cable with a very low capacitance/unit of length.

==========================

** You sure ?

Scope probes become very wide band when terminated with a 50ohm dummy load.

FYI the "secret" of low C co-ax is the use of very THIN centre conductors.

.... phil

Unfortunately since the E field goes like 1/r, as the centre conductor shrinks, the E field near there grows, so you only win logarithmically as you go towards thinner centre conductors.

You can get higher impedance with slow-wave networks, e.g. a helical centre conductor, but that does nothing to reduce capacitance.

Unless you're connecting to something in a very hostile environment, the usual way round this is to put some circuitry at the business end of the cable, as in a FET scope probe.

Cheers

Phil Hobbs

RG178 has about 10 pF lower capacitance per unit length than RG174, but also significantly greater attenuation in dB/100m across-the-board from

100 MHz to 1 GHz, both into 50 ohms. RG178 has very thin center conductor, significantly more than 1/10th as thin as RG178 I think, its loop resistance is 2x RG174.

More than 1/10th as thin as RG174, rather

If you are NOT making an oscilloscope probe, RG-62a/u has the lowest capacitance of commonly available coax cable at 12.8pf/ft:

It's commonly available online: It's also MUCH stiffer than oscilloscope probe coax cable and might not be suitable for whatever you're doing.

Oscilloscope probe coax cable is typically 9 or 10pf/ft. However, I couldn't find a bulk source for this type of cable or a detailed data sheet. Try searching for "low capacitance coaxial cable".

Thanks. I found the Belden 9857, RG-63, with 9.7 pF/ft. Hope I can find a supplier without purchasing a huge quantity!

=======================

** Depending on the app - is "bootstrapping" the co-ax shield an option ??

Could almost eliminate parallel C from the game.

..... Phil

You can get some here for about $ 1.39 a foot in small lengths. I have not ordered from him, but have bought from him many times in the past in person at the hamfest flea markets.

I bought some low capacitance cable off ebay a while back. It was some that had connectors on each end. Just cut the connectors off as I did not need that type.

Just curious, but why do you need it?

To connect a photodiode to an amplifier. PHD + cable capacitance causes an opamp to oscillate and causes the series noise of this opamp to create an equivalent input current noise in capacitance*2*pi*frequency.

So the obvious question is: "Why don't you move the amplifier closer to the photodiode?"

Then in addition, there are a whole lot more improvements you can make. Ask Phil Hobbs, and/or read his book.

On a sunny day (Fri, 10 Sep 2021 16:23:49 +0200) it happened Jean-Pierre Coulon snipped-for-privacy@cacas.pam.oca.eu> wrote in <alpine.WNT.2.21.999.2109101621110.15816@dhcp3-6>:

Why not drive it into a low impedance like a transistor base After all it is just electrons. The tranny will give you current gain. Did that with a vidicon in the sixties, worked perfectly

Vaccuum, radiation, cryonic temperatures?

So, solutions would include: (1) double-shielded wiring, with the inner shield driven to match the center wire (no capacitive current, because the inner shield is at same pottential as the signal wire) (2) buffer amplification at the source (there are MOSFETs that work well at cryogenic temperatures, and in vacuum) (3) use of a virtual-ground current receiver, i.e. a transconductance type amplifier: just an opamp (-) input with feedback resistor and signal wire, and (+) input connected, possibly through a resistor, to signal GND (but that can also be a nonzero bias voltage, if bypassed well)

Vacuum, at least, isn't very compatible with most coaxial cables (except the rigid/semirigid ones).

try Trompeter TWC-124-1A TWC-124-2

Maybe, but we haven't been told. Most of the things that would really destroy the electronics would also destroy the photodiode.

Do you have Phil Hobbs' book? He analyzes pd amps in detail.

There is a complex tradeoff of speed and noise, and pd+cable capacitance are critical. Cascoding, namely loading the pd with a low impedance, is one way to reduce the effects of the capacitance; it keeps the tia from seeing all the pd+cable capacitance and getting all unstable.

It's impressive how many data sheets and appnotes about photodiode amps just miss cascoding, and grossly compromise bandwidth.

Every time I see this question I think window line. I'm sure someone will complain about that though.