Steering Pump Moan Noise

1
Steering Pump Moan Noise
2
The Problem

• Noise from a Power Steering Pump
• is multi tonal
• has background masking
• is speed dependant
• Need for a Standard Test
• easy to use
• repeatable
• single value ( Effective Tonal Noise)
• Noise is a buy parameter

3
Test Criteria

• Operational range of pump
• Different vehicle conditions
• Microphone position

4
Vehicle Variations

• Multiple runs over identical conditions
• 3 runs found sufficient
• Simple and repeatable test
• use stationary engine run-up test with loaded
  steering
• Vehicle stationary, steering turned to achieve
  constant 50Bar at pump outlet
• Engine speed raised from idle to 3000rpm slowly
  during recording
• Prosig System provides monitoring of Pressures
  and automatic start/stop of recording at
  specified engine speeds

5
Microphone Position

• Physical aspects
• driver head height
• highs and lows (nodes anti-nodes)
• Single standard position ?
• No, too variable
• Binaural ?
• Still variable
• Selected 6 microphones - in a sphere

6
The Test

• 18 time history signals
• Six microphones and three runs
• Data is speed dependant
• waterfall analysis ( A weighted)
• dominant orders (10th, 20th 30th for example
  pump)
• Tonal Noise
• 1 RMS Order cuts
• Background Noise
• 1/3 Octave (23) Order cuts

7
Waterfall
8
1 Cut (Noise)
9
23 Cut (Background)
10
The Data

• 18 time history signals (21 with Tacho)
• 3 vehicle runs 6 microphones
• 18 Waterfalls
• typically 2048 point FFTs every 20RPM
• 3 Dominant Orders
• Orders 10, 20, 30 (example pump had 10 blades)
• Extract 1 and 23 Order cuts
• each order for each waterfall
• 108 Signals (54 pairs) !!!!

11
The Analysis

• Vehicle variability
• Average of the 3 runs
• Take mean of all orders on all microphones
• Both 1 and 23 cuts
• Reduces to 36 signals (18 pairs)
• 6 positions
• 3 dominant orders
• 1 tonal noise 23 background noise

12
Averaged Noise(1 cuts)
13
Averaged Background(23 cuts)
14
The Analysis

• We now have
• Six microphones
• Orders 10, 20, 30
• Next we.
• Take the maximum across microphones

15
Maximum Noise1cuts, 10th Order
16
Maximum Background23 cuts, 10th Order
17
The Analysis

• We now have data
• on noise and background
• for orders 10, 20, 30
• averaged across three runs
• with maximum from all microphones
• Now reduced to 6 signals (3 pairs)

18
The Analysis

• To complete the analysis we
• reduce the background by 3dB (masking)
• subtract background from noise
• This leaves 3 signals (1 per dominant order)
• Finally...
• define everything above 0dB as heard
• sum all values above 0dB together
• Gives one value as the
• EffectiveTonal Noise
• Correlates well with subjective ratings!

19
Effective Tonal Noise
20
End