I am looking to see if anyone can recommend a class D amplifier with a large bandwidth going up to 100kHz with less than -3dB drop. It should be a bridged design taking +/-55V inputs. I would like to power it with a Vicor dual supply that I have that provides 55V at 11.1A per supply. Monochannel.

 

Don't know about 100KHz but TI has lots of class D stuff like the TAS5624A which might be a start. It's more for audio but try a parametric search.
Amazon has various chinese class D amp boards which are closely derived from the manufacturer's evaluation board ($ vs $$$$ from TI) which might get you started for cheap.

Good luck.

 

go for 500kHz and you could just about become a radio station. Jeez. what is lowest freq you require? and if you say 20Hz I am gonna throw the internet at you

 

go for 500kHz and you could just about become a radio station. Jeez. what is lowest freq you require? and if you say 20Hz I am gonna throw the internet at you

LOL I know, its a tough requirement. I do have signals I have to support though that run in the mid 90s. The lowest I have to support is 9kHz

 

Maybe a more feasible idea would to try to get something to run from 9-50kHz and come up a seperate solution for the higher signals. I wouldn't need to sweep in the higher range as they are at fixed frequencies.

 

that sounds much more attainable commercially.
now getting 300 watts at those low range radio freq around the 75-100kHz area is another very high hurdle.
building something to do that is something very few people would be capable of doing and not being a standard freq range for an amplifier makes it unlikely to be available in the markets. so I ask again; 300 watts?

 

maybe you could search for old military surplus sonar gear?

 

Amplifier Research will have anything you want, but will be very pricey. What is your power requirement and your load impedance?

Oops, just saw your power requirement in the title.

Load impedance???

 

According to Maxim, typical class D switching freq is 250 KHz to 1.5 MHz for a 20 KHz signal, or a signal to carrier ratio of ratio from 12 to 1 to 75 to 1. Doing that with a 100 KHz signal bandwidth will be tough. 250 KHz would be just over the Nyquist limit, so fidelity would suffer and output filtering would be a bear. What kind of distortion or harmonic products can you tolerate?

ak

 

My load can be adjusted somewhat but will fall in the range from 5-14 ohms. I have to double check what my loads are exactly for each frequency to find out if I really need 300W at the higher frequencies. Harmonics should be at least 40dB down from the primary. I don't have particular number in mind for THD, but I want as clean a signal as possible. Its funny that you say "old" sonar gear. I am working on making redesigning "old" gear so that it can be smaller, more efficient and more capable. Right now I am only capable of putting out short duration bursts of 200ms from charged caps while still maintaining a level of about 50Vpeak to feed my output transformer. More typically I put out a burst just under 50ms. This has all been done with a class AB amp. I wanted to have the capability of playing longer signals so I spec'd a power supply to be able to eliminate those caps and deliver continuos current. Currently I output at fixed frequencies and have a seperate tuning circuit for each frequency. I want get away from that. I also wanted the capability to change frequency on the fly and maintain my output level. Right now each load is tuned for resonance and if I go too far out I would be dumping too much reactive power back to the amplifier. This lead me down two paths so far. 1) Trying to come up with an automatic variable inductor based on feedback, which I am working on. 2) Get a more efficient amplifier design, which has me learning about class D. So even if I am badly matched, my thought is, I won't heat up the amp nearly as much. Thanks for the input so far. Its got my wheels turning.