Parts Bin

I rarely resuse parts. And I keep breadboards, in case they might be useful again some day.

ftp://jjlarkin.lmi.net/BreadBoards.jpg

John

"John Larkin" a écrit dans le message de news: snipped-for-privacy@4ax.com...

Have just received the protos of my 8mm dia sensor all stuffed with 0201 caps, diodes and resistors. Nothing marked at all...

Well, I like them (breadbords) for some quick tests but you have to be careful. The last thing I wanted to test was a proof of concept for a modified twin T rej filter + buffer, to measure distorsion on a 50kHz ppm distorsion level, medium power amp. The breadboarded circuit worked so well (~0.1ppm HD) that I didn't bother to go to PCB and all the first run was tested with that.

The next one will be some BP filter/preamp combo ideas for my IR sensor. Got some real neat ideas to test for 4th order tunable, high Q, BPF with 200M transresistance "gain", all that for under 20cents. (hey Joerg, beat that :-)

For such "simple" things, those breadbords allow to shoot first time for the good PCB... Worth it IMHO. Oh, and they are so bad that they really teach you what's important in a layout (or it won't work :-)

OK, they don't do ps though :-)

-- Thanks, Fred.

I find that dead bug is at least as fast as white protoboards, because I don't have to futz with as many wires. (It's dramatically faster than perf board, even the fancy Vector 8007 ones.) John L. has some irrational attachment to keeping the chips right-way-up, but I think that since God gives me 14 standoffs for free in each DIP package, it's theologically questionable not to use them that way. ;)

My dead bug protos also have low stray capacitance, so they work fine up to about 1 GHz. One analogue photoreceiver I built that way had about

190 ps edges, measured by shoving the TDR pulse from a Tek SD-24 sampling head into a diode laser biased right at threshold.

Cheers

Phil Hobbs

(Owner of one nearly pristine white protoboard and a few boxes full of gnarly dead bug protos)

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

I solder as many pins as possible to ground, unbent. Live bug is nice that way.

John

office

the

back

Better hf behavior if soldered upside-down.

ftp://jjlarkin.lmi.net/0805_res_fix.JPG

ftp://jjlarkin.lmi.net/0805_normal.JPG

ftp://jjlarkin.lmi.net/0805_flipped.JPG

John

DIP???? I can't remember the last time I saw one of them.

You need to get out more. ;)

If I were doing FPGAs, I wouldn't need as many protos either. I do have one box that I built 20 years ago and still occasionally use. It has a DIP socket and a bunch of nice metal pins--it's for mocking up sequencing using 22V10s. There are lots of places were micros aren't the right medicine, especially ones needing really good timing accuracy.

Being able to iterate fast is important when you're doing stuff you don't fully understand yet. Being a researchy sort, that's where I usually work.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

I still remember this very theoretically oriented German classmate who would start to analyze poles and zeros when it was clearly a loose connection in the plastic breadboard.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

"Spehro Pefhany" a écrit dans le message de news: snipped-for-privacy@4ax.com...

Maybe all he wanted to do was pluging a pole in the loose hole...

Capacitance between two five contact strips is 2.5pF IRRC. Not a big deal for a lot of stuff. I once had to reduce that, using one shielding guard strip between two others, which gets you pretty low parasitics (don't remember the figure).

-- Thanks, Fred.

office

the

back

Why? Is the resistive material on the top?

...or you do. ;-)

While I was hired to do FPGAs, I've done a lot more analog stuff than digital over the last couple of years. Through-hole is a no-no (except connectors).

Which is why Larkin uses live-bug prototypes. We have a prototype machine but it's too much trouble to use. I either plan a second pass or have prototype boards made. It's cheap.

I gave up on those plastic breadboards in college when my 709s kept letting their magic smoke out. I didn't use one again until last week. Then I remembered why I gave up on them. ;-)

Nah, I use SMT for actual boards, and also frequently use them in dead bug work. SOIC and SOT packages aren't difficult to use with leaded components, as long as you're careful. You can stagger SOIC leads like saw teeth, for instance. I like DIPs because they're quick to work with--the pins are almost bulletproof. 0805s are pretty easy to work with in protos as well, although they sometimes crack.

Sure, but much much slower than dead bug. At least an order of magnitude for me. Also the protos are very rarely complete instruments, so there'd be no gain in making custom boards. Frequently there are several completely different methods for doing some job, and it isn't clear which one is best in real life.

A recent example is an AC tweaking network for a tarted-up laser noise canceller that I'm working on.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

If I need to prototype something that gets into "dead-bug vs live-bug", I often like to get a couple boards from expresspcb.com - should the IC(s) be "live-buggable" or home-solderable (including SOIC and the similar SOP as far as I am concerned).

All-too-often, I run into layout sensitivities due to switching frequencies in recent years getting to 100-500 KHz range even for voltage regulators, LED drivers and battery chargers!

I remember even, soldering fine copper wires afterwards-clipped-short onto ICs for "live bug" prototyping and even repairs to prototypes, some of which were built by others on basis of my design!

--
 - Don Klipstein (don@misty.com)

The most recent time I found a significant performance difference between "breadboard" and an etched PCB, the difference was in resistance or inductance, not capacitance.

Keep in mind if "relevant impedance" is greatly outside the ~50-300 ohms range, things can get tricky at higher frequencies. If you need to get trickier due to high frequency, even if frequency is still so low that you don't have to bother with a "Smith chart" or "stripline" or the like (typically 10's of MHz or higher), please get some idea what the relevant inductances and capacitances are.

If "relevant impedance" is greatly below 50 ohms, then inductances of so much as wires or PCB traces can easily be significant even with length of conductor a few percent of a "wavelength" or less.

If "relevant impedance" is much over 600 or maybe even much over

300 or "characteristic impedance" ohms, then capacitance matters a lot at high frequencies.

--
 - Don Klipstein (don@mi9sty.com)

piece

;-)

office

the

of

back

plastic

electronics

as

Yes. The electricity is happier if it can stay in the plane of the traces, and not have to climb up and down the end caps. You can buy resistors without the end cap plating; they are really good, mounted resistive-element-down.

John

We had one of those pcb milling machines. Agree, more trouble than it's worth.

John

We don't stock two sets of components. What I prototype with is almost always something we're using for another product or at least will be used. One exception is LEDs. I've learned that it's important to see them before buying. ...particularly green ones.

Sure. Prototype both on the same board. If it's simple I'll do it with Larkin's method. If not, I'll have a prototype done. If the boss is in a cheap mood I'll risk a another pass; his choice.

I'd actually use the prototype machine more but the layout guy hates it for some reason. He probably thinks he's going to be put in charge of running it. He's probably right. ;-)

None of my stuff is that sexy. ;-)

I was shown a product we make with the stupid thing. I doubt they've sold one since I started there, but still! Simple PCBs are *cheap*.