Best way to get a function generator signal to a breadboard?

[Pasky]

Gonna be picking up my first function generator, not quite sure how I'm going to get a BNC cable to a bread board for testing and what not.  I know one way would be to have something with a long lead and use an oscilloscope probe and attach it to it, but what if I want to use both of my probes?

Any ideas?  Thank you.

[Simon]

By another oscilloscope probe or depending on what you want to do there is nothing stopping you making up your own plug and pair of wires for low-frequency. I don't know what sort of circuits you intend to play with and if you need to have cable of a certain impedance. I don't know what it was used for but I picked up a box of junk being thrown out by my local talking newspaper and it included a BNC plug with a plastic body which broke the two connections out into two separate screw terminals. I would assume they were for audio use.

[alanb]

I would buy a bnc plug and some coaxial cable and make up my own lead. If you use oscilloscope probe leads beware. They are resistive even on the 1x setting and unless the input on the bread board that you are feeding into is high impendence you will attenuate your signal.

[Pasky]

Okay, thank you for the replies.  I was just curious if there was actually anything specific for this task before I go and gut some cables.

[liquibyte]

I just got these to see if I could output my scope to an audio card.  I wonder if what Simon described is what these are?  They're used for video connections I believe.

Edit for a pic.

[Simon]

Well there will be the correct cable for the job but it depends on what the job is. Your signal generator has a certain output impedance and your circuit on the breadboard will also have an input impedance so the cable will need to be suitable. And of course as the frequency goes up things get a bit more touchy. So really the question you need to ask is what cable do I need to do such and such.

[Simon]

I just got these to see if I could output my scope to an audio card.  I wonder if what Simon described is what these are?  They're used for video connections I believe.

Yes that sort of thing but much older and chunkier and it actually has two thumbscrews with a very large wire opening under each screw. Probably something out of the 50s or whenever BNC connectors first came about.

[Retep]

You could get one of these: http://www.banggood.com/BNC-Male-Plug-Q9-to-Dual-Hook-Clip-Test-Probe-Cable-Leads-p-925841.html


Since we are talking about a function generator and breadboard I assume we are not talking about high frequency stuff. In that case the cable is not that critical and you could also make a cable for yourself that you can directly insert into the breadboard , like I did.

[AG6QR]

Any of these solutions can work, but the performance at higher frequencies may be less than stellar, due to impedance mismatches and stray capacitance/inductance.  But a breadboard itself has pretty awful parasitic capacitance and inductance at high frequency.  If you stick to low enough frequencies where the breadboard works well, you can use practically any method to get a signal to the breadboard.

One word of caution: the idea in the original post to use an oscilloscope probe is probably the most trouble-prone.  An oscilloscope probe has a lot of resistance in its center conductor.  This is intentional, in order to dampen ringing that would otherwise happen at the impedance mismatches on the ends of the cable.  The resistance is typically something around 100 ohms or so, which is significant relative to the output impedance of the signal generator (but pretty trivial relative to the megaohm input impedance of the scope for which it was designed).  In short, an oscilloscope probe may get a signal to your breadboard, but it may not be at the magnitude and impedance you're expecting, and this can cause confusion.

One nice solution I've heard of doing is to take a bit of sheet aluminum, roughly as long as your breadboard, and make a 90 degree bend in it.  Mount the bottom of it solidly to the bottom of your breadboard.  You now have one "wall" of a chassis.  You can drill holes and mount any sort of chassis mount connectors, including BNC, 1/8" audio, USB, RS-232, barrel connectors for power, whatever you want.  The wall of the chassis carries the mechanical stress of the cables, preventing accidental tugs from unplugging things and messing up your breadboard.  Solder some short solid-core wires to the connectors and plug them into the breadboard wherever you want.

But most people just use some sort of cable that ends in bare wires, alligator clips, micrograbbers, or something like that.

[Pasky]

Excellent.  Thanks again for the information regarding impedance and capacitance, tips, and the suggestions.  I won't be working with anything high speed, so looks like I'll be playing around with bare wires.

[T3sl4co1l]

I typically use:

BNC cable (random length or as appropriate)
F-F coupling (or F-M-F tee, because I have a bunch laying around; ignore the 90 degree part)
Male BNC to binding posts adapter (Pomona 1296, etc.)
Hookup wires secured in binding posts; insert in breadboard as needed.

This sucks beyond say 20MHz or so, because the binding posts are just as they appear; there's a bit of metal inside, looping from post, back down to the BNC part, for each post.  So the inductive loop is large, or alternately, the transmission line impedance steps up suddenly.  Because those wires are buried inside the plastic molding, you can't effectively null the loop by bringing the hookup wires back along the body; it helps, but doesn't eliminate it.  So it sucks for high sensitivity purposes as well (radiation sensitivity, poor CMRR / shielding).

But you probably aren't breadboarding much beyond 20MHz (< 20ns risetime), so that's fine.  For the circuits where you are, I suggest an "edge launch" BNC, or something like that, using "dead bug" construction.  You can take something like this,
http://www.digikey.com/product-detail/en/5-1634503-1/A97581-ND/1755969
and solder three colinear pins (two opposing ground pins and the center pin) to the PCB surface, after cutting out a trace for the center pin.  The nickel plated body doesn't solder very well, but with some heat and flux, you can solder directly to the body, which allows you to make a bottom side ground contact, in addition to the two ground pins on top.  Example:
http://seventransistorlabs.com/Images/5687_Bottom.jpg
http://seventransistorlabs.com/Images/5687_Glowy.jpg

You could also mount these vertically, if you don't mind drilling some holes and cutting clearance on your deadbug breadboard.

You could also try some of these,
http://www.digikey.com/product-detail/en/1274727-1/A116816-ND/4730610
or http://www.digikey.com/product-detail/en/361V509E/991-1034-ND/2355572
which are properly made for edge mounting on 1/16" (1.6mm) stock.

If you don't mind drilling a much larger hole, you can also mount the panel style connectors right into a piece of copper clad.  Impedance control won't be so good (the pin sticks out the back end too far to transition smoothly into something like microstrip or wire-over-ground-plane), but the shielding will be even better than these examples.

Even if you're not doing high speed right now, the advantages in shielding may come in handy, especially if you get interested in things like millivolt level signals, or radio.

Oh, and remember -- bandwidth is only what you allow.  If you allow unlimited bandwidth... expect all the problems associated with that bandwidth!  A standard 74HC logic gate produces nanosecond range edges.  If you don't want the problems associated with that, be mindful to filter the signal wherever necessary (usually by a series resistor at the driving pin, or an R+C lowpass) so those switching edges don't bite you later!  This matters, even just for -- indeed, especially for, crappy LM358 range op-amps -- interference in the MHz+ range is rectified by the input stage and screws up your signals.  Don't forget to filter (and ESD/surge protect) signals from the outside world; assume everything outside of your circuit is out to get you, at any possible frequency.

Tim

[Howardlong]

FWIW, and while I realise this is probably unnecessary for the OP,  I traditionally used U.FL and W.FL connectors for RF probing points, usually DNP for production units, but increasingly I have started using Murata MM8030-2610RK0, rated to 11GHz, which are nice because they are pennies and have an integrated switch, so by inserting your test lead they can automatically break the path allowing you to hook up VNAs and spectrum analysers in circuit and perform intrusive testing without having to unsolder stuff for isolation purposes.

[Ryano]

I picked up one of these awhile back. Seems to work as well as can be expected.



http://www.ebay.com/itm/BNC-Breadboard-Quick-Connect-Plugin-with-3-6-BNC-Cables-GREAT-PRICE-/160766154416

[dan.soethe]

Most function gens don't go above 1Mhz. So impedance makes little difference.

[Pasky]

I picked up one of these awhile back. Seems to work as well as can be expected.


http://www.ebay.com/itm/BNC-Breadboard-Quick-Connect-Plugin-with-3-6-BNC-Cables-GREAT-PRICE-/160766154416

Yup, saw these!

EDIT:  Apparently his ebay store is cheaper than his website  ,

He has some without the cables included

http://www.ebay.com/itm/EDU-SPECIAL-3-BNC-Connector-PLUG-INTO-YOUR-Breadboad-adapter-Kit-/161101911587?pt=LH_DefaultDomain_0&hash=item25826c1623

Here as well but only 1 and a bit pricey:

http://www.gravitech.us/bnc50ohmbrbo.html