Active Eurorack Busboard Part 3

THR power supply board

 

Time travel

Let's jump into a timeline a little bit. Before I finish writing about all of my adventures while designing this system I'd like to write about its final form, the one, that is hitting the market right now.

Quick start guide

Many people probably landed here just to use their board or to judge how hard it is to use them, So here is a simple a guide on how to set up your new THR:

• Mount the board either on a test bench or in a case. If the mounting is to be permanent consider the thermal requirements first.
  
• Connect your power supply to the two-pin connector. Don't search for polarity markings, there aren't any, but the board is reverse polarity protected so if it doesn't work just try to reverse the  connection (just in case - is closer to the edge, + is closer to the capacitor).
  
• Make sure that all 3 LEDs are lit up. If neither is lit probably there's a problem with the input power. If some are working and some are not or only blinking, that indicates an overload condition or a short circuit at the output.
• Measure and make sure that all 3 output voltages correspond to the desired ones
• Connect your device / busboard
• Have fun!
  

Board overview

The board has one input power connector which accepts an 18-24V input and two 4-pin output connectors for powering up to two busboards or other loads of your choice. The board outputs +-12V and 5V delivering 2A on each of the positive rails and 1A on the negative rail.

3 mounting holes are available for securing the board to a case or a test bench.

prototype board after all stress testing and torture

Protection overview

• Input reverse polarity protection. As much as I believe that people should be able to route 2 wires properly, even I reverse them from time to time. So, the board is protected from reverse voltages of up to 30V, while wasting no energy when proper polarity is applied.
• Output reverse polarity protection. As I said in one of my previous posts when someone, somehow shorted the positive and negative rails one of them will shut down first possibly reversing polarity and killing sensitive modules. To make sure this won't happen 3 diodes are included to make sure that if shorted, the rails will maintain their polarity.
• Output overcurrent and short circuit protection. I don't think I need to explain why they're crucial. A screw which fell inside the busboard, a missed patchcable, or too many modules. At least one of these things has happened to us at some point in time.
• Thermal shutdown. As cooling and airflow management is akin to black magic and this board packs a lot of power in a tiny footprint it's prone to overheating. This protection is in place to make sure it doesn't cook itself to death.

Thermal considerations

Even though I tried designing the board to be as efficient as possible it is still not 100% efficient; therefore, it generates some waste heat. The negative rail converter is roughly 80% efficient and the positive rail converters are approximately 90% efficient. This means that under typical operation (+5V,+12V @ 2A, -12V @ 1A) the board generates over 6W of heat with a 24V power supply and over 7W with an 18V power supply.

No one wants fans in their device. They're loud, prone to breaking and are a perfect way for dust to get inside. So, here are a few suggestions on how to cool the PSU passively.

If your case is made from metal, probably aluminum, you've hit the jackpot.  Aluminum transfers heat so well that your cooling problem reduces to transferring heat from the board to the case. I suggest 3 solutions:

• Thermal pads - putting thermal pad and securing the board to the case, while applying the right pressure to the pad. Keep in mind that the capacitor's connectors and leads are sticking out, so make sure either the thermal pad is smaller, located in the center of the PCB, or it's electrically insulating (most of it is electrical insulation).

• I don't recommend any crazy products like thermal grizzly, just cheap, generic ones. The board is much bigger and it produces less heat than "typical" application requires. Thermal pads that I found are 3mm thick and have a thermal conductivity of 6W/mK. If applied only in the center of 43x20mm rectangle, it'll make board merely 5.8K hotter than case.

• Air coupling - if you mount the board just above a flat surface of the case (1-2mm away), it'll improve cooling significantly. You'll still have to allow some airflow around, but air movements will transport much more heat thanks to nearby cold metal. Keep in mind that the board has pins at the bottom which limit minimum distance. This method mostly suits for the low profile variant that can be mounted really, really close to the case.

• Metal spacers - 3 mounting holes on the board are actually connected to the ground plane that distributes heat around the board, hence they're perfect for cooling the board. By using short copper/aluminum spacers, heat can be transferred from PSU to the metal it is mounted to.

For those who prefer the aesthetics of the wood, I have some bad news: cooling is harder. Now all the heat must be taken away by airflow around the board. Hence, more thought has to be put into it

• Mounting board horizontally is better than vertically thanks to the chimney effect increasing convection
• Lifting the board even by 1cm allows air to cool both sides a bit. The top side releases heat mostly through components which can only do so much good. The bottom side is designated primarily for distributing heat so it's crucial to cool it too, I'd go as far as to say:
• If you're limited on space and/or can't mount the board for good cooling, I suggest ordering reversed variant with all connectors on the bottom side. Now mounting it upside down, it will increase cooling, especially if you attach the included radiators!
• Experiment! 

If you have a thermal measuring device like a thermocouple or a thermal camera make sure that the middle of the bottom side part of the board (marked in red) is below 81C.

thermal reference point

To summarise, it's possible to cool the board passively, but it requires open airflow around most of the board and lifting it up. If you cannot cool down the board at a full load, there's one crude solution left. With a smaller load, so is the heat generated, to that effect, cooling becomes easier. Try spliting the case into two smaller ones, and powering them separately with 2 PSUs. At least consider slowly spinning noctua fan - it's quiet and it increases air flow a ton.

One should not be scared to experiment with thermals. All ICs have thermal shutdown so nothing bad will happen. Thanks to it, even when insufficient cooling is applied. It should be noted that the hotter the electronics are the shorter their lifespan. This is especially true for included electrolytic capacitors, so try cooling them as best as you can. One of the ICs even specifies that it can work up to 165C, but above 125C, its lifespan is shortened.

OEM variations and types

Board comes in two main variants for two types of cooling:

• Case cooling comes with thermal pads and all connectors on the top
• Air cooling comes with aluminum radiators and connectors on the bottom

Air cooled variant (radiators not shown)

 

Case cooled variant (case simulated with big aluminum block)

As there is a wide range of power supply needs product is highly customizable. Many modifications are possible, of which the most popular are:

• Changing voltages. Board can output a wide range of both positive and negative voltages. Anything between 3V and Vin-1V can be selected on positive rails, and anything from -1V up to -36V is possible (though some of the range with reduced current capability)
• Reversed design. As mentioned above, sometimes it may be beneficial to mount the board upside down to help with cooling. Obviously, in the current version, it would be very uncomfortable to use it. Power connectors would have screws at the bottom and the ecaps require the board to be elevated by a good bit. You can counteract this by buying a variant with the connectors and ecaps soldered on the bottom
• Lower profile. For designers of slim cases, the height of electrolytic capacitors might be unacceptable. There are few possibilities to deal with this. Capacitors can be bent 90 degrees under the board, or removed completely. As expected, every decision has certain downsides (otherwise they wouldn't be included in the design).
• Folded capacitors. Height is lower, and ecpas can be used as spacers so mounting is simplified to 4 pieces of double sided tape. Sounds too good to be true? Unfortunately so. When the ecaps are in the better contact with the board, they operate at temperatures closer to the board's temperature. If cooling is not increased to compensate, then their lifespan is shortened.
• Ecaps replaced with MLCCs. This makes the board much smaller, but at an increased price. High density ceramic capacitors are simply more expensive than ecaps (around 10 euro per board)
• Removed capacitors. If design can tolerate tripping of power supply ripple, then this is the simplest solution. Just make sure that higher ripple voltage is not a problem.
  

Power source

There aren't many requirements for the power source, but a single most important is be of a sufficient quality. When it inevitably fails, it won't throw mains power at the output. The board requires a power supply in the 18-26V range, with sufficient wattage to satisfy the board’s power draw at a maximum load, while having some headroom left  (assuming 80% efficiency for negative rail, 90% for positive rails and 30mA of quiescent consumption). All cheap 24V power bricks and most old laptop power supplies (usually 18 or 19V) should work. Bear in mind that this board is optimized for a 24V input so at a low input voltage (18V) the board has a slightly lower efficiency and consequently, it requires a better cooling. Additionally, negative rail has a slight decreased amperage to around 0.8A.

Noise levels on all ports

Worst case from 200mA to maximum output power. Due to the EU energy saving regulations, the positive rails gets slightly worse at small loads. All measurements were made at the 24V input.

Input noise level: 19.8mVrms 72mVpp

-12V noise level: 3.21mVrms 40.8mVpp

+12V,+5V noise level: 19.7mVrms 68mVpp

Peak to peak values (Vpp) are considerably higher than the RMS values (Vrms) due to residual transients from switching. They're mostly gone once connections to the busboard, busboard, ribbon and module input capacitance are included.

Ordering guide

If you'd like to order PSU or you have any questions feel free to contact me on social media, on contact form on this blog, or by email:

my email: michal # elektryk # hotmail # ending

where # is (dot) and ending is standard .com