Lo-Z Probe

[prenato]

Sometimes also know as the "Howard Johnson" probe, though I'm not sure he was the first to come-up with the idea:

http://paulorenato.com/joomla/index.php?option=com_content&view=article&id=93:praise-for-the-lo-z-probe&catid=4:projects&Itemid=4

Works surprisingly well...

Paulo

[robrenz]

Nicely done!

[prenato]

Thanks Rob!

[notsob]

And heres Doug Smiths version

http://www.emcesd.com/1ghzprob.htm

[BravoV]

Thanks for sharing ! 

This popular cheap probe, its proven really good, another version by our fellow Jahonen (Janne) with its characteristic plotted using HP 8753D VNA from 30 kHz to 6 GHz and it's rising edge beats the 7 GHz Agilent Infiniimax probe  -> http://koti.mbnet.fi/jahonen/Electronics/DIY%201k%20probe/ , this link was discussed here -> https://www.eevblog.com/forum/projects/cheap-wideband-121-oscilloscope-probe-for-logic-signal-applications/

InfiniiMax Probe .. from the appearance it self, I'm just too scared to look for it's price .. 

[vk6zgo]

Sometimes also know as the "Howard Johnson" probe, though I'm not sure he was the first to come-up with the idea:

http://paulorenato.com/joomla/index.php?option=com_content&view=article&id=93:praise-for-the-lo-z-probe&catid=4:projects&Itemid=4

Works surprisingly well...

Paulo

I'm sorry,but after the preamble about measuring a 50 Ohm source,the obvious assumption is that his PCB also has an output Z of 50 Ohms.
If so,why would he need a probe  at all?

A plain coax lead with an SMA one end & a BNC the other,plus a 50 Ohm termination,would be the normal way to do this.

OK,there are spots where you do need to probe,but his example was a poor one.

[saturation]

Nice.  Here's details on HoJos probe by the man himself:

http://www.sigcon.com/Pubs/straight/probes.htm

[alm]

I'm sorry,but after the preamble about measuring a 50 Ohm source,the obvious assumption is that his PCB also has an output Z of 50 Ohms.

His example of the 50 ohms source was to show the effect of capacitive loading of high-Z passive probes. This effect will be even worse, and the resistive probe will be more superior, at higher source impedances.

looking the docu made by people like Janne, Douglas and this made me think that most aftermarket passive probes out there are just audiophoolery stuffs.

Lo-Z probes have some advantages, but are much less versatile. Try subjecting one to mains voltages or using it in an audio circuit (usually fairly high impedances). A high-Z passive probe is much more complex to make, since you can't just use a simple transmission line. There is also the issue of mechanical integrity. How long is that resistor at the end of a coax cable going to last if it is used every day?

[codeboy2k]

Lo-Z probes have some advantages, but are much less versatile. Try subjecting one to mains voltages or using it in an audio circuit (usually fairly high impedances). A passive probe is much more complex to make, since you can't just use a simple transmission line. There is also the issue of mechanical integrity. How long is that resistor at the end of a coax cable going to last if it is used every day?

... I'll be waiting for robrenz to apply his skills to making a low-Z probe like these ones, that is also mechanically sound!

[PA4TIM]

looking the docu made by people like Janne, Douglas and this made me think that most aftermarket passive probes out there are just audiophoolery stuffs.

Lo-Z probes have some advantages, but are much less versatile. Try subjecting one to mains voltages or using it in an audio circuit (usually fairly high impedances). A passive probe is much more complex to make, since you can't just use a simple transmission line. There is also the issue of mechanical integrity. How long is that resistor at the end of a coax cable going to last if it is used every day?

A lo-Z probe is not a replacement for a passive general purpose probe. It. A way to lower capacitance without the disadvantage of fet probes ( limmited voltage ratings and psu need)
It is used in high speed probing. Not for gerenal use. I played a bit with them too. They really perform very good. I used it to probe a 500 ps pulser, a 250MHz squarwave and it gave very good results. And as a compare: I have a 500 MHz HP FET probe, 1GHz philips and a Tektronix active probes, three commercial 1GHz  hi-Z probes ( strange thing is they are called hi-Z but are made the same way, only with higher resistance so higher Z (100x probe) so where is the border you go from high to low-Z, is it not better to talk about high Resistive but low reactance probes ?
In real high speed use they solder a probe adapter straight to the pcb and that holds the low-Xc probe
That is an other advantage of the DIY probe. Just solder the resistor to the pcb.

By the way, I found that a small capacitor from coax centre to shield improved the performnce. Also the type of resistor. Very funny, tried a bunch off odd choises like a carbon composite that worked very well ( no guaranty they all do) and a wirewound with two trimmers was one of the best, but you had to tune it for other frequencys  ;-)

[vk6zgo]

I'm sorry,but after the preamble about measuring a 50 Ohm source,the obvious assumption is that his PCB also has an output Z of 50 Ohms.

His example of the 50 ohms source was to show the effect of capacitive loading of passive probes. This effect will be even worse, and the resistive probe will be more superior, at higher source impedances.

Yeah,I know,but my point was that a source  with a 50 Ohms output Z was a poor choice of example,as there is a better way to look at such an impedance.
I have noticed in the past that some less experienced folk on this forum will "fall in love" with some technique & use it instead of a a more appropriate one.
My comment was directly mainly to this section of the community.

looking the docu made by people like Janne, Douglas and this made me think that most aftermarket passive probes out there are just audiophoolery stuffs.

Lo-Z probes have some advantages, but are much less versatile. Try subjecting one to mains voltages or using it in an audio circuit (usually fairly high impedances). A passive probe is much more complex to make, since you can't just use a simple transmission line. There is also the issue of mechanical integrity. How long is that resistor at the end of a coax cable going to last if it is used every day?

The resistor choice is fairly critical,too,as some film capacitors look like inductors at higher HF & VHF frequencies,although as PA4TIM has demonstrated,compensation is possible.

[grenert]

By the way, I found that a small capacitor from coax centre to shield improved the performnce. Also the type of resistor. Very funny, tried a bunch off odd choises like a carbon composite that worked very well ( no guaranty they all do) and a wirewound with two trimmers was one of the best, but you had to tune it for other frequencys  ;-)

That's interesting:  There was an article in Nuts and Volts in November 2010 which had a recipe for a Lo-Z probe, and it specifically said to use a carbon resistor!  I wonder what characteristics make for a good resistor for this application.

[SeanB]

Non inductive cut carbon or metal film works well, but they are expensive, as they are adjusted during manufacture by sandblasting the material away all around the unit or by using a flat substrate and eroding it across the whole surface before conformal coating it. Hand work that takes time. Capable of very good response at high frequencies.

Carbon composition has tolerance problems, though you can adjust them a little with a hacksaw to get them higher in value closer to the desired value. Then seal with epoxy. They also have a limited frequency response and increase resistance with frequency as skin effect increases.

[PA4TIM]

Carbon composite is not good for much but they have two interesting features.
They are slightly capacitive, this lowers there resistance with increasing frequency. That is, according to literature, i have not read about skineffect in resistors. So I must measure for the total  effect . But they have a voltage cooficient and terrible tempco so i will not advise them.

They are good for pulsed loads and high voltage, but that is not interesting for high frequency low level probing.

But the capasitive part is interesting maybe in some cases function as compensation.

[tggzzz]

Sometimes also know as the "Howard Johnson" probe, though I'm not sure he was the first to come-up with the idea:

http://paulorenato.com/joomla/index.php?option=com_content&view=article&id=93:praise-for-the-lo-z-probe&catid=4:projects&Itemid=4

Works surprisingly well...

Paulo

Here's a Tektronix publication from 1969 describing a "Z0 probe", and it makes no claim for originality; see http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf and (for more general material) https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Certainly Tek and HP were suppling such probes back then.

[w2aew]

Sometimes also know as the "Howard Johnson" probe, though I'm not sure he was the first to come-up with the idea:

http://paulorenato.com/joomla/index.php?option=com_content&view=article&id=93:praise-for-the-lo-z-probe&catid=4:projects&Itemid=4

Works surprisingly well...

Paulo

Here's a Tektronix publication from 1969 describing a "Z0 probe", and it makes no claim for originality; see http://www.davmar.org/TE/TekConcepts/TekProbeCircuits.pdf and (for more general material) https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Certainly Tek and HP were suppling such probes back then.

The Tek P6150 (500 ohm input Z) and P6158 (1k ohm input Z) are still made today by Tek.  Very useful - I've used them in several of my videos (the one that illustrates the filter functions of an RF up-converter comes to mind).

[TimFox]

The problem with high-value carbon composition resistors at high frequencies is their internal construction.
The internal material is a mixture of powdered carbon and powdered talc, with less carbon at high resistance values.
The talc grains form tiny capacitors inside the resistor volume, which short out the carbon locally.
Therefore, if you model the device as a resistor and capacitor in parallel, the parallel resistance falls at high frequencies.
I measured this effect a long time ago with 1 megohm A-B traditional resistors at 5 to 20 MHz, and the effect was dramatic.

[dadler]

So not DIY but I managed to get these on eBay for cheap, and for the convenience I think they are awesome.

Maybe one can find a deal if they look around?

[mazurov]

I usually solder 0603 resistors tombstone-style where I need to probe then solder a coax to the upper end of one of them and ground nearby.

[Ivan7enych]

I've made this LoZ probe and checked how it works on spectrum analyzer + tracking gen (signal hound bb60c + TG44).

1st attempt was made with simple 1\8 wt 1k resistor. The response rises high up to +12dB on 4GHz

2nd attempt was done with 4 smd resistors, soldered on each other to minimize parasitic capacitance, 51 + 120 + 220 + 560 = 951om
It appears much better. But I don't like the 2dB ripple across all frequency range.

I've added 20dB attenuator from minicircuits between probe and analyzer, it almost removes the ripple.
It seems, that my analyzer has not exactly 50om termination and reflects some energy back to probe, but  the probe end is not 50om matched at all, and reflects everything back again.

I think, this probe requires very good 50om impedance matched input on oscilloscope (or analyzer) end to minimize the ripple.

Another variant is to make probe end 50om matched (add 56om resistor between coax shield and center wire on probe end). I'll try to do this.

[tggzzz]

I've made this LoZ probe and checked how it works on spectrum analyzer + tracking gen (signal hound bb60c + TG44).

Which LoZ probe? Many different ones have been described, and the details matter! Would a photo give us any useful information?

1st attempt was made with simple 1\8 wt 1k resistor. The response rises high up to +12dB on 4GHz
2nd attempt was done with 4 smd resistors, soldered on each other to minimize parasitic capacitance, 51 + 120 + 220 + 560 = 951om
It appears much better. But I don't like the 2dB ripple across all frequency range.

Thanks for showing the results. Of course the 2dB imperfection has to be weighed against the alternatives.

A quick spice simulation of the probe plus analyser input can reveal the theoretical response, and you can compare your practical circuit against that. Tweaking various values gives a feel into what is important.

[Ivan7enych]

Which LoZ probe? Many different ones have been described, and the details matter! Would a photo give us any useful information?

The simplest type,
SMA connector + 70cm of rg316 cable + a resistor (1k or 950om as a sum of 4 smd resistors)

[joeqsmith]

I tried to make some of these embedding the resistors inside of the coax.  Not the best results. 

https://www.eevblog.com/forum/testgear/fifty-ohm-probes/15/

[Gyro]

Just to add my build to the mix:

The only novel feature they bring to the party (apart from a bit of gold plated bling) is that they include inserts from a turned pin IC socket which allows me to easily unplug and replace the tip resistors (normally 1k for ~20:1 with 50R through-terminators) for different impedances or damage. The ground contacts could probably be a bit shorter but apparently less of an issue with Zo probes.

I don't make any claims for frequency flatness but they did allow me to view the outputs of some 100MHz crystal oscillators which simply refused to function with a Tek P6139A (8pF) passive probe!

[joeqsmith]

Looks good.   TGZZZ posted that EDN article that I never even noticed until I reposted.     Things get buried so fast on this site.     I have never came up with one that would work very well over a GHz. 

I picked up some new old stock of LeCroy PP061 resistive probes that I have no documentation for.  These were very inexpensive.