, as suggested earlier, add together.
If both were the same value then the bass roll off frequency would move up one octave. Decide if this is a problem. You can alleviate this by ensuring one of the caps is at least twice the size of the other, choose the poorest one to replace if you need to go that route. One of my preamps has a pair 10uF PP, that will never need replacing.
If you intend to run multiple parallel power amps from the same preamp source then all the power amps except one MUST have a dc blocking cap on the input. This prevents interaction of the input offset voltage between the paralleled input stages.
Both sides of the input and output caps should be connected to signal ground. You can add 1M0 to 2M2 at the RCA connectors to achieve this since it is very often omitted from amps and preamps. Croft are one company that are now recommending removal of these extra resistors from their preamps to improve sound quality (by removing excess loading from their direct coupled high impedance valve output stage).
Allright, input caps are useful and they can in fact contribute to our general safety and such, we should learn to love it.
Or should we? I know of many many people that have goine to great lengths just so they can "mod out" the output cap of their preamps. Beit DC servos, transistor matching, buffers, etc. Isn't that all ruined by a input cap on the power amp?
I was wondering if it is fairly standard practice to take out the input cap on the power amps if purist "no caps in the signal path" approaches were used in the preamp. I guess this applies to the likes of people that use their Gilmore Dynalo's as preamps.
Or should we? I know of many many people that have goine to great lengths just so they can "mod out" the output cap of their preamps. Beit DC servos, transistor matching, buffers, etc. Isn't that all ruined by a input cap on the power amp?
I was wondering if it is fairly standard practice to take out the input cap on the power amps if purist "no caps in the signal path" approaches were used in the preamp. I guess this applies to the likes of people that use their Gilmore Dynalo's as preamps.
Hi eVITAERC,
I think removing all caps in the signal path is misguided and just plain silly. That's like wishing there was no money and we'd get whatever we wanted. Pure Communism. (I don't believe in pure Capitalism either)
The small signal caps in the input circuit will have the least effect on sound quality. There are far worse things than these small caps. Speakers for one, circuit design for another.
As I posted earlier, a matched Krell pre and power needlessly killed a woofer (from a new PSB Stratus Gold). The customer was none too pleased. Why did the amp not have a coupling cap? I doubt you could hear it if it was there as long as you used a good film cap.
So everything in life is a balance and compromise to some extent.
-Chris
I think removing all caps in the signal path is misguided and just plain silly. That's like wishing there was no money and we'd get whatever we wanted. Pure Communism. (I don't believe in pure Capitalism either)
The small signal caps in the input circuit will have the least effect on sound quality. There are far worse things than these small caps. Speakers for one, circuit design for another.
As I posted earlier, a matched Krell pre and power needlessly killed a woofer (from a new PSB Stratus Gold). The customer was none too pleased. Why did the amp not have a coupling cap? I doubt you could hear it if it was there as long as you used a good film cap.
So everything in life is a balance and compromise to some extent.
-Chris
AndrewT said:
I'm glad you brought this issue. But I noticed that they usually use bigger capacitance than necessary so the chance for this problem should be minimized. But then again, it seems that the theoritical value for -3dB at 10Hz (also in feedback decoupling) is never sufficient (when the system allows for lows reproduction)!? Is this somekind of "headroom" issue? And with shunt resistor between pre output cap and amp input cap, is the Ct=1/(1/C1+1/C2) still apply?
AndrewT said:
So what is the purpose of connecting both sides of the caps to ground? (I have been wondering about this mega-ohm resistors. It doesn't seem like an impedance issue). If the Croft thinks that removing the resistors is better soundwise, why shouldn't it be?
My stochino amp (under construction) in the input is using 47K (shunt), 48uF (series), 1K5 (series), 4K7 (shunt), 820p (shunt). Will the 1K5 be useless?
Hi Jay,
I think I can shed some light on this. Andrew can kick me if I'm wrong here.
It's not the frequency cutoff but rather the phase shift that might worry some people. Depends on the speakers and program material. A true perfectionist would want to keep the cutoff points 10X from the band of interest at each end of the spectrum.
Referencing the other sides of the capacitors to ground through 1M0 or higher resistors will hopefully prevent pops and bangs as the capacitors are suddenly charged or discharged. This may happen if the impedance at the input suddenly changes if the capacitor is allowed to float. Some muting circuits do this.
-Chris
I think I can shed some light on this. Andrew can kick me if I'm wrong here.
It's not the frequency cutoff but rather the phase shift that might worry some people. Depends on the speakers and program material. A true perfectionist would want to keep the cutoff points 10X from the band of interest at each end of the spectrum.
Referencing the other sides of the capacitors to ground through 1M0 or higher resistors will hopefully prevent pops and bangs as the capacitors are suddenly charged or discharged. This may happen if the impedance at the input suddenly changes if the capacitor is allowed to float. Some muting circuits do this.
-Chris
Hi Jay,
That's 2 Hz (20 / 10). Many designers feel they need to go way down there to, say 0.2 Hz or lower. This may make it tough to get good decoupling in the power supply.
Given that most speakers cut off -3 dB above 40 Hz, as do many rooms, I am normally not that concerned about getting down to 2 Hz. A filter at 20 Hz often cleans up the sound since even there you are just pushing the woofer back and forth. That's in the real world with a turntable.
Now if you have a room and speakers that will support 20 Hz, and some of us do, with program material that gets down there, you may want that response down to 2 Hz or below. That would not be the majority of people.
-Chris
That's 2 Hz (20 / 10). Many designers feel they need to go way down there to, say 0.2 Hz or lower. This may make it tough to get good decoupling in the power supply.
Given that most speakers cut off -3 dB above 40 Hz, as do many rooms, I am normally not that concerned about getting down to 2 Hz. A filter at 20 Hz often cleans up the sound since even there you are just pushing the woofer back and forth. That's in the real world with a turntable.
Now if you have a room and speakers that will support 20 Hz, and some of us do, with program material that gets down there, you may want that response down to 2 Hz or below. That would not be the majority of people.
-Chris
Hi Anatech and others,
I will never kick you. Even if you disagree with me. It is always better to hear a variety of views and then evaluate.
If the builder wants to become a designer he can choose to take on board the various info and make his own compromises. Or he can simply regurgitate what someone else provides and never learn anything.
Sorrry to all our female readers for the continued use of he.
I will never kick you. Even if you disagree with me. It is always better to hear a variety of views and then evaluate.
If the builder wants to become a designer he can choose to take on board the various info and make his own compromises. Or he can simply regurgitate what someone else provides and never learn anything.
Sorrry to all our female readers for the continued use of he.
Hi Andrew,
In the English language, the term "he" can be taken as male or female as the case may be.
At least when I went to school.
Actually, I was referring to the fact that Jay was responding to your post. I jumped in. This gives you the chance to give me a kick for butting in. I don't know if you agree with my answer.
You generally have some thoughtful things to say.
At DIY, we are all designers from time to time. Even if what we do is wrong, we will remember the mistake that way!
Cheers, Chris
In the English language, the term "he" can be taken as male or female as the case may be.
At least when I went to school.Actually, I was referring to the fact that Jay was responding to your post. I jumped in. This gives you the chance to give me a kick for butting in. I don't know if you agree with my answer.
You generally have some thoughtful things to say.At DIY, we are all designers from time to time. Even if what we do is wrong, we will remember the mistake that way!
Cheers, Chris
Hi Anatech,
slightly off topic:-
Glad to see we disgree. Here in Scotland HE and SHE mean something completely different and they can never be interchanged. Nor should they "Vive l'difference" French speakers are welcome to jump in and correct my spelling.
Yes, your reply seemed to follow closely enough to my understanding that there was no advantage in coming back to support the same and/or similar view.
slightly off topic:-
Glad to see we disgree. Here in Scotland HE and SHE mean something completely different and they can never be interchanged. Nor should they "Vive l'difference" French speakers are welcome to jump in and correct my spelling.
Yes, your reply seemed to follow closely enough to my understanding that there was no advantage in coming back to support the same and/or similar view.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.