Background info: About a week ago I finished my first DIY audio project; the adorable Mini³ designed by Ti Kan of AMB labs. I chose to build that specific model because there is lots of documentation on it, the parts are easily obtainable and portability is great (fits in pocket). The construction went without a hassle, apart from some slight issues with the positive battery contacts.
After my little project I got the confidence to start on something bigger. Instead of a pocket-sized portable amplifier I will be building a full size stereo headphone amplifier. I collected some money together from birthday funds, and I was ready to go. I first needed to select the right project to build though, and following will be my process of choosing.
 |
| The insides of a magnificent Beta22 |
Design specifications of my ideal DIY headphone amplifier
Tube vs. Solid state: There are a huge amount of DIY amplifiers available for building. Perhaps even more than commercial ones. The first step in narrowing the choices down is usually deciding on the design topology of the circuit. The main two branches in which they divide are tube and solid state amplifiers. For the audio-illiterate I'll briefly explain what the two are and their pro's and cons.Tubes: they are glass light bulb like components that can be used to amplify an audio signal. This technology dates back before the first transistors, and are what made the first computer possible. It's outdated, but there are a few reasons people still use it.
 |
| The Littledot MK IV, a good example of the beauty of tubes |
- They look pretty
- They sound pretty
- They are expensive
1) The looks are actually a very important factor. They give of a very nice glow, and look aesthetic when turned off as well.
2) They add a bit of harmonic distortion as well as making the sound slightly warmer. This makes them sound better in some cases, although in other cases it's actually not good at all.
3) The price of things turns audiophiles on. Surely a $200 tube outperforms a $4 chip. The underlying science says 'no', but many audiophiles have more faith in the underlying psuedo-science apperently.
Solid state: Solid state amplifiers work with transistors instead of the old fashioned triodes (tubes). Technically speaking they outperform tube amps in both price and quality. However they lack the appeal tubes have.
I personally chose to built a solid state amplifier as it's both cheaper and of higher quality. I'm wise enough to have learned that for amplifiers and DAC's measurements say more than subjective rambling of scientifically illiterate audiophiles. Headphones and speakers are a totally different story on the other hand, but I'll leave that for another time.
Discrete vs. IC's: A circuit is discrete when it does not contain any integrated circuits (chips). Some audio purists believe that all IC's will have bad effects on the audio signal, much worse than discrete components will have. However this has no basis in reality, and this myth comes from the 70-80's when op-amps were still a new technology in audio. We've come a long way since then and now op-amps are generally better than discrete components when it comes to amplification.
As a result I won't mind building an op-amp based amplifier, despite the fact that op-amps are a pain in the *** to solder. However if a discrete amplifier just appeals to me more while having similar specs to one based on op-amps I won't mind either way.
Power supply: If you want to build a proper amplifier you have
to pay attention to the power supply as well. There are two reasons why this is:
1. stable output voltage and current. The amplifier always puts out the exact same voltage under any load. If the power supply can't supply enough current, clipping of the audio signal will occur at higher volumes. This is very bad, and should be avoided at all costs. Furthermore the voltage needs to stay constant as well; if the AC input voltage changes or has sudden spikes, the DC voltage output should remain constant. If this is not the case then pops will be audible during playback due to AC peaks, and circuits can become unstable if they assume a constant voltage on the power supply. (e.g. bias voltages)
2. Low ripple current/voltage. When AC is converted into DC you want the output to be a perfect DC signal with two rails, i.e. a positive and negative output. Unregulated power supply currents have a large amount of ripple voltage that changes the signal in an audible way. If the ripple voltage is more than -100dB after the gain stage it can be audible.
These two problems can be solved in two ways.
1. Batteries. This is the most simple solution. This is because batteries behave like near perfect power supply given that the load impedance is significantly larger than the internal resistance of the battery. This has two disadvantages however; the first is the fact that batteries drop in output voltage over time. This means that the circuit needs to be designed such that it does not rely on a constant power supply voltage. The second disadvantage is the fact that batteries need to be replaced after 500-1000 recharge cycles, which can easily happen within a year of heavy use. Another major advantage however is that it permits portability, but this is not necessary in this project.
2. Regulated power supplies. These kind of power supplies first convert the AC signal into a set unregulated power and then use some sort of regulation circuit. The performance of this highly varies with different designs and costs. The main advantage of regulated power supplies over batteries is the higher output voltage and current, lower output impedance, constant voltage under all circumstances and lower ripple voltage if designed properly. The big disadvantages are cost and portability, but in terms of performance regulated power supplies are a clear winner over batteries.
As a guideline I want to have a power supply with ripple voltages of less than 1mV over the entire audio band. An output impedance of <0.1 Ohm. A tracking rail design to prevent power on/off 'thumb' that can potentially harm drivers.
Balanced vs. Single ended: Some audio amplifiers are balanced, which means that each channel has two amplifiers dedicated to it with reversed polarity. The advantage of this is that the two signals can be combined in such a way that all noise picked up in a cable is canceled out against each other. This is useful if you either have a very low power signal (e.g. a microphone) or a very long cable (e.g. cables used on stage). Another advantage is that you have twice the output power, but this brings with it the disadvantage of having twice the output impedance - more on that later.
In commercial applications there is no real advantage in using balanced audio, and considering the fact that it costs twice as much I will limit myself to single ended, i.e. not balanced.
Virtual ground vs. 'real' ground: Some amplifiers have a dedicated third channel for the ground, instead of using the chassis and/or ground plane of the circuit board. I don't know exactly how this works, but for those interested it's discussed in this arcticle by NwAvGuy. Bottom line: it's more expensive and there is no real advantage to it. In fact it only increases cross talk, instead of supposedly reducing it. I'm also a bit skeptic towards NwAvGuy's article, as he does have the tendency to blow things up a little.
And there is a real use for using a virtual ground: if you have a single battery power supply, then you need some way of splitting the rails. This can be done by creating a virtual ground and splitting a 9V DC signal into a 4.5V and -4.5V DC signal relative to the virtual ground. Relative to the real ground it will be 9V and 0V. This won't be needed if you have two batteries, since you can use each battery for a separate rail. At least this is what my logical interpretation of a virtual ground is.
Output impedance: Output impedance is the exactly what its name says: the impedance on the output signal. This impedance acts as a voltage divider, meaning that it will soak up parts of the output signal depending on the impedance of the load. Say you have an amplifier with a 8 ohm output impedance (very high!), and you connect a 16 ohm load to it, e.g. an IEM.
1/4th of the power output will be lost top the output impedance leaving only 3/4 of the power for the headphone! This wouldn't be so bad if the impedance of a headphone/speaker was constant throughout the entire spectrum. Truth is that it's not, look at the diagram of the impedance vs. frequency diagram of Ultimate Ears TF10 - a normal balanced armature IEM. There is a lot of variance of impedance with frequency, therefore having a high output impedance amplifier will give more power to some frequencies than others. This will alter the sound in a bad way.
Long story short: keep output impedance as low as possible. 2 ohms is a good maximum to hold on to in my opinion. This also rules out tubes even further as tube amps typically have a very large output impedance. You can hook them up to an output transistor, but this increases the price even further.
Budget: Now that we made some clear requirements let's start by naming a budget and getting all the available amps from there. I will also list a small number of well known amplifiers I won't choose for whatever reason.
My budget is around $400-$500. Despite the fact that I live in Europe I still compare audio components in dollars for convenience, i.e. most components are priced in dollars and not in euros.
The following is a list of amplifiers available with a brief pros and cons
I can rule out the Kumisa III on the basis of sound quality, and the M³ can be ruled out on the basis of power output. The Beta22 and the O2 remain.
The Beta22 on the other hand costs about $320 + shipping on parts and another $150 for enclosure and panels.
They differ in price by about a factor 4, therefore the Beta22 needs to obviously outperform the O2 in order for it to be a viable option.
Build difficulty The O2 is a smaller build with less parts, but it has in contrary to the Beta22 multiple chips. Chips are more difficult to solder, but having done two on my Mini3 I know that it's still fairly straightforward. If I order pre-drilled panels then the build of the O2 will be fairly easy. In fact I will probably be done within four hours.
The Beta22 on the other hand is far larger and features many more parts. Building it and acquiring the parts needed to populate the board is far easier on the other hand since GlassjarAudio offers a kit, while the O2 probably needs to bought via a group buy.
There is additionally work with the Beta22 with the wiring. Since this amplifier does not have a dedicated enclosure I will need to wire the boards to the I/O components on the enclosure.
Looking at it from a perspective of how much I will learn from each project: From the O2 I will learn nothing at all, since the same things need to be done as I did with the Mini3
From the Beta22 I will learn far more. I will learn how to work boards into a non-dedicated enclosure. I will learn how to order and design front panels to suit my needs, as well as learning how to paint a chassis. I will learn how to work with trimpots and adjust quiescent current and DC offset. And finally I will learn how to work with hook-up wires.
Aesthetics The Beta22 will without doubt win on this area. The O2 is made for one specific enclosure, while the Beta22 can be fit into any enclosure I desire. The chassis will add extra work and costs, but given the increase in aesthetics it can be quite worth it.
Extra features I have investigated the possibility of using the Beta22 as a preamp with multiple outputs and inputs, as well as multiple headphone outputs. No surprises: it's all possible! Very convenient, as that would mean I can just built a good quality power amp and be done with my speaker rig as well. Multiple outputs is very useful as well for comparing headphones, although the volume will vary because I can use only one potentiometer. Two potentiometers is not possible because the potentiometer is before the gain stage.
The O2 is a slight disappointment when it comes to this. It has one input and one output, both 3.5mm TRS jacks. It's not possible to change this as there simply isn't enough room for the extra parts. It would be possible with a larger enclosure, but that is not possible either due to the O2's reliance on the chassis ground of the enclosure.
Sound quality I suspect the Beta22 to be the winner here, but that is irrelevant. This is because they both offer a fully transparent sound, i.e. they amplify the signal without changing the signal in any audible way. I highly doubt whether there will be an audible difference in sound quality between the two.
The verdict In the end I'm leaning towards the Beta22 despite it's larger price that isn't justified by a higher sound quality. The Beta22 will be a larger learning experience, it has more features and it will become far more beautiful. It is ironic that in the end I, an audiophile, select an amplifier not on the basis of sound quality but on other specifications.
2) They add a bit of harmonic distortion as well as making the sound slightly warmer. This makes them sound better in some cases, although in other cases it's actually not good at all.
3) The price of things turns audiophiles on. Surely a $200 tube outperforms a $4 chip. The underlying science says 'no', but many audiophiles have more faith in the underlying psuedo-science apperently.
Solid state: Solid state amplifiers work with transistors instead of the old fashioned triodes (tubes). Technically speaking they outperform tube amps in both price and quality. However they lack the appeal tubes have.
I personally chose to built a solid state amplifier as it's both cheaper and of higher quality. I'm wise enough to have learned that for amplifiers and DAC's measurements say more than subjective rambling of scientifically illiterate audiophiles. Headphones and speakers are a totally different story on the other hand, but I'll leave that for another time.
Discrete vs. IC's: A circuit is discrete when it does not contain any integrated circuits (chips). Some audio purists believe that all IC's will have bad effects on the audio signal, much worse than discrete components will have. However this has no basis in reality, and this myth comes from the 70-80's when op-amps were still a new technology in audio. We've come a long way since then and now op-amps are generally better than discrete components when it comes to amplification.
As a result I won't mind building an op-amp based amplifier, despite the fact that op-amps are a pain in the *** to solder. However if a discrete amplifier just appeals to me more while having similar specs to one based on op-amps I won't mind either way.
Power supply: If you want to build a proper amplifier you have
to pay attention to the power supply as well. There are two reasons why this is:
1. stable output voltage and current. The amplifier always puts out the exact same voltage under any load. If the power supply can't supply enough current, clipping of the audio signal will occur at higher volumes. This is very bad, and should be avoided at all costs. Furthermore the voltage needs to stay constant as well; if the AC input voltage changes or has sudden spikes, the DC voltage output should remain constant. If this is not the case then pops will be audible during playback due to AC peaks, and circuits can become unstable if they assume a constant voltage on the power supply. (e.g. bias voltages)
2. Low ripple current/voltage. When AC is converted into DC you want the output to be a perfect DC signal with two rails, i.e. a positive and negative output. Unregulated power supply currents have a large amount of ripple voltage that changes the signal in an audible way. If the ripple voltage is more than -100dB after the gain stage it can be audible.
These two problems can be solved in two ways.
1. Batteries. This is the most simple solution. This is because batteries behave like near perfect power supply given that the load impedance is significantly larger than the internal resistance of the battery. This has two disadvantages however; the first is the fact that batteries drop in output voltage over time. This means that the circuit needs to be designed such that it does not rely on a constant power supply voltage. The second disadvantage is the fact that batteries need to be replaced after 500-1000 recharge cycles, which can easily happen within a year of heavy use. Another major advantage however is that it permits portability, but this is not necessary in this project.
2. Regulated power supplies. These kind of power supplies first convert the AC signal into a set unregulated power and then use some sort of regulation circuit. The performance of this highly varies with different designs and costs. The main advantage of regulated power supplies over batteries is the higher output voltage and current, lower output impedance, constant voltage under all circumstances and lower ripple voltage if designed properly. The big disadvantages are cost and portability, but in terms of performance regulated power supplies are a clear winner over batteries.
As a guideline I want to have a power supply with ripple voltages of less than 1mV over the entire audio band. An output impedance of <0.1 Ohm. A tracking rail design to prevent power on/off 'thumb' that can potentially harm drivers.
Balanced vs. Single ended: Some audio amplifiers are balanced, which means that each channel has two amplifiers dedicated to it with reversed polarity. The advantage of this is that the two signals can be combined in such a way that all noise picked up in a cable is canceled out against each other. This is useful if you either have a very low power signal (e.g. a microphone) or a very long cable (e.g. cables used on stage). Another advantage is that you have twice the output power, but this brings with it the disadvantage of having twice the output impedance - more on that later.
In commercial applications there is no real advantage in using balanced audio, and considering the fact that it costs twice as much I will limit myself to single ended, i.e. not balanced.
Virtual ground vs. 'real' ground: Some amplifiers have a dedicated third channel for the ground, instead of using the chassis and/or ground plane of the circuit board. I don't know exactly how this works, but for those interested it's discussed in this arcticle by NwAvGuy. Bottom line: it's more expensive and there is no real advantage to it. In fact it only increases cross talk, instead of supposedly reducing it. I'm also a bit skeptic towards NwAvGuy's article, as he does have the tendency to blow things up a little.
And there is a real use for using a virtual ground: if you have a single battery power supply, then you need some way of splitting the rails. This can be done by creating a virtual ground and splitting a 9V DC signal into a 4.5V and -4.5V DC signal relative to the virtual ground. Relative to the real ground it will be 9V and 0V. This won't be needed if you have two batteries, since you can use each battery for a separate rail. At least this is what my logical interpretation of a virtual ground is.
Output impedance: Output impedance is the exactly what its name says: the impedance on the output signal. This impedance acts as a voltage divider, meaning that it will soak up parts of the output signal depending on the impedance of the load. Say you have an amplifier with a 8 ohm output impedance (very high!), and you connect a 16 ohm load to it, e.g. an IEM.
 |
| The impedance-frequency curve of a UE TF10 |
Long story short: keep output impedance as low as possible. 2 ohms is a good maximum to hold on to in my opinion. This also rules out tubes even further as tube amps typically have a very large output impedance. You can hook them up to an output transistor, but this increases the price even further.
Budget: Now that we made some clear requirements let's start by naming a budget and getting all the available amps from there. I will also list a small number of well known amplifiers I won't choose for whatever reason.
My budget is around $400-$500. Despite the fact that I live in Europe I still compare audio components in dollars for convenience, i.e. most components are priced in dollars and not in euros.
The following is a list of amplifiers available with a brief pros and cons
- Cavalli-Kan Kumisa III - Fully discrete apart from the opamps used in the power supply. The power supply does not have a transformer, and the RMAA measurements are not that good compared to others.
- M³ - Three channel discrete amp using MOSFETs. I don't like the third active ground channel, however the measurements are reference quality. (All excellent's on RMAA). The price is very good as well.
The biggest drawback is the fact that the output power is not great enough to power cans such as the HD 650 or orthos like the LCD-2 and HE-600 - β22 + σ22 power supply (Beta22) - Fully discrete with the option to choose 2 channels, 3 channels (active ground), 4 channel (balanced), or 6 channels (balanced with active ground). The power supply is excellent and complies to all my requirements with a huge amount of headroom left over. The measurements are reference quality.
The output power is very high - it will even power small speakers if run balanced.
The only drawback is that it's touching the boundaries of the budget at around $350 in parts and $100 for a good chassis. - Objective2 (O2) - Portable with reference quality measurements. The price is very low at under $100. It has a cult status of being the cheapest amp available that is fully transparent, i.e. 'a wire with gain'.
The drawback is the battery operated power supply, but THD+N seems inaudible at normal frequencies so this should not be a too large problem.
I can rule out the Kumisa III on the basis of sound quality, and the M³ can be ruled out on the basis of power output. The Beta22 and the O2 remain.
The Beta22 versus the O2 - a battle of the specs
Price The O2 is by far the cheapest. It costs around $120 for wall wart, enclosure, parts, knob and panels, including shipping.The Beta22 on the other hand costs about $320 + shipping on parts and another $150 for enclosure and panels.
They differ in price by about a factor 4, therefore the Beta22 needs to obviously outperform the O2 in order for it to be a viable option.
Build difficulty The O2 is a smaller build with less parts, but it has in contrary to the Beta22 multiple chips. Chips are more difficult to solder, but having done two on my Mini3 I know that it's still fairly straightforward. If I order pre-drilled panels then the build of the O2 will be fairly easy. In fact I will probably be done within four hours.
The Beta22 on the other hand is far larger and features many more parts. Building it and acquiring the parts needed to populate the board is far easier on the other hand since GlassjarAudio offers a kit, while the O2 probably needs to bought via a group buy.
There is additionally work with the Beta22 with the wiring. Since this amplifier does not have a dedicated enclosure I will need to wire the boards to the I/O components on the enclosure.
Looking at it from a perspective of how much I will learn from each project: From the O2 I will learn nothing at all, since the same things need to be done as I did with the Mini3
From the Beta22 I will learn far more. I will learn how to work boards into a non-dedicated enclosure. I will learn how to order and design front panels to suit my needs, as well as learning how to paint a chassis. I will learn how to work with trimpots and adjust quiescent current and DC offset. And finally I will learn how to work with hook-up wires.
Aesthetics The Beta22 will without doubt win on this area. The O2 is made for one specific enclosure, while the Beta22 can be fit into any enclosure I desire. The chassis will add extra work and costs, but given the increase in aesthetics it can be quite worth it.
 | |||||
| One of the many examples of how beautiful the Beta22 can be |
 |
| The in my opinion not so attractive Objective2 |
The O2 is a slight disappointment when it comes to this. It has one input and one output, both 3.5mm TRS jacks. It's not possible to change this as there simply isn't enough room for the extra parts. It would be possible with a larger enclosure, but that is not possible either due to the O2's reliance on the chassis ground of the enclosure.
Sound quality I suspect the Beta22 to be the winner here, but that is irrelevant. This is because they both offer a fully transparent sound, i.e. they amplify the signal without changing the signal in any audible way. I highly doubt whether there will be an audible difference in sound quality between the two.
The verdict In the end I'm leaning towards the Beta22 despite it's larger price that isn't justified by a higher sound quality. The Beta22 will be a larger learning experience, it has more features and it will become far more beautiful. It is ironic that in the end I, an audiophile, select an amplifier not on the basis of sound quality but on other specifications.
Design choices of the Beta22
Power supply isolation A second option is whether or not to isolate the power supply in a separate enclosure. This will reduce the interference of the large electromagnetic fields generated by the transformer of the power supply. However I doubt that this effect will be very huge, and if it is I can always try to use a bit of insulation on the boards to limit it. The main reason not to do this is the need for two enclosures, while enclosures are not cheap at all.
This also introduces the issue of what power transformer to use. There are two main types: laminated cores and toroidal transformers. I will be using a toroidal transformer, and this is because they have a smaller external magnetic field. Since the transformer will be in the same enclosure the electromagnetic interference caused by the transformer is a serious issue, and should be limited as much as possible. Using a laminated core, while providing less ripple current and more efficiency, means having more electromagnetic interference, and possible interference reaching an audible level.
Inputs and outputs Yet another important option is the inputs and outputs. I want to have two inputs and two outputs. This way I can connect both my turntable and my DAC to the amp. It also allows me to have two headphone outputs, which is very nice for comparing the two. The volume won't be equalized on the other hand since there is one potentiometer. It would be possible to have to potentiometers by adding them post gain stage, but I doubt that would do much good to the sound quality, I might need to consult a professional on this. Furthermore I also want a preamp output, or a loop-out. This is essentially an output that skips the gain and potentiometer stage and directly links the output to the input. This is very useful for using it as an input selector for a power amplifier, which I might be doing somewhere in the future when I upgrade my speaker rig.
The in and outputs need to be selectable by a switch. I have done the needed research and came to the conclusion I will be needing a DPDT toggle switch. I might be ordering those along with the knobs from a separate vendor, since looks are very important. Mouser on the other hand does not offer a lot of attractive choices.
I made a diagram of how I want the boards and inputs/outputs configured.
Jacks There are two main types of output headphone jacks: XLR and TRS. TRS, or 6.3mm tip-ring-sleeve is by far the most common, but it has a very large disadvantage: when the plug is inserted or pulled out it creates a momentary short circuit between ground and the right channel. If the amplifier does not have any current limiting output protection circuitry then this short circuit can damage both the amplifier and the drivers of the headphone. Unfortunately the Beta22 does not have such output protection, and therefore you need to turn the amplifier off completely before moving the plug. If not the output MOSFET's will get damaged, and if repeated many times even destroy them completely.
It should be pretty obvious that this is not wanted.
Luckily there are two saviors present! Locking type TRS jacks and XLR jacks. The former is special type of jack that locks the plug in place, preventing any accidental movement of the plug. This does not remove the issue that if you want to change headphones that you need to turn off the amplifier first. For this latter requirement we have the XLR jacks. They are simple sturdy three pin jacks that provide far less ground problems. The only disadvantage is that the most headphones either need a cable retermination or an XLR adapter. But luckily this won't amount to much more than a few bucks. The only requirement of such an adapter is that it locks in the TRS plug pretty tight, because pulling out the plug from the adapter will still cause a short circuit.
As a result I chose to have one TRS plug, and one XLR plug selectable by a DPDT toggle switch. In this way I will still be able to compare headphones, and use the TRS plug whenever needed. If I ever decide to DIY my own cables this also allows me to make a nice XLR cable, instead of the ugly and unfashionable 6.3mm TRS plugs.
The jacks of the inputs and outputs are important as well. The famous company called Neutrik offers very high quality jacks, but I honestly doubt whether or not it's going the change the sound quality. The only great thing is that they make very durable jacks that lock your headphone or RCA cable tightly in place. However some cheaper alternatives also do this. I think I'll go with something decent and gold plated, but not too fancy. $5 per RCA pair or XLR/TRS jack seems a nice budget
Enclosure Another very important option is the enclosure. Since I'm on a tight budget after the high cost of all the components I will go with something fairly simple. The Par-Metal 20 series seems an excellent option. They cost about $45 without shipping unanodized and unpainted. I think it says that you need to add another $25 if you want vent holes and pre-drilled panels. I will definitely do this as last time I tried to do the panel holes myself it was a complete disaster.
I have no idea as to what the shipping costs will be. I'm hoping around $50, but it might just as well be a lot more than that. The exact Par-Metal model I would order is the 20-12123N. It's 12" by 12" by 3".
Apart from Par-Metal there is also the supplier THL-Audio. They offer some very nice enclosures from Taiwan. The great thing about Taiwan is that they ship with EMS or other Asian postal services. These are far cheaper than DHL, TNT, UPS, etc. on international shipping. However they don't offer custom pre-drilled panels and are slightly more expensive. Despite the fact that their enclosures look a lot more pretty I will only buy from them if they are significantly cheaper since I would have to drill the panels myself.
I have the tools to do this, but it's a lot harder than you'd imagine. Precision is less essential in this case than with the Mini3 though, since the components are panel-mount. This means that even if the hole is of by a little I can still mount the components in, it will just look less pretty.
However if I do it properly this won't even be necessary. What I can do is draw the holes precisely using a protractor and calipers and then drill the holes with ~2mm headroom and file the last bit out by hand. This is a lot of work given the 15 holes I need to do this way, but it's worth it since the method is far more risk free.
Currently I have decided to order from THL-audio, unless something ridiculous comes up. This is because I haven't heard from Par-Metal in two weeks after I quoted them for shipping costs to the Netherlands. Very disappointing, if you ask me. THL-audio isn't the quickest in responding either, but so far they have responded within a three days every time.
I quoted the price for the PRAL-2 case, which fits my needs perfectly. They said it would cost 2800 TWD, or $92.50. They didn't mention the shipping costs yet, despite me dropping of my shipping details. I hate the fact that contact is so difficult with these people, but I guess that's what you get for ordering from a Taiwanese company.
THL-audio also offers some very nice looking knobs, which I might look into if I'll be ordering from them in any case. Partspipe is nice, but not on part with these beauties.
Using Inkscape I created the following guide for the the front and back panels if the Par-metal.
I made it accurate to the nearest millimeter and put it on an A3 PDF as well.
This link is the PDF with numerical guides.
This one is without.
Potentiometer The rest of the options are small details. One of which is the potentiometer. AMB recommends using Alps RK271, also know as the "Blue Velvet". It's a high quality potentiometer with a very smooth feel. I have heard many good things about it, and many hi-fi pre-amps tend to use this particular potentiometer. I will be using the the 50kOhm stereo version, as supplied by GlassJarAudio. The only downside is the fact that it costs a whopping 17 bucks per piece. I don't know if it's worth it performance wise, but I'll at least get something nice in return.
Knob The knob of the potentiometer is also a very important detail. I'll be ordering most parts through GlassjarAudio, who in turn orders it from Mouser. Since Mouser does not stock any fancy knobs fitting the thread size of the RK271 (8.0mm thread, shaft is 6.0mm), I'll have to find them somewhere else. Someone on the AMB forum's suggested the ebay seller 'partspipe' and they ship a couple nice knobs. I will need one for my Mini3 too, as well as switches for my Beta22. I can order all these three things from this vendor in one go. The best part is that they offer free world wide shipping.
I'm going to choose this knob for the Beta22, this knob for the Mini3, and four of these switches for the Beta22. I choose four DPDT switches, while in reality I'll be using two DPST switches as well. In principle you can just use a DPDT switch and hook up nothing to one of the sides to turn it into a DPST switch.
I will edit this blog post whenever I have some more information to be added. I will also discuss my choices with an audio engineer in a few days, so I might change my mind on several choices. I'm looking forward to building this, and I should enjoy it, given the hours of research I have spent on this topic.
The amount of channels: The Beta22 has many options as to how you are exactly going to build it. First is the amount of channels. You can have 2, 3, 4 or 6 channels, but since I'm just going to run it single ended with a passive ground I will stick with 2 channels, especially considering the fact that each extra channel will cost an extra ~$100 on parts. The only advantage of using 4 channels would be the ability to use the Beta22 as a speaker amplifier, but I'll have to forfeit that one.
 |
| A toroidal transformer |
This also introduces the issue of what power transformer to use. There are two main types: laminated cores and toroidal transformers. I will be using a toroidal transformer, and this is because they have a smaller external magnetic field. Since the transformer will be in the same enclosure the electromagnetic interference caused by the transformer is a serious issue, and should be limited as much as possible. Using a laminated core, while providing less ripple current and more efficiency, means having more electromagnetic interference, and possible interference reaching an audible level.
Inputs and outputs Yet another important option is the inputs and outputs. I want to have two inputs and two outputs. This way I can connect both my turntable and my DAC to the amp. It also allows me to have two headphone outputs, which is very nice for comparing the two. The volume won't be equalized on the other hand since there is one potentiometer. It would be possible to have to potentiometers by adding them post gain stage, but I doubt that would do much good to the sound quality, I might need to consult a professional on this. Furthermore I also want a preamp output, or a loop-out. This is essentially an output that skips the gain and potentiometer stage and directly links the output to the input. This is very useful for using it as an input selector for a power amplifier, which I might be doing somewhere in the future when I upgrade my speaker rig.
The in and outputs need to be selectable by a switch. I have done the needed research and came to the conclusion I will be needing a DPDT toggle switch. I might be ordering those along with the knobs from a separate vendor, since looks are very important. Mouser on the other hand does not offer a lot of attractive choices.
I made a diagram of how I want the boards and inputs/outputs configured.
 |
| A diagram of the wiring. (subject to change) |
 |
| TRS plug (left) and three-pin XLR plug (right) |
Luckily there are two saviors present! Locking type TRS jacks and XLR jacks. The former is special type of jack that locks the plug in place, preventing any accidental movement of the plug. This does not remove the issue that if you want to change headphones that you need to turn off the amplifier first. For this latter requirement we have the XLR jacks. They are simple sturdy three pin jacks that provide far less ground problems. The only disadvantage is that the most headphones either need a cable retermination or an XLR adapter. But luckily this won't amount to much more than a few bucks. The only requirement of such an adapter is that it locks in the TRS plug pretty tight, because pulling out the plug from the adapter will still cause a short circuit.
As a result I chose to have one TRS plug, and one XLR plug selectable by a DPDT toggle switch. In this way I will still be able to compare headphones, and use the TRS plug whenever needed. If I ever decide to DIY my own cables this also allows me to make a nice XLR cable, instead of the ugly and unfashionable 6.3mm TRS plugs.
The jacks of the inputs and outputs are important as well. The famous company called Neutrik offers very high quality jacks, but I honestly doubt whether or not it's going the change the sound quality. The only great thing is that they make very durable jacks that lock your headphone or RCA cable tightly in place. However some cheaper alternatives also do this. I think I'll go with something decent and gold plated, but not too fancy. $5 per RCA pair or XLR/TRS jack seems a nice budget
I have no idea as to what the shipping costs will be. I'm hoping around $50, but it might just as well be a lot more than that. The exact Par-Metal model I would order is the 20-12123N. It's 12" by 12" by 3".
Apart from Par-Metal there is also the supplier THL-Audio. They offer some very nice enclosures from Taiwan. The great thing about Taiwan is that they ship with EMS or other Asian postal services. These are far cheaper than DHL, TNT, UPS, etc. on international shipping. However they don't offer custom pre-drilled panels and are slightly more expensive. Despite the fact that their enclosures look a lot more pretty I will only buy from them if they are significantly cheaper since I would have to drill the panels myself.
I have the tools to do this, but it's a lot harder than you'd imagine. Precision is less essential in this case than with the Mini3 though, since the components are panel-mount. This means that even if the hole is of by a little I can still mount the components in, it will just look less pretty.
However if I do it properly this won't even be necessary. What I can do is draw the holes precisely using a protractor and calipers and then drill the holes with ~2mm headroom and file the last bit out by hand. This is a lot of work given the 15 holes I need to do this way, but it's worth it since the method is far more risk free.
 |
| The THL-audio PRAL-2 aluminium pre-amp case |
I quoted the price for the PRAL-2 case, which fits my needs perfectly. They said it would cost 2800 TWD, or $92.50. They didn't mention the shipping costs yet, despite me dropping of my shipping details. I hate the fact that contact is so difficult with these people, but I guess that's what you get for ordering from a Taiwanese company.
THL-audio also offers some very nice looking knobs, which I might look into if I'll be ordering from them in any case. Partspipe is nice, but not on part with these beauties.
Using Inkscape I created the following guide for the the front and back panels if the Par-metal.
I made it accurate to the nearest millimeter and put it on an A3 PDF as well.This link is the PDF with numerical guides.
This one is without.
Potentiometer The rest of the options are small details. One of which is the potentiometer. AMB recommends using Alps RK271, also know as the "Blue Velvet". It's a high quality potentiometer with a very smooth feel. I have heard many good things about it, and many hi-fi pre-amps tend to use this particular potentiometer. I will be using the the 50kOhm stereo version, as supplied by GlassJarAudio. The only downside is the fact that it costs a whopping 17 bucks per piece. I don't know if it's worth it performance wise, but I'll at least get something nice in return.
 |
| The knob I plan on buying for the Beta22 |
I'm going to choose this knob for the Beta22, this knob for the Mini3, and four of these switches for the Beta22. I choose four DPDT switches, while in reality I'll be using two DPST switches as well. In principle you can just use a DPDT switch and hook up nothing to one of the sides to turn it into a DPST switch.
I will edit this blog post whenever I have some more information to be added. I will also discuss my choices with an audio engineer in a few days, so I might change my mind on several choices. I'm looking forward to building this, and I should enjoy it, given the hours of research I have spent on this topic.
Geen opmerkingen:
Een reactie posten