I ended up ordering the DP-60HS through a company called "Global Media Pro" out of New Zealand. None of Pintek's US sources carry it, though Cal Test may start after I directed their attention to it and the unique place it holds in the market for small signal development work (as opposed to the usual high voltage isolation). I was a little squeamish about ordering from a company that I'd never heard of, and doesn't appear to be a test equipment vendor most of the time. but an email exchange with Pintek in Taiwan confirmed GMP is an authorized distributor so I went to my bank and wired them USD 352.00 (including USD 46.00 in shipping).
Literally overnight, I got my order confirmation. Later that day, I got my tracking number. The tracking number worked in Taiwan's postal service website, and then amazingly it worked in the USPS website too. They estimated 4-5 business days for delivery and that's exactly how it turned out. I'm very happy with GMP and would order from them again without hesitation, unless Cal Test brings in the DP-60HS on their own.
The DP-60HS was nicely boxed for shipment. It does not have an option to run on a battery, so it includes a dedicated 9V wall wart power supply. It also includes two banana plug test leads and two sets of tips for them. One set is a huge and lengthy dual-jaw grabber style pair, and the other is a huge set of alligator clips. I think they just include the typical tips they use for their high voltage units because it's unlikely I'd use these for anything this probe would measure. For my first tests, I used a set of micrograbber-to-bananas which worked perfectly. They also include a short BNC-BNC cable to connect the probe to your scope.
The probe is mind-numbingly simple to use. It has only a single four position control, Off/x10/x5/x1, which serves as both the power and range switches. It has an LED to indicate power and another to indicate overrange. That's it. And frankly that's enough, because you don't really need a whole lot more even on an active probe. What you're after is on the scope screen.
I'm working on a lot of galvanically isolated stuff right now, plus a bunch of capacitor charging applications, so this probe looked perfect for those jobs. I've needed a diff probe many times in the past but the cost was always a deterrent; but now it's almost a necessity and finally finding one that had a x1 range tipped me over. Besides, who doesn't like new toys once in a while?
First up, I did a diff measurement of the grounds on opposite sides of a galvanically isolated power supply that is on the bench. This revealed some interesting things that weren't visible with our traditional grounded scope probes, such as the amount of differential 60Hz they're picking up and the amount of switching power supply noise that is there when a grounded probe isn't connected (read: under normal operating conditions!). We suddenly have some possible answers to some things we've been seeing, AND some ideas about how to address them. Worth it right there.
Next, I tried measuring the floating/isolated supply within one of our breadboards. This worked perfectly too, and frankly gave a lot more visibility to some things we knew were there but never could quite see very well.
During this process I also discovered something I had never needed before: My Rigol DS4000 series scope has a whole suite of native probe ratios, happily including the oddball 5x middle range of the DP-60HS. I was a little concerned about all the mental gymnastics I'd have to be doing, but instead I just dialed in the 5x range on the Rigol when I was using it on the diff probe and everything read natively on the screen. So I'm lazy... it's one less thing, you know.
Then I tried the killer test: Measuring a very small cap that is used for timing purposes within one of our circuits. We charge and discharge this cap on a regular basis, about every 500mS. Actually seeing what this charging waveform looks like is something we've always needed to do, but a grounded scope probe cannot work because its ground has too much effect on the circuit. This yielded some extremely odd results. Initially, when the diff probe was connected, the circuit's operation was utterly unaffected (first time ever!) but the display on the scope screen was unreadable; all sorts of square waves and other artifacts that, if they were actually there, would render the circuit inoperable.
But then came the really weird part: After leaving the DP-60HS connected for more than about 30 seconds, the charging voltage started drooping, until the circuit saturated and stopped working. This looked exactly like the probe had added a bunch of capacitance to the circuit, and that capacitance had finally saturated. But what was going on during that first ~30 seconds? Disconnecting the probe instantly restored normal operation, and leaving it connected for less than ~30 seconds had no effect. But then the effect ramped up and killed off the circuit. This is 100% repeatable and occurs on all ranges of the probe.
Granted this is a high impedance environment (constant current charging a cap to obtain a linear ramp) working into a very small capacitance. But why the ~30 second delay? More investigation is needed for sure.
One other unexpected artifact: Isolation power supplies have a figure of merit called "isolation capacitance" that is a measurement of the effective capacitance across the galvanic isolation barrier. One reason for the diff probe was because connecting a normal probe to the isolated circuitry would result in an effective increase in the isolation capacitance, yielding a net reduction in the AC isolation (greater capacitance = lower AC reactance). Our expectation was that a fully isolated, fully differential probe would eliminate this effect because no ground reference could exist.
To my complete surprise, unfortunately, the diff probe still causes the effect! Connecting even ONE (either one!) of the leads of the diff probe to the isolated circuit instantly generates the effective increase in isolation capacitance (i.e. less isolation). I am at an utter loss to explain this, but I'm on the hunt. First clue: Disconnecting the BNC cable to the scope (and its earth ground reference) eliminates the effect, which strongly suggests the probe isn't as "isolated" as one would expect.
The typical way to build probes like this is to use a differential or instrumentation amp so the inputs aren't directly related to ground. However, if you're doing a pure differential measurement there is often not a path for the amplifier input bias current, so you're forced to use a high value resistor to either ground or the supply rail so the input bias current has a path. (It's always bothered me that this basically violates the very isolation and high input impedance that you're usually looking for when using an in-amp.) In the case of probes, you also often want a known and predictable input impedance and a fixed resistor would yield that as well.
If they did indeed employ bias current resistors, and those tie to the same ground as the BNC heading out to the scope, that would exactly explain the behavior I'm seeing. Substituting an actual discrete 1M resistor, I can duplicate almost exactly the effect of connecting this diff probe. The next step will be to crack the case open and see if I can reverse engineer what they did at their input stage.
A future step may be to design up a TRULY isolated differential probe. Getting DC response poses a challenge; one way would be to have an A/D front end and optocouple the data stream across the isolation barrier, such that there is absolutely positively ZERO galvanic connection from front to back. Sigh... another project for another day.
I'll continue to post as we gain experience with the DP-60HS. So far it works well in most of the applications we intended for it. Granted some of our stuff is a little esoteric so I might be a bit harsh in my expectations, but when I think "differential" I think "utterly unreferenced" or "truly floating". Apparently that's not necessarily the case.
Home
Help
Search
Login
Register
