Active Probe vs Differential Probe

The device you are measuring should have some sort of GND. Does your battery powered scope have a GND stud?

My bench scope has an earth ground stud..mmmm It's line powered and earth grounded.

Guess I could run a earth ground wire from the scope to the earth ground on the DUI to reduce some CM noise.

D from BC

A 10:1 fet probe can be stunningly cleaner than a 10:1 passive probe, even if it's not differential. High frequency junk and fuzz pretty much go away. You can often do good probing without even using the ground clip, which seems strange but often works. With a good fet probe, you don't even have to touch some signals... just get close.

John

Wow! I didn't know. I've never used an active probe. If that's the case, I suppose the greatest use for a differential probe is for testing non-isolated electronics.

I'll dig up some differential probe manuals and see if I can find it's uses and benefits.

D from BC

That's why I was amazed at GGB's Picoprobe line.

DC to 26GHz (for their Model 35) and a single probe point with no ground connection. I still shake my head at that.

Robert H.

The amazing spec is .05pf input capacitance! What are methods to achieve that? Mike

I can't even get near that with a OPA356 GBW=200Mhz FET input op amp with Cin~1pF.

D from BC

I think the fet probes are just a jfet source follower and a current source. So it runs open loop, so to speak. Lots of BW, fair linearity. I never did any jfet chips, but I would imagine they are robust regarding ESD, so there is no protection device to add capacitance.

One of the primary benefits of differential probes is to see what's actually going on with circuits in high noise environments or where there is a common mode signal that is to be ignored. Non-isolated electronics is just a special case of measurements with a common mode signal. ;-)

At low frequencies you can do a similar thing with two probes and the scope set to A-B. This gets dicey fast and the two probes had better be well matched.

--
Keith

We have a Tek TPS2024. That's a 4-channel, 200 MHz scope that has all four channels and the trigger individually isolated. You can clip the "ground" lead of a probe to the cathode of a triac that's 240 volts AC off ground, and probe the gate, and it just works. Fabulous scope.

John

Hi,

If you have a standard 1000x ~40kV probe can you do differential measurements at those voltages with that on this scope? (not planning to do this, just curious)

Does this scope completely eliminate the need for differential probes?

How about fet probes, if they don't even need their ground leads connected, no worries about input isolation there, but I suspect fet probes aren't used at high voltage?

some possible drawbacks of this scope:

1. "TPS2PWR1" software sold separately, not sure of the cost, tek.com website said it is 0 in CAN$, but that supposedly includes 4 0 P5120 probes, so I think that price was wrong (low)?

1. low record length of 2.5K points (is this an issue in real life?)

2. only 1/4 VGA LCD

1. no USB interface (has compactflash (1GB max) though, can this be used to compensate for the low internal 2.5k memory?

some pluses:

2. GS/s sample rate for the 200MHz scope, while some other manufacturers have lower sample rates for their 200MHz scopes.

1. battery operation

2. the input isolation

cheers, Jamie

Isn't four grand a little steep for a 200MHz scope?

--
Keith

Depends on what you're working on. One of our nmr gradient amps sells for 10K, and people buy them three at a time. This scope is fabulous for measuring fet gate drivers that are riding on a bus that's 180 volts off ground.

A good spectrum analyzer or arbitrary waveform generator can cost well over $100K. And a year's worth of an engineer's time can cost twice that.

Compared to people and net product sales, equipment like this is downright cheap. If it lasts 10 years, that's about $400 per year. Figure in taxes, it's closer to $200.

John

Well, a good car is 30 grand, but isn't four grand for a 200MHz scope a little steep?

...and that's a good thing! ;-)

Yes, but a $1K scope (or $200 on eBay) is even cheaper. A thousand of those and you can save enough to buy another engineer. ;-)

Seriously, I'm not against spending money, something that corporations are very bad at (I'm stuck with a single 19" display), given the realities you point out above. I just think the price of this particular scope is a bit "large". I remember purchasing

7904s, then 7104s, when they were upwards of $20K (not 2007 $$ either), but there wasn't anything else in that class.

--
Keith

I think John's point is that, by the time you properly outfit a cheaper scope with decent isolated input adapters (and include all the engineering time in doing so), $4k is starting to look pretty cheap.

Have you looked at the products John offers? IMHO his prices are entirely reasonable, so I have to believe he has a decent grasp on what's worth the money and what isn't.

The thing with CRTs/LCDs and computers often really just is a status game: It often literally is, "My LCD is bigger than yours, chump!"

---Joel

Differential probes aren't all that spendy.

Yes, many times and have been suitably impressed. I'm not bitching at John for spending his money, rather at the pricing of that particular scope. There isn't any way I'm buying one, though I'm scopeless.

Often it is, particularly with management types (even when they're the makers of said equipment). OTOH, I'll guarantee you that a couple of 20", or larger, displays will improve my productivity by a tad more than a minute a day. Enough so, that I'd buy the monitors myself were I allowed to (and knew I'd get them back). I have far better computer equipment at home than on my desk at work. My time at work is far more expensive though.

--
Keith

At 200MHz they are! A single Tektronix P5205 differential probe, for instance (which is only 100MHz too), retails for over $1k, and even third-party probe (e.g.,

are still many hundreds.

OK, understood. Anything from Tek or Agilent certainly does have a pretty hefty mark-up to it... they just don't know how to not do it, I guess. :-)

Yeah, sad when that happens.

---Joel

I mostly use 1x or 10x probes. What this scope is best for is looking at small voltages riding on a baseline that's way off ground, like an opamp riding on a 200 volt power bus that is jumping all over the place.

Not entirely, but mostly.

A fet probe with unconnected ground clip still measures signals relative to earth ground.

The scope came with four extra-insulated probes. I have no use for the software.

Seems fine to me. Since we're mostly scoping power/analog stuff, not digital data streams, the record length is plenty.

Plenty good enough, given the adc resolution.

If I want a scope picture, I freeze the trace and shoot it with a digital camera. That allows me to stick a post-it on the bezel explaining context, and include that in the pic.

Hi,

If you have a USB oscilloscope and a battery powered laptop, can you use a normal scope probe to do these high common mode voltage measurements as the scope with floating inputs can do, at least with one channel? Same for a battery powered scope or a scope without a ground plug?

cheers, Jamie

John Lark> >