That will require a bloody big transformer.
And many large filtering caps.
Not cheap unfortunately.
The current peaks will be less than a milli-second. I doubt that heating will be a problem with music.
What breaks ribbons is puffs of air from handling the assembly. Magnepan ships their tweeters with a flexible rubber magnetic strip covering the tweeter to prevent air flow.
If the ribbon breaks, it is very quick and easy to replace. In the picture, you can see the screws/nuts holding small plates in place. Loosen the nuts. Slide out the remnants of the broken ribbon. String a new ribbon. Tighten the nuts. Back in business.
Maybe not. The average power may be very low. The caps may charge fast enough to supply the next peak without needing heavy iron. The tweeter likes to be crossed over no lower then 3kHz and does not like a gentle crossover slope.
Keep in mind that a tweeter supplies less of the acoustic watts of a typical speaker system than a woofer or a midrange. Typical music spectral content falls at 10dB / decade when looked at from 20Hz to 20 kHz if i recall correctly.
I am thinking that a few paralleled conventional TO220 MOSFETs may work in the complementary Citation configuration. Mouser has some TO220 N and P FETs that are in the 3-ish milli-ohm RDSon range and are rated around 200A. I have ordered a few of each to characterize.
I can not say for sure but something is off. As a comparison, we usually weld 0.8mm stainless steel plates at 10-20A, your previous comment about 300-400A will melt the tip of 1/4" steel rod on contact. Initial testing with much lower current say 20A-30A makes a lot of sense. you can easily add more power later.
Isn't speaker sensitivity normally measured at 1m distance? Like xxdB at 1W at 1m.
That would give better perspective and comparison figures.
Jan
That will be the goal of the first attempt; to create more than 20A into the ribbon. It will just be low duty cycle 3kHz pulses to see where the amp clips and how well it works into low a impedance load.
You can measure sensitivity with an amp and a calibrated mic, etc. You can also measure the T/S parameters and the sensitivity falls out the calculations.
I chose to measure the T/S parameters of the midrange. I have tested these drivers outdoors with a friend's LMS and they are quite flat to 3kHz. The acoustic test I performed was to compare the near field sensitivity of the two drivers. Near field measurements are less prone to room interaction. Measuring at 1m requires either an anechoic chamber or outdoors where there are no reflections. I chose the path of least resistance. (pun intended)
Your welders are continuously on when in use. Music is something like 10dB peak to average. In the treble, the power is much less than in the bass region.
The large current capability will be for transients like when a drummer is riding the cymbal with the drum stick.
Then 300A estimate makes no sense.
Consider a 300A into 0.5 ohm resistive load will deliver 180kW of instantaneous power or roughly 130kW RMS, 260kW stereo for any frequency. In your estimation, how small of power transformer is good enough to supply the amplifier? Would, in you estimation, a 36kW be enough for stereo? Do you have any idea of the weight of a 36.000VA transformer? A toroidal unit would be at about 150lbs. Nice industrial grade size equipment I would never think of using at home.
Something is seriously off. Measured SPL could be too low or not applicable for higher current, or something else. Check with 20A-30A source as is, then tweak your ribbon to yield better sensitivity. Going the power plant route would quickly get you into fantasyland nightmare.
However, I have to admit that some of us here are what NP in 2002 exactly say "We are fearless amplifier builders. A little heat doesn't scare us at all." Should you succeed (however unlikely), your 260kW amplifier system for with the pair of ribbon tweeters is monumentally impressive. People will compare you to Sakuma san. You may have to prepare for the hundreds probably thousands of people coming yearly to visit you for a listen. Just remember there are 5 of us from 5 countries already show interest in your fantastic thread.
Answers in context below...
The amp will not be 260kW. It will be 300A(or more if possible) with some reasonable compliance voltage, optimistically 20V p-p. That makes peak power 3 kw. Average power will be dictated by mechanical, electical and thermal engineering. 200 watts average seems reasonable for starters. Also finding suitable capacitors and allowing the di/dt to flow will be a challenge.
Then 300A estimate makes no sense.
***The 300A is not going to be RMS at 20 Hz. It will be peak for something less than 1 mS.
The 1/2 ohm tweeter is the fat diaphragm. The skinny diaphragm is more than 10X less resistance. ***
Consider a 300A into 0.5 ohm resistive load will deliver 180kW of instantaneous power or roughly 130kW RMS, 260kW stereo for any frequency. In your estimation, how small of power transformer is good enough to supply the amplifier?
***The power transformer only needs to supply average power. Peak power comes from local capacitors. This is similar to what bypass capacitors do digital circuits. ***
Would, in you estimation, a 36kW be enough for stereo? Do you have any idea of the weight of a 36.000VA transformer? A toroidal unit would be at about 150lbs. Nice industrial grade size equipment I would never think of using at home.
****Again, you are thinking average instead of peak. That said, there is such a beast. It is made by a company called Ballanine. It can do 100A RMS down to DC but the compliance voltage is 4V. The amp costs $10k, is very heavy and is way overkill for what I want to accomplish. ****
Something is seriously off. Measured SPL could be too low or not applicable for higher current, or something else. Check with 20A-30A source as is, then tweak your ribbon to yield better sensitivity. Going the power plant route would quickly get you into fantasyland nightmare.
However, I have to admit that some of us here are what NP in 2002 exactly say "We are fearless amplifier builders. A little heat doesn't scare us at all." Should you succeed (however unlikely), your 260kW amplifier system for with the pair of ribbon tweeters is monumentally impressive. People will compare you to Sakuma san. You may have to prepare for the hundreds probably thousands of people coming yearly to visit you for a listen. Just remember there are 5 of us from 5 countries already show interest in your fantastic thread.
The amp will not be 260kW. It will be 300A(or more if possible) with some reasonable compliance voltage, optimistically 20V p-p. That makes peak power 3 kw. Average power will be dictated by mechanical, electical and thermal engineering. 200 watts average seems reasonable for starters. Also finding suitable capacitors and allowing the di/dt to flow will be a challenge.
Please perform integral of 20khz 300A sinusoidal current for 1mS of the so called peak signal into 0.5 ohm to calculate power dissipated and tell me that it is way off the rms calculated value of ~130kW.
So what is the resistance of the ribbon you want to use for the amp? Is it decided yet?
Off course it is possible, not easy but no unsolvable problem. But don't expect any board to pass that kind of current. High current will be hard wired connection, a wire with at least 3/4" diameter copper core is recommended to minimize losses at 300A for speaker connection. You need to specify load resistance and how much more current you want. 1000A enough?
This is no joke, I am willing to help once you decide what you really want. Sensible or not. Whenever you are ready. No charge for design. Make sure you are serious and prepared to cover the cost of development and build. I would even build it for you if you have no problem paying for shipping it from Indonesia to your place. Large transformers can be custom ordered economically here.
As stated, the initial goal is to have +/- 10V rails so that the amp can play into a 30 milli-ohm load if desired. I ordered a 10-milli-ohm shunt to use as a load. Its a 4-wire shunt which will allow capturing the current peaks.
300A would not be for 0.5 ohm. 300A would be for 10 to 30 milli-ohm loads. 10V peak at 30 milli-ohms yields 333 amps peak. So, 300 to 400A is the goal. 10 milli-ohms would be considered the floor.
I will let you know when I have more details worked out. I have ordered some output MOSFETs to test. They are 40V, RDSon ~ 3 milli-ohms, 120A continuous drain current, 480A pulsed drain current. The first iteration will be built on a protobard to work out biasing, etc.
The output device connections will have to be solid copper of some sort. The tweeter needs this as well. There is all kinds of contact resistance in the tweeter prototype that has to be eliminated.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.