Part Selection Guide
Some of the parts below are optional. For most of these parts, you simply leave them out if you don't want them. If a jumper is required, it will be mentioned.
Resistors
5% carbon film resistors are the default choice. Resistor accuracy isn't a big deal in this circuit since the output voltage is adjustable. You might choose instead to use metal films because of their lower drift and noise.
All of the resistors should be 1/4W types, except for R2 which should be 1/2W in most situations. When R1 is 120 Ω, the current across R2 is about 10mA, so R2 must be under 2.5 KΩ if you want to use a 1/4W unit. When R1 is 100 Ω, the current is 12.5mA, so R2 should be under 1.6 KΩ if you want to use a 1/4W resistor. Even when 1/4W is sufficient, you may want to use a 1/2W unit to keep heat in this resistor down to improve long-term accuracy.
If you want to use the Vishay Dale RN/CMF series resistors, the RN55s are 1/4W and the RN60s 1/2W at the temperatures you should see in this power supply. They're specified for lower wattage because their temperature range is so high.
R1
This resistor sets the current used by the regulator to operate. There is always 1.25V across this resistor, so you pick its value to put about 10mA through this path. While some LM317 type regulators will operate with as little as 5mA, they are spec'd to require 10mA in the worst case situation. Better safe than sorry...
R2
This resistor sets the regulator's output voltage in conjunction with VSET.
The output voltage formula is:
There is a table on the schematic with several useful R1 and R2 values, assuming that you use a 500 Ω VSET trimmer.
If you want to calculate your own values, then assuming R1 is 120 Ω:
This will put your desired voltage right in the middle of the adjustment range.
R2 is normally a 1/2W resistor, for reasons stated above.
VSET
This pot allows you to adjust the output voltage over a small range.
You could jumper across R2 and just put a 2.5K pot here to get full adjustment from approximately 1.5V to 32V, but these trim pots are not designed to tolerate frequent adjustments. The idea is to have a small adjustment range so you can dial in the exact output voltage you want. Once it's set where you want it, you can dab a bit of nail polish or paint on the top of the pot to keep the value from changing.
If you don't need any adjustment, you can jumper across this pot.
R3
This sets the current through the LED. The LED is more than just a power indicator. It also keeps a small, steady draw on the power supply, which can improve performance in some cases. It also helps discharge the filter cap when the power is removed. If you don't want a power LED, add R3 anyway, and just jumper across the LED.
You should consider setting the current through this path to be 5 to 20mA, even though that will make the LED rather bright. This will ensure quick discharge of the filter caps.
Capacitors
C1, C2, C3, C4 (ceramic)
These caps suppress any noise generated by the bridge as the diodes in it switch on and off. The exact value to use here is not terribly critical. 100 pF is a reasonable value. If you choose a different value, I'd say smaller is better.
These caps are optional.
Largest Part Size: 4mm × 2.5mm
C5 (electrolytic)
This is the main filter capacitor.
If you're using an AC transformer directly with the TREAD board, this cap's voltage tolerance should be at least twice the rated voltage output of your transformer. For instance, if you have a +/-15V transformer, you should use 63V caps. The reason is, when the transformer is lightly loaded, its voltage will go up by as much as 40%. Also, the peak voltages put out by the rectifier bridge will be about 1.4× the RMS voltage. Taken together, 2× the voltage tolerance is required for safety.
If, instead, you're using an unregulated AC-DC wall wart with the board, this cap's voltage tolerance can be much lower. Simply measure the wall wart's output voltage without any load on it and make the cap a bit higher in voltage than that. Or, leave it out entirely: the unregulated supply will have an output capacitor already, and its size may be sufficient.
Largest Part Size: 18mm diameter.
C6 (film)
This is a small, fast film cap to lower the ESR of the main filter cap. Its value is not critical, and it can be left out without a big penalty.
Largest Part Size: 5mm × 7.5mm
C7 (tantalum)
This cap bypasses the regulator's adjustment pin, increasing the regulator's ripple rejection. You could put a 100 µF electrolytic here instead of the default tantalum, but the tant will work reliably over a greater temperature range and it will last longer.
This cap is optional.
Largest Part Size: 5mm diameter
C8 (electrolytic or tantalum)
This cap does a bit of post-regulation filtering.
For the LM317 it can be as big as you'd like. A 1 µF or so tantalum or a 22 µF of so electrolytic is a good base value. Performance will improve a bit with larger values. You can leave it out if you want.
For LDO regulators (LM1086, LT1085...) a cap is required here and it must have an ESR of 1 to 2 ohms to guarantee the regulator's stability. Since higher value caps have lower ESRs (all else being equal), stick to 1 µF for tantalums and 22 µF or so for electrolytics.
This cap is optional if you're using the LM317.
Largest Part Size: 5mm diameter
Diodes
B1
This is a monolithic rectifier bridge.
The standard part here is a W01G. Other names for this part are the NTE5304, RC202 and RB152. Some of these are simple round things, but I prefer the type with a flatted side, because it makes it easy to get the orientation right.
If you can't find this part, a 1KAB has the same pin spacing and pin ordering. Due to the case shape, however, the snubber caps will prevent it from sitting all the way down on the board.
If the TREAD's power source is a DC supply (for instance, an unregulated wall wart), this part is not necessary. If you leave it out, however, you will have to hook the power supply up to bypass the bridge position; the simplest way to do this is to hook the power source to the C5 pads, which is also not needed in this situation. You can still add the bridge if you want; it just drops the voltage a bit, which is harmless in most instances, but unnecessary.
D1, D2
These diodes protect the regulator when the capacitors around the regulator discharge. Generic 1N400x types will work. 1N4001s are probably sufficient, but make them 1N4002s or higher to be safe.
These diodes are required if you add C7 and C8. If this supply will drive a capacitive load (such as a headphone amplifier's power section) you still need D1 at least. My view is, these are 4-cent parts...why leave them out if there's even the possibility that they may help?
LED
This is the power indicator LED. See R3 for details.
Choosing a Regulator
There are many choices for regulators that will work in this design, ranging from a 40 cent off-brand LM317 to a $10 industrial grade LT1084IT. The standard audiophile reaction to these facts is that the LT1084 must be the best choice. The fact is, there are tradeoffs among these choices, so more expensive isn't always better, even when money is not a concern.
The lowly LM317 has many advantages: it's cheap, easy to find, docile, and its performance is very close to that of its more expensive family members. If you want to spend a little more on your regulator, get the National Semiconductor version instead of an off-brand clone, or get the high-spec LM317A variant instead of the plain LM317.
All the other regulators that work in this board are "low drop-out" types, meaning that the voltage drop across the regulator is lower than for the standard LM317. The downside of these LDO regulators is that they are sensitive to the impedance on their output. This means you carefully take into account the characteristics of the power supply's output capacitor, the wire between the power supply and the load, and any power rail capacitors in the circuit being powered. For example, powering a PPA that has 2000 µF of rail capacitance (C1) with an LDO TREAD will most likely fail to work. For more details on this, read the datasheet for the regulators you're considering. If you don't want to read datasheets, you should stick with the LM317.
Another problem with LDOs is that they only provide a benefit when there is a low voltage drop across them. If your configuration puts a 3V drop across the regulator, an LDO has absolutely no advantage over an equivalent standard regulator.
If an LDO regulator will work in your situation and you want the small advantages it will provide, the cheapest option is the LM1086. This gets you the LDO feature plus a tiny performance boost over an LM317. The next step up from that is Linear Technology's version of the same design, the LT1086. Beyond that, there's the 3A and 5A versions of this family (LM1085, LT1084, etc.); you probably don't need the higher current, but to provide that higher current they necessarily have lower output impedance, which improves things a bit further still.
Again, I must caution you about spending exorbitant amounts of money on the regulator IC. There comes a point where the extra money you can spend on a better regulator IC would be better spent on a different, inherently better power supply design. This circuit gives the best value with ICs from the low end of the cost range.
Choosing a Power Source
The easy way to go here is to use some kind of wall wart, rather than a traditional transformer. You can find AC-AC wall warts, which are just a transformer encapsulated in a box with a wall plug and a low-voltage power cord coming out. More common are unregulated AC-DC wall warts, which follow the transformer with a diode bridge and a filter cap, to give DC with a fairly high ripple component. The advantages of using a wall wart are several:
- It puts the transformer far away from the regulator circuit, so there will be less induced AC hum in the output.
- It greatly reduces your exposure to the dangers of AC line wiring.
- The casework is simpler: a low-voltage barrel jack is easier to add to a case than an AC power inlet module.
- It's an easy way to make the powered device world-power compatible: just pick up an appropriate local wall wart when traveling, or use a world-power compatible wall wart.
If you choose the AC-DC type of wall wart, there are a few additional advantages over the AC-AC type:
- You can leave the diode bridge and filter cap out when building the TREAD.
- There's even less chance of induced AC hum, since much of the AC ripple is taken care of within the wall wart.
Using a bare AC transformer isn't without its advantages, however. The primary ones are:
- Putting the transformer in the same case as the regulator and using an AC power cord is neater.
- You can get "dual secondary" type transformers, which you can use to make a dual supply. See below.
- If you believe audiophile power cords have any benefit, this lets you use the IEC inlet you need to accept these cords.
If you're going to use an AC-AC wall wart or a plain AC transformer, the current rating should be at least 120% of the current you will be drawing from it. The voltage question is a little trickier. Instead of explaining all the details here, I've wrapped all it all up into my Power Supply Parameter Estimator.
To make a dual supply, you need a dual-secondary transformer and two TREAD boards, configured like so:
|
It's common for dual-secondary transformers to also have dual primaries, as shown in the schematic above. For ~240V power systems, you connect AC line to 1, AC neutral to 4, and 2 and 3 together. The secondary side is the same for all power systems.
Choosing a Heat Sink
The size of heat sink you need is a function of the amount of heat that the regulator will be putting out. I won't go into details here, as my Power Supply Parameter Estimator pretty much covers it. You put in your configuration details, and it will tell you how hot the regulator will get. If it says your configuration will get too hot, you have several choices: lower the power source's voltage, raise the output voltage, lower the load current, or use a bigger heat sink. I'd try those in that order. "Bigger heat sink" is the lowest priority, because it attacks the symptom instead of the cause. Be certain that the cause can't be treated before going after the symptom.
You may be tempted to bolt the regulator's tab to the case of the circuit your TREAD is powering to get a "free" heat sink, but I recommend against it. The first problem is that the tab on an LM317 family regulator is tied to the output voltage, so you will need to use an insulator kit to make this work. This adds thermal resistance, so it reduces the effectiveness of the "heat sink". The second problem is that most cases don't have a high surface area to volume ratio, as a proper heat sink does. Many people make the mistake of conflating massiveness with heat sink effectiveness. The effectiveness of a heat sink is a function of surface area, not mass. A heavy case with flat sheet metal walls isn't an especially good heat sink.
Miscellaneous Hardware
If you're using an AC transformer, I recommend that you add an IEC power inlet of some sort, rather than a captive AC power cord. I recommend the Qualtek 723W or the Schurter 6200 series, since they have built-in fuse holders.
If you're using a wall wart, you'll probably need a power jack of some sort. Most wall warts use 2.1/5.5mm or 2.5/5.5mm barrel jacks.
You'll notice three holes in the board labelled TP1 through TP3. These are test points, useful for testing the power supply. I like wire loop type test points, so I can grab onto them with my meter's grabber leads. Keystone type 5005 through 5009 wire loop test points work well here. You can also use the pins cut from SIP pin strips, or make loops from resistor lead cuttings, or leave the holes empty to form a crude DMM probe "socket". I include some of the Keystone test points with the TREAD kits, but not with the bare boards.
This space intentionally left blank. :)