Questions about the various different probes for logic analyzers and o-scopes

[Spork Schivago]

Hi,

I'm fairly new to the electronic world.   I've taken a college level digital electronics class, I've taken a basic high school electronics class.   I'm trying to learn about circuits, how to repair them, how to build them, but it's going a bit slow.   We have a baby now and I don't have as much time as I did before.

I have an old HP / Agilent / KeySight 16702B logic analyzer and I have an old Hitachi V-1065A 100 MHz bandwidth dual-channel oscilloscope.   Today, in the mail, I received a 16760A timing and state analyzer module.   I need to order some flying lead sets and probes for it.   I think after spending a few hours googling, I'm even more confused than I was before.

I see for the logic analyzer module they make soft-touch probes, differential soft-touch probes, single-ended probes and differential probes.   For my oscilloscope, I know they also make differential probes.   For the life of me, I cannot figure out what the difference is between all of these.

I googled differential signal and found this on Wikipedia:

Code: [Select]

Differential signaling is a method for electrically transmitting information using two complementary signals. The technique sends the same electrical signal as a differential pair of signals, each in its own conductor. The pair of conductors can be wires (typically twisted together) or traces on a circuit board. The receiving circuit responds to the electrical difference between the two signals, rather than the difference between a single wire and ground. The opposite technique is called single-ended signaling. Differential pairs are usually found on printed circuit boards, in twisted-pair and ribbon cables, and in connectors.

But I'm still a little confused.   Using two complementary signals....what exactly does that mean?   Does that mean if we have an IC that uses differential signaling, on the PCB, one trace might send +5V, the other might send +3V, and the differential signal would be +8V?    Or +5VDC and -3VDC and the signal would be +2VDC?

For the oscilloscope, I see words like passive and active.   Not 100% sure what they mean either.   From what I was reading, it might have something to do with impedance.   If the component under test has a high impedance (greater than 1Mohm), I would want to use an active probe, if it's 1Mohm or under, a passive probe.   I'm still a bit confused and any help from someone more knowledgable than me would be greatly appreciated.

Thank you!

[w2aew]

Differential signals consist of a pair of signals, where each wire is used to carry the opposite logic state of the other.  The "information" is carried in the "difference" between the wires.  The main advantage is noise immunity, since noise couples to the pair of wires equally, and does not affect the difference voltage.  Typically used for high speed logic, where the logic swings are small.

Here is a video I made on differential probes:

[danadak]

These might help -

http://www.edn.com/design/analog/4335883/Single-ended-or-differential-That-is-the-question


http://www.bitscope.com/blog/DJ/?p=DJ01A


http://www.tek.com/document/application-note/making-single-ended-measurements-differential-probe



Regards, Dana.

[Spork Schivago]

Differential signals consist of a pair of signals, where each wire is used to carry the opposite logic state of the other.  The "information" is carried in the "difference" between the wires.  The main advantage is noise immunity, since noise couples to the pair of wires equally, and does not affect the difference voltage.  Typically used for high speed logic, where the logic swings are small.

Here is a video I made on differential probes:

I need some clarification still.   When you show the speakers, you say with differential signaling, if there's noise, it'll be on both lines and we won't hear it.   I don't understand that.   If both lines have the same noise, how does the noise get ignored by the component?   This is where I'm getting hung up.

Also, I haven't finished the video yet, I'm almost done, I'm thinking for the time being, because I'm just starting out, I might want a single-ended probe and not a differential right now.   Turns out the flying lead adapters for my 16760A are really really really expensive.   It's going to be hard saving up for either the single-ended or the differential.   Right now, I'm just using the logic analyzer and oscilloscope to learn, but eventually, if things go right, I'll be using them to diagnose and fix problems.   If it was you, would you be saving up for the differential flying lead set or the single-ended?

[Spork Schivago]

I'm going to have to wait a little bit before I read those!   I've read sooooo much today, I feel like my eyes are going to explode!   Thank you though for supplying the links, I'm sure they're going to be informative!

I'm going to step away from the question for a little bit now and concentrate on my CH341A.   Maybe when I come back later, my mind won't be so jumbled!

It's all starting to click though, which is good.   That video the first guy posted, that was extremely helpful and I really like how there was a video!

[w2aew]

I need some clarification still.   When you show the speakers, you say with differential signaling, if there's noise, it'll be on both lines and we won't hear it.   I don't understand that.   If both lines have the same noise, how does the noise get ignored by the component?   This is where I'm getting hung up.

The noise voltage appears identically on both wires, not affecting the voltage difference between the wires.  When the noise appears equally on both wires, it is said to be "common mode".  The speaker responds to the voltage difference between the wires.  In other words, since the speaker is connected between the wires, it only "sees" the differential signal and not the common mode signal (because that causes both terminals of the speaker to change in concert with each other).

[tggzzz]

I need some clarification still.   When you show the speakers, you say with differential signaling, if there's noise, it'll be on both lines and we won't hear it.   I don't understand that.   If both lines have the same noise, how does the noise get ignored by the component?   This is where I'm getting hung up.

One line has V_p=+V_signal+V_noise, the other has V_n=-V_signal+V_noise. The receiver measures V_p-V_n = 2V_signal

Other benefits are that you can reduce the supply voltage and still have the same received voltage swing => lower power dissipation, and that currents cancel out reducing EMC problems.

[james_s]

To build on what has been said already, you are probably not going to need differential probes until you get to something fairly specialized, and by that time you'll know you need them. Some common applications you might encounter differential signals are things like Ethernet, HDMI, many forms of SCSI, SATA, computer related stuff mostly. For the time being don't sweat it, you can go a long way with just some basic reasonable quality probes.

[Spork Schivago]

Oh, I think I finally see!

One line would have a +2.5VDC pulse, for instance, right?   The other line would have a -2.5VDC pulse.   The device would see +2.5VDC - -2.5VDC = +5VDC.   But if both lines have +0.3VDC noise, would it be,
one line has +2.5VDC + 0.3VDC = +2.8VDC, the other line would have -2.5VDC + 0.3VDC = -2.2VDC.   The device would see +2.8VDC - -2.2VDC = +5VDC.

Is that correct?   I kept on thinking on line would have +2.8VDC, the other line would have -2.8VDC.   I haven't gotten a lot of sleep lately because of the baby.   Still trying to get on some sort of schedule, but I think I finally understand!!!!  Unless I'm wrong, then I'm totally lost again!

[mtdoc]

Oh, I think I finally see!

One line would have a +2.5VDC pulse, for instance, right?   The other line would have a -2.5VDC pulse.   The device would see +2.5VDC - -2.5VDC = +5VDC.   But if both lines have +0.3VDC noise, would it be,
one line has +2.5VDC + 0.3VDC = +2.8VDC, the other line would have -2.5VDC + 0.3VDC = -2.2VDC.   The device would see +2.8VDC - -2.2VDC = +5VDC.

That's basically it.

But when you say "one line would have a +2.5 VDC pulse" you need to ask: +2.5VDC relative to what?  To earth ground?  Likewise, -2.5VDC relative to what?

You could just as easily measure +5VDC with a differential probe placed between 2 voltages that where at +100 VDC and +95 VDC relative to earth ground. Or -10 VDC and -15 VDC relative to ground, etc.

The differential probe will only tell you the relative difference between 2 points and nothing about their potential relative to ground.    If not careful you might be "shocked" to discover this fact. So to speak..

[Spork Schivago]

Oh, I think I finally see!

One line would have a +2.5VDC pulse, for instance, right?   The other line would have a -2.5VDC pulse.   The device would see +2.5VDC - -2.5VDC = +5VDC.   But if both lines have +0.3VDC noise, would it be,
one line has +2.5VDC + 0.3VDC = +2.8VDC, the other line would have -2.5VDC + 0.3VDC = -2.2VDC.   The device would see +2.8VDC - -2.2VDC = +5VDC.

That's basically it.

But when you say "one line would have a +2.5 VDC pulse" you need to ask: +2.5VDC relative to what?  To earth ground?  Likewise, -2.5VDC relative to what?

You could just as easily measure +5VDC with a differential probe placed between 2 voltages that where at +100 VDC and +95 VDC relative to earth ground. Or -10 VDC and -15 VDC relative to ground, etc.

The differential probe will only tell you the relative difference between 2 points and nothing about their potential relative to ground.    If not careful you might be "shocked" to discover this fact. So to speak..

So, if a component is expecting a differential signal, and it receives +100VDC down one line and +95VDC down the other, then the device would only see +5VDC, right?   Does the device need to be able to handle +100VDC or just the +5VDC?   I'm trying to understand now how I could implement differential signaling, if I wanted to.   If I have a PIC with two A2D channels, if I send +100VDC to one pin, +95VDC to the other and subtract the voltages in code on the PIC, unless that pic can support that high of voltage on those pins, it'd fry it, right?

What happens if I have an LED that has a voltage drop of +1.5VDC and I put a differential probe between the two leads?   What would the differential probe show me?   I wish I had a differential probe for my oscilloscope to play around with, I bet I'd learn a lot more quickly about this.

[iainwhite]

So, if a component is expecting a differential signal, and it receives +100VDC down one line and +95VDC down the other, then the device would only see +5VDC, right?   Does the device need to be able to handle +100VDC or just the +5VDC?

The answer depends on whether the 'device' is referenced to ground, or not.
In your example of the PIC, I would assume that the PIC has a power supply that has, say +5V and a ground. This means that the PIC  has a ground reference, and therefore it will be dealing with a signal input of +100V relative to ground and would fry.

In the case of a differential probe,  the probe is dealing with the voltage with reference to ground internally, and only presenting the result at it's output - to use your example, the probe is dealing with +100V and +95V on it's inputs, but only presenting outputs of +5V and 0V.
If you look at Alan's (W2AEW's) video above at approx 4:15 you will see that his Tek THDP0200 probes have a limit of 1000V (relative to Ground) on each input, and a relative limit of 1500V between the two inputs.   However, the output of these probes into the scope will be nowhere near that level - i believe they have 500x attenuation, so the max output will be +3V for an input differential of 1500V

Note: I am not a qualified EE, hope the above is correct, but I'm sure someone will correct me if not.
 

[tggzzz]

Oh, I think I finally see!

One line would have a +2.5VDC pulse, for instance, right?   The other line would have a -2.5VDC pulse.   The device would see +2.5VDC - -2.5VDC = +5VDC.   But if both lines have +0.3VDC noise, would it be,
one line has +2.5VDC + 0.3VDC = +2.8VDC, the other line would have -2.5VDC + 0.3VDC = -2.2VDC.   The device would see +2.8VDC - -2.2VDC = +5VDC.

That's basically it.

But when you say "one line would have a +2.5 VDC pulse" you need to ask: +2.5VDC relative to what?  To earth ground?  Likewise, -2.5VDC relative to what?

You could just as easily measure +5VDC with a differential probe placed between 2 voltages that where at +100 VDC and +95 VDC relative to earth ground. Or -10 VDC and -15 VDC relative to ground, etc.

The differential probe will only tell you the relative difference between 2 points and nothing about their potential relative to ground.    If not careful you might be "shocked" to discover this fact. So to speak..

So, if a component is expecting a differential signal, and it receives +100VDC down one line and +95VDC down the other, then the device would only see +5VDC, right?   Does the device need to be able to handle +100VDC or just the +5VDC?   I'm trying to understand now how I could implement differential signaling, if I wanted to.   If I have a PIC with two A2D channels, if I send +100VDC to one pin, +95VDC to the other and subtract the voltages in code on the PIC, unless that pic can support that high of voltage on those pins, it'd fry it, right?

For all devices, always read and understand everything you see in the datasheet. Manufacturers tend to put too little information in them; the corollary is that all information counts.

In this case you would be looking for the absolute maximum ratings, the common-mode and differential mode voltage ranges.

What happens if I have an LED that has a voltage drop of +1.5VDC and I put a differential probe between the two leads?   What would the differential probe show me?   I wish I had a differential probe for my oscilloscope to play around with, I bet I'd learn a lot more quickly about this.

RTFDS for the info noted above.

There is no such thing as "0V"; there are only voltage differentials.

Be careful with the shield on scope inputs: it is connected to mains protective ground ("earth"). If you touch it to a part of your circuit that is not at that potential, current will flow - and that current may damage your circuit or your scope. That's why you never ever disconnect a scope's earth lead. See

[Spork Schivago]

So, if a component is expecting a differential signal, and it receives +100VDC down one line and +95VDC down the other, then the device would only see +5VDC, right?   Does the device need to be able to handle +100VDC or just the +5VDC?

The answer depends on whether the 'device' is referenced to ground, or not.
In your example of the PIC, I would assume that the PIC has a power supply that has, say +5V and a ground. This means that the PIC  has a ground reference, and therefore it will be dealing with a signal input of +100V relative to ground and would fry.

In the case of a differential probe,  the probe is dealing with the voltage with reference to ground internally, and only presenting the result at it's output - to use your example, the probe is dealing with +100V and +95V on it's inputs, but only presenting outputs of +5V and 0V.
If you look at Alan's (W2AEW's) video above at approx 4:15 you will see that his Tek THDP0200 probes have a limit of 1000V (relative to Ground) on each input, and a relative limit of 1500V between the two inputs.   However, the output of these probes into the scope will be nowhere near that level - i believe they have 500x attenuation, so the max output will be +3V for an input differential of 1500V

Note: I am not a qualified EE, hope the above is correct, but I'm sure someone will correct me if not.

Thanks for the clarification, I understand now.   I did watch the video, but I'm still confused with the limits for that probe.   1,000v relative to ground, I understand that now.   1,500v limit between the two inputs, I don't understand that.   Does that mean if one input has +5VDC on it, and the other input has 1,505VDC on it, the probe would see the 1,500VDC?   I don't see how it can handle that but it can't handle anything over 1,000VDC in reference to ground.

[Spork Schivago]

Oh, I think I finally see!

One line would have a +2.5VDC pulse, for instance, right?   The other line would have a -2.5VDC pulse.   The device would see +2.5VDC - -2.5VDC = +5VDC.   But if both lines have +0.3VDC noise, would it be,
one line has +2.5VDC + 0.3VDC = +2.8VDC, the other line would have -2.5VDC + 0.3VDC = -2.2VDC.   The device would see +2.8VDC - -2.2VDC = +5VDC.

That's basically it.

But when you say "one line would have a +2.5 VDC pulse" you need to ask: +2.5VDC relative to what?  To earth ground?  Likewise, -2.5VDC relative to what?

You could just as easily measure +5VDC with a differential probe placed between 2 voltages that where at +100 VDC and +95 VDC relative to earth ground. Or -10 VDC and -15 VDC relative to ground, etc.

The differential probe will only tell you the relative difference between 2 points and nothing about their potential relative to ground.    If not careful you might be "shocked" to discover this fact. So to speak..

So, if a component is expecting a differential signal, and it receives +100VDC down one line and +95VDC down the other, then the device would only see +5VDC, right?   Does the device need to be able to handle +100VDC or just the +5VDC?   I'm trying to understand now how I could implement differential signaling, if I wanted to.   If I have a PIC with two A2D channels, if I send +100VDC to one pin, +95VDC to the other and subtract the voltages in code on the PIC, unless that pic can support that high of voltage on those pins, it'd fry it, right?

For all devices, always read and understand everything you see in the datasheet. Manufacturers tend to put too little information in them; the corollary is that all information counts.

In this case you would be looking for the absolute maximum ratings, the common-mode and differential mode voltage ranges.

What happens if I have an LED that has a voltage drop of +1.5VDC and I put a differential probe between the two leads?   What would the differential probe show me?   I wish I had a differential probe for my oscilloscope to play around with, I bet I'd learn a lot more quickly about this.

RTFDS for the info noted above.

There is no such thing as "0V"; there are only voltage differentials.

Be careful with the shield on scope inputs: it is connected to mains protective ground ("earth"). If you touch it to a part of your circuit that is not at that potential, current will flow - and that current may damage your circuit or your scope. That's why you never ever disconnect a scope's earth lead. See

Thanks for the info.   When I read datasheets, there's still a lot of things I don't understand.   I think once I get a better understanding of the fundamentals though, then maybe some of the information in the datasheets that I currently don't understand will make more sense.   For capacitors, stuff like tangent of loss angle and then the frequencies at what caps were measured at, that I don't get.   I think once I finish learning about electricity and how it works though, maybe I'll have a better understanding.   There's a lot of information out there and there's little time now that we have a baby, but every day, I try to learn something new.

[tggzzz]

Thanks for the info.   When I read datasheets, there's still a lot of things I don't understand.   I think once I get a better understanding of the fundamentals though, then maybe some of the information in the datasheets that I currently don't understand will make more sense.   For capacitors, stuff like tangent of loss angle and then the frequencies at what caps were measured at, that I don't get.   I think once I finish learning about electricity and how it works though, maybe I'll have a better understanding.   There's a lot of information out there and there's little time now that we have a baby, but every day, I try to learn something new.

Good attitude! Be glad that there are many things left for you to find out With many aspects of life it is necessary to quickly determine what you don't need to fully understand.

Nobody understands everything at once, and it is sometimes difficult for me to remember what it is like as a beginner. I do remember having some embarrassing misunderstandings, but fortunately I've forgotten what most of them were. Even now I come across items in datasheets that I don't understand. And where RF is concerned, even the concepts are a challenge - that's why I'm starting doing it now.

Have fun with your little person.

[iainwhite]

  1,500v limit between the two inputs, I don't understand that.   Does that mean if one input has +5VDC on it, and the other input has 1,505VDC on it, the probe would see the 1,500VDC?   I don't see how it can handle that but it can't handle anything over 1,000VDC in reference to ground.

The limit to ground (1000 VDC) can be positive or negative.
So you could have +600V on one probe and -600V on the other for a differential of 1200V
Your example with 1505V would be outside the 1000V limit for one probe.

[Spork Schivago]

  1,500v limit between the two inputs, I don't understand that.   Does that mean if one input has +5VDC on it, and the other input has 1,505VDC on it, the probe would see the 1,500VDC?   I don't see how it can handle that but it can't handle anything over 1,000VDC in reference to ground.

The limit to ground (1000 VDC) can be positive or negative.
So you could have +600V on one probe and -600V on the other for a differential of 1200V
Your example with 1505V would be outside the 1000V limit for one probe.

I understand now!   I got you.   Thank you for explaining that!

So we could have +900VDC on one probe, -600VDC and it'd be supported, because no probe goes over 1,000VDC in reference to ground, and the differential voltage would be +1,500VDC, which would be supported.   Right?

[iainwhite]

So we could have +900VDC on one probe, -600VDC and it'd be supported, because no probe goes over 1,000VDC in reference to ground, and the differential voltage would be +1,500VDC, which would be supported.   Right?

Right!
Glad to help. 

I would recommend browsing other videos at W2AEW youtube - lots of good stuff there, plus I like all the older HP and Tektronix gear he uses.  I convinced myself I want to buy an analog Tek 465 or 475 sometime (when the wife isn't looking)

[bitseeker]

For the oscilloscope, I see words like passive and active.   Not 100% sure what they mean either.   From what I was reading, it might have something to do with impedance.   If the component under test has a high impedance (greater than 1Mohm), I would want to use an active probe, if it's 1Mohm or under, a passive probe.   I'm still a bit confused and any help from someone more knowledgable than me would be greatly appreciated.

Hi Spork,

I don't think this question was covered yet. Yes, it has to do with the input impedance. The oscilloscope's input impedance is typically 1 MOhm (or switchable between 1 MOhm and 50 Ohm, for example) and that's usually high enough not to significantly alter the signal in the circuit that you want to probe. However, if you're probing something approaching 1 MOhm, then the scope will load down the circuit and alter the signal.

An active probe has electronics in it that (1) provides even higher input impedance to minimize loading down the circuit, (2) has less capacitance to further minimize altering the signal being measured especially those with high frequency components and/or short rise/fall times, (3) captures the signal very close to the circuit thereby eliminating performance degradation that could occur along the cable on its way to the scope and (4) may use differential signaling between the probe and scope to eliminate interference/noise.

[tggzzz]

I convinced myself I want to buy an analog Tek 465 or 475 sometime (when the wife isn't looking)

My preference is for a 485  (I do have 2*465 + 1*475, and I need to divest myself of them).