Sorry for raising this subject as I know there are masses of posts on this, but I have NEVER found one that asks the following. Does the loudspeaker's crossover inductor (very large value) override the small value of the Zobel one, thus making it unnecessary? In simple terms, the reason for including one is that if the amp drives a capacitive load it could oscillate? Surely 99% of all loudspeakers are mostly inductive, with the large value inductor right at the start of the x-over filter?... I use a an LJM design that does not use any O/P coil, and it runs perfectly. Thanks for any replies.
Last edited:
You also have the loudspeaker cable to consider. It consists of two conductors with an insulator in between, just like a capacitor, although it also forms a wire loop and therefore also has inductance.
The impedance of a loudspeaker depends much on how the crossover was built.
It migh have just the big unductor in series with woofer, thus offering major inductive reactive load, but might, and normally has, other components in parallel to drive the mid-range and tweeter. A good crossover design can have a neutral reactive impedance.
Look at the latest loudspeaker I built last weekend - the measurement shows that it is not predominantly inductive, instead, it's more neutral except in the speaker resonant frequency range where it alternates between inductive, resistive and capacitive.
I have had an amp without an output coil parallel a high wattage resistor, pick up sports talk AM radio from the speaker line. Happened when the ruling elder added 30 feet speaker wire and moved the reeds speaker to the opposite side of the stage from the Allen organ. Installing 10 turns around a parallel 10 ohm resistor inside the amp chassis fixed the problem. AM radio station was 3 miles away.
The potential oscillation we talk about in this situation is high frequency oscilation, that is beyond audio frequency. The speaker impedance plot is not that useful.
Conventional wisdom is to put an inductor no bigger than 5uH. It completely prevents the oscillation in the case of capacitive load, although the chance of capacitive load is rare at the first place.
Conventional wisdom is to put an inductor no bigger than 5uH. It completely prevents the oscillation in the case of capacitive load, although the chance of capacitive load is rare at the first place.
Thanks very much everyone. You have convinced me it is better to have a coil on the output. I was putting together another LJM L20 V 9.2 amp (it sounds so clean and transparent) and wanted to be sure that a coil is a sensible mod to add. I can see the Douglas Self "Blameless" amp has one and the L20 is a very close copy to it. What diameter and gauge should I make it please? Also do I need the 10 Ohm resistor INSIDE the inductor. I can see the blameless amp does not.... it is at the side
Attachments
Last edited:
BTW, the RC net work should be put at the speaker side of the inductor, not the the amp side. I know most people won't agree on this, but this is true.
Investigating Zobel network.
All below are simulating without the speaker load. Somehow a tiny capacitor is added to the output.
Putting RC on the amp side, not stable.
Putting RC on the speaker side, a little overshoot, not that bad.
Without RC, not stable.
Without the inductor, only RC, it is marginal stable.
People keep copying and pasting Zobel, but mostly are wrong, including me.
All below are simulating without the speaker load. Somehow a tiny capacitor is added to the output.
Putting RC on the amp side, not stable.
Putting RC on the speaker side, a little overshoot, not that bad.
Without RC, not stable.
Without the inductor, only RC, it is marginal stable.
People keep copying and pasting Zobel, but mostly are wrong, including me.
Thanks very much jdxking. This would make adding a coil quite a bit trickier in terms of needing to cut the PCB track rather than just adding it to the o/p terminal... Wonder what others think?
10 ohm 5 watt wirewound resistor is convenient diameter to wind output inductor around. Need not be inside. Can use AA battery or china marker as a coil form. 8 to 14 turns of 22 ga wire or thicker for 100 w amp. 800 watt amps use 16 ga wire. Need not be magnet wire, insulated wire works. Peavey says 14 turns of 16 ga wire 1 cm diameter is 800 uH.
On the Allen organ, I drilled one hole for # 6 screw through the back near the speaker output. One terminal of coil+resistor soldered to output terminal inside the amp. Combination held to the metal wall by a tie-wrap with screw hole. Wire from amp board that used to go to speaker output terminal is soldered to the input end of the coil+inductor.
More secure mounting provisions necessary for amps that are set up, torn down, and transported every night. Organ amps stay put most of their lives.
On the Allen organ, I drilled one hole for # 6 screw through the back near the speaker output. One terminal of coil+resistor soldered to output terminal inside the amp. Combination held to the metal wall by a tie-wrap with screw hole. Wire from amp board that used to go to speaker output terminal is soldered to the input end of the coil+inductor.
More secure mounting provisions necessary for amps that are set up, torn down, and transported every night. Organ amps stay put most of their lives.
The RC network is to keep the amplifier loaded at high frequencies. It should be before the inductor.
Thanks very much indianajo for all the information. I've checked the new boards and have room near the central output tag to fit the coil and resistor. Much obliged...
Piezo tweeters act as a capacitor that can cause an amplifier to oscillate.
Normally speakers have a small resistor in series with such tweeters but some can have missed.
Any way a speaker cord has an inductance of about 0,2uH per foot so with more then 10 foot speaker cables it should be safe without a coil.
Fast amplifier needs less inductance. I tested a small Mosfet amp that was stable with 1uF capacitors and just 1 foot speaker cable.
Normally speakers have a small resistor in series with such tweeters but some can have missed.
Any way a speaker cord has an inductance of about 0,2uH per foot so with more then 10 foot speaker cables it should be safe without a coil.
Fast amplifier needs less inductance. I tested a small Mosfet amp that was stable with 1uF capacitors and just 1 foot speaker cable.
Hi, if you mentioned my post, these frequencies are around the resonance of the cabinet.
It's increadible what the impedance curve can show - it's like an X-Ray 🙂
The inductor is needed for electrostatic speakers that present a capacitive load.
Ed
Ed
Ah, panel resonances. Presumably, your frequency sweep for that measurement was quite slow to allow thos resonances time to build up and appear. But yes, it's amazing what the electrical impedance curve shows. A lot of full-range drivers have a blip around 1kHz-2kHz that I put down to mechanical impedance mismatch between coil former and cone.
Yes, I use a slow sweep time of 10.9sec compared to some very quick sweep I see in some videos on YT.
