So I have been seeing questions about impedance of a speaker when it comes to tube amps, mainly if the impedance of a cabinet when not matched to an amp of the same impedance, will it damage the amp. Another question I commonly see is if you need to half the impedance on the amp's selector switch if you pull out two of the tubes. Well I plan on explaining everything.
For some basic electronics theory. All semiconductors or amplifiers need to have a load on them. A load is resistance that limits maximum current and controls gain. With power amps, the load is the speaker, however most speakers have a very low impedance. If you were to look at the load required for a 6L6 for a typical 120 Watt amp, would be in the ballpark of 1042 ohms from plate to center tap. There is a slight incompatibility as described below:
Tube: High Voltage, Low Current
Speaker: Low Voltage, High Current
Since high current would destroy a tube and high voltage would destroy a speaker, a transformer is used to match the signals. A transformer can convert voltage to current, and current to voltage depending on the setup. A step down transformer takes the high voltage of the vacuum tubes from about 350v down to about 44v. The stepdown ratio is thus in the ballpark of an 8:1 stepdown. For every 8v on the input, you get 1V at the output. For current you get the opposite, for every 1mA available at the input, you get 8mA available at the output. As a result, the load that is placed on the transformers output will have a multiplication effect on the load.
If you have a 16 ohm speaker plugged into a the output of the transformer then to the input of the transformer will be 1042 ohms. So to two 6L6 tubes in parallel, when the amp is operating in class B, the speaker actually looks like a 1042 ohm speaker and to the speaker it looks like the amp is pumping out about 44v out and 44v into a 16 ohm speaker is indeed about 120 Watts.
The impedance selection taps allow the user to plug into different cabinets. Each tap has a different winding ratio so that you can use different impedance cabinets but still maintain that 1042 ohms to a parallel set of tubes. Due to the winding ratio change when a different impedance is selected, the voltage decreases. This is good because with a lower impedance cab, you need less voltage. The lower voltage combined with the lower speaker impedance also causes an inverse increase in current, keeping the power dissipation to 120 Watts. Keep in mind the effective speaker load to the tubes will as a result be left unchanged.
So what happens when say you have an amp set to 8 or 4 ohms being plugged into a 16 ohm cabinet? It is simple, the load to the tubes looks larger and since the voltage to the tubes never changes, the current has fallen and so has power. An easy way to get less power out of your amp is simply to have the amp set on a lower impedance than the cabinet your are plugging it into. With a 16 ohm cabinet, the 8 ohm setting on the amp you get your half power, and the 4 ohm setting will get you quarter power. How much quieter is quarter power? Well in order to get half the perceived volume from your amp you need to reduce the power by one-tenth, or 12 Watts. Even at quarter power, or 30 Watts, the amp will still be way loud. The benefit to this though is that you achieve power amp distortion at much lower volumes. As a rule of thumb the cabinets total load should not be more than 4 times higher than the impedance the amp is set for, so for the different impedance settings, don't go any higher than this:
Amp -> Speaker
4 ohms -> 16 ohms
8 ohms -> 32 ohms
16 ohms -> 64 ohms
Anything higher can cause the inductance kickback voltage to exceed the winding enamel's breakdown voltage and cause destruction to the power transformer and power tubes. This is the same failure when an amp is played with no load or rather an "infinite load".
What happens when you go the opposite direction? While it is not advised to do so as the amp will over dissipate, due to the pentode's constant current characteristic, which limits the total amount of current regardless of voltage, the increase in power is not linear like the decrease of power when you use a larger load. So if you had the amp on a 16 ohm setting and plugged in an 8 ohm cabinet, the amount of power increase will only be slightly higher than if you used a 16 ohm cabinet. Keep in mind thought that this will cause a shortened life of the tube and transformer and most likely the transformer will be over saturated which will cause overheating which mitigates the working life of all transformers. Not only that, but when you over dissipate the tubes, they actually loose gain due to their load line characteristics, so you will notice a drop in output volume as the efficiency of the amp drops dramatically. More and more of that power is just wasted as heat and not sound.
What about pulling tubes, will that effectively reduce power? Not a chance. While if you do pull a tube from each side will cause a very drop in total power due to tube resistance and tube current limiting, the drop is too small to notice and over dissipates the two tubes left in the amp. Even with two tubes pulled, since the load of the transformer has not changed and remember it is the load that determines how much power is pulled, the two tubes will be trying to pull 120 Watts, remembering though that the tubes do have a current limiter characteristic, there will be a decent power drop, but the amp will still be dissipating over 30 Watts per tube reducing tube life. Transformer life is not reduced in this scenario because it won't be over dissipated, just the tubes will be.
So what is the moral of the story? Within reason an amp can have an impedance mismatched, even if it means the amp and tubes are being over dissipated. While the amp or tubes won't blow up, the life of the tubes and possibly the transformer will be reduced, so users be warned. If less power and a tighter amp is desired, pulling two opposing tubes and setting the amp to 4 ohms when using a 16 ohm cab will turn a 120 Watt beast into a two 6L6 amp 30 Watt little jam amp.
Until next time!
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Tuesday, April 23, 2013
Saturday, March 2, 2013
Headroom and dB explained
There have been questions floating around the internet asking about tracking levels, what the meter readings (in dB) mean among other things. The internet is a place full of this information but it is very scattered and incomplete so I figured I would set some things in line.
For starters we have to talk about what a decibel is and what they mean in relation to a signal. A decibel (dB) is a unit of comparing a value to another value, in our case, a voltage reference compared so a signal. When we talk about the signal strength in line level audio which is used in professional and consumer based electronics, the input or output signals are not measured in volts, but with a system that allows you to compare the nominal, or average signal amplitude to other systems as well that may not be directly compatible, as such it allows for quick comparison of ratios and signal strength that may be difficult when analyzing voltage or power alone. As the decibel is a value comparison to another value the question becomes, "what is the reference for line level dB"?
The unit is noted as dBu, or "decibel unloaded" and its reference point is the voltage required to dissipate 1mW into a 600 ohm load, or in other terms, about 0.7746v. The equation for calculating decibels is as follows:
dB=20[log(Vo/Vi)]
Where:
Vi = Input or reference votlage
Vo = Output voltage or signal voltage
This means that 0dB is the same as the reference voltage. If the input signal is less than 0.7746V or 0dBu, then the decibel value will be negative and if the input signal is greater than 0.7746v or 0dBu, then the decibel value will be positive. Negative means signal loss, positive means signal gain, zero means no change in signal magnitude. If you have a gain of 2 for example (put 1v in and you get 2v out) you have a gain of roughly +6dB. If you had a gain of 10 as another example, you would have a gain of 20dB.
Most balanced line inputs have a nominal rating of +4dBu which roughly represents 1.228v. Nominal means that is the target signal amplitude that you would be aiming for that would be considered "proper tracking levels". A balanced input means there is a hot, neutral and ground connection, used to reduce noise over the cable run and increase gain. If you connect an unbalanced signal to the input, you will get -6dB of gain reduction as we have an output signal that is half the original signal, the equation 20[log(1/2)]=-6. External noise can also increase by as much as 60dB. The 60dB figure comes from the fact that the common mode rejection, a fancy name for saying how much a noise signal can be attenuated, on average attenuates noise signals up to 60dB and higher with better audio equipment. Without a balanced connection that same noise level is unimpeded.
Now how does this nominal 4dB of the line inputs interface have anything to do with the dB scale in your DAW. Well the dB scale in a daw is a negative range, this is because its 0dB reference is the maximum headroom of the system before clipping. All Analog to Digital Converters, while having a nominal value. Like mentioned before this nominal value is the average signal strength that the manufacture deems you should track your levels at. As a result, all ADCs have much more headroom than that and will typically peak and clip any aditional signal after 24dBu. If you have an ADC that has a maximum input level of 24dBu, a signal peaking at 24dBu will register as peaking 0dBFS in the DAW. In other words the DAW is telling you that the ADC is maxed out and any additional signal will cause is to clip.
"Decibel fullscale", or dBFS is simply the name given to digital equivalents to actual analog signals, and this is the measuring unit in all DAWs. This means that 0dBFS is equal to the maximum headroom level of the audio interface, the line input and in most cases, the line output as well.
Now how does this relate to a line out? Professional line out devices used in audio recording all the way down to small USB interfaces have a range usually in the 12dBu to 30dBu output before clipping. The bigger professional DACs are all 24dB. Lets say that the line output is rated for a maximum output of 24dBu, if a signal that is peaking 0dBFS in the DAW, it will come out as 24dBu. These are extremes though because again the proper tracking levels are listed as the nominal level designated by the manufacture which for compatibility, professional line level is nominal 4dBu as mentioned. This means that if you track your lvels nominal for 4dBu, never raise the faders in the DAW over 0dB, then the line outputs will come out nominal 4dBu as well. The signal will be less than 4dBu if the fader for that track in the DAW was turned down as is typical in the mixing process. There is a lot of headroom left over as a result. If we have ADCs and DACs that have a maximum rating of 24dBu and each track is at most peaking at 4dBu nominal, that means we have 20dB of headroom. The reason this is done is that in Digital to Analog Converters (DACs), the reconstruction circuitry and associated amplifiers become less linear or pick up more Total Harmonic Distotrtion (THD) the closer the signal from the DAW reaches 0dBFS. To keep the signal clean and free of coloration from the ADC and DAC process, the accosiated amplifiers in the converters must have as much headroom as possible..
This is where we get into consoles, their headroom and VU meters. A VU is the same as a dBu in terms of the reference voltage but with an analog console 0VU is the same as 4dBu. In a console you can see up to 24dBu of headroom which leaves you a good 20dB of headroom left over after it receives a 4dBu (or 0VU) signal. Obviously you need this headroom when you sum all the tracks together, you shouldn't be anywhere close to distorting the master bus. If you are working "In the Box" and want to give yourself as much headroom as you would with an analog console, you would take the average amount of headroom you would have left with that analog gear and subtract that from your DAW or, take that 18dB to 20dB out from the 0dBFS. This is why 0VU is commonly referred to as being the same as -18dBFS in the DAW. This value is by no means scientific, as 0dBFS is always the maximum output of the ADC and DAC being used and in less than professional equipment ranges quite dramaticlly. The -18dBFS peak of the tracks in the DAW is usually a safe bet and a good starting point. As always depending on what you are using you will have to adjust your track levels and the input levels on your analog gear in some cases.
The way that the ADC, DAC and console setup in signal amplitude would be as such: if the audio was tracked peaking at 4dBu on a nominal 4dBu ADC, it would register in the DAW peaking at -18dBu. If the DAC has an output of 24dBu that was sent to a track in the console, the console will register a track signal peaking 0VU and when all the tracks are summed together, should not exceed the 24dBu headroom of the master bus. Our tacking levels are happy, the Fullscalse values are low enough to keep THD of the DAC low while the signal strength to the console is adequate. We have a happy system.
There is one unfortunate side to this though, if you are tracking in 16-bit, peaking tracks at -18dBFS is quite low and they will lack dynamic and have an increased noise floor as well as possibly having a somewhat digital sound. Some digital plugins and amp sims may not have enough signal to function properly as well. To remedy this, tracking in 24-bit ensures that you have more than plenty of headroom, resolution and noise. A 16-bit track has only 65,535 different points representing the signal where a 24-bit track 16,777,215 different points representing the signal. Of course that increase of resolution will not only give you enough headroom to burn, it will be overkill and actually provide you with punchier mixes even if you export to 16-bit as the final format.
If you are using some plugins that still aren't working correctly with the tracks peaking at -18dBFS, then using the preFX trim control in your DAW's channel strip will allow you to adjust as needed. It can also be used to bring down tracks that are too hot and clip or cause malfunction to certain plugins, specifically analog simulated plugins.
Even if you are not going to use analog gear and will stay completely "In the Box", having a good habit of tracking levels is always wise, especially if you use analog simulated plugins which impress that equivalent headroom limit of the real analog gear, basing its headroom in the real world on 0dBFS. If you are using an console simulator, then to get it to color the sound as if you were really pushing that console track to about 0VU, you need to keep the signal in your DAW peaking around -18bBFS and never peak higher than -14dBFS which is the same as 4VU.
Until next time!
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
For starters we have to talk about what a decibel is and what they mean in relation to a signal. A decibel (dB) is a unit of comparing a value to another value, in our case, a voltage reference compared so a signal. When we talk about the signal strength in line level audio which is used in professional and consumer based electronics, the input or output signals are not measured in volts, but with a system that allows you to compare the nominal, or average signal amplitude to other systems as well that may not be directly compatible, as such it allows for quick comparison of ratios and signal strength that may be difficult when analyzing voltage or power alone. As the decibel is a value comparison to another value the question becomes, "what is the reference for line level dB"?
The unit is noted as dBu, or "decibel unloaded" and its reference point is the voltage required to dissipate 1mW into a 600 ohm load, or in other terms, about 0.7746v. The equation for calculating decibels is as follows:
dB=20[log(Vo/Vi)]
Where:
Vi = Input or reference votlage
Vo = Output voltage or signal voltage
This means that 0dB is the same as the reference voltage. If the input signal is less than 0.7746V or 0dBu, then the decibel value will be negative and if the input signal is greater than 0.7746v or 0dBu, then the decibel value will be positive. Negative means signal loss, positive means signal gain, zero means no change in signal magnitude. If you have a gain of 2 for example (put 1v in and you get 2v out) you have a gain of roughly +6dB. If you had a gain of 10 as another example, you would have a gain of 20dB.
Most balanced line inputs have a nominal rating of +4dBu which roughly represents 1.228v. Nominal means that is the target signal amplitude that you would be aiming for that would be considered "proper tracking levels". A balanced input means there is a hot, neutral and ground connection, used to reduce noise over the cable run and increase gain. If you connect an unbalanced signal to the input, you will get -6dB of gain reduction as we have an output signal that is half the original signal, the equation 20[log(1/2)]=-6. External noise can also increase by as much as 60dB. The 60dB figure comes from the fact that the common mode rejection, a fancy name for saying how much a noise signal can be attenuated, on average attenuates noise signals up to 60dB and higher with better audio equipment. Without a balanced connection that same noise level is unimpeded.
Now how does this nominal 4dB of the line inputs interface have anything to do with the dB scale in your DAW. Well the dB scale in a daw is a negative range, this is because its 0dB reference is the maximum headroom of the system before clipping. All Analog to Digital Converters, while having a nominal value. Like mentioned before this nominal value is the average signal strength that the manufacture deems you should track your levels at. As a result, all ADCs have much more headroom than that and will typically peak and clip any aditional signal after 24dBu. If you have an ADC that has a maximum input level of 24dBu, a signal peaking at 24dBu will register as peaking 0dBFS in the DAW. In other words the DAW is telling you that the ADC is maxed out and any additional signal will cause is to clip.
"Decibel fullscale", or dBFS is simply the name given to digital equivalents to actual analog signals, and this is the measuring unit in all DAWs. This means that 0dBFS is equal to the maximum headroom level of the audio interface, the line input and in most cases, the line output as well.
Now how does this relate to a line out? Professional line out devices used in audio recording all the way down to small USB interfaces have a range usually in the 12dBu to 30dBu output before clipping. The bigger professional DACs are all 24dB. Lets say that the line output is rated for a maximum output of 24dBu, if a signal that is peaking 0dBFS in the DAW, it will come out as 24dBu. These are extremes though because again the proper tracking levels are listed as the nominal level designated by the manufacture which for compatibility, professional line level is nominal 4dBu as mentioned. This means that if you track your lvels nominal for 4dBu, never raise the faders in the DAW over 0dB, then the line outputs will come out nominal 4dBu as well. The signal will be less than 4dBu if the fader for that track in the DAW was turned down as is typical in the mixing process. There is a lot of headroom left over as a result. If we have ADCs and DACs that have a maximum rating of 24dBu and each track is at most peaking at 4dBu nominal, that means we have 20dB of headroom. The reason this is done is that in Digital to Analog Converters (DACs), the reconstruction circuitry and associated amplifiers become less linear or pick up more Total Harmonic Distotrtion (THD) the closer the signal from the DAW reaches 0dBFS. To keep the signal clean and free of coloration from the ADC and DAC process, the accosiated amplifiers in the converters must have as much headroom as possible..
This is where we get into consoles, their headroom and VU meters. A VU is the same as a dBu in terms of the reference voltage but with an analog console 0VU is the same as 4dBu. In a console you can see up to 24dBu of headroom which leaves you a good 20dB of headroom left over after it receives a 4dBu (or 0VU) signal. Obviously you need this headroom when you sum all the tracks together, you shouldn't be anywhere close to distorting the master bus. If you are working "In the Box" and want to give yourself as much headroom as you would with an analog console, you would take the average amount of headroom you would have left with that analog gear and subtract that from your DAW or, take that 18dB to 20dB out from the 0dBFS. This is why 0VU is commonly referred to as being the same as -18dBFS in the DAW. This value is by no means scientific, as 0dBFS is always the maximum output of the ADC and DAC being used and in less than professional equipment ranges quite dramaticlly. The -18dBFS peak of the tracks in the DAW is usually a safe bet and a good starting point. As always depending on what you are using you will have to adjust your track levels and the input levels on your analog gear in some cases.
The way that the ADC, DAC and console setup in signal amplitude would be as such: if the audio was tracked peaking at 4dBu on a nominal 4dBu ADC, it would register in the DAW peaking at -18dBu. If the DAC has an output of 24dBu that was sent to a track in the console, the console will register a track signal peaking 0VU and when all the tracks are summed together, should not exceed the 24dBu headroom of the master bus. Our tacking levels are happy, the Fullscalse values are low enough to keep THD of the DAC low while the signal strength to the console is adequate. We have a happy system.
There is one unfortunate side to this though, if you are tracking in 16-bit, peaking tracks at -18dBFS is quite low and they will lack dynamic and have an increased noise floor as well as possibly having a somewhat digital sound. Some digital plugins and amp sims may not have enough signal to function properly as well. To remedy this, tracking in 24-bit ensures that you have more than plenty of headroom, resolution and noise. A 16-bit track has only 65,535 different points representing the signal where a 24-bit track 16,777,215 different points representing the signal. Of course that increase of resolution will not only give you enough headroom to burn, it will be overkill and actually provide you with punchier mixes even if you export to 16-bit as the final format.
If you are using some plugins that still aren't working correctly with the tracks peaking at -18dBFS, then using the preFX trim control in your DAW's channel strip will allow you to adjust as needed. It can also be used to bring down tracks that are too hot and clip or cause malfunction to certain plugins, specifically analog simulated plugins.
Even if you are not going to use analog gear and will stay completely "In the Box", having a good habit of tracking levels is always wise, especially if you use analog simulated plugins which impress that equivalent headroom limit of the real analog gear, basing its headroom in the real world on 0dBFS. If you are using an console simulator, then to get it to color the sound as if you were really pushing that console track to about 0VU, you need to keep the signal in your DAW peaking around -18bBFS and never peak higher than -14dBFS which is the same as 4VU.
Until next time!
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Saturday, January 19, 2013
Tranformerless Active DI and Reamp circuits
A couple of days ago I talked about the myths behind active vs passive circuits. I also said that part two would be out to talk about the newest Corvus Audio Product, the DI+, a transformeless simultaneous active DI and Reamp box.
The transformer is a great device, it can isolate grounds, step up or step down voltages, can change the effect of a load on a circuit. For many years, the passive DI setup was the most common way to manufacture DI and Reamp Devices. They are simple, they only have one component, connected to jacks and a ground lift. The unfortunate effect of these passive DI units is that the voltage conversion is intended for going into a mic preamp. There are a few problems with this.
-The voltage is stepped down to mic level, only to be amped again throught the preamp, increasing noise
-The mic preamp has a low input impedance
-Mic preamps are rarely color or distortion free
-The final input impedance to the guitar is still relatively low
-The transformer has a high distortion at low frequencies
The big problem is that the transformer overcompensates the voltage drop needed to get the pickup to line level, it has to do this because the larger the turns ratio is, the more the original load's effect load to the pickup is magnified, meaning, if you have a small load, in order to get that load to look much larger to the guitar pickup, you need a large ratio on the transformer. Since the average of 28K to 47K load of a line level device is still to low for the modest small ratio transformer (the guitar pcikups load would still be too low) a larger reduction in voltage is needed to get the signal voltage low enough to be capable of going into a mic preamp. A mic preamp's impedance is even lower, usually in the 1K region, however due to the large step down of the transformer, the effective load to the pickup is usually around 470K. This is still way to low as most guitar amps have a load of 1M.
There is another problem with transformers in general. From one winding to another, they act like a capacitor, they do not let low frequencies through. Unfortunately, storing a low frequency charge requires a large transformer, much larger than what it in a typical DI device. The result is that the cutoff point for the high pass filter is in the audible range and does effect the perceived sound.
One huge problem with passive DIs is that with a mic preamp load they tend to only load the pickup of about 470K, when the parallel out is being used to monitor through the amp, the amp's impedance is now in parallel with the transformer load to the mic preamp. 470K in parallel with 1M in parallel with the volume pot, lets say 500K for a passive pickup is 195K. 195K with the DI versus the 333K if it went straight into the amp. Try this, plug your DI in to a mic pre, turn it on and plug it into your interface, plug the parallel out into an amp and plug your guitar into the DI, play through it, then remove the XLR cable from the DI, removing the transformer from the mic pre, the tone coming out of the amp will come to life as long as the DI is not plugged into the mic pre. Even a small load change can have a large effect on the tone. It is also effect on the tone going to the mic pre as well and consequently, the tone you are recording
To combat the load problem of passive DI's some companies began loading the transformer with an active buffer. The winding ratio would remain the same so the stepdown voltage would remain the same. The benefit of adding a DC power supply and active components was that the load to the guitar would be the same when the amp was plugged into the parallel out, the loat would raise to that as to what was typical without the load. There still remains the problem though of using the transformer, as it is still producing a bandpass filter that is causing distortions at both high and low frequencies.
So why are transformers used even in active devices? The designers want an easy way to electronically isolate the signal. That or they like keeping the people ignorant to the fact that transformers are not the only way you can isolate grounds. An easy way to isolate ground is by not having a ground connected to a jack if you know that jack will always be in a setup where there will be an earth ground connected elsewhere. You only need one earth ground connection to an active device, no other grounds are needed, the shielding of the cable is actually supposed to be left open on one side and not connected to ground on two ends. The only reason guitars are connected to both ends is that the guitar itself does not have an earth ground connection and gets it from the amp. In case of the DI, we know that we will always be using the DI output so if we permanently remove the ground of the parallel out, the guitar amp is not grounding to the DI as well as the computer/interface/mic pre. This removes the ground loop and is good because the ground shielding from the parallel out to the guitar amp is only connected to the guitar amp, any noise that is present across the signal gets grounded to the amp, the amp is not earthed to another device, and the guitar itself has an earth connection through the recording device connected to the DI.
Still to this day all manufactures are connecting all the jacks to ground and using transformers to isolate the grounds. This reduces linearity of the signal and increases the cost of the DI dramatically. All this just because the designer wants to have everything connected to ground, even when the best option is to have everything not isolated and connected to only one earth. Then you can remove the transformer and get the benefits of the linearity of a signal that has no direct capacitance or inductance.
We can further increase the design by dealing with the noise level and coloration of going through a mic preamp. Since the guitar pickup's level is only slightly higher than line level, if we use an active buffer, a volume control and an unbalanced to balanced converter, we can attenuate the exact level needed for that pickup to the line level desired, we removed the massive attenuation to mic pre level and we have removed the need for the colorful mic pre. We now have a line level DI output that can amplify all the way to DC, ultra linear, virtually no distortion. But we are not done.
I said before that 9v supplies cannot correctly power guitar pickups. To combat this a pump charge converter wired in a basic configuration or even a voltage multiplier setup, can provide the output voltage necessary from a 9V supply. All Corvus Audio products that run on a 9v adapter can supply a clean output voltage of 17Vpp, meaning with the DI, no pickup would ever distort the device, and the line input coming from the DI would distort way before with the volume on maximum.
Lets go ahead and take a look at the Corvus DI+
The benefit of this DI is that it allows you to use it as a DI and Reamp at the same time. There are separate connections for the DI and the reamp, unlike some devices that can do both. With other devices, if you want to go from DI to reamp, you have to switch cables or switches/settings on the device every time. The Corvus DI has two separate circuits, one for the DI and one for the Reamp, which outputs are mixed together using a mixer buffer that goes to the amp output. There is no parallel output, you have an amp output that is a mix of the buffered raw guitar and the signal from the reamp. With a simple send routing setup, you can record your DI monitoring through your amp. You can turn the send in your DAW on and press play, and the DI track from your DAW will instantly come out of your amp, without needing to touch your DI box.
There is one ground lift switch that is to be engaged when you are using both the DI and reamp portions at the same time, if you are using just the DI or just the reamp, the ground lift should be off.
Here is a little block diagram of the internals of the DI, as you can see the guitar buffer and the Reamp outputs are being mixed together to one amp out. To top it off there are TWO amp outputs, yes the DI+ has a ground separated Y splitter for running two amps at the same time. The DI+ can be used as a standalone active buffered Y splitter so you can record, practice or even gig live with two amps at the same time, without needing to use the reamp or the DI features.
The DI output is active unbalanced to balanced using the standard TRS cable, no more XLR. The reamp input is also active balanced to unbalanced using the TRS cable. No more dealing with mic preamps, except for the guitar input, all inputs and outputs of the DI are line level, ready to go direct into your converters and outboard gear.
This device will most likely cost $160 USD including shipping to the US and $180 USD including shipping to the EU. They may actually be cheaper than that when they are available to buy within the week.
Until next time.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
The transformer is a great device, it can isolate grounds, step up or step down voltages, can change the effect of a load on a circuit. For many years, the passive DI setup was the most common way to manufacture DI and Reamp Devices. They are simple, they only have one component, connected to jacks and a ground lift. The unfortunate effect of these passive DI units is that the voltage conversion is intended for going into a mic preamp. There are a few problems with this.
-The voltage is stepped down to mic level, only to be amped again throught the preamp, increasing noise
-The mic preamp has a low input impedance
-Mic preamps are rarely color or distortion free
-The final input impedance to the guitar is still relatively low
-The transformer has a high distortion at low frequencies
The big problem is that the transformer overcompensates the voltage drop needed to get the pickup to line level, it has to do this because the larger the turns ratio is, the more the original load's effect load to the pickup is magnified, meaning, if you have a small load, in order to get that load to look much larger to the guitar pickup, you need a large ratio on the transformer. Since the average of 28K to 47K load of a line level device is still to low for the modest small ratio transformer (the guitar pcikups load would still be too low) a larger reduction in voltage is needed to get the signal voltage low enough to be capable of going into a mic preamp. A mic preamp's impedance is even lower, usually in the 1K region, however due to the large step down of the transformer, the effective load to the pickup is usually around 470K. This is still way to low as most guitar amps have a load of 1M.
There is another problem with transformers in general. From one winding to another, they act like a capacitor, they do not let low frequencies through. Unfortunately, storing a low frequency charge requires a large transformer, much larger than what it in a typical DI device. The result is that the cutoff point for the high pass filter is in the audible range and does effect the perceived sound.
One huge problem with passive DIs is that with a mic preamp load they tend to only load the pickup of about 470K, when the parallel out is being used to monitor through the amp, the amp's impedance is now in parallel with the transformer load to the mic preamp. 470K in parallel with 1M in parallel with the volume pot, lets say 500K for a passive pickup is 195K. 195K with the DI versus the 333K if it went straight into the amp. Try this, plug your DI in to a mic pre, turn it on and plug it into your interface, plug the parallel out into an amp and plug your guitar into the DI, play through it, then remove the XLR cable from the DI, removing the transformer from the mic pre, the tone coming out of the amp will come to life as long as the DI is not plugged into the mic pre. Even a small load change can have a large effect on the tone. It is also effect on the tone going to the mic pre as well and consequently, the tone you are recording
To combat the load problem of passive DI's some companies began loading the transformer with an active buffer. The winding ratio would remain the same so the stepdown voltage would remain the same. The benefit of adding a DC power supply and active components was that the load to the guitar would be the same when the amp was plugged into the parallel out, the loat would raise to that as to what was typical without the load. There still remains the problem though of using the transformer, as it is still producing a bandpass filter that is causing distortions at both high and low frequencies.
So why are transformers used even in active devices? The designers want an easy way to electronically isolate the signal. That or they like keeping the people ignorant to the fact that transformers are not the only way you can isolate grounds. An easy way to isolate ground is by not having a ground connected to a jack if you know that jack will always be in a setup where there will be an earth ground connected elsewhere. You only need one earth ground connection to an active device, no other grounds are needed, the shielding of the cable is actually supposed to be left open on one side and not connected to ground on two ends. The only reason guitars are connected to both ends is that the guitar itself does not have an earth ground connection and gets it from the amp. In case of the DI, we know that we will always be using the DI output so if we permanently remove the ground of the parallel out, the guitar amp is not grounding to the DI as well as the computer/interface/mic pre. This removes the ground loop and is good because the ground shielding from the parallel out to the guitar amp is only connected to the guitar amp, any noise that is present across the signal gets grounded to the amp, the amp is not earthed to another device, and the guitar itself has an earth connection through the recording device connected to the DI.
Still to this day all manufactures are connecting all the jacks to ground and using transformers to isolate the grounds. This reduces linearity of the signal and increases the cost of the DI dramatically. All this just because the designer wants to have everything connected to ground, even when the best option is to have everything not isolated and connected to only one earth. Then you can remove the transformer and get the benefits of the linearity of a signal that has no direct capacitance or inductance.
We can further increase the design by dealing with the noise level and coloration of going through a mic preamp. Since the guitar pickup's level is only slightly higher than line level, if we use an active buffer, a volume control and an unbalanced to balanced converter, we can attenuate the exact level needed for that pickup to the line level desired, we removed the massive attenuation to mic pre level and we have removed the need for the colorful mic pre. We now have a line level DI output that can amplify all the way to DC, ultra linear, virtually no distortion. But we are not done.
I said before that 9v supplies cannot correctly power guitar pickups. To combat this a pump charge converter wired in a basic configuration or even a voltage multiplier setup, can provide the output voltage necessary from a 9V supply. All Corvus Audio products that run on a 9v adapter can supply a clean output voltage of 17Vpp, meaning with the DI, no pickup would ever distort the device, and the line input coming from the DI would distort way before with the volume on maximum.
Lets go ahead and take a look at the Corvus DI+
There is one ground lift switch that is to be engaged when you are using both the DI and reamp portions at the same time, if you are using just the DI or just the reamp, the ground lift should be off.
Here is a little block diagram of the internals of the DI, as you can see the guitar buffer and the Reamp outputs are being mixed together to one amp out. To top it off there are TWO amp outputs, yes the DI+ has a ground separated Y splitter for running two amps at the same time. The DI+ can be used as a standalone active buffered Y splitter so you can record, practice or even gig live with two amps at the same time, without needing to use the reamp or the DI features.
This device will most likely cost $160 USD including shipping to the US and $180 USD including shipping to the EU. They may actually be cheaper than that when they are available to buy within the week.
Until next time.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Tuesday, January 15, 2013
Active vs Passive
There is a pretty interesting phenomenon that is still taking place in the world of modern music, musicians are paying large amounts of cash for passive components, usually because there is this thing going around the music world that makes musicians think that active components are bad. Even worse the typical op amp is regarded by audiophiles and choosy gear snobs as the most unmusical device there is on the planet. None of this can be further from the truth, but the reality is, a lot of companies are still using transformers in things they can use well designed op amps in so they can charge an arm and a leg and the buyer feels it is justified.
Same goes for active circuits in general, I hear it all the time, designers staying away from, even bragging about not using active circuitry, as if the active circuit was this evil thing that sucked all the life, vibe and character out of the tone. I recently heard this about a passive, foot switchable gain adjustment that was in pedal form, as in, it sits between your guitar and your amp. Now the designer mentioned about not having active circuits just pure raw tone, so let analyze with a little bit of electrical engineering math. Now a typical gain knob in an amp is 1M, so lets just say that this gain knob stompbox is indeed 1M. Now if we use that in conjunction with the guitar's volume pot is in parallel with the gain knob, is also in parallel with the amp's 1M input. At complete maximum, both the volume of the guitar and the gain control, you would have an input impedance to the pickup of only 166Kohms. Now the typical humbucker output impedance ranges from 12K to about 20K, meaning that this 166Kohm input impedance simply is not high enough.
So what happens to the sound. For starters the pickup cannot deliver the current being drawn from it by the load. With a very high quality low capacitance cable, with a high powered humbucker, the low pass filter being created is just above 20KHz with a 6m or 21ft cable. Obviously with a longer cable or cheaper cable, you are now attenuating high frequencies. Naturally with a guitar that has a 500K volume pot straight into an amp that has a 1M input impedance, you could use up to 60ft (~19m) of cable before you would even begin getting into audible attenuation of the high frequencies.
Now what would happen if a well designed op amp based buffer was used. The pickup would be seeing roughly 4-17Gohms, yes you heard right, 17 GigaOhms with a low noise FET input audio op amp. This is much higher than the guitar amp's impedance but the relative loading difference between the two in terms of how the pickup operates is <0.1%. The large plus is that the output impedance of the opamp is also much lower than the pickup, meaning it can deliver the current needed.
So for the most part why are active circuits hated by the tone snobs? Well here are a few myths and points made:
Noisy
More complex
Robs tone
Op amps distort badly
The big reason true bypass is still a big buzzword is because most pedal companies make horrible input buffers and suck the tone out of the guitar. Typically you won't see a pedal that has an input impedance higher than 470K and even that is relatively high. Figure the zones even less, then when you figure that the load is in parallel with the volume pot, you can get effective input impedance that are almost as low as the pickups output impedance (YIKES!!!). This does give the buffer a bad name and is one of the big reasons some really do hate active circuits.
"They are more noisy"
Sure, any semiconductor adds more noise and generally requires additional passive components which add their own noise as well. Semi conductor noise and passive component noise is amplified by the amplifiers being used, the more the gain, the worse the noise is amplified. One thing to keep in mind that if you have clean amplification, no clipping, the signal to noise ratio stays the same, the noise ratio only increases in clean circuits only when additional circuitry is added, both passive and active. Transistors and op amps of yesteryear may have been noisy, but today, with high grade FET input audio transistors and op amps, their noise level reaches that lower than even passive components such as resistors. As for as buffers are concerned, with a gain equal to or less than 1, the noise being added to the circuit are humanly impossibly to hear, even when put into a high gain amp that has a cumulative gain >6 Million.
"They are more complex"
Again another true, the more you add to a circuit the more that could go wrong. In most cases with a well designed active circuit, the benefits outweigh the very slight increase of failure probability.
"They rob tone"
Now this can be true with a bad design. Remember when I was talking about how impedance could effect frequency response? If too high a load is used on a semiconductor the small amounts of capacitance in the board or circuit could cause a high pass filter that could get into the audible range. Again this is bad and has a lot to do with horrible sounding active circuits. Also, poorly designed input buffers can just like the guitar pickup example have negative effects on the previous stage.
"Op amp distortion is awful sounding"
This comes down to again the design and the quality of the chip. There have been articles all over the place that talk about how "swapping op amps in your pedals can make the tone better is a myth" and for clean headroom based applications, including distortion pedals that use diode clipping this is completely true. However, there is a condition that happens with cheap op amps. For those who don't know an op amp is an amplifier that can amplify DC voltages hand have superior input and output impedance compared to a transistor. The op amp has two inputs, an inverting and non inverting, the output of the amp is the difference of the two inputs. One main signal input usually goes to the non inverting (+) and a feedback loop sends a small amount of the output signal to the other input, the inverting input (-). When the input signal (+) exceeds the supply voltage to the amp, the input voltage continues to increase, however, the output can only go as high as the input, it clips off. Now the two inputs are not "virtually" the same and the gain dramatically bucks around like an angry bull, swinging dramatically making all sorts of weird output signals as long as the input signal stays above the supply voltage. This is not a problem with more expensive audio op amps as they distort just like a transistor or vacuum tube will.
9v battery adapters and supplies aren't typically enough to power pedals as most guitar pickups, line level signals, mic preamp output signals etc can all exceed that by merely a few volts causing distortions, this is why EMG pickups distort slightly when ran on 9v and goes away with 18v. In all pedal deign, a good designer can implement a voltage multiplier which can turn a 9v adapter into just about any voltage desired. As a matter of fact all new Corvus Audio pedals that run off of 9v adapters can give a clean output of 17Vpp, meaning you will almost never distort the signal unless you were doing some crazy stuff that I wouldn't recommend, but even if you did, the op amps will distort just like a transistor, nice and musical, not as musical as a 12AX7, but still musical non the less.
It all comes down to good design, as good design with high quality components and forward thinking means that certain active designs can far surpass their passive counterparts.
In part two, we will be going over the design philosophy of the Corvus Audio DI+, the transformerless active simultaneous DI and Reamp unit.
Until next time.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Same goes for active circuits in general, I hear it all the time, designers staying away from, even bragging about not using active circuitry, as if the active circuit was this evil thing that sucked all the life, vibe and character out of the tone. I recently heard this about a passive, foot switchable gain adjustment that was in pedal form, as in, it sits between your guitar and your amp. Now the designer mentioned about not having active circuits just pure raw tone, so let analyze with a little bit of electrical engineering math. Now a typical gain knob in an amp is 1M, so lets just say that this gain knob stompbox is indeed 1M. Now if we use that in conjunction with the guitar's volume pot is in parallel with the gain knob, is also in parallel with the amp's 1M input. At complete maximum, both the volume of the guitar and the gain control, you would have an input impedance to the pickup of only 166Kohms. Now the typical humbucker output impedance ranges from 12K to about 20K, meaning that this 166Kohm input impedance simply is not high enough.
So what happens to the sound. For starters the pickup cannot deliver the current being drawn from it by the load. With a very high quality low capacitance cable, with a high powered humbucker, the low pass filter being created is just above 20KHz with a 6m or 21ft cable. Obviously with a longer cable or cheaper cable, you are now attenuating high frequencies. Naturally with a guitar that has a 500K volume pot straight into an amp that has a 1M input impedance, you could use up to 60ft (~19m) of cable before you would even begin getting into audible attenuation of the high frequencies.
Now what would happen if a well designed op amp based buffer was used. The pickup would be seeing roughly 4-17Gohms, yes you heard right, 17 GigaOhms with a low noise FET input audio op amp. This is much higher than the guitar amp's impedance but the relative loading difference between the two in terms of how the pickup operates is <0.1%. The large plus is that the output impedance of the opamp is also much lower than the pickup, meaning it can deliver the current needed.
So for the most part why are active circuits hated by the tone snobs? Well here are a few myths and points made:
Noisy
More complex
Robs tone
Op amps distort badly
The big reason true bypass is still a big buzzword is because most pedal companies make horrible input buffers and suck the tone out of the guitar. Typically you won't see a pedal that has an input impedance higher than 470K and even that is relatively high. Figure the zones even less, then when you figure that the load is in parallel with the volume pot, you can get effective input impedance that are almost as low as the pickups output impedance (YIKES!!!). This does give the buffer a bad name and is one of the big reasons some really do hate active circuits.
"They are more noisy"
Sure, any semiconductor adds more noise and generally requires additional passive components which add their own noise as well. Semi conductor noise and passive component noise is amplified by the amplifiers being used, the more the gain, the worse the noise is amplified. One thing to keep in mind that if you have clean amplification, no clipping, the signal to noise ratio stays the same, the noise ratio only increases in clean circuits only when additional circuitry is added, both passive and active. Transistors and op amps of yesteryear may have been noisy, but today, with high grade FET input audio transistors and op amps, their noise level reaches that lower than even passive components such as resistors. As for as buffers are concerned, with a gain equal to or less than 1, the noise being added to the circuit are humanly impossibly to hear, even when put into a high gain amp that has a cumulative gain >6 Million.
"They are more complex"
Again another true, the more you add to a circuit the more that could go wrong. In most cases with a well designed active circuit, the benefits outweigh the very slight increase of failure probability.
"They rob tone"
Now this can be true with a bad design. Remember when I was talking about how impedance could effect frequency response? If too high a load is used on a semiconductor the small amounts of capacitance in the board or circuit could cause a high pass filter that could get into the audible range. Again this is bad and has a lot to do with horrible sounding active circuits. Also, poorly designed input buffers can just like the guitar pickup example have negative effects on the previous stage.
"Op amp distortion is awful sounding"
This comes down to again the design and the quality of the chip. There have been articles all over the place that talk about how "swapping op amps in your pedals can make the tone better is a myth" and for clean headroom based applications, including distortion pedals that use diode clipping this is completely true. However, there is a condition that happens with cheap op amps. For those who don't know an op amp is an amplifier that can amplify DC voltages hand have superior input and output impedance compared to a transistor. The op amp has two inputs, an inverting and non inverting, the output of the amp is the difference of the two inputs. One main signal input usually goes to the non inverting (+) and a feedback loop sends a small amount of the output signal to the other input, the inverting input (-). When the input signal (+) exceeds the supply voltage to the amp, the input voltage continues to increase, however, the output can only go as high as the input, it clips off. Now the two inputs are not "virtually" the same and the gain dramatically bucks around like an angry bull, swinging dramatically making all sorts of weird output signals as long as the input signal stays above the supply voltage. This is not a problem with more expensive audio op amps as they distort just like a transistor or vacuum tube will.
9v battery adapters and supplies aren't typically enough to power pedals as most guitar pickups, line level signals, mic preamp output signals etc can all exceed that by merely a few volts causing distortions, this is why EMG pickups distort slightly when ran on 9v and goes away with 18v. In all pedal deign, a good designer can implement a voltage multiplier which can turn a 9v adapter into just about any voltage desired. As a matter of fact all new Corvus Audio pedals that run off of 9v adapters can give a clean output of 17Vpp, meaning you will almost never distort the signal unless you were doing some crazy stuff that I wouldn't recommend, but even if you did, the op amps will distort just like a transistor, nice and musical, not as musical as a 12AX7, but still musical non the less.
It all comes down to good design, as good design with high quality components and forward thinking means that certain active designs can far surpass their passive counterparts.
In part two, we will be going over the design philosophy of the Corvus Audio DI+, the transformerless active simultaneous DI and Reamp unit.
Until next time.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Monday, January 14, 2013
Black Box Theory
The Corvus Audio webpage has been up for some time. In that time I never thought to run a blog where I can give out informationals to dispel a lot of myths that go around in the music world as well as give advice, tips and tricks that I learned first hand the hard way. Over on the Andy Sneap, Ultimate Metal sub forum, it would seem too arrogant for me to post an informational about a topic that may have been discussed before to have a thread that was informational but overall trivial and left a discussion of electronic gurus from all levels of experience to fight out how they understood electronics.
I don't care to write books to make a profit, I always wanted to give the information that I had for free, mostly because my knowledge in the music world came to me at no cost and I have learned additional things a long the way. The best thing I realized I could do was to make a blog on my very website, hosted by Blogger of course, to allow me to give our such free information.
The first topic is a rather psychological one, one of the belief that visual representations can change perception of aural stimuli. When I first started building stompboxes this was never something I had in mind, but as time went on I developed this love for the Black Box. What I mean is, when I first started making stompboxes, I wanted to make them different colors, different graphics and sized to the electronic components. After hand painting my first two boxes white, I realized that it took too much time without proper equipment, in my case I wanted to power coat, but the company that I got the boxes from offered every enclosure in a textured flat black. When I first put everything together, I absolutely loved the look. The problem was that because black was used by a pedal, I couldn't use it again for something I was more proud of, something I felt black really deserved.
Then it hit me, use the same enclosure with the same power coat from the factory for all the pedals.
I realized that this was something different. To date, no other company has made stompboxes in only one color, let alone black and to make them the same size, with the same simplistic graphic means that the only way to tell one pedal from the other is by the the small labeling noting what the pedal is. The layouts for all pedals being the same, means they all look the same, a very simple plain, but still aesthetically pleasing appearance. The user won't be enticed to like or dislike the sound based on its appearance, certain colors effecting the way they perceive the sound. Down to the overly simplistic programmer or electrical engineering font, the only sense of attitude is from the Corvus Audio font.
The enclosures still have a vibe about them, something that does look good to see, but they have an emotionally neutral vibe, and as far as I am concerned, that is the way I like it.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
I don't care to write books to make a profit, I always wanted to give the information that I had for free, mostly because my knowledge in the music world came to me at no cost and I have learned additional things a long the way. The best thing I realized I could do was to make a blog on my very website, hosted by Blogger of course, to allow me to give our such free information.
The first topic is a rather psychological one, one of the belief that visual representations can change perception of aural stimuli. When I first started building stompboxes this was never something I had in mind, but as time went on I developed this love for the Black Box. What I mean is, when I first started making stompboxes, I wanted to make them different colors, different graphics and sized to the electronic components. After hand painting my first two boxes white, I realized that it took too much time without proper equipment, in my case I wanted to power coat, but the company that I got the boxes from offered every enclosure in a textured flat black. When I first put everything together, I absolutely loved the look. The problem was that because black was used by a pedal, I couldn't use it again for something I was more proud of, something I felt black really deserved.
Then it hit me, use the same enclosure with the same power coat from the factory for all the pedals.
I realized that this was something different. To date, no other company has made stompboxes in only one color, let alone black and to make them the same size, with the same simplistic graphic means that the only way to tell one pedal from the other is by the the small labeling noting what the pedal is. The layouts for all pedals being the same, means they all look the same, a very simple plain, but still aesthetically pleasing appearance. The user won't be enticed to like or dislike the sound based on its appearance, certain colors effecting the way they perceive the sound. Down to the overly simplistic programmer or electrical engineering font, the only sense of attitude is from the Corvus Audio font.
The enclosures still have a vibe about them, something that does look good to see, but they have an emotionally neutral vibe, and as far as I am concerned, that is the way I like it.
As always, check out Corvus Audio for new products and be sure to like us on Facebook.
Subscribe to:
Posts (Atom)