Let's say we have this preamp. At the output we have this low-pass type network, a cap and a grounded resistor. (From what I can tell, this isn't always present in ss gear.) The cap, as I understand it, is there to block DC.
In a phono preamp, having a low-pass filter is not a bad thing. The only sub 20Hz information the cart will pick up is rumble. But the problem with the output coupling cap and grounded resistor is that you usually need relatively high values. If I'm to use a coupling cap, I want to use something like a silver mica.
I'm trying to understand why you "have" to have this kind of network. Ditching the resistor eliminates the roll-off. And placing the cap after the resistor can still produce the, in phono stages, welcome roll-off while keeping the cap value in the picas. So what is going on here?
Could somebody briefly explain the reason/background for this network? As well as explaining why ditching the resistor (using only a DC block cap) or placing the cap after the resistor would be a bad idea.
While we're at it, what is the main cause or source for DC? Is any technology (tubes, ss) more susceptible? I know DC is a reality in SE tube amps.
In a phono preamp, having a low-pass filter is not a bad thing. The only sub 20Hz information the cart will pick up is rumble. But the problem with the output coupling cap and grounded resistor is that you usually need relatively high values. If I'm to use a coupling cap, I want to use something like a silver mica.
I'm trying to understand why you "have" to have this kind of network. Ditching the resistor eliminates the roll-off. And placing the cap after the resistor can still produce the, in phono stages, welcome roll-off while keeping the cap value in the picas. So what is going on here?
Could somebody briefly explain the reason/background for this network? As well as explaining why ditching the resistor (using only a DC block cap) or placing the cap after the resistor would be a bad idea.
While we're at it, what is the main cause or source for DC? Is any technology (tubes, ss) more susceptible? I know DC is a reality in SE tube amps.
hi
Anything you add to an amplifier output has got an impedance.
This is like a resistance seen and it is frequency dependent.
A loudspeaker can have lower impedance at 200 hertz and high impedance at 50 hertz.
The nominal impedance for a load is most often given for 1 kHz.
This is the resistance for AC signals,
and can be different than for only DC current.
8 ohm loudspeaker has got ~7-8 ohm impedance.
A power amplifier input, attached to preamplifier output
has got an input impedance, often like 22k - 47k.
Headphones can have 16-32 or 300 or even 600 ohm impedance.
So an output cap has got some impedance to work into,
depending on what is attached to the output.
Sure there are many more topics in the past, that try to explain this.
If we can search and find them
I tell a little bit about formula to use to get different rolloff frequency in this post:
http://www.diyaudio.com/forums/showthread.php?postid=1024640#post1024640
I am sure somebody will fill in with more and better information here.
Regards
lineup
Anything you add to an amplifier output has got an impedance.
This is like a resistance seen and it is frequency dependent.
A loudspeaker can have lower impedance at 200 hertz and high impedance at 50 hertz.
The nominal impedance for a load is most often given for 1 kHz.
This is the resistance for AC signals,
and can be different than for only DC current.
8 ohm loudspeaker has got ~7-8 ohm impedance.
A power amplifier input, attached to preamplifier output
has got an input impedance, often like 22k - 47k.
Headphones can have 16-32 or 300 or even 600 ohm impedance.
So an output cap has got some impedance to work into,
depending on what is attached to the output.
Sure there are many more topics in the past, that try to explain this.
If we can search and find them
I tell a little bit about formula to use to get different rolloff frequency in this post:
http://www.diyaudio.com/forums/showthread.php?postid=1024640#post1024640
I am sure somebody will fill in with more and better information here.
Regards
lineup
Hi,
the in line cap followed by the grounding resistor is a high pass filter.
The DC and very low frequencies are blocked.
The input impedance of the next stage must be paralleled with the grounding resistor to find the "load" that the cap forms a filter with.
There may also be a further DC blocking cap at the input of the next unit (receiver).
Let's suppose you have a 1uF blocking cap and a 1M0 resistor on the output of the source unit.
Now the receiver has a DC coupled input with a 22k ohm impedance setting resistor to ground.
The 1uF sees the 22k//1M0 which is equivalent to 21.526k, the 1M0 has very little effect.
The -3db frequency of the high pass filter is 1/2 Pi R C=1/(2*3.14159*22526*1*10^-6)=7.1Hz The effect of this filter is audible up to about 70Hz.
You then have a choice to raise the RC time constant. Either increase the capacitor or increase the input impedance. That's why you often find 50K used at the front end of receivers.
If the receiver also has a DC blocking capacitor then the two caps are in series and the effect of both need to be checked.
Lets make the arithmetic easy and say the receiver also has a 1uF cap. The effective capacitance is 0.5uF and the -3db frequency is now up to 14.2Hz and it's effect now extends up to 140Hz. Not much low bass left in the signal now.
2.2uF for both source and receiver moves you in the right direction, but I would fit a good quality 10uF polypropylene at the output of the pre-amp and hope that the power-amp has at least a sensible cap to match the chosen input impedance. I aim for F(-3) of between 1Hz and 2Hz (RC=80 to 90 mS).
the in line cap followed by the grounding resistor is a high pass filter.
The DC and very low frequencies are blocked.
The input impedance of the next stage must be paralleled with the grounding resistor to find the "load" that the cap forms a filter with.
There may also be a further DC blocking cap at the input of the next unit (receiver).
Let's suppose you have a 1uF blocking cap and a 1M0 resistor on the output of the source unit.
Now the receiver has a DC coupled input with a 22k ohm impedance setting resistor to ground.
The 1uF sees the 22k//1M0 which is equivalent to 21.526k, the 1M0 has very little effect.
The -3db frequency of the high pass filter is 1/2 Pi R C=1/(2*3.14159*22526*1*10^-6)=7.1Hz The effect of this filter is audible up to about 70Hz.
You then have a choice to raise the RC time constant. Either increase the capacitor or increase the input impedance. That's why you often find 50K used at the front end of receivers.
If the receiver also has a DC blocking capacitor then the two caps are in series and the effect of both need to be checked.
Lets make the arithmetic easy and say the receiver also has a 1uF cap. The effective capacitance is 0.5uF and the -3db frequency is now up to 14.2Hz and it's effect now extends up to 140Hz. Not much low bass left in the signal now.
2.2uF for both source and receiver moves you in the right direction, but I would fit a good quality 10uF polypropylene at the output of the pre-amp and hope that the power-amp has at least a sensible cap to match the chosen input impedance. I aim for F(-3) of between 1Hz and 2Hz (RC=80 to 90 mS).
Thanks guys. I had to rethink before posting, and then time flow by.
Maybe I'm looking for a simple answer when there is none. It still bothers me. I mean, say I put in all this effort making a direct-coupled tube whatever doing everything to avoid caps and then, wham, toss it out by adding an output cap.
Maybe chips are different, having low capacitance. This active crossover, for example, has no output caps.
And HIGH pass. Doh!
Maybe I'm looking for a simple answer when there is none. It still bothers me. I mean, say I put in all this effort making a direct-coupled tube whatever doing everything to avoid caps and then, wham, toss it out by adding an output cap.
Maybe chips are different, having low capacitance. This active crossover, for example, has no output caps.
And HIGH pass. Doh!
Hi,
but it has inherent DC offset on the outputs and these offsets will vary with temperature, whether caused by ambient conditions, heat generated within the equipment or by internal heat generated within the ICs.
The high pass filters have DC blocking for offsets coming in from previous stages, but the low pass filters aggregate all the offsets as the signal passes towards the final output.
BTW. that crossover had better have socketed IC's because the standard tl07Xs are pretty poor at driving reactive loads, even small capacitances on the outputs can upset them.
but it has inherent DC offset on the outputs and these offsets will vary with temperature, whether caused by ambient conditions, heat generated within the equipment or by internal heat generated within the ICs.
The high pass filters have DC blocking for offsets coming in from previous stages, but the low pass filters aggregate all the offsets as the signal passes towards the final output.
BTW. that crossover had better have socketed IC's because the standard tl07Xs are pretty poor at driving reactive loads, even small capacitances on the outputs can upset them.
I will have to regroup on this issue. But I have a related question.
I got some 600:600Ohm matching transformers for my PlayStation as CD player. I had already modified the output on the PSX, now consisting of a coupling cap and grounded resistor. The idea was to replace the caps and resistors with a transformers.
I made one crucial mistake. I never tested the transformers before installing them, being sure I had got everything right. It didn't go as planned. The sound was very low with lots of hum. Since I never tested the transformers, if I got the leads right and things like that, I didn't know what was wrong. It could be anything. So I had to remove them. I later hooked RCA connectors to the transformer leads and connected them between the PSX and amp. Works perfect. From that I can conclude that I had connected them right in the first place.
My problem now is that I don't know what went wrong. And I don't want to do it all over again without making sure it will work.
So my question is, shouldn't it be possible to replace the output cap and resistor with a transformer?
I got some 600:600Ohm matching transformers for my PlayStation as CD player. I had already modified the output on the PSX, now consisting of a coupling cap and grounded resistor. The idea was to replace the caps and resistors with a transformers.
I made one crucial mistake. I never tested the transformers before installing them, being sure I had got everything right. It didn't go as planned. The sound was very low with lots of hum. Since I never tested the transformers, if I got the leads right and things like that, I didn't know what was wrong. It could be anything. So I had to remove them. I later hooked RCA connectors to the transformer leads and connected them between the PSX and amp. Works perfect. From that I can conclude that I had connected them right in the first place.
My problem now is that I don't know what went wrong. And I don't want to do it all over again without making sure it will work.
So my question is, shouldn't it be possible to replace the output cap and resistor with a transformer?
phn said:
Depends on the transformer, the circuit and how you hook it up.
You probably have DC between the point where you connected the transformer and ground, or wherever you connected the other side of the transformer primary. Not good for the transformer or the circuit that proceeds it.
That DC blocking cap keeps the DC where it belongs, out of the following stages - like the amp or the transformer primary.
Got a schematic of the PS1 output?
In short, the original out put (caps and muting resistors) is replaced with a DC block cap and grounded resistor.
http://dogbreath.de/PS1/output/output.html
http://dogbreath.de/PS1/output/output.html
panomaniac said:
I'm back.
I really know nothing about the transformers other than they are "standard" (whatever that means) pro line/balancing transformers made by Haufe.
Anyway, the transformer output works perfect when I use coupling caps. But when I remove the caps, the sound gets very low with hiss as loud as the music.
This is apparently caused by DC. Is there a way to solve this? Or am I forced to use coupling caps?
The original question was about a coupling capacitor between a pre-amplifier and I presume, a power amplifier (not stated). For a meaningful calculation one needs the output impedance of the pre-amp as well as the input impedance of the power amplifier. These values are mostly specified.
Our good friend AndrewT did an excellent analysis of the basics. But with a -3 dB cut-off at 7,1 Hz, the effect at 70 Hz will be -0,86 dB. Still audible??
Anyway, any coupling time constant below 5Hz should suffice. Phn, you also do not need expensive mica capacitors. A lot has been said about magic capacitor effects (none quantative); it has been shown that poly-ester is perfectly capable of fitting the bill.
You could also certainly use the capacitor after the R (just make sure that that R is not shunting any d.c. to common without a blocking capacitor.) The blocking capacitor will then just form a high-pass network with the power amplifier input impedance. AndrewT gave this as some 22K; I would hope that it is rather higher, but do not have much experience of commercial equipment.
Regards
Our good friend AndrewT did an excellent analysis of the basics. But with a -3 dB cut-off at 7,1 Hz, the effect at 70 Hz will be -0,86 dB. Still audible??
Anyway, any coupling time constant below 5Hz should suffice. Phn, you also do not need expensive mica capacitors. A lot has been said about magic capacitor effects (none quantative); it has been shown that poly-ester is perfectly capable of fitting the bill.
You could also certainly use the capacitor after the R (just make sure that that R is not shunting any d.c. to common without a blocking capacitor.) The blocking capacitor will then just form a high-pass network with the power amplifier input impedance. AndrewT gave this as some 22K; I would hope that it is rather higher, but do not have much experience of commercial equipment.
Regards
Johan, thanks for the reply. I still don't quite get it. But I'm not forcing it. If I knew why a 600Ohm load is more difficult to drive than a 10k load I might have been closer to the answer. I will get there, eventually.
Yeah, I was being stupid. I wanted to avoid starting a new thread and ended up more or less ruining this one for everybody by going OT. I will have to start a new thread anyway. I want to have the PlayStation up and running. Another lesson learned.
As for micas. I won't spend money on boutique caps. I don't consider micas boutique caps. And they can be found at reasonable prices. They may cost 2-3 times as much as poly caps. But since we talk low-value caps and relatively small sums of money, it doesn't matter that much.
Yeah, I was being stupid. I wanted to avoid starting a new thread and ended up more or less ruining this one for everybody by going OT. I will have to start a new thread anyway. I want to have the PlayStation up and running. Another lesson learned.
As for micas. I won't spend money on boutique caps. I don't consider micas boutique caps. And they can be found at reasonable prices. They may cost 2-3 times as much as poly caps. But since we talk low-value caps and relatively small sums of money, it doesn't matter that much.
Hi.Untill now, i understood that you want to replace the capacitor and resistor of the output of you PS with a transformer.The cap is there not to pass the DC component of the signal.Am i already familiar with the circuit?phn said:
If you want to use transformer instead of the cap and resistor, first of all you must know the potential of the DC you want to get rid of.This is important, becouse the transformer can carry specific max value of DC passing across the primary winding.After that value, the transformer will reach its maximum limits and start to act like a crazy - get warmer and so on...
That value depends on the type of the transformer's `iron`and the air gap between the windings and the `iron`.
The reason why a smaller load is more difficult to drive is because the current passing through the 600ohms load is a lot more that through the 10k's.The current, supplied by the unit before the load - let name it Current1, has it's maximum,and you must try to keep the value of the current passing through the load lower that that of Current1.Using a smaller load lowers the value of the DF(Demphing Factor) - not good for the sound.The DF is the result of this:
Rload divided by Ro.Rload is the resistance of the load,Ro - the output resistance of the unit before the load.
Hope my post was usefull to you.
Freestyler, thanks. The Demphing Factor is new to me. I don't think I'm able to apply it in this case.
Being stuck with these "non-working" trafos for a month, I finally started a new thread today. Now I have two threads on basically the same subject.
My current trafos are clearly not up to the job. From what I understand I need something in the line of SE interstage or output trafos, except with 600:600 or so Ohms. They will need to be big and expensive. They will most likely be too big to be mounted on the PlayStation. That sucks. But at least I know why the current ones don't work. That's a step forward. Now I need to ditch the line trafos.
Being stuck with these "non-working" trafos for a month, I finally started a new thread today. Now I have two threads on basically the same subject.
My current trafos are clearly not up to the job. From what I understand I need something in the line of SE interstage or output trafos, except with 600:600 or so Ohms. They will need to be big and expensive. They will most likely be too big to be mounted on the PlayStation. That sucks. But at least I know why the current ones don't work. That's a step forward. Now I need to ditch the line trafos.
Excuse my mistake - it is Damping in English,we spell it demphing in Bulgarianphn said:Freestyler, thanks. The Demphing Factor is new to me. I don't think I'm able to apply it in this case.
Being stuck with these "non-working" trafos for a month, I finally started a new thread today. Now I have two threads on basically the same subject.
My current trafos are clearly not up to the job. From what I understand I need something in the line of SE interstage or output trafos, except with 600:600 or so Ohms. They will need to be big and expensive. They will most likely be too big to be mounted on the PlayStation. That sucks. But at least I know why the current ones don't work. That's a step forward. Now I need to ditch the line trafos.
Yes - the trafs you have now maybe aren't suitable for the aplication.I don't know your PS's size - I never had such a thing(computers,computers)
If I was you - a would prefer better capacitors (AgTa) - it is a PlayStation at all - far away from a CD Player, and calculationg a transformer is not an easy deal.But you can also try out your first idea(trafos) - it will take you a lot more time, but the result may be better!Good luck with the project
FreestylerBG said:
Excuse my mistake - it is Damping in English,we spell it demphing in Bulgarian
"Demphing" did throw me. But I got a couple hits when I googled, including this: "The hypotesa of the demphing of the instability and the building of the stationary periodical gratings of the refractive-index is suggested." Who am I to argue with Harvard?
http://adsabs.harvard.edu/abs/1998SPIE.3485..440B
To use ohms law to illustrate... imagine your amp needed a 1V input to reach full volume.
In fact imagine you have 2 of those amps.
One with an input impendance of 600R, and one with 10K
To put 1V over 600R it would need 1V at 0.0017A
To put 1V over 10K would need only 1V at 0.0001A This amp would be a much "easier" load on a preamp or any preceding device in the signal chain.
In fact imagine you have 2 of those amps.
One with an input impendance of 600R, and one with 10K
To put 1V over 600R it would need 1V at 0.0017A
To put 1V over 10K would need only 1V at 0.0001A This amp would be a much "easier" load on a preamp or any preceding device in the signal chain.
Nordic, thanks. I figured it was straight math/physics. I sort of had it figured before (more current goes to ground or something such with a small load), though I had no hard figures. Working in Spice has taught me quite a bit, while at the same time preventing me from learning the fundamental stuff.
Johan, I know electronics is simple, as simple as math. I personally think it's hard. But that has to do with me and not electronics.
Johan, I know electronics is simple, as simple as math.
I personally think it's hard. But that has to do with me and not electronics.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.