Complicated psuedo-scope question

[Rob11966]

Hi Folks,

I have a question for those that know a bit about scopes.

I currently have an old 'Ad Instruments' Powerlab which is a very nice high quality digital amplifier. It's original application was not for testing electronic circuits but I thought that it may be useful to me.

The input impedance of the main part of the amp is 1 M ohm/47pf with a maximum input voltage of +/-10V. It has a standard BNC connector.

So I figured that my 10:1 oscilloscope probe could be plugged directly into this and I could measure voltages up to 150V.

My understanding with the 10:1 probe was that it is a voltage divider, attenuated the voltage by a factor of 10. Being cautious by nature, I thought that I should test this prior to attempting above. When I used the 10:1 probe on the mains voltage (240V, expecting to see an attenuated voltage of 24V) and measured the voltage on a DMM the voltage measured around 120V! I tried a couple of different probes and all were the same. Now I presume that the 10:1probes act in conjunction with the input impedance of the scope so the scope 'sees' a 10th of the actual voltage. I am thinking that my digital Powerlab should interface with the 10:1 probe in a similar fashion?

I am just having trouble getting past the concept that there is a lot more voltage at the end of the probe than I anticipated and I may exceed the input voltage of the powerlab?

Does anyone have any thoughts on this?

Cheers,

Rob

 

[SY]

Was the probe terminated with 1M? That's critical. Also, 47pF may be too much for you to adjust the 10x probe for good HF response, but it's sure worth a try.

 

[cliffforrest]

How did you connect the DVM? That is critical!

What is the DVM input impedance?

 

[BZed]

If you DMM is like most the input impedance is 10M. Your probe is designed to work with a scope that has a 1M input. For 10/1 the probe will have a series resistance of 10M. As you can see if the DMM is 10M with a 10M series R of the probe then you have a divide by 2 in that configuration.

I noticed that your amp has a +/- 10 vdc limit on the input. You said that you thought you would be able measure up to 150v. With a 10/1 probe and a +/- 10volt limit you can measure +/- 100 vdc not 150, so be careful testing high voltages.

BZ

 

[Rob11966]

Hi Sy and Cliff,

Thanks for the responses. No, the probe was not terminated with 1M ohm, I just clipped the alligators on directly. I figured that this was the key to the problem and that the scopes input impedance was an integral part of the probes 10:1 attenuation.

So assuming that the Powerlab has an input impedance of 1 M ohm then it would be safe to test voltages of up to 150V? What I don't understand is this - lets say for arguments sake that I am measuring 150V with a 10:1 probe. If I use my DMM to measure the voltage at the end of the probe it may read around 75V. Therefor, irrespective of what is happening in the oscilloscope or powerlab, the PD at the end of the cable is 75V. Does this not exceed the +/- 10V input of the Powerlab?

Sorry for being stupid about this.

Rob

 

[SY]

The voltage at the end of the probe depends absolutely on the load. Remember the voltage divider equation? V1/Vtot = R1/(R2 + R1), where R1 is the load resistance (expected to be 1M) and R2 is the series resistance of the probe. To reassure yourself that this is true, clip a 1M resistor across the probe's BNC plug, attach your DMM across the resistor, then connect the probe to the 240V source. I'll bet you see 24V.

 

[Rob11966]

Yes absolutely Sy, I am not going to argue about that. But when this particular device quotes an input voltage of +/- 10V, the input voltage will be exceeded in the example that I quoted above. Should I load the probe with 1 meg prior to plugging into the powerlab?

Rob

 

[SY]

Why? The 1M input impedance will take care of that for you. If you want to do belt-and-suspenders, tie reverse-biased diodes from the "hot" part of the input to each of the (presumably) bipolar power rails as extra protection. But that's likely not at all necessary.

 

[jjman]

I picked up a couple 100:1 ("100x") probes on ebay recently. With only +/- 10v for the max input 100x probes may be the easiest way to go.

Here's the stats on the Tektronix 100x probes that I know of. They use BNC.

MHZ # PF
250 P5100 7-30
30 P6002 20-50
30 P6005 20-50
50 P6007 20-47
120 P6009 8-47
?... P6057. .?

 

[Rob11966]

Why? The 1M input impedance will take care of that for you. If you want to do belt-and-suspenders, tie reverse-biased diodes from the "hot" part of the input to each of the (presumably) bipolar power rails as extra protection. But that's likely not at all necessary.

As I was assuming that the designers of the powerlab took this into account when they quoted the max input voltage. This powerlab was never designed to accept an oscilloscope probe (10:1) or otherwise. Oscilloscopes for example quote an input voltage of around 300V with is very different from my powerlab. I have no doubt you are correct Sy - I am just struggling with the concept. By way of example, I would not hook up my bench power supply to this unit with 75V output, so how is this any different fom the example of the 10:1 probe example measuring 150V - there is still 75V at the end of the cable.

BZed, thanks, it was a typo. The input voltage is actually +/- 15V but I used 10 to makes the calcs easier for the discussion and then confused myself.

Hi JJman, I thought of that and it's a good idea but I am struggling with the concept at the moment.

Rob

 

[SY]

FWIW, I built an interface box for my soundcard that works just this way; it's a buffer running off of +/-15V supplies and a 1M input impedance with output to my soundcard. I use a 1x, 10x, or 100x probe, depending on voltage level and required measurement impedance. Works like a charm and, despite lots of measurements on high voltage tube circuits, no failures in the 5 years or so I've been using it.

 

[Rob11966]

FWIW, I built an interface box for my soundcard that works just this way; it's a buffer running off of +/-15V supplies and a 1M input impedance with output to my soundcard. I use a 1x, 10x, or 100x probe, depending on voltage level and required measurement impedance. Works like a charm and, despite lots of measurements on high voltage tube circuits, no failures in the 5 years or so I've been using it.

Sy, it sounds like I might need to build one of these. Aside from the 1 M resistor, should I include any capacitance as well, keeping in mind that the powerlab has 47uf already? Maybe 1 in series to reduce the overall input capacitance down to around 25 or so?

Rob

 

[SY]

Well, remember, if you buffer, the Powerlab is completely isolated from the circuit under test. I added a small cap to equal the input capacitance of my scope (can't remember the number and I'm too lazy to walk upstairs and look but I think it was something like 27pF). It's less important for you because of the relatively high input Z of the Powerlab, but my soundcard was something like 10-15k.