EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: geostep on November 28, 2019, 02:42:42 pm
-
Hi All,
I scored a Tektronix TDS 540C 4 channel 500 MHz oscilloscope on eBay. It's on route to me now. However I'm fairly certain there are no probes included with it.
A quick look around shows that high quality passive probes will set you back $400 - $500 apiece. That's way out of my hobbyist budget.
Do you guys have recommendations for 500 MHz passive probes that are not obvious junk but are not budget busters either?
It's a 4 channel scope, but at this stage I'll probably just get a pair. The scope and probes will be put to general use, nothing exotic.
Thanks for your input!
- George
-
https://www.picotech.com/accessories/passive-oscilloscope-probes (https://www.picotech.com/accessories/passive-oscilloscope-probes) 180 each. I'm sure there's something a bit cheaper if you find the OEM.
https://www.staubli.com/en/connectors/test-and-measurement/high-frequency-measurement/touch-protected-passive-probes/ (https://www.staubli.com/en/connectors/test-and-measurement/high-frequency-measurement/touch-protected-passive-probes/) seem to go for 80-100 depending on accessories.
-
Hi George:
Let me Caveat by saying I have not used their Oscilloscope probes, but the Multimeter Probes are excellent. It would stand to reason that their scope probes are probably decent too.
https://probemaster.com/5900-series-oscilloscope-probes/
These too are in the $180.00 range brand new. Might be worth a look. I am looking for 1GHz probes myself at the moment, and even on the used market probes are very pricey. If I were interested in 500MHz probes I would give these a try based on my experience with their other probe products.
Rich
-
If you really need 500MHz probe, forget passive probes... To see how useful are these just calculate their impedance when the tip capacitance is 10pF. Measuring at such bandwidth isn't going by just tapping here or there with the probe anyway.
For such bandwidth you either need a Low-Z probe (reasonably easy to DIY) or a real active probe.
Usefulness of passive probes tops around 200-300MHz.
-
Beware that active probes are easily damaged and virtually impossible to repair so it's easy to end up buying junk if you're not careful. The P6139A is the passive probe that would have come with that scope, while trying to actually measure up to 500MHz with a passive probe has the caveats mentioned above they are nice probes. Just don't buy probes from China ebay sellers, I did that recently, received hacked together garbage and had to file a claim through ebay to get a refund.
-
I bought some new old stock resistive probes and have made my own. These are fairly inexpensive. You may want to have a look at used active probes. I have a very old Tektronix active probe and a couple of differential probes that I picked up for a decent price.
-
Hi All,
I scored a Tektronix TDS 540C 4 channel 500 MHz oscilloscope on eBay. It's on route to me now. However I'm fairly certain there are no probes included with it.
A quick look around shows that high quality passive probes will set you back $400 - $500 apiece. That's way out of my hobbyist budget.
Do you guys have recommendations for 500 MHz passive probes that are not obvious junk but are not budget busters either?
It's a 4 channel scope, but at this stage I'll probably just get a pair. The scope and probes will be put to general use, nothing exotic.
Thanks for your input!
- George
What are you going to be probing?
If ordinary digital logic then that can probably tolerate a loading of 500ohms//0.7pF (*10), or 5000ohms//0.7pF (*100). Strongly consider a "low" impedance Z0 resistive divider probe. They are easy to make.
However, that requires a scope with a 50ohm input. It isn't clear to me whether the tds540 has a proper 50ohm input (like the tek485) or merely has 50ohm resistor in parallel with 15pF//1Mohm (like the tek24x5 and many others).
-
Thanks for all the insightful replies so far! I'll be mainly using it for digital logic. I don't *really* need 500 MHz of bandwidth, just like how many times have you gone 180 MPH in your car just because the speedometer tops out at 180. ;D ;D
But the scope was priced right and has four channels which made it desirable. And I do want to see how well it will perform. Can I really only get 3 db of drop at 500 MHz? No, I haven't looked at the specs for the scope or probes to see what the losses are as one approaches 500 MHz. I want to do that myself and see what results I can get.
And yes, I have been resisting the temptation to get a P6139A or P6139B probe from eBay. Who knows what you will actually get. And I don't have the proper gear to test out a probe, I have to take it on faith it will perform as advertised.
Also, I like the low Z idea that has been mentioned. I'l be looking into it.
- George
-
This is probably the thread you want for constructing Z0 probes...
https://www.eevblog.com/forum/projects/lo-z-probe/ (https://www.eevblog.com/forum/projects/lo-z-probe/)
I humbly suggest that the ones in reply #23 look particularly nice. :D
Don't forget that, when constructing such probes, you need a 50 ohm BNC through-terminator at the scope end (I don't remember the 540C having a built-in 50R termination option - I may be wrong though).
-
Can you use Z0 probes with a 50 Ohm terminator into a 1Meg scope input?
-
Can you use Z0 probes with a 50 Ohm terminator into a 1Meg scope input?
Ofcourse. With a 50Ohm terminator (either as a feed-through or terminator on a tee).
-
And yes, I have been resisting the temptation to get a P6139A or P6139B probe from eBay. Who knows what you will actually get. And I don't have the proper gear to test out a probe, I have to take it on faith it will perform as advertised.
I don't see why you are worrying so much. As long as they are not mechanically damaged, they should be fine. Also, you can test them with the scope's calibrator.
-
Wasn't there an issue with very early P6139s developing cable shorts? I don't know if that was a P6139 / P6139A change?
http://w140.com/tekwiki/wiki/P6139 (http://w140.com/tekwiki/wiki/P6139)
I never experienced any issues myself - other than the rubber on one going sticky... https://www.eevblog.com/forum/repair/tek-p3169a-_probe-rot_-repair/ (https://www.eevblog.com/forum/repair/tek-p3169a-_probe-rot_-repair/)
-
Can you use Z0 probes with a 50 Ohm terminator into a 1Meg scope input?
Ofcourse. With a 50Ohm terminator (either as a feed-through or terminator on a tee).
Except that it will actually be 50ohm//20pF or whatever the scope's input capacticance is. 20pF@500MHz is 15ohms.
And since the transmission line isn't properly terminated, there will be reflections and the response will be frequency dependent.
-
Thanks for all the insightful replies so far! I'll be mainly using it for digital logic. I don't *really* need 500 MHz of bandwidth, just like how many times have you gone 180 MPH in your car just because the speedometer tops out at 180. ;D ;D
Exactly! So for 'daily driver' use, I'd recommend a cheap set of 100 MHz probes. Whatever you get for higher frequencies, keep them safely stashed away until you really need them.
Also, I like the low Z idea that has been mentioned. I'l be looking into it.
Rather than go the DIY route, I ended up with a pair of HP 10020A low impedance resistive probes. They have <1 pf capacitance, variable divider ratios from 1:1 to 100:1 and a frequency response of 700 MHz. They sometimes show up on ebay and aren't usually too expensive - until now! :palm: :-//
-
Thanks for all the insightful replies so far! I'll be mainly using it for digital logic. I don't *really* need 500 MHz of bandwidth, just like how many times have you gone 180 MPH in your car just because the speedometer tops out at 180. ;D ;D
Exactly! So for 'daily driver' use, I'd recommend a cheap set of 100 MHz probes. Whatever you get for higher frequencies, keep them safely stashed away until you really need them.
But throw away the 6 inch ground lead, since that will introduce a resonance at @100MHz (150nH//20pF). Use the curly spring ground, and a groundplane.
Also, I like the low Z idea that has been mentioned. I'l be looking into it.
Rather than go the DIY route, I ended up with a pair of HP 10020A low impedance resistive probes. They have <1 pf capacitance, variable divider ratios from 1:1 to 100:1 and a frequency response of 700 MHz. They sometimes show up on ebay and aren't usually too expensive - until now! :palm: :-//
I have three :) But two of then are 10:1 only :(
I even used them as inspiration for a 3D printed "bayonet" for a "high" impedance *10 probe, to learn about 3D modelling and materials - and reduce the inductance. The results (https://entertaininghacks.wordpress.com/2016/09/17/scope-probe-accessory-higher-frequency-results/) were good. (N.B. the Tek485 has a real 50ohm input, not 50ohm//15pF!)
-
Can you use Z0 probes with a 50 Ohm terminator into a 1Meg scope input?
Ofcourse. With a 50Ohm terminator (either as a feed-through or terminator on a tee).
Except that it will actually be 50ohm//20pF or whatever the scope's input capacticance is. 20pF@500MHz is 15ohms.
But there will be some series inductance too. A properly designed 1M Ohm input isn't purely capacitive at higher frequencies by design. Otherwise it wouldn't make sense to have >100MHz bandwidth on a 1M Ohm oscilloscope input.
Besides that your circuit is also wrong. At the input the impedance from the divider towards the cable should be 50 OHms as well; that is likely the cause for the reflections in your simulation.
-
Thanks for all the insightful replies so far! I'll be mainly using it for digital logic. I don't *really* need 500 MHz of bandwidth, just like how many times have you gone 180 MPH in your car just because the speedometer tops out at 180. ;D ;D
Exactly! So for 'daily driver' use, I'd recommend a cheap set of 100 MHz probes. Whatever you get for higher frequencies, keep them safely stashed away until you really need them.
But throw away the 6 inch ground lead, since that will introduce a resonance at @100MHz (150nH//20pF). Use the curly spring ground, and a groundplane.
Totally agree! Some cheap probes *don't* include the springs so check carefully!
-
But throw away the 6 inch ground lead, since that will introduce a resonance at @100MHz (150nH//20pF). Use the curly spring ground, and a groundplane.
Don't throw it away, it is useful. Just know not to use it when you are trying to measure high frequency or fast rise time signals. Proper probing technique is a whole subject in itself.
-
Can you use Z0 probes with a 50 Ohm terminator into a 1Meg scope input?
Ofcourse. With a 50Ohm terminator (either as a feed-through or terminator on a tee).
Except that it will actually be 50ohm//20pF or whatever the scope's input capacticance is. 20pF@500MHz is 15ohms.
But there will be some series inductance too. A properly designed 1M Ohm input isn't purely capacitive at higher frequencies by design. Otherwise it wouldn't make sense to have >100MHz bandwidth on a 1M Ohm oscilloscope input.
You'll have to explain that, preferably with a schematic to show the location and magnitude of that inductance.
Besides that your circuit is also wrong. At the input the impedance from the divider towards the cable should be 50 OHms as well; that is likely the cause for the reflections in your simulation.
The circuit is correct. See, for example, fig 3.2 in http://w140.com/tekwiki/images/6/62/062-1146-00.pdf (http://w140.com/tekwiki/images/6/62/062-1146-00.pdf)
If the capacitor is removed, the ripple disappears.
If the transmission line delay/length is changed, the frequency/spacing of the peaks changes as you would expect with a (very imperfect) comb filter.
Try the simulation yourself, and see.
-
Here is the difference between using a 50 Ohm input or 1M Ohm input (at 500MHz bandwidth) on a 66MHz square wave from a 50 Ohm source.
A cheap 50 Ohm feed-through:
[attachimg=1 width=800]
A 50 Ohm terminator on an SMA tee:
[attachimg=3 width=800]
Direct 50 Ohm:
[attachimg=2 width=800]
As can be seen the signals are not dramatically different.
-
Keep your eyes on HP 10441A or especially 10441B probes...very nice...I have a bunch, sometimes you get the entire "kit" of clips, tips, etc for $25 or so, I even got one for $10 once (new, in bag, with accessories). I only buy when it includes the tophat. Another time, I got 5 of this series probe for about total of $87 and they also came with a big bagful of extra gnd clips, top hats, etc,either a few or all 5 were new in sealed bags.
sometimes you see this exact same thing for $10 to $30 (5 for $87 was a once-in-a-decade deal)
https://www.ebay.com/itm/Agilent-HP-10441A-Oscilloscope-Probe/123724232534?epid=1900713148&hash=item1cce89f756:g:HUAAAOSwBt5ZD3Mr (https://www.ebay.com/itm/Agilent-HP-10441A-Oscilloscope-Probe/123724232534?epid=1900713148&hash=item1cce89f756:g:HUAAAOSwBt5ZD3Mr)
-
I've seen probes similar to the HP 10441 and I've always been puzzled by one thing. The specs state that the input impedance of the scope must be 1 Mohm in parallel with 6 - 9 pf. How many scopes have capacitance that low? I have a LeCroy 9384 1 GHz scope and even it has an input capacitance of 15 pf. If I was using a 10441 probe, I couldn't compensate it. :-//
-
I've seen probes similar to the HP 10441 and I've always been puzzled by one thing. The specs state that the input impedance of the scope must be 1 Mohm in parallel with 6 - 9 pf. How many scopes have capacitance that low? I have a LeCroy 9384 1 GHz scope and even it has an input capacitance of 15 pf. If I was using a 10441 probe, I couldn't compensate it. :-//
You wouldn't guess, HP scopes usualy made such way to save on two way attenuator.
-
R&S RTB2004 has 1MΩ||9pf input impedance.
-
Here is the difference between using a 50 Ohm input or 1M Ohm input (at 500MHz bandwidth) on a 66MHz square wave from a 50 Ohm source.
...
As can be seen the signals are not dramatically different.
Source characteristics, especially risetime? Scope characteristics, especially bandwidth? Lead length/characteristics?
Ensure the source has a risetime of <<3.5ns. Easily achieved with modern logic such as 74lvc which is <1ns and (with care) can be <300ps.
Then either do a frequency sweep or, more easily, use the standard technique of a PRBS (doesn't have to be very long) and an eye diagram with scope on infinite persistence.
-
The limitations of high impedance passive probes at 500 MHz have been discussed so I will skip that. Given their limitations, I would try to avoid spending a lot of money on them and instead save for a suitable active or low-Z probe. I have however had good results with TEXAS probes which are resold by many oscilloscope manufacturers. The 500 MHz ones with readout support can be found on Ebay (https://www.ebay.com/itm/One-new-500MHZ-Oscilloscope-clip-probe-w-readout-pin-TX6250/220306076099) for like $70 each.
Getting the best performance out of them requires a fast reference level pulse generator and coaxial probe tip adapter so their high frequency transient response can be calibrated to the oscilloscope. But again unless you are using them under ideal conditions which means 25 ohm source impedance and coaxial to probe tip connection, you will never get best performance anyway. A calibration source does make for a sanity check though.
-
The fact that the TEXAS probe is a dedicated X10 probe is a good sign (even if its input capacitance is a bit high at 10.5pF). You are far more likely to achieve believable characteristics on a dedicated X10 probe than any of the switchable X1/X10 ones.
-
I've never understood why most of the cheap probes are switchable. I've never wanted to use one in 1X mode, on the rare occasion that I want 1X I can just use a plain piece of coax directly into the scope. The switch is just extra complexity, extra capacitance, an extra opportunity for something to be flaky.
-
I would recommend 200Mhz 10$ P2200 probe from aliexpress. It easily goes up to 460Mhz -3dBm. TEXAS ones seems to be a little dodgy, It has three regulation capacitors and you need to configure it properly to have adequate frequency response. As you can see from pictures it is impossible to use straight from the factory. All probes and oscilloscope were set to 10X .
-
I would recommend 200Mhz 10$ P2200 probe from aliexpress. It easily goes up to 460Mhz -3dBm. TEXAS ones seems to be a little dodgy, I has three regulation capacitors and you need to configure it properly to have adequate frequency response. As you can see from pictures it is impossible to use straight from the factory. All probes and oscilloscope were set to 10X .
My findings are similar. Eg. I have a Tek P6106 (250MHz) and Tek P6105 (100MHz) vintage probes, these are modular probes, and they behave mostly the same only that the compensation box is more refined (complicated) on the P6106, so you can set up the probe response more carefully, while the P6105 had a little oversoot at real high frequencies, their ristime is basically the same. The same was observed with the even cheaper P6100 chinese probes, their ristime is good but one may not be able to dial in compensation properly. I think it is essential for universal (that is not meant for one specific scope family) high bandwidth passive probes to have more than a single point adjustment compensation.
-
The P2200 looks reasonably good, it would be even better if they eliminated the useless 1x/10x switch that most of the Chinese probe makers seem obsessed with.
I'm also disappointed in the small number of aftermarket probes that include the Tek readout pin. Once you're already molding connector shells the cost of including the readout pin ought to not be much. There are millions of scopes out there that use it.
-
TEXAS ones seems to be a little dodgy, It has three regulation capacitors and you need to configure it properly to have adequate frequency response. As you can see from pictures it is impossible to use straight from the factory. All probes and oscilloscope were set to 10X .
Texas has multiple models covering the same frequency ranges. I have had good results with the 250 MHz Texas probes which have the black compensation box with two adjustments and they were right on for my 300 MHz DSO.
The probes that they OEM for others are identical as far as I can tell except that the compensation box has the adjustment holes covered with the label. Keysight sells them as their N28xx series of probes with the 500 MHz probe being the N2890A or N2843A.
-
TEXAS ones seems to be a little dodgy, It has three regulation capacitors and you need to configure it properly to have adequate frequency response. As you can see from pictures it is impossible to use straight from the factory. All probes and oscilloscope were set to 10X .
Texas has multiple models covering the same frequency ranges. I have had good results with the 250 MHz Texas probes which have the black compensation box with two adjustments and they were right on for my 300 MHz DSO.
I did not said they were bad, its just they require extra calibration after purchase.
Here is how they look like after 15 min of playing with adjusting capacitors. After all this measurements I do not see any point in expensive brand probes.
-
I did not said they were bad, its just they require extra calibration after purchase.
The calibration is more important for higher bandwidth probes, say 300 MHz and higher, where the high frequency compensation has to be matched to the oscilloscope model's high impedance input. That is why Tektronix would release different letter versions of the same probe with the same bandwidth; they were for different oscilloscope models and mixing them up would spoil the high frequency transient response.
But like I said, this only matters when you use a 25 ohm terminated source with a coaxial connection to the probe tip because that is how probes are specified and calibrated. If you use the high impedance passive probe in any other configuration like probing common fast logic which is not 50 ohms with the spring tip adapter, everything goes out the window and you will not achieve the probe bandwidth anyway at least with a high impedance passive probe. So calibration is less relevant for most people.
I mostly use my calibration gear to make sure the probe is functioning properly because otherwise this is very difficult to know.
-
I did not said they were bad, its just they require extra calibration after purchase.
Here is how they look like after 15 min of playing with adjusting capacitors. After all this measurements I do not see any point in expensive brand probes.
It is not clear what you are measuring. What is your signal source, and how is it connected to the probe?
-
I did not said they were bad, its just they require extra calibration after purchase.
Here is how they look like after 15 min of playing with adjusting capacitors. After all this measurements I do not see any point in expensive brand probes.
It is not clear what you are measuring. What is your signal source, and how is it connected to the probe?
Signal source is R&S SMIQ03E signal generator, probe is connected to output via bnc to 6mm tip adapter.
-
Also here is example of Agilent 1165A miniature probe which 100 times more expensive if purchased as new. I use it most of the time because it is very soft and comfortable to use.
-
I did not said they were bad, its just they require extra calibration after purchase.
Here is how they look like after 15 min of playing with adjusting capacitors. After all this measurements I do not see any point in expensive brand probes.
It is not clear what you are measuring. What is your signal source, and how is it connected to the probe?
Signal source is R&S SMIQ03E signal generator, probe is connected to output via bnc to 6mm tip adapter.
Without the feedthrough termination, the source impedance will be 50 ohms halving the potential bandwidth. The stub length of the adapter is a problem also.
The real coaxial probe tip adapters used at 500 MHz and higher combine the feedthrough termination as part of the adapter so it is located as close to the probe tip as possible. Tektronix used to sell them but I do not think anybody makes them available now. You can sometimes find old Tektronix ones on Ebay. These are good to 4 GHz or even higher.
A good alternative if you can find it is a coaxial T adapter which is designed to interface directly with the probe tip in sampling applications. This allows the termination to be located remotely. Having a 50 ohm transmission line on either side provides the 25 ohm source at the probe tip.
You can make a pretty good probe tip adapter using stripline construction with a probe tip to printed circuit board socket and a pair of 100 ohm surface mount resistors for the termination. This sort of socket allows the pin on the end of the probe tip to be directly inserted into the printed circuit board. They look very much like a Peltola connector (http://w140.com/tekwiki/wiki/Peltola_connector).
-
Without the feedthrough termination, the source impedance will be 50 ohms halving the potential bandwidth. The stub length of the adapter is a problem also.
Bandwidth of what? Generator can go up to 3.3Ghz.
High impedance passive probes are actually strip of restive wire around 300 Ohms so they are universal and do not cause much of reflections due impedance mismatch.
Why 25 Ohm output is so important for high impedance passive probes testing?
-
Without the feedthrough termination, the source impedance will be 50 ohms halving the potential bandwidth. The stub length of the adapter is a problem also.
Bandwidth of what? Generator can go up to 3.3Ghz.
High impedance passive probes are actually strip of restive wire around 300 Ohms so they are universal and do not cause much of reflections due impedance mismatch.
Why 25 Ohm output is so important for high impedance passive probes testing?
The source impedance combined with the tip capacitance limits bandwidth. A parallel terminated 50 ohm output is 25 ohms. So you get different results if the 50 ohm output is not itself terminated and probes are specified and calibrated with a 25 ohm source produced from a terminated 50 ohm source.
This is also why high bandwidth high impedance oscilloscope inputs have lower input capacitance. 20 picofarads was common at 100 and 200 MHz but lower input capacitance is required for higher bandwidth and 500 MHz is about as high as a 1 megohm input can support.
-
This is probably the thread you want for constructing Z0 probes...
https://www.eevblog.com/forum/projects/lo-z-probe/ (https://www.eevblog.com/forum/projects/lo-z-probe/)
I humbly suggest that the ones in reply #23 look particularly nice. :D
Nice one. Been thinking about making something similar. Thinking about using Crystek SMA housing (SMA-KIT-1.5MF) for probe handle. Need to find some kind longer probe pin or needle that fits and extends out of the SMA bushing for the probe tip. So that naybe the resistor could be inside the housing. Though that would make the probe pin a few cm long after the resistor and I couldn't tell right now what that would effect.
-
Sewing needles and pins tend to be difficult to solder but I have had good results using a sharpened piece of music or piano wire which is high carbon steel.
-
Sewing needles and pins tend to be difficult to solder but I have had good results using a sharpened piece of music or piano wire which is high carbon steel.
But don't try to cut piano wire with standard diagonal cutters!
Piano wire can be bent into useful shapes. Repeated bending rapidly causes work hardening, and the resulting fracture might be sharp enough without specifically sharpening it