Voltage Regulator Test Results. Graph 1: Low Voltage Dropout Characteristics, 1.2A load. Graph 2: Thermal Performance, 2.3A load

Transcription

1 Voltage Regulator Test Results Test performed by: Chris Bajorek (RCG: SoaringDude) Test date: 01/18/2012 last update: :30am added pulsed load test results Product tested: Novak 5465 Regulator type: Linear Input voltage: 2-S Li-Po / 4-6 NiMH/NiCd ( 26v max per the datasheet) Output voltage: 6.0v fixed Load, continuous: 5.0A (assuming adequate exposure to ambient air) Load, surge: 7.5A typical (from IC data sheet) Weight: 12.7g Dimensions: 17.8 x 25.4mm Website: Regulator IC used: Micrel 29502BT Minimum load current: 10 ma (from IC data sheet) Product submitted by: Robert-CSD (RCG) Conditions set for all tests: Voltage regulator = 6.0v (fixed) Graph 1: Low Voltage Dropout Characteristics, 1.2A load Vin DROPOUT PERFORMANCE: Very good. As Vin drops below 6.3 the regulator falls out of regulation while still delivering continuous voltage and current. Below dropout was always ~0.4v less than Vin. Vin- Vin- Degrees (C) Picture above: Voltage regulator size compared to a quarter and a 1450mAh LiFe 2S Hyperion battery pack. Graph 2: Thermal Performance, 2.3A load Time (min) Open air Restricted air THERMAL PERFORMANCE: Ideally we could translate the above measured temperatures into junction temperatures and know whether we are exceeding the regulator IC limits. Unfortunately this is more complex than time permits so the goal is to get a relative understanding of different thermal conditions under an unusually high continuous load. Bottom line is this: under normal competition servo loads the Novak reg will do just fine even though the board does not use a heat sink. However it is highly recommended that you mount the unit with as much open fuselage air as possible surrounding it.

2 PROS: very small size, rock solid 3A to 4A regulation performance, fair price at $27. PULSE LOAD PERFORMANCE: (for comparative pulse load test graphs see separate report elsewhere on RC Info Share). See test details below. This critical test generates short very high current demands such as those found in a typical 6-servo competition sailplane in high-g maneuvers The ideal regulator in our application could handle such loads with minimal voltage sag. The Novak unit's high current performance was found to be fair. Beyond 3A loads it experienced a higher percentage of sag than many other units tested. On the plus side it did not impart any fatal characteristics and the load-on and -off transitions were fast and clean. However if your goal is to use a regulator that gives you among the best in high current transient performance this is not the one. CONS: (a) high current pulse performance is not among the best, (b) 28ga output wires are too small and cause a loss of 0.2v at a 2.3A load. They should be replaced with 22ga or 20ga wires to the receiver for best performance. (c) The included switch works fine but unit has a small current drain (10uA typ, 0.5mA max) even when switch is off. We prefer a switch in series with the battery lead. CONCLUSION: the Novak should work safely and well for 6-servo gliders with the caveat that during peak high current demands the regulator will not be able to deliver as much voltage or current as other units in the same price class. Knowing the regulator IC enabled us to see the undocumented (by Novak) operational limits. The circuit that was implemented by Novak was direct from the manufacturer's datasheet. EFFICIENCY: at a nominal battery pack voltage of 6.6v and an average current of 0.6A the regulator's efficiency is 89%. Most linear regs should run in this general ballpark. Table 1: Quiescent Current Test Vin Rload- ohms Iin- ma Pvr- mw 7.0 none Conditions Calculate VR power consumption with no load Table 2: High limit input voltage load test short duration Vin Rload- ohms Iout- A Pvr- W % % % % % Eff.- % Conditions 30 seconds at Pvr is power dissipated by VR each load unit % V drop caused by 28ga output wires. Probing direct to the IC's output pin 4 (using a pin to punch through the shrink wrap cover) resulted in the going from 5.74 to If the IC's ground pin were similarly contacted with the voltmeter neg. probe the full output of 5.95v would have been measured. Table 3: Nominal input voltage load test short duration Vin Rload- ohms Iout- A Pvr- W % % % % % Eff.- % Conditions 30 seconds at Pvr is power dissipated by VR each load unit

3 % V drop caused by 28ga output wires. Same notes as previous test. Table 4: Low Vin dropout test Vin Rload- ohms Iout- A Vin Conditions =6.0v Keep in mind wiring is also dropping some voltage. Table 5: Thermal Test: Open air Time (min) Vin Rload- ohms Iout- A Temp- C Conditions =6.0v VR on table facing up. Thermal probe taped to top of VR Load disconnected here Table 6: Thermal Test: Restricted air Time (min) Vin Rload-ohms Iout- A Temp- C Conditions =6.0v VR on table wrapped in 2 layers of thin dense foam. Thermal probe taped to top of VR (inside the foam).

4 Load disconnected here Date of test: 03/26/12 Table 7: Pulse Load test: Vin (no load) (load) Rloadohms Iload- A (calc) Vsag (v) Vsag (%) Vin - (load) Vbench set to 6.0v, Vbatt providing all voltage for pulsed loads Pulse Load test notes: 1. A range of loads were pulsed on and off to simulate real world hi-g servo loads using an actual LiFe 2S battery pack. 2. Target currents range from 30mA to as high as 9A depending on where the regulator's was set (lower 's yield lower load currents). 3. Vin = input v to regulator, =output v from regulator. 4. Vin drops as load current increases due to the 1450mAh LiFe battery internal resistance. 5. Vin in the results table above is measured during a brief period when each load is manually switched on. 6. measurements are taken from captured digital storage oscilloscope traces during load pulsing. Pulse Load Digital Storage Scope Screenshots: Notes: Yellow trace = from regulator Red trace = Load ON pulse signal Screen 1: Note that the voltage spikes seen on the leading and trailing edges of the load pulses are due to the fast load switching from the test bed MOSFETS and are not issues with the regulator.

5 Screen 2: