Overhaul maintenance of super-tweeter FOSTEX T925A and further signal fine tuning thereafter
Hello friends, Happy New Year from Japan!
The super-tweeter (ST) Fostex T925A was launched in audio market in April 1994, and it has been keeping good reputation until today. It is amazing Fostex still has
T925A in their product lineup, current price tag is JPYen 55,000 inclusive of 10% sales tax (somewhat cheaper in several web stores).
View attachment 176150
Please refer to
my post here for the unique physical alignment (positioning) of T925As in my system setup.
I purchased a pair of T925A in 1996, and have been using with great satisfactions over 25 years. Last month, early December 2021, I contacted with Fostex Company, with no expectation of positive response though, inquiring the possibility of overhaul maintenance of my T925As.
They very quickly responded; "We thank you and very proud of your long-year loyalty of T925A. We are very happy to give overhaul maintenance of your T925As!" Then my T925As were hospitalized at Fostex service center for a week, and came back to my home in 10 days. Fostex completely disassemble T925A, cleaned-up every parts, measured, and replaced some of them, including the inner thin connection wires, with new ones.
I very carefully checked the refreshed T925As confirming the better sound quality, and then intensively simulated/calculated the high-pass (low-cut) capacitor configuration in SP level signal going into T925As; now I decided using series of 3 microF and 10 microF capacitors as shown here;
View attachment 176151
The series of 3 microF and 10 microF make total capacitance of 2.308 microF giving further -6 dB/Oct high-pass (low-cut) at
ca. 8,611 Hz for the SP level signal going into T925A (impedance 8 Ohm) which already had -12 dB high-pass (low-cut) at 6,000 Hz by digital software crossover EKIO. (As for the use of 22 Ohm parallel resistors in the lines, please refer to
my post here and
here.)
All of the other elements of the configuration remain unchanged;
View attachment 176153
Then, firstly, I measured the frequency response (FR) of the SP high-level signals
after the capacitors, just
before going into SP drivers, just like
I did measured as shared here, this time using another Hi-to-Low converter Audio-Technica
AT-HLC130 , and TEAC
TASCOM US-1x2HR Audio Interface, and Adobe Audition 3.01 on the second Windows 10 Pro PC;
View attachment 176304
The FR of SP high-level signals to go into the super-tweeters (STs) and tweeters (TWs) were "measured" in good conformity with my simulation/calculation;
View attachment 176183
I also measured the SP high-level signals to go into beryllium squawkers (SQs), woofers (WOs), and the line-level signal to go into active sub-woofers giving this total and individual FR curves;
View attachment 176156
Please note that the sensitivity (efficiency) of T925A is rather high in 108 dB (1m, 1W2.8V), and this means T925A sings 15 dB louder than the Yamaha 3cm Beryllium dome tweeter JA-0513 of 93 dB (1m, 1W2.8V) with a same gain/volume input signal.
As usual, I also measured
the actual room air sound at listening position using the measurement microphone BEHRINGER ECM8000, just like I shared
here, and
here through
here;
View attachment 176157
The total and individual SP driver air sound FR curves at my listening position, with L & R SPs singing, were "measured" as shown in this diagram;
View attachment 176158
Finally, the latest shape of the best tuned total FR;
View attachment 176159
The above latest total FR as of January 2, 2022, is essentially identical to the one I shared
here, but now I have the refreshed sound of T925A with better/improved variable gain control.