P. Short
07-01-2007, 09:33 PM
Over the last few days I've spent a bit of time working with Wayne about the transformer board. The initial focus of the discussion was adding an RS232-RS485 converter on the board, as well as an extra RJ45 connector to make it easy to use this board to provide power for the second or later controllers in the daisy-chain. At this point I am happy with these aspects of the design.
Then, we noticed that the board was just a tiny bit larger than the 3.8"x2.5" ExpressPCB mini-board specs, which in turn led to a closer examination of the transformer dimensions. In this process, we ended up choosing a 6.3VAC transformer followed by a full-wave rectifier.
That brings up a problem...that of ensuring a high enough voltage on the regulator of the Renard64 without exceeding the temperature rating. This problem generally applies to the Renard64 with the PWM code and all of the LEDs mounted when power is brought in through the RJ45 connector. It is not an issue if the non-PWM code is used, or the current is otherwise kept below 100 mA (when you can use an LP2950 LDO regulator).
The problem is this. With a 6.3VAC transformer, the peak output voltage is 8.8V at full load (6.3V * 1.4). There are three diodes in the path to the regulator, which would drop 1.8V total if the diodes are Schottkey diodes like the 1N5819. This leaves 7V on the regulator, which is exactly what is needed. This leaves no margin for voltage drop in the cables, and no margin for a drooping AC line (such as when you turn on 50A or 100A worth of Christmas lights). At the same time, this causes the regulator to dissipate 2W (if the current is 1A).
So, for the moment, I am investigating this situation. I'll get the transformer and see what it actually puts out, and measure the actual current draw of the Renard64 that I have here. And look for other parts or solutions to the problem.
For now, my recommendation is to use a 5V external power supply for the Renard64 (either a PC supply, a lab supply, or an open-frame power supply), if you want to use the PWM version of the firmware. Otherwise, use the older transformer board design if you don't need the PWM firmware.
--
Phil
Then, we noticed that the board was just a tiny bit larger than the 3.8"x2.5" ExpressPCB mini-board specs, which in turn led to a closer examination of the transformer dimensions. In this process, we ended up choosing a 6.3VAC transformer followed by a full-wave rectifier.
That brings up a problem...that of ensuring a high enough voltage on the regulator of the Renard64 without exceeding the temperature rating. This problem generally applies to the Renard64 with the PWM code and all of the LEDs mounted when power is brought in through the RJ45 connector. It is not an issue if the non-PWM code is used, or the current is otherwise kept below 100 mA (when you can use an LP2950 LDO regulator).
The problem is this. With a 6.3VAC transformer, the peak output voltage is 8.8V at full load (6.3V * 1.4). There are three diodes in the path to the regulator, which would drop 1.8V total if the diodes are Schottkey diodes like the 1N5819. This leaves 7V on the regulator, which is exactly what is needed. This leaves no margin for voltage drop in the cables, and no margin for a drooping AC line (such as when you turn on 50A or 100A worth of Christmas lights). At the same time, this causes the regulator to dissipate 2W (if the current is 1A).
So, for the moment, I am investigating this situation. I'll get the transformer and see what it actually puts out, and measure the actual current draw of the Renard64 that I have here. And look for other parts or solutions to the problem.
For now, my recommendation is to use a 5V external power supply for the Renard64 (either a PC supply, a lab supply, or an open-frame power supply), if you want to use the PWM version of the firmware. Otherwise, use the older transformer board design if you don't need the PWM firmware.
--
Phil