Is a 100MHz analog bandwidth scope adequate for all audio work?

[Speakerguy]

I am considering buying a scope for audio and embedded systems work. I am heavily leaning toward the DPO3000 series from Tektronix. I would like the 300MHz version, but I was really hoping to keep this to 5-6k.

The DPO3000 has 2.5GS/s real time sampling (no time division multiplexing or anything). I like it because it's a DPO, has a real deep memory, can decode I2C/SPI busses, and is nice and large and easy to see/use. Essentially I love the feature set; I just don't know if I should settle for the lowest bandwidth model.

Any suggestions here?

 

[RadioRon]

In my opinion, you probably won't run out of bandwidth for your work. Any analog stuff you are monitoring will be well within this bandwidth. The embedded digital processors used for analog work are probably not the fastest around and from my experience, even a 100MHz analog scope is fast enough to work with clock rates up to 30 Mhz or so. Most of the guys that I have worked with doing embedded stuff have been using 200 MHz Tek 2000 series scopes with good results and many of them are working at clock rates between 20 MHz and 60 MHz. So I think you will be very comfortable with that 3000 series at 300 Mhz.

 

[Nigel Goodwin]

For audio work you don't need anywhere near that high a spec, but if you can afford it the higher the spec the better, for the one occasion every few years where you may need a higher spec scope.

 

[Brian218]

15-20 years ago, I bought a 40 MHz dual trace scope, and I still have not outgrown it. For +99% of the time, it works great for me.

At my work, we use the newer 'high performance" scopes, but I've seldom had any need for a scope with more bandwidth than 40 MHz.

As previously stated, get the best o-scope you can comfortably afford, although you may never actually need a higher end scope, they are very nice to have!

 

[Speakerguy]

I can think of two instances where a 100MHz scope might not be enough: looking at I2S signals (up to 27MHz digital signals), and looking at the edges of PWM amp signals to fix overshoot and ringing so I don't muck up the FM band. I'm debating if these are worth the extra cost, and still thinking if there is anything else I might ever need to use it for.

 

[Dean Huster]

Two of the best audio scopes I've worked with is the Hewlett-Packard 130C (500KHz bandwidth) and the Tektronix 432/434 (25MHz bandwidth). Both have clean, sharp traces and the 434 has split-screen bistable storage. However, the 424/434 were also one of Tek's most horrible designs from a mechanical standpoint -- extremely difficult to work on and high-failure items such as front panel lamps take a major disembowelment to repair.

It's good that you're looking at the digitizing rate rather than the highest frequency the scope will display. The inferior scopes have low S/s rates and fail to tell you that they're using recurring samples to fill in the data whereas Tek tends to spout out their single-shot performance. Even their low-end TDS220 scopes had 1 and 2GS/s rates while the inferior products usually run in the KS/s and MS/s areas.

For audio work, sometimes high bandwidth can get in the way. High bandwidth scopes usually have higher minimum v/div sensitivities (e.g., 5mV/div) while the low-bandwidth scopes often get as low as 500µV/div. The trace noise on a high bandwidth scope can be very annoying. It can usually be reduced by kicking in a "bandwidth limit" function, but typically 20MHz is still the lowest bandwidth to which one of those scopes will limit.

I don't always adhere to the "buy as much scope as you can possibly afford" theory. There's no point in overbuying on an instrument that may be outdated and difficult to support in 10 years. Why pay hundreds of thousands on a high-performance European sports car that's capable of 180 mph when there's no place you can conveniently and safely drive at those speeds?

Dean