In the old days with through-hole ic's, you had no choice but to learn how to count pins backwards. But I definitely agree - live is better than dead. One alternative is those litle smd adapters for ic's.
Another is to glue ic's to a small piece of pcb as a standoff and straighten the pins so they don't touch the groundplane. Then solder components and wires directly to the pins. This may be a better choice for higher frequencies.
Earlier you said you scoped it and the input to the LS163 looked fine. What were you looking at?
-- Rick C
Oops, wrong board.
Some of the pins of a DIP, GND and often some others, can be soldered to the plane; bend the others horizontal and wire them up. But who uses DIPs any more?
Tubes counted pins clockwise as seen from below. Early ICs were packaged in round TO cans and followed that convention, which is why pins are now CCW as seen from the top.
The dumb thing they did, to DIPs and to SO parts, was to make the pattern symmetric, so you can put them in two or four different ways.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Although the plastic plug-in proto boards are indeed not very good for fast logic they are actually good enough for some applications. One such plastic board was used to prototype my Altoids jelly bean mixer:
After the circuit was debugged the components were taken out of the proto board and popped into a strip board. The mixer works perfectly. Problem solved. No hum whatsoever.
Thank you,
-- Don Kuenz, KB7RPU
.
4000 series CMOS particularly at 6 Volts, is mud slow. The rise/fall times may be too slow for a TTL chip. A Schmitt trigger device between the CMOS a nd TTL will likely fix the problem provided the Schmitt unit is TTL.
Those are the adapters I was talking about. The big problem is the log leads that give trouble at higher frequencies.
You can do the same with op amps, ecl, other fast logic by mounting them on a small pedestal. This makes for much shorter leads, better suited to high frequencies.
Very good explanation.
It's much more likely to be the pull-down that's not strong enough. TTL needs a much stronger pull-down than pull-up, which is the reason for HCT in the first place. HC is symmetrical, HCT has a stronger pull-down (and maybe a lower logic-zero level).
10k into 10pF is -3dB at 1.6MHz - and you can get half that capacitance from a single DIP pin, with no trace connected. To pull an LS TTL input down to 0.4V, it needs to be much stronger than -3dB. If you have to do this, use a 470R pull-up not 10K.Open and run your mouse pointer along the 10k-ohm line (increase max R) to read off capacitance vs frequency (this will be the -3dB point).
Clifford Heath.
alternative
frequencies.
Nah, dead bug rules. If you do it right, it's far faster and easier than pe rf board. Back when chips were labelled with actual paint, maybe live-bug h ad some points in its favour, but it has a critical disadvantage: it weaken s the pins badly.
With dead bug you get 8 or 14 or 16 nice strong standoffs per chip, which m akes it a breeze to bend passives into a nice tight 3D layout. Solder one e nd, crank it around to where you want it, and solder the other end. (Pro t ip: reflow the first joint when you're done, to relieve the spring tension. )
I don't use DIPs in products, but I stock a lot of them for protos. I do a lot of proof-of-concept things, for which dead bug is easily 10 times faste r than doing a PCB. The stray capacitance is comparable to a 4-layer SMT bo ard--the larger components are compensated by air's lower dielectric const ant and the 3D layout.
Cheers
Phil "the only good bug is a dead bug" Hobbs
Who is talking about perf board? I'm talking copperclad.
Mounting the ic right side up has no effect on the strength of the leads.
Mount the ic on a small pcb standoff and glue it to the copperclad. The leads are straightened to keep them from touching the copperclad.
For dip ic's, straighten the leads and cut off the thin section that normally goes through the hole. This gives the wide stub as an excellent anchor for component and wire connections.
Don't stress the component or ic lead by cranking it around. Mount it to bridge the connection. Solder one end, then the other, then retouch the first connection to relieve any stress.
Use Manhattan style standoffs to support the components. The standoffs can be purchased online, or simply made from a pcb hole punch. A 3/16" or 1/4" diameter standoff will have about 1/2pF capacitance to the copperclad ground. You can glue them close together to connect SMD caps and resistors so you don't need o use wires for connections.
Live bug gives the same speed advantage, with faster lead recognition and fewers misconnects during build and operation. You can use the regular SMD ic's that you use in production.
For ic's with tiny lead spacing, take one of the headers shown in JL's proto board. Use a sharp tin snip to cut off most of the adapter except about 1/8" for the leads. Solder the ic to the adapter, then glue the adapter to the copperclad.You now have a much stronger support to mount components and wires, and the ic is right-side up.
The increase in stray capacitance is negligible.
Especially the original CD4000 series. The HEF4000 series (Philips, NXP, Nexperia) is better.
If you want to get at the parts, you have to flip the board over, and first disconnect all the cables and stuff. It's hard to probe or IR image the bottom of a powered-up board.
I'm not afraid of bugs in plain sight. If you solder them down, they can't crawl away.
Fast things:
Complicated thing, didn't work:
High voltage:
-- John Larkin Highland Technology, Inc lunatic fringe electronics
It was quicker than laying out a board, making a BOM, ordering stuff, waiting, assembling. This was a demo for a customer who appreciated getting it tomorrow.
I do slow stuff too. Synchros, thermocouples, DACs, relays even.
But SF? Most people here just type Java or SQL or something. There's not much real (by my standards) electronics being designed here.
I can Dremel and solder pretty fast, and it's fun. Most of my parts are surface-mount, so I couldn't use one of those plastic gadgets anyhow. And they confuse me.
The nice thing about an FR4 proto is that it's rugged, and you can write on it, document it, and keep it. All of our protos, PCB or hand-built, and all of our experiments, get permanently documented for future reference.
I can't say who is happier. Life's been good to me so far. I'm way ahead of my PCB layout and firmware guys, so I have time to branch out, experiment, learn things that have no immediate use. Those things often turn out to be good to know.
I'm not sneering at anyone. I do slow stuff too, switchers and opamps and mosfets and picoamps. I just don't like those plastic proto things.
I discovered a while back, at some discomfort, that nylon is leaky as hell. I was trying to measure femtoamps and it wouldn't let me.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
But nobody here is using them!!!!!!1111
HCT's only difference is the input threshold (the gate thresholds are tweaked with doping).
HC was designed to be output-compatible with TTL. TTL's pull-up is a bit weak, and saturates around 3.5V, whereas HC has a strong pull-up to full rail voltage. Both perfectly acceptable for driving TTL inputs. :)
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
Sounded like Neon admitted to it, above.
The speed advantage of dead/live bug consists partly in the ease of attaching the chips to the Cu clad board with Krazy Glue or the equivalent.
It sure does if you glue the chips to the board--you have to bend them at least 90 degrees and probably more.
You seem to be describing a far more elaborate method using spacers and stuff--have you actually built something that way? I find it hard to imagine doing anything at all complicated that way.
Way too slow. Mounting the IC upside down takes about three seconds, and the leads automatically don't short to anything. Maybe 12 seconds all told if you use dikes to gouge the package at the pin 1 end, as I usually do, and write the part number on the board with Sharpie.
Unnecessary.
You obviously don't try to achieve a tight layout, then. I don't think I've ever broken a TH resistor or capacitor by bending it too hard. They're super strong. You can cook the leads off a capacitor if you're not careful, but they don't break.
Way too complicated. You get 14 nice strong standoffs per chip, free, in three seconds, if you use dead bug.
Not so much. SOIC leads are way too wimpy for the job. I use breakout boards for parts that don't come in DIP or TO92, but it's a huge pain by comparison--at least 20 times slower.
I invite you to try it the other way. Your purported method is at least three times slower than mine on a good day.
Depends on the application. Last year I did a super cheap 100-ps class TDR with ChesterW, using 0603 and 0402 class SMT parts in mid-air. It took a few times longer to get right than DIPs, but it worked great--the PCB came out just the same.
Cheers
Phil "Dead bug rules" Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
I use white proto-board too - it is quick and easy for non-critical, slow, non precision stuff. Encourages experimentation too. As long as you keep in mind its limitations it saves a lot of time.
piglet
That guy swiped one of my breadboard pics! Not even my best work.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
That bling board is definitely bad for the eyes. ;)
Where's the laser?
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
The gold can be blinding. And it's hard to photograph. That was an attempt at a hysteretic-switching wideband 5-amp programmable current source. It wasn't stable enough vs low enough ripple to be useful.
It's driving some resistors and diodes in the middle as dummy loads.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.