Why Oscilloscope Probes are expensive

[JustAnotherGuy999]

it's just a cable with a cap and resistor, why is it expensive?
for example Tektronix P2220 1X/10X 200 MHz costs $117.90

[retrolefty]

it's just a cable with a cap and resistor, why is it expensive?
for example Tektronix P2220 1X/10X 200 MHz costs $117.90

 Part is name brand and published and trusted specifications. However cheap Asian probes are and will be available that may vary in quality but some should work within their specification.

 This might make a decent 'shoot-out' amoung different price class probes Vs 'Brand' priced and see how well they all perform?

[JustAnotherGuy999]

they are still expensive "in my opinion" compared to the components they contain.
I mean why people don't just "hack" a probe using a wire a cap and a resistor, there something i don't know

[TimFox]

it's just a cable with a cap and resistor, why is it expensive?
for example Tektronix P2220 1X/10X 200 MHz costs $117.90

If you assemble a probe with only a 9.0 megohm resistor, a trimmer capacitor, a length of RG-58/U cable, and a BNC plug, it won't meet the 200 MHz specification.  Try it:  it would be a cheap experiment.

[tggzzz]

it's just a cable with a cap and resistor

No, it isn't. Their construction is surprisingly subtle and complex.

To start to understand, see the references in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

[JustAnotherGuy999]

TimeFox can you explain plz what does "it won't meet the 200 MHz specification" mean

[PA4TIM]

You think it is just a cable, capacitor and resistor  Do not forget the switch a cheap 1/10x probe has. Good probes are not switchable so that should make them even cheaper 

A good 10x probe for , lets say, a 100 MHz scope must have a higher bandwidth. It has often 2 or more adjustable components in the compensation  compensation network(s). That network is more as just a resistor an a cap. Often several resistors, inductors and capacitors. But also relative cheap parts. The probe body is a thin wall metal tube for shielding purpose. The probe cable is a complex part. It is 1 strand of very thin wire. You can see this more or less a distributed network on its own. Besides that is must not break easy. from 99% of the broken probes I had, the cause was the cable.
Besides that there is often the 1x, 10x, 100x recognition mechanisme in the bnc.

Then the tip, these can bend and break. But in good probes they are replaceable and for sale. This is not a universal part.

A good probe is often repairable, a cheap probe often not. I have repaired several probes with succes.  See some pictures from probe guts: http://www.pa4tim.nl/?p=5549

By the way, an expensive probe can cost a lot more as 117 dollar and one of the increasing factors is bandwidth.

[JustAnotherGuy999]

thanks guys

[AlfBaz]

Dave did a video where he "probes" into probes. He shows the peculiar construction of the cable in it
I think it's this one

[Tom45]

Your question is actually a good one because scope probes do indeed look like very simple devices. But there is an amazing amount of technology that goes into the design and manufacture of such an apparently simple device.

Besides Dave's video that was referenced, you can go straight to Tektronix, the inventors of high frequency scope probes:

http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf

Whether or not you understand all of the Tekronix probe document, you should at least get a feel for just how much goes into the design of a high frequency scope probe.

[JustAnotherGuy999]

thanks

[mmagin]

People often look at things like this and forget that the good quality product has good quality mechanical parts, and the mechanical parts are very often much costlier than the electronic components.   In general, good quality test leads (and scope probes) are far more durable and will continue to perform correctly for longer than the cheapest approach.  This even comes down to simple things like the plating on hooks, probes, and alligator clips, the springs specified, the design of switches, etc.  And it costs even more to ensure uniform quality of those things across every production run of the product.

Also, it's just not made in the quantities of consumer electronics, so you can't hope to get as much for your money as you do in say, your smartphone.

[Howardlong]

I really liked this article http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

Which explains why rolling your own hi-z probe ain't so easy.

[saturation]

Low cost Chinese passive probes are decent so long as you know and work within their limits.  Its 60+ year old technology.  There's a lot of data in the eevblog archives, here are some choice ones. 

https://www.eevblog.com/forum/reviews/scope-probe-what-the-difference-between-100mhz-and-500mhz/msg395742/#msg395742

Note the frequency response to actual sweeps between an Agilent probe and a nameless cheap one:

https://www.eevblog.com/forum/reviews/cheapest-100mhz-oscilloscope-probes-hands-on-review/msg100171/#msg100171

[pickle9000]

Not a bad question.

[Mechatrommer]

it's just a cable with a cap and resistor, why is it expensive?

because the knowledge to build it. try build it, once your get the knowledge, it will be cheaper to diy... just dont expect fancy looking probe.

[andyturk]

I have repaired several probes with succes.

Hey PA4TIM,

I've got a bad Rigol RP3500. It just seems to be open. Any suggestions on how to find the problem?

[klunkerbus]

I won't defend a $117 probe, but I worked for a test equipment company in the 70's and led the oscilloscope development team. On a new scope we were releasing, I was surprised at the hassles we ran into with the associated probe. Coming up with a physical design was daunting, between meeting the upper end of the 1KV voltage spec for the probe, the tooling costs of molded parts, and my god, finding a cable that would not be microphonic when you touched it or dragged it across the bench. Who wants a scope trace that jumps around as the probe cable is moved?  Seriously ended up having a cable house manufacture custom cable for the probe... As a pretty green engineer at the time, that was quite the eye opener on how subtle components can really bite you in the behind.  The rest of the oscilloscope design was pretty straightforward. 

EDIT: Also... Striving for a small design that still felt well from a human factors standpoint.  Obtaining a flat frequency response.  Providing the means to adjust compensation.  Coming up with a means of providing a sharp point tip that had some hardness to it.  Coming up with a clip accessory that would do at least a fair job across a wide range of applications. 

[PA4TIM]

Hey PA4TIM,

I've got a bad Rigol RP3500. It just seems to be open. Any suggestions on how to find the problem?

Yes, the first problem is, that it is a Rigol.  >:-)

Use TDR to find the location of the open. It will probably be the cable and most chance it is at the end just before the point it enters the probe.
Then you have to take apart the probe and that can be a big problem. If you can take them apart you have to cut the cable and remove the crimp-ring that holds the shield of the cable. Then you have to find a way to connect the core of the cable to the probe. If you are lucky, you can solder the material, if not you need to find an other way. Then crimp a new collar to the "connector" used to couple the cable to the probe body. If there is one. I had two 100 MHz Rigol probes that lasted less as a year. I could not take them apart and was not able to repair them. But  they performed so sad  that I probably did not try it hard enough.

[TimFox]

TimeFox can you explain plz what does "it won't meet the 200 MHz specification" mean

Attached is the result of a Spice simulation of the AC response for a simple probe. 
The circuit comprises:
1.0 megohm and 20 pF scope input (typical values at panel BNC)
6 nsec of 50 ohm coax (approximately 4 feet of RG-58/U at 66% velocity), which has 120 pF capacitance at low frequency.
9.0 megohm and 15.5555 pF (140 pF/9) between probe point and coax input.
Note that the attenuation at low frequency is 20 dB, as expected, but we get lower attenuation above about 20 MHz, and serious resonant problems above 50 MHz.  These resonances are in the improperly-terminated 50 ohm coax.
With 2 feet of cable (3 nsec), the effects kick in around 100 MHz, and still mess up the circuit at 200 MHz, as shown in the second graph.
Note that Spice does not allow a frequency-dependent attenuation in the transmission-line model.  In reality, the resonances won't be quite this strong, due to loss in the coax at high frequencies.

[Mechatrommer]

I've got a bad Rigol RP3500. It just seems to be open. Any suggestions on how to find the problem?

with press at different probe cable location, i can pinpoint where the break is, luckily (or unluckily?) its near the BNC connector, so i split them into 2 and i will reinstall new BNC connecter. but later when i have time for it...

[andyturk]

Thanks Mechatrommer and PA4TIM!

You guys inspired me to actually take my broken probe apart and see what was wrong. Turns out it wasn't "open", the inner conductor wire was actually shorting against the metal probe shield.

There were some extra loops of the zig-zag conductor, so I desoldered it, cut it to an appropriate length and pushed a tiny bit of insulation inside the hole to keep the conductor off the shield. It works like a champ now. 

[Jbliss]

Probably because there is quite a bit of design that goes into them

[kunzem]

yep design plus the quality assurance is what you pay for, i dont want to relay on a 10$ probe from some bunker in whang hang loaw when i want reliable results in my tests. (sure material wise they are overpriced x100, but thats not what you pay for).

[borjam]

The probes are certainly the weakest point of affordable oscilloscopes. I've got a Rigol DS1074Z and the probes are really crappy. Just tapping them with the finger shows how flaky they are.

But of course if I get four decent probes at $100 each, well, it's more than 50% of the price of the oscilloscope itself. Anyway it would be great to have a shootout of reasonably cheap but modestly reliable probles, if that thing exists after all!