My first LAB but what cable and what not?

[nctnico]

The solderless breadboards come in a wide variety of qualities. I have used them in the past but connecting all the wires becomes tedious quickly and you need to use clean component leads. Nowadays I just make a small PCB which is also easier for SMD components. I rarely use thru hole components.

[tggzzz]

I don't know what to tell you, maybe you're doing it wrong? Maybe you had unrealistic expectations? Maybe you had a cheap or faulty breadboard? Maybe you've never actually used one? I have no idea, all I can say is that millions of people like myself have successfully used solderless breadboards for decades without problems.

Ah yes, the "I haven't had problems so there aren't problems" argument. But many have, and many do get subtle problems.

I suggest you do a few LTspice simulations where you include parasitic inductance and capacitance. I'll do one for a digital circuit.

Consider the ground bounce in a 1" wire (i.e. includes hookup wire and bond leads inside the ICs), the rule of thumb is 1mm of wire has 1nH of inductance, so 25nH. In most solderless breadboard lashups the wires are much longer, so I'm being generous.
Assume 8 digital outputs driving eight 5pF loads (i.e. 40pF in total).
Assume a leisurely 10ns transition time.
Assume each driver has 50ohms output resistance (modern jellybean CMOS is more like 7ohms, so I'm being generous!), so eight in parallel are 6.25ohms.

In the picture below, V1+R1 are the output drivers, L1 is the ground lead inductance, C1 is the load capacitance. Vg represents the ground bounce that is seen somewhere in the circuit (exactly where depends on what you regard as zero volts).

As you can see, even this crude simulation indicates spikes sufficient that the analogue voltage might not be correctly interpreted as a digital signal.

N.B. such things were a problem for some octal buffers, even when correctly mounted on well-designed PCBs!

I invite you to do something similar for analogue circuits, e.g. a simple comparator or amplifier.

[_Wim_]

I'd start with 4mm banana leads (silicone jacket) with stackable connectors and BNC cables. You can buy decent BNC cables with RG316 coax from Ebay very cheaply. Rg316 is thin and flexible so it is easy to work with. I often cut these cables up to connect a circuit to a function generator or scope (without probe). Furthermore some BNC T splitters and BNC to banana (and reverse) will come in handy.

Just buy cables. Making them yourself is a waste of time.

I think the above is very good advice, would have recommended almost the same.

I would also add that for "normal" use (no Ghz, KV,nA,nV...) many of the low cost adapters, cables etc from china are very useable. I personally would recommend to buy some small collections of the various items in the attached picture (this is partly what nctnico, I added some rca to bnc adaptors for some audio stuff, some male bnc-bnc connectors that are very useful to attached to the end of a bnc-bnc cable to attach any of the other adaptors and the bnc to screw terminal adaptors).

As these are so cheap (25€ gets you already a nice collection) it is very handy to have them around, because when you realize you need some during a project, you do not want to wait on ordering these.

[MarkF]

I would just start with a small assortment and see what you tend to use.
Here is a list of cables I have on hand.  You might want some BNC T's and 50 Ohm termination later on.

Test Lead Set & Alligator Clips
TL-21 Mini Grabber for IC Test Leads
Banana to Mingrabber Set
Banana to Alligator Set
Banana to Banana Set
BNC Test Leads Set

I would also get a few Solderless Breadboards.

830 Point Solderless Prototype PCB Board Kit  (a better brand than these)

[tautech]

I would also get a few Solderless Breadboards.

If you restrict bread boarding to just proving a 'building block' then a couple is enough for most projects.
That way daisy chaining all your proven blocks together at the PCB design stage generally works fine.

Here's an example of what can be done with a few:   


https://www.eevblog.com/forum/projects/cloning-a-tandy-trs-80-model-1/msg2101375/#msg2101375

[hwj-d]

I would also get a few Solderless Breadboards.

If you restrict bread boarding to just proving a 'building block' then a couple is enough for most projects.
That way daisy chaining all your proven blocks together at the PCB design stage generally works fine.

Here's an example of what can be done with a few:   


https://www.eevblog.com/forum/projects/cloning-a-tandy-trs-80-model-1/msg2101375/#msg2101375

That's nothing. Look, what you can real do:

[james_s]

It's quite possible to build rather complex circuits that work fine. They don't have to be a complete rat's nest either. Since I'm able to prototype circuits on them without problems, all I can guess regarding those who can't make them work is that they're using inferior breadboards or using them wrong or for the wrong sort of projects.

http://searle.hostei.com/grant/6502/Simple6502.html

http://searle.hostei.com/grant/spaceInvaders/index.html

[FriedMule]

Oh oh and then some ketchup:-)

I really like your advices, I can understand that below 1GHz it is not so important what I get, as long it is not made down in the bicycle garage south of Hong Lo.

I hope that it will be ok if I'll look at your links and then return with a list of "what I thin I need", so you can judge if that is ok?

[FriedMule]

I have found some cable that should be great:-)
The decription is: CBL ASSY BNC PLUG RG142 Cable Assembly Coaxial BNC to BNC Male to Male

Do not know if that is enough for anyone to say if that type of coaxial cable is fine for general usage?

[nctnico]

RG142 is too thick and sturdy which makes it a nuisance to use in test setups. I really recommend to get RG316 cables for general purpose use (<1GHz).

[hwj-d]

It's quite possible to build rather complex circuits that work fine. They don't have to be a complete rat's nest either. Since I'm able to prototype circuits on them without problems, all I can guess regarding those who can't make them work is that they're using inferior breadboards or using them wrong or for the wrong sort of projects.

http://searle.hostei.com/grant/6502/Simple6502.html

http://searle.hostei.com/grant/spaceInvaders/index.html

Yes, ok. For the very first.
But in 2019, for the second, take KiCad, osh a pcb for such things. 

[tggzzz]

It's quite possible to build rather complex circuits that work fine.

Yes, it is possible to cross the Rockies/Alps/Himalalays on a push bike. That doesn't mean it would be a good choice, especially for some that is learning to ride a bike!

They don't have to be a complete rat's nest either. Since I'm able to prototype circuits on them without problems, all I can guess regarding those who can't make them work is that they're using inferior breadboards or using them wrong or for the wrong sort of projects.

And what happens when a beginner's circuit doesn't work as they expect? Is it the breadboard, the technique, the circuit, or the beginner's understanding of the circuit?

Those questions (and problems) can be minimised by using better implementation techniques, especially manhattan and rat's nest.

[FriedMule]

RG142 is too thick and sturdy which makes it a nuisance to use in test setups. I really recommend to get RG316 cables for general purpose use (<1GHz).

I am glad for your help. If we take the cable I did tought as ok, how much thinner is the RG316?

[tggzzz]

I am glad for your help. If we take the cable I did tought as ok, how much thinner is the RG316?

That is easily found here.

What's just as important is the flexibility. You wouldn't want hardline coax

[nctnico]

It's quite possible to build rather complex circuits that work fine.

Yes, it is possible to cross the Rockies/Alps/Himalalays on a push bike. That doesn't mean it would be a good choice, especially for some that is learning to ride a bike!

Still the problems with ground bounce are not a big problem for slower logic. In the past they build large logic circuits using wire-wrap. Not prototypes but production hardware! I recall messing around with a magnetic tape spool machine controller to hack it a little bit.

[tggzzz]

It's quite possible to build rather complex circuits that work fine.

Yes, it is possible to cross the Rockies/Alps/Himalalays on a push bike. That doesn't mean it would be a good choice, especially for some that is learning to ride a bike!

Still the problems with ground bounce are not a big problem for slower logic. In the past they build large logic circuits using wire-wrap. Not prototypes but production hardware! I recall messing around with a magnetic tape spool machine controller to hack it a little bit.

In the mid 80s there were some notorious TTL ics w.r.t. ground bounce - even on PCBs with groundplames. They weren't fast!

The devil is in the details, and solderless breadboards prevent decent construction details.

[tggzzz]

It's quite possible to build rather complex circuits that work fine.

Yes, it is possible to cross the Rockies/Alps/Himalalays on a push bike. That doesn't mean it would be a good choice, especially for some that is learning to ride a bike!

Still the problems with ground bounce are not a big problem for slower logic. In the past they build large logic circuits using wire-wrap. Not prototypes but production hardware! I recall messing around with a magnetic tape spool machine controller to hack it a little bit.

Yes indeed. The computer I used for my final year undergrad project, a Modcomp4, was wirewrapped. And I still have wirewrap kit, and occasionally use it.

But that doesn't validate your point.

You probably don't remember the "surprising" technology people had to resort to, e.g.

• Rogers MicroQ flat capacitors placed under ICs, often retrofitted when problems arose
• IC sockets with capacitors built into them

and I doubt anybody would use those with solderless breadboards!