By Pete Basel    9-12-2008
Classic SPICA TC-50 Loudspeaker Restoration
Part 2 - Crossover and Driver Measurements
Detailed MeasurementsMore detailed measurements were taken to determine what is required to realign the woofer notch filters, and to better match the sensitivities. Often, a large measured difference is caused by the accumulation of several different errors. A difference of more than 3 dB is seen between the two systems in the 2 to 3 kHz range and it was found that about 1 dB is due to the basic woofer sensitivities being different. System B's woofer crossover shows about 1 dB more baffle step compensation, accounting for 1 more dB. Finally, the woofer resonant peak has moved and is lower in amplitude indicating that the notch filter must be re-tuned in order to eliminate excessive attenuation at the upper end of the woofer passband.
Individual Driver Acoustic Output Through Crossover
The drivers were disconnected, one at a time, so that the output from each could be measured independently. The frequency response plots follow for systems A and B:
Systems A and B tweeter outputs overlaid in order to show how well they match. The crossover point is about 3 kHz, and the match is quite good from 3 to 12 kHz:
Fig. 2.1    Systems A and B Tweeter Frequency Response Through Crossover
Systems A and B woofer outputs overlaid in order to illustrate the reduced output from the B woofer previously noted. The piston range is mismatched by about .5 to 1 dB, this will be corrected by the addition of a bucking magnet on woofer B. The diminished output from woofer B is even worse from 1.5 to 3.5 kHz where it reaches a maximum of about 4 dB. This will be corrected, as much as possible by retuning the woofer notch filter:
Fig. 2.2    Systems A and B Woofer Frequency Response Through Crossover
Fig. 2.3    System A Woofer and Tweeter Frequency Response Through Crossover
Fig. 2.4    System B Woofer and Tweeter Frequency Response Through Crossover
Crossover Loaded Transfer Functions
The loaded crossover transfer functions were measured in system by bringing test leads out from the terminals of each driver. These transfer functions will provide a baseline check against the rebuilt crossovers. Note that the tweeter response dips around the resonance notch frequency as I also noted in the simulations. It can be seen that this dip appears to match a peak in the acoustic output, probably due to cabinet diffraction, that can be seen in Fig. 2.10 around 800 to 900 Hz. Frequency response plots follow for systems A and B:
Fig. 2.5    System A Tweeter Crossover Loaded Frequency Response
Marker #1 in the following curve shows the crossover notch frequency as 3.65 kHz, very close to the system service code of 3.7. However, note that the peak in woofer A has shifted, probably due to aging of the cone treatment, and is now lower, at 3.19 kHz as shown in Fig. 2.11.
Fig. 2.6    System A Woofer Crossover Loaded Frequency Response
Fig. 2.7    System B Tweeter Crossover Loaded Frequency Response
Marker #1 in the following curve shows the crossover notch frequency as 3.65 kHz, very close to the system service code of 3.7. However, note that the peak in woofer B has also shifted and is now lower, at 3.39 kHz as shown in Fig. 2.13.
Fig. 2.8    System B Woofer Crossover Loaded Frequency Response
Woofer A and B crossover frequency responses are overlaid in the following figure. The B woofer crossover provides about one dB more attenuation around 2 kHz which contributes to the diminished system output in that range. The cause of this difference will be determined and corrected when the individual crossover components are measured.
Fig. 2.9    Systems A and B Overlaid Woofer Crossover Loaded Frequency Response
Driver Acoustic Output Driven Directly
Frequency response plots follow for the tweeter and woofer outputs - systems A and B, driven directly bypassing the crossover. These measurements were taken with the felt and grille in place, on the optimum listening axis:
Fig. 2.10    System A Tweeter Frequency Response - Directly Driven
Marker #2 in the following curve shows the peak in woofer A as 3.19 kHz which does not match the woofer service code of 3.7. The peak in woofer A has shifted lower, probably due to aging of the cone treatment.
Fig. 2.11    System A Woofer Frequency Response - Directly Driven
Fig. 2.12    System B Tweeter Frequency Response - Directly Driven
Marker #2 in the following curve shows the peak in woofer B as 3.39 kHz which does not match the woofer service code of 3.8. The peak in woofer B has also shifted lower.
Fig. 2.13    System B Woofer Frequency Response - Directly Driven
The following curve shows woofer A as having a peak that is about 3 dB higher than woofer B. This indicates that the drivers have shifted significantly from their original specification since woofer B has a higher service code peak of 10 dB and woofer A of only 8 dB. The peak in woofer A should be 2 dB lower than that in woofer B however it measures 3 dB higher:
Fig. 2.14    Systems A and B Woofer Responses Overlaid - Directly Driven
Fig. 2.15    Systems A and B Tweeter Responses Overlaid - Directly Driven
Driver Thiele and Small Parameters
System A Tweeter Thiele and Small ParametersSPICA TC-50 SYSTEM: 13593A Sept 22, 2008 1" Audax Tweeter: Markings: TWO25D04 TTP0010A2 120974W 22304 Rdc 3.22 ohms Fs 837.2 Hz Re (LAUD) 3.48 ohms Qe 1.03 Qm 4.04 Qts .821 Lead in wires are solid. Front plate thickness: 3mm, front plate bevels down about ~1.5mm Back plate thickness: 3mm VC wind height: 1.5 mm VC ID: 1.0" VC OD: 1.02"
Fig. 2.16 System A Tweeter Free Air Input Impedance
System A Woofer Thiele and Small ParametersSPICA TC-50 SYSTEM: 13593A Service Code = H 3.7-8 Sept 18, 2008 6.5" Audax Woofer: Markings: Polydax Made in France REF: HIF17JVX4(ohm PN: 120076P SN: 0330200506 H 3.7-8 written in blue pen on magnet Rdc 3.45 ohms Pd 14.5 cm (manufacturer spec) Fs 36.5 Hz Vas 47.2 liters Re (LAUD) 3.76 ohms Qe .49 Qm 1.64 Mms 15.5 grams no .451 % SPLref 88.5 dB Bl 5.23 T-m Qts .376 Cms 1.23 mm/N
Fig. 2.17 System A Woofer Free Air Input Impedance
System B Tweeter Thiele and Small ParametersSPICA TC-50 SYSTEM: 13593B Sept 22, 2008 1" Audax Tweeter: Markings: TWO25D04 TTP0010A2 120974W 22304 Rdc 3.30 ohms Fs 838.1 Hz Re (LAUD) 3.71 ohms Qe .98 Qm 2.71 Qts .720
Fig. 2.18 System B Tweeter Free Air Input Impedance
System B Woofer Thiele and Small ParametersNote that this is a replacement woofer that does not match the system service code of H 3.7-8: SPICA TC-50 SYSTEM: 13593B Service Code = S 3.8-10 Sept 22, 2008 6.5" Audax Woofer: Markings: Polydax Made in France REF: HIF 17JVX 4* 2C A12 120076P 29812 S 3.8-10 written in red pen on magnet Rdc 3.28 ohms Pd 14.5 cm (manufacturer spec) Fs 34.6 Hz Vas 47.0 liters Re (LAUD) 3.62 ohms Qe .544 Qm 1.86 Mms 17.2 grams no .345 % SPLref 87.4 dB Bl 4.99 T-m Qts .421 Cms 1.23 mm/N A bucking magnet was temporarily added to woofer B, and the woofer was retested with the following results. The goal is to increase the drivers sensitivity; without the magnet it is about 87.4 and with 88.2 dB - only .3 dB less than woofer A, as compared to 1.1 dB less without the bucking magnet. Note that these measurements are difficult to make and would have to be repeated with averaging to get a more accurate result. Other expected differences include: motor strength - Bl increases from 4.99 to 5.41 T-m A woofer: 5.23 T-m Basic Efficiency no increases from .345 % to .418 % A woofer: .451 % Driver Qts (from Qe) decreases from .421 to .365 A woofer: .376 Fs 34.9 Hz Vas 47.3 liters Re (LAUD) 3.65 ohms Qe .461 Qm 1.76 Mms 16.9 grams no .418 % SPLref 88.2 dB Bl 5.41 T-m Qts .365 Cms 1.23 mm/N
Fig. 2.19 System B Woofer Free Air Input Impedance
More to come as time permits.