Rob Redding wrote:

I hope you'll let us know how you get on with the Velleman PCS500..

Yes, of course. Unfortunately, the parcel was on transportation the last weeks. So, it took some days for the reply.

I have an older PC, just seleceted for this purpose: Pentium, 166MHz, Windows 95.

I made the setup, connected PCS500 to PC and everything worked fine, from beginning. Everything necessary for running the PCS500 is delivered, so, you don't need to buy additional stuff, like transformer, a.s.o. Only probes are very poor, not suited to make serious measurings in the MHz range. But I also buied a probe (fully isolated) from same manufacturer.

Whole device looks a little bit cheap: BNC connectors are made of plastics, for having isolation, but I think they will not resist many pluggings and unpluggings. So, probes should rest connected, if possible.

On the screen everthing can be handled with mouse, but for some fine settings (scroll bars) arrow keys are necessary. Autoset function is a little bit slowly working. I think this is, because all controlls have to be transmitted via parallel port.

I missed automatically measuring of some signal parameters, like Vpp or frequency. The 'only' you get automatically is True RMS and dB value. But you have a cursor function: While moving the two bars voltage, frequency and dt is displayed. But if signal is changing you get only right information about voltage and frequency, when you again shift cursor bars.

Triggering is a little bit problematic when dealing with signal showing some jitter: Edge on which is triggered and whole signal sometimes jumps some pixels left. But dirty triggering you can also see with much more expensive DSOs, right?

Vertical resolution (8bit) is ok for me. 12bit would be better of course. But what me a little bit irritated at the first look is rather poor horizontal resultion. When displaying square waves some edges suddenly jump one pixel to the left or right. This is stressing a little bit, because this 'edge walking' looks strange and unnatural. Here I would wish that from the 4096 samples there are more used to increase resolution of displayed screen. But if you set time base of horizontal deflection so, that only one period is displayed over whole screen then 'edge walking' does not stress any longer. It's only strange as long as you do not know, what's going on...

1GHz/sec sampling mode an spectrum anaylzer feature are really imposing!!!

After first test run, I opened enclousure, of course. To my surprise I didn't found any mcu. Only digital circuitry and a big GAL. At inuts there is ADS830 from BurrBrown, which makes 60Ms/sec with 8bit resultion. PCB is four layer, I think, well soldered with no additional fixings. Only one decoupling capacitor is soldered after manufacturering.
PCB has splitted massplanes for analog and digital circuitry, but I would have designed layout a little bit differently...
Interfacing is done via optocoupler isolation to parallel port of PC. I think some data lines are used to fabricate 5V for 74HC244 and some 6N136. Optocouplers are sitting in sockets. That's ok, when they are a failing one fine day. There's absolutely no electronic part connected between both grounds, means no capacitance (for high frequency decoupling), no high impedance resistance (for electrostatical discharge) and no varistor or transzorb for protecting optos in case of heavy electrostatical charge. So, being able to easily remove optos might be very useful.
I think it would be advantegously to have all involved mains supplies sitting in the same plug console (sorry for bad english!), additionally having some overvoltage protection (those mains plug consoles having some lightning protection).

Did I forgot something?

Well, I'm so satisfied with PCS500 that I will additionally buy this PCG10 device, which is a very comfortable function generator. I do need sinus burst a lot of times, because I'm doing lot's of audio and analog developments.

Bye,
Kai