So I just bought an Audio Electronics Supply AE-2 solid state preamplifier. It's well-constructed (made by Cary), in near mint condition with original box, and is a nice sounding piece of gear. It does have an unusual feature: AC and DC outputs for the amp. According to the manual:
"The AE-2 utilizes a single gain stage based on an N-channel silicon gate TMOS power FET transistor. The AE-2 is operated in a class A single-ended mode...The amplified output signal is taken from the drain MOSFET. The output audio signal is routed in two paths. One is through a 10 MFD coupling capacitor to provide AC coupling to the power amplifier. The second path is diret from the gate of the MOSFET with a DC potential of 20VDC along with the amplified audio signal to the DC output RCA connector on the rear of the AE-2."
So I know DC is bad news for speakers, but I have heard of DC-coupled amps (which I assume would have input coupling capacitors to filter the DC) and wonder what are the advantages? I read that most amps now are AC coupled, so what would would happen if I hooked into the DC output? A soft fizzle of voice coil melt? Exploding transistors? Smoking resistors? Streaks of blue flames? Or are "modern" amplifiers capable of handling DC on the input?
Any experience or knowledge about DC-coupling is welcome.
"The AE-2 utilizes a single gain stage based on an N-channel silicon gate TMOS power FET transistor. The AE-2 is operated in a class A single-ended mode...The amplified output signal is taken from the drain MOSFET. The output audio signal is routed in two paths. One is through a 10 MFD coupling capacitor to provide AC coupling to the power amplifier. The second path is diret from the gate of the MOSFET with a DC potential of 20VDC along with the amplified audio signal to the DC output RCA connector on the rear of the AE-2."
So I know DC is bad news for speakers, but I have heard of DC-coupled amps (which I assume would have input coupling capacitors to filter the DC) and wonder what are the advantages? I read that most amps now are AC coupled, so what would would happen if I hooked into the DC output? A soft fizzle of voice coil melt? Exploding transistors? Smoking resistors? Streaks of blue flames? Or are "modern" amplifiers capable of handling DC on the input?
Any experience or knowledge about DC-coupling is welcome.
Fmjunkie,
Firstly, let us not mix things up. The d.c. that is bad for the loudspeaker refers to the output of an amplifier - a "so-called" d.c. input has nothing to do with this (in a manner of speaking).
To come to d.c. coupled amps. If I may be straight: It is an overrated term bringing more grief than advantages - actually, I cannot think of a single advantage. One listens to signals in the audio band: Nominally 20 Hz - 20kHz. These days with ultra (another overrated word) low frequencies eminating from certain DVDs the low frequency margin may have to be lower. (That depends on your taste and is a different subject.)
I belong to the group that justifiably prefers everything outside the audible band to be gotten rid of. (Again that is a separate subject, concerning an amplifier's ability or rather lack of it to deal with signals outside the band it was designed for.) A d.c. coupled pre-amp 🙁 - except for snob value I cannot think why. The specified 10µF output coupling cap is already quite large; it would depend on the internal circuitry and the main amp input impedance whether such a value has merit. Again a schematic would be required to answer that. This seems like a simple single stage amplifier.
But you have it, so enjoy it, and have no further concern regarding d.c. coupling. Use the a.c. output, or d.c. in series with a smaller cap, depending on your power amplifier input impedance. (E.g. if that impedance is say 47K, you can get away with a 330nF polyester capacitor; the 10µF is likely to be an electrolytic, not desirable.)
Firstly, let us not mix things up. The d.c. that is bad for the loudspeaker refers to the output of an amplifier - a "so-called" d.c. input has nothing to do with this (in a manner of speaking).
To come to d.c. coupled amps. If I may be straight: It is an overrated term bringing more grief than advantages - actually, I cannot think of a single advantage. One listens to signals in the audio band: Nominally 20 Hz - 20kHz. These days with ultra (another overrated word) low frequencies eminating from certain DVDs the low frequency margin may have to be lower. (That depends on your taste and is a different subject.)
I belong to the group that justifiably prefers everything outside the audible band to be gotten rid of. (Again that is a separate subject, concerning an amplifier's ability or rather lack of it to deal with signals outside the band it was designed for.) A d.c. coupled pre-amp 🙁 - except for snob value I cannot think why. The specified 10µF output coupling cap is already quite large; it would depend on the internal circuitry and the main amp input impedance whether such a value has merit. Again a schematic would be required to answer that. This seems like a simple single stage amplifier.
But you have it, so enjoy it, and have no further concern regarding d.c. coupling. Use the a.c. output, or d.c. in series with a smaller cap, depending on your power amplifier input impedance. (E.g. if that impedance is say 47K, you can get away with a 330nF polyester capacitor; the 10µF is likely to be an electrolytic, not desirable.)
Using the AC or DC coupled output !
The AC coupled output should have a high value resistor (say 470K to 1meg) to ground on the output side of the cap. That ground references the output (otherwise with it open circuit and unconnected you would still measure 20 volts DC there).
Which output to use. If the following stage (such as a power amp) has a non polarised cap at its input then using the DC coupled output eliminates the 10uf cap from the signal chain. You must ensure that the following cap is non polarised or if an electrolytic, that its polarised correctly. The AC output is presumed safe for any connection (but it should have the resistor fitted that I mentioned earlier). That resistor both ensures that no damaging pops and bangs are produced as the preamp is connected. 20 volts charged on a 10uf cap could conceivably permanently damage or degrade input devices on the following stage (such as degrading the noise figure) if connected "live". It also equalises (or more correctly ground references) the possibility of having two caps in series and one being an electroylitic and developing a reverse bias.
The AC coupled output should have a high value resistor (say 470K to 1meg) to ground on the output side of the cap. That ground references the output (otherwise with it open circuit and unconnected you would still measure 20 volts DC there).
Which output to use. If the following stage (such as a power amp) has a non polarised cap at its input then using the DC coupled output eliminates the 10uf cap from the signal chain. You must ensure that the following cap is non polarised or if an electrolytic, that its polarised correctly. The AC output is presumed safe for any connection (but it should have the resistor fitted that I mentioned earlier). That resistor both ensures that no damaging pops and bangs are produced as the preamp is connected. 20 volts charged on a 10uf cap could conceivably permanently damage or degrade input devices on the following stage (such as degrading the noise figure) if connected "live". It also equalises (or more correctly ground references) the possibility of having two caps in series and one being an electroylitic and developing a reverse bias.
I find it hard to believe that such an irresponsible "feature" can actually be sold to the general public.
The combination of the wrong output and a dc coupled power amp will without any doubt squeeze out all the smoke from the bass drivers coils. Seriously ugly.
The designers probably could not afford a decent coupling cap within the price bracket of the unit and this is the "solution" they came up with.
The "feature" is unusual.
The feature is potentially damaging and not user proof.
One MUST understand what has been offered and recognise which output can be used with other equipment.
The feature is potentially damaging and not user proof.
One MUST understand what has been offered and recognise which output can be used with other equipment.
I have a preamp with a 10uF output cap.
The preamp is described by the manufacturer as having no electrolytics anywhere in it.
I opened it and checked. Even the PSU is built up solely from plastic film capacitors.
That 10uF was criticised by some HiFi reveiwers because it potentially could reduce the bass frequency response into some equipment.
Since all my equipment is Rin>=47k//RF capacitance, I was not concerned.
Equipment with Rin<<47k may result in a little loss of extreme low bass content.
I guess that was why
has the option to bypass.
Don't use the DC coupled output of your preamp on a power amp until you have the power amp schematic diagram, understand it, and know there is no problem. DC on input voltage of a power amp won't usually cause output DC, but not all power amps have an input capacitor. My Dynakit ST70 tube amp has a DC coupled input and would likely produce no sound with 20 VDC on the input.
Sounds like most of us are on the same page...There is no coupling capacitor on the DC output which means since it comes from the drain of the MOSFET, that signal has everything in it: the good, bad, and downright ugly. The manual states that if the DC output is used, the amplifier must be designed to have an input coupling capacitor. It recommends the companion SET-II amplifier that can be operated with DC coupling.
The manual also states the frequency response as DC - 85KHz. If I am putting everything you all are saying together with what I can gather in the manual, the DC output--by bypassing the 10uF cap--allows the listener (if he or she is a blue whale or a canary) to hear frequencies all the way from DC (0Hz) to 85 KHz. If that is the case, I wonder what it would sound like if we could actually hear all that stuff?
At any rate, it sounds like I might be able to mod that DC output to block DC, but still expand the frequency range beyond 20Hz - 20KHz. It would probably be a waste of time, but it is tempting. Does anyone know a formula (Johan?) I could use in which I'd plug in the impedance of an amp, frequency response, etc. and get a cap value that would protect the amp against DC?
The manual also states the frequency response as DC - 85KHz. If I am putting everything you all are saying together with what I can gather in the manual, the DC output--by bypassing the 10uF cap--allows the listener (if he or she is a blue whale or a canary) to hear frequencies all the way from DC (0Hz) to 85 KHz. If that is the case, I wonder what it would sound like if we could actually hear all that stuff?
At any rate, it sounds like I might be able to mod that DC output to block DC, but still expand the frequency range beyond 20Hz - 20KHz. It would probably be a waste of time, but it is tempting. Does anyone know a formula (Johan?) I could use in which I'd plug in the impedance of an amp, frequency response, etc. and get a cap value that would protect the amp against DC?
What is practical is easily determined. 10 uf output cap, about $.10. 100 uf cap, about $.30. 1000 uf cap, about $1, and unless a very high quality 1000 uf cap is bought, it might leak enough DC current to cause bias problems. The bigger the coupling capacitance, the lower the crossover frequency. For the actual formula, look up vector capacitance of the parallel RLC circuit formula in a textbook or wikipedia. Personally having both amps with no caps in the flow, and amps with caps in the sound flow, that sound about the same, I am not worried about one 90 degree phase shift.
Fmjunkie,
The formula connecting frequency, capacitance and impedance is
f = 1/(2*pi*frequency*capacitance*impedance)
from which I derived the handy 159/f*R*C =1
Thus whatever one wants, one takes 159 and divide by the other two. (For R in K.ohm, f = Hz and C = µF) The 159 is actually 159155... but for practical purposes one stops at 159
This gives the point where response is 3dB different - 3dB down in the case of a coupling capacitor and next-stage input impedance. Thus, if one is looking for a response 3 dB down at 10 Hz feeding into an impedance of 100K, one finds the cap to be 0,159µF or nearest.
BUT:
That R is the total impedance in the circuit, i.e. the input impedance of the fed stage plus that of the feeding stage. If as said the output is from the drain of the MOSFET, that feeding impedance will not be zero. I am not sure what it might be, but one can count on several K.ohm (in that sense feeding through a 10µF cap is rather nonsense - but never mind ....)
A rather long-winded explanation, but I hope to have made it clear. As said by others, you can find analyses of R.C circuits and such on the internet for further education.
The formula connecting frequency, capacitance and impedance is
f = 1/(2*pi*frequency*capacitance*impedance)
from which I derived the handy 159/f*R*C =1
Thus whatever one wants, one takes 159 and divide by the other two. (For R in K.ohm, f = Hz and C = µF) The 159 is actually 159155... but for practical purposes one stops at 159
This gives the point where response is 3dB different - 3dB down in the case of a coupling capacitor and next-stage input impedance. Thus, if one is looking for a response 3 dB down at 10 Hz feeding into an impedance of 100K, one finds the cap to be 0,159µF or nearest.
BUT:
That R is the total impedance in the circuit, i.e. the input impedance of the fed stage plus that of the feeding stage. If as said the output is from the drain of the MOSFET, that feeding impedance will not be zero. I am not sure what it might be, but one can count on several K.ohm (in that sense feeding through a 10µF cap is rather nonsense - but never mind ....)
A rather long-winded explanation, but I hope to have made it clear. As said by others, you can find analyses of R.C circuits and such on the internet for further education.
Thanks guys. I might give it a shot if I have some free time, but I may simply leave the DC output as is and find a nice cap for the AC. My time and money will probably be better spent there, especially if as indianajo says, "his amps sound about the same".
Tell us the impedance you want to work into and at what frequency and we can tell you what value cap is needed.
A couple of examples,
Into 10k and response -3db down at 1Hz needs 16uf. 600ohm and 1Hz needs 265uf and 50k and 1Hz needs 3uf
You get the idea 🙂
I agree fmjunkie; it is as simple as putting a say 390 - 470nF in series, and devote your attention to more worthwhile matters. (As said before, only also a say 470K resistor over the output to keep it at zero potential at switch-on and avoid charge-up through the main amplifier, depending.)
- Status
- Not open for further replies.