Tony tests out the Gryphon Quasar HV BEC

A couple weeks ago at the field, Timmy and I had a short discussion about BECs. He decided to send me a Gryphon Quasar HV BEC for testing.

There have been a few reports in the past were the original Quasar users have claimed the BEC failed, and damaged their electronics. The manufacturer, Gryphon RC, revised the BEC, making it more robust. The revised unit is more tolerant to high frequency input ripples (caused by the ESC during PWM switching), and high frequency and high current load transients (high power brushless servos etc).

The unit Timmy sent me was the v2, revised unit. It can be seen below.
http://www.solarfreaks.com/bec/001.jpg

Anyways... The specs for the unit are:
Input Voltage Range : 11V ~ 75.6V (3~18 Cells)
Output voltage Range : 5.2V, 6.0V, 6.8V, 7.4V, 8.4V Selectable
Voltage ADJ : 4.8V~7.7V ADJ (Potentiometer)
Output current(Peak) : 20A
Output current(Cont) : 10A
Low voltage alarm setup : 10V~75V (Available GDB-1010 or GDB-1030)
Size : 12mm(H) x 31mm(W) x 52mm(L)
Weight : 20g (Excluding Cables)

The first thing i did was pop it open to have a sneak peak inside.
http://www.solarfreaks.com/bec/02.jpg

The PCB layout looks nice and neat, and the unit looks well assembled. The PCB has a nice gold finish, which not only prevents corrosion, but looks cool too .

The PWM controller IC is an LM5088, from National Semiconductor. I know a few guys on the electronics forums that have used it successfully in multiple circuit designs. The only downfall of this IC is its a non-synchronous controller, so at higher input voltages we will have more power lost (heat) across our free-wheeling diode in the converter. Synchronous designs use two FETs as opposed to one FET and one Diode in the converter. The non-Synchronous design, however, has the advantage that it wont blow up if you accidentally plug a receiver pack in conjunction with the BEC, and it wont adsorb reverse voltages. So we can say its a more idiot proof design.

After analysing the circuit, i thought 10Amp continuous load current at 75V input is a bit too optimistic. Especially with only that tiny red heatsink. So the testing begins...

I wanted to test a few things:
1) Output noise at 10Amp load with 50V input (12S Lipo)
2) Temp rise at 10Amp load with 50V input after 1 min. (lets face it, our servos are never going to draw more than 10Amps for more than a minute straight)
3) Response to input voltage transients.
4) Response to Output Load transients.
5) Voltage regulation at Peak Current (20 Amps).
6) Short Circuit Situation

Testing Methods:
1) Applied a 10Amp load (using my Variable load - Based on my MOSFET design, as can be seen in the pic), and measured the noise using my oscilloscope.
2) Applied a 10Amp load and measured the temp rise after 1 min.
3) I used my current limited power supply to power the BEC, Connected a FET across the input terminals of the BEC, and used my function generator to apply 0->5V noise at the FET gate, causing voltage ripples on the supply.
4) I actually tested the unit on a set of outrage torq servos, and quickly pulsed the cyclic with the servos under load, and used my oscilloscope to monitor the output voltage. I also loaded the BEC with some halogen down-lights.
5) Monitored the voltage during 20Amp peaks using my oscilloscope.
6) Just shorted the output leads.

Results:
1) Only 40mV low frequency ripple, and a 350mV high frequency ripple. This is pretty good. The high frequency noise will be filtered away by the ceramic caps in your Rx and Servos, so they will only see the low freq ripple of 40mV. And remember this is at 10Amp load. The Voltage was rock solid at 7.4Volts (+/- 0.02volts).
http://www.solarfreaks.com/bec/1.jpg

2) Only 34°C rise after 1 min @ 10Amp load with a 50V supply. I quickly gained confidence that the unit will easily handle the current at the rated 75 volts. Looks like the diode used is a later generation schottky diode, with low forward voltage drop. Not bad at all..
http://www.solarfreaks.com/bec/2a.jpg
http://www.solarfreaks.com/bec/2b.jpg

3) With input voltage ripples, High and Low freq, the output remained solid at 7.4V. The picture below shows the dual tracking powersupply in series mode, giving us ~50V to supply the BEC.
http://www.solarfreaks.com/bec/3.jpg

4) My servos wouldnt draw more than 6.5 Amps from the BEC, even under stall. The voltage was rock solid.
http://www.solarfreaks.com/bec/4.jpg

So i connected some downlights to the BEC. The voltage dipped slightly... then quickly rose. This dip is probably because down-lights are basically a short circuit when cold.
http://www.solarfreaks.com/bec/4b.jpg

5) At 20Amps the voltage at the BEC terminals was 7.39Volts. The voltage on the end of the wire was 7.32Volts. The unit has a Dual feed going to Rx. This minimises the droop and also provides redundancy.
http://www.solarfreaks.com/bec/5.jpg

6) The ULTIMATE TEST... The unit survived a direct short for 20 seconds. I directly shorted the outputs and measured the peak current. My Clamp meter outputs 1mV per 1Amp, and the Fluke registered 27.75Amps peak under short circuit. The BEC limited the output current by dropping the voltage, and saved itself.
http://www.solarfreaks.com/bec/6.jpg

Basically, this unit is very efficient, regulates perfectly, and is very robust too. I never had a chance to see or test version 1 of this BEC, so i cant comment on it. However, after a combination of efficient and robust operation during testing, i have gained confidence in the unit and it has earned a place in my Goblin. Nice one.