PS I wrote that I always do a nearfield measurement, to fine tune such a sub or speaker. This of course will take care of the DIYS bipolar problem of not having exactly the half capacity of the polars used. If you don't have a measuring mike, that may be a problem, but then you do no speaker design anyway.
The test track where I condemned the sound of back to back polarized caps was Steinway grand piano, including some solo top octave. At 1 vac into SP2-XT speakers. Obviously there will be no 8000-14000 hz content in a high filtered subwoofer.
About to pull the trigger on some 2740 uF polyprophylene caps to put series the SP2. https://www.digikey.com/en/products/detail/kemet/C44UOGT7274M32K/14324241 $148 ea. Should be 2 ohms impedance at 29 hz. A CS800x hulk to rebuild with decent DC protection relays would be cheaper, but use a lot more electricity than a 70 w/ch M-2600. Plus 16 On semi MJ21193/94 would be ~$160 alone. Will report how it sounds.
Last edited:
Per post #59, I hooked up one SP2 speaker through two used 5000 uF polar capacitors series plus to plus. I used the input that left the Peavey crossover in circuit. As far as I can tell on orchestral music from the FM radio, there is no difference to a direct wire at 1 Vav. My solo top octave Steinway piano LP (Peter Nero Young & Warm & Wonderful When I Fall in Love track) was stolen, so I will not be doing that test. Bass was okay on ZZ Top Eliminator. SP2(2004) are flat to 60 hz, -3 db at 54.
Last edited:
To avoid other's experiments with 2500 uF series speaker caps, about the 4th song into ZZ Top Eliminator the M-2600 amp went silent with the light still on the power switch. There are thermal switches on both sides cutting off transformer AC input. Although the exterior heat sinks were room temperature, I presume heat is what took out the sound. CD player kept running and after I turned the amp power off and restrung the direct wire speaker cable (about 4 minutes) sound returned. M-2600 does have 2 inductor series the outputs to reduce sensitivity to capacitative loads. Not good enough the experiment demonstrates.
Digikey cart entry for 2740 uF polyprophylene caps zeroed out. I'm examining my inventory to build a R J Keene nfet speaker disconnect, which is cheap but about 12 parts per channel. A dual flip flop to remember the fault, a power up set, a complimentary optocoupler to rapidly discharge the gate of the fet in a fault, a red green LED and LED driver run the parts count up. Some power supply modifications for the flip flop LEDs etc. Most expensive part the Panasonic APV1122 photovoltaic optocoupler, in DIP package. Yes other part # are cheaper, but I will not process surface mount. What is spooky, I have never seen a commercial amp schematic with 2 of these included.
Digikey cart entry for 2740 uF polyprophylene caps zeroed out. I'm examining my inventory to build a R J Keene nfet speaker disconnect, which is cheap but about 12 parts per channel. A dual flip flop to remember the fault, a power up set, a complimentary optocoupler to rapidly discharge the gate of the fet in a fault, a red green LED and LED driver run the parts count up. Some power supply modifications for the flip flop LEDs etc. Most expensive part the Panasonic APV1122 photovoltaic optocoupler, in DIP package. Yes other part # are cheaper, but I will not process surface mount. What is spooky, I have never seen a commercial amp schematic with 2 of these included.
Last edited: