What do you use - prototyping.

[BenKenobi]

What do people here use for prototyping circuits on a 'temporary' bench lash up.

I have a variety of 'breadboards' some are decent, others are fit only to stop the table rocking or door closing, same goes for the jumper cables. Is their a gold standard in these things - it seems very hit and miss. I've got one really nice one  but the contacts in it seem to be going loose and I can't remember where I got it, looks like this

http://www.ebay.co.uk/itm/5X-2860-tie-point-solderless-breadboard-Including-jumper-wire-S-/172881587757?hash=item28408bae2d:g:QqwAAOSwOudZxmTv

It's really good because I can put voltage into it from different power sources this lets me mix n match my 5V devices and my 3.3V devices. At some point I guess I'll want to be converting prototypes into PCB designs but I'm a bit clueless on that score at the minute.

And what do you use in terms of test points, I'd like to use some sort of pin but have yet to find one that works, I usually use a jumper cable but this can be unreliable and its messy.

I'm looking for ways to work on projects, build them up then put a lid on them and move onto another without breaking them down so that I can come back to them (I tend to break projects down into smaller project sections and then combine), I also want a tidy way to add test points.

[Yansi]

It very depends on the exact job to be done. Some circuits can be breadboarded, some are better to be soldered on the veroboard (typically higher voltages and/or currents). Some circuits can be rapid prototyped using a piece of copper clad board where you mill out some squares for the solder joints. Some circuits cannot be tested without properly etched bords (RF, microwave).

I use all of these above.

[Benta]

I never use solderless breadboard. Far too much effort troubleshooting due to bad contacts, bad routing etc.
I use 90% veroboard or 10% air mockup, always soldered.

[BenKenobi]

I don't do a lot of high frequency stuff but I am experiencing the shortcomings in breadboard use, I find the soldering up of stuff a hassle - and I always muck it up.

I marvel at how tidy some people seem able to work but I have problems transferring schematic to reality so I tend to breadboard a lot.

I have used Vero but I find the copper tracks very limiting and end up cutting tracks and putting jumpers all over the place and it is useless for anything with an IC or two, I've had some that wouldn't take solder readily.

As you say reliable contacts on breadboard are really challenging and utterly useless for anything with digital communication.

Just looking for a reliable cost effective method, I've searched and searched for different methods and come up empty.

[Benta]

I have used Vero but I find the copper tracks very limiting and end up cutting tracks and putting jumpers all over the place and it is useless for anything with an IC or two, I've had some that wouldn't take solder readily.

When I say Veroboard, I don't use the ones with continuous tracks, but rather with individual solder pads or with IC layout and 3-hole strips from each IC pin. Works great for me.

[tggzzz]

If you use modern logic, are you so sure you don't have high frequencies? Calculate the di/dt when switching, and calculate the voltage drop across a wire; assume 1nH/mm. Consider 74lvc, which has sub-nanosecond risetimes, and significant energy above 700MHz and even 1GHz.

See Jim William's AN47 - easily findable via google.

Other search terms: manhattan, pittsburgh, island, dead-bug, live-bug. Any of those is better than solderless breadboards.

[BenKenobi]

Even on a quick scan that 'AN47' looks like a really interesting doc - and I thought my breadboarding was untidy - but I do get the concept of why ( I think ) , I'm starting to think that just buying an oscilloscope (not even got it yet) is going to confirm that my circuits are likely sound but my test reg construction isn't - perhaps the device isn't the defect after all.

I definitely need to work on my prototyping ...

[joeqsmith]

Sometimes its cardboard tubes and Reynolds Wrap.

[bd139]

I've built a frequency counter which was good to 35MHz on a breadboard. It depends what you're doing, how fast it's going, where signals are routed etc. The input stage was not on the board and was dead bug. The first counter and gate was on the board. Signals slightly ugly but it worked.

I actually build quite a few things on solderless boards first to see if they work first before I invest in throwing the parts in the bin if they don't work. Yes they have problems, yes they change circuit characteristics slightly, but at the low end of the frequency spectrum it makes little difference, say under a couple of MHz. Reliability isn't a big problem unless you buy crappy boards or keep them for a long time.

Recently I built a very narrow 600Hz CW audio filter to add to a receiver. The whole thing started off as a total mess on a breadboard:



Evolved into an even large three pole tangled mess. Eventually it pops out the other end looking relatively decent:



The above is actually stripboard, not padboard/matrix board. Plan was to build it with low spec parts and see if I can get it in spec. 603Hz centre frequency, Q of 11 and 1.8dB of gain at the centre = win.

[BenKenobi]

And the prototype became the final ...

So how do you define a good board from a bad one, as I said I've got some that look fine but I only use them to keep the door open ..

Who makes the good stuff in terms of track cutters for Vero - mine seem fine for a couple of holes then I get the Dremel out ... who makes good Vero, good breadboards, good interconnect cables etc.

I even bought the tools to make my own cables, crimps, pins, various headers but for the life of me I can't get the hang of that - I destroy more pins than I make successfully.

[bd139]

Here's my recipe:

Good boards: Wisher, 3M. That's it. All the other vendors are crap. There's a thread here where we go into detail about good ones and crap ones: https://www.eevblog.com/forum/chat/breadboard-maintenance-tips/ - main thing is you have to throw them away sometimes.

Track cutters: If I want a compact layout I'll use an x-acto knife to break the traces by hand. If I am not too bothered I'll use a 5mm hss drill bit with a bit of insulation tape around it. The track cutters you get tend to dig holes too deep when you just want to cut the copper.

Good vero boards: I use Roth Electronik ones from RS. They are FR4 (glass fibre) substrate with tinned copper. Not crappy paper boards that aren't tinned. http://uk.rs-online.com/web/p/stripboards/5185932/

Good jump wires: I make my own. I buy 22 AWG wire from Tayda Electronics here http://www.taydaelectronics.com/hardware/cable-wire/awg-22-red-hook-up-wire-1ft-30cm-solid.html (it's really good!) and some T-strippers (these are expensive but worth it): http://uk.rs-online.com/web/p/cable-strippers/0243112/

I use the Molex connectors here for board interconnects: http://www.taydaelectronics.com/connectors-sockets/wafer-housing-crimp-terminal/serie-2500-2-54mm.html ... crimped with some crappy pliers from Machine Mart. No funky tools required really. They may or may not be genuine Molex but at that price they represent very good value for money.

Note Tayda takes a couple of weeks to turn up but they're generally better value than the local suppliers here in the UK.

[BenKenobi]

All good stuff, it is clear that prototyping isn't necessarily a pretty business ..

[tggzzz]

I'm starting to think that just buying an oscilloscope (not even got it yet) is going to confirm that my circuits are likely sound but my test reg construction isn't - perhaps the device isn't the defect after all.

Be aware that scope probes are part of the circuit; FFI, see https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

AN47 is a classic; you will see repeated references to it. Make sure you read the appendices.

[bd139]

Agree with AN47.

Also this is a gem, even though it’s 1960s vintage: https://youtu.be/ACUbBUydgUQ

[BenKenobi]

Be aware that scope probes are part of the circuit

I know - scarily so, working out just how much they influence it is frankly beyond me right now. The real irony here is that 'studying' what I needed in an Oscilloscope and how to use an Oscilloscope has me reading so much stuff that I'd not even considered - remember I'm a computer science type not an electronics type. Although I already understand many of the concepts as the majority of my work is based in the real time world so being able to measure and detect things without the presence of the tester affecting the outcome of the test is always a challenge - and not only with scope probes.

I find it somewhat intimidating, I knew I had to have good joints and connections, I know a fair bit about EMC and such, but actually seeing the extent to which this can go is insane - some of the probe leads cost as much as the scope !!!!  - I'll never be able to justify that no matter how 'correct' it is, how far is too far for an amateur.

I posted a link to this vid on another thread - just a simple wire NOT 

[Howardlong]

Coincidentally I did a video today about the breadboard and prototyping methods that I use which should be up in the next couple of hours (uploading now *** Sorry, I had to re-upload, something went wrong with the transcoding ***).

https://youtu.be/5Og9MUkIeHk

I use several methods, depending on the circuit. Even if the end result will be on a custom PCB, I do a lot of unit testing using the various techniques before designing and committing a complete design to a board house.

Solderless breadboards, certainly, I use them all the time, together with breakouts for SMDs. I agree with others, there are some crap ones out there. Somewhat counter-intuitively, I get mine from Maplin here in the UK, they are pretty reasonable compared to some of the cheapo crap on ebay and Amazon. I do buy the link wires from Amazon in bulk, they're not the best but they're good enough usually.

For most breakout boards for standard SMDs like SOIC/TSSOP/MSOP/QFN/QFP/SOT23 etc I go to ebay and buy in bulk and keep a reasonable stock as not a day passes without me using one or two.

For some more SMD stuff like sensitive analogue and mixed signal, I use Schmartboards which have an integrated ground plane mitigating many effects that breadboarding suffers.

Some breakouts for BGAs and slightly more exotic packages are available from Proto-Advantage, but they don't have the ground plane that Schmartboards do.

For basic soldering, I've used veroboard, particularly Tripad, for decades and continue to do so, and I mount a hybrid of both through hole and discrete SMD parts on them.

More recently after a post a few months ago here, I have started using BusBoard SMTpads surface mount boards with solid copper groundplanes and holes for optional ground vias. These are free-form like veroboard and better support things like switch mode power supplies and some analog and mixed signal stuff thanks to their solid ground planes. They're available in both 0.1" pads and 0.05" pads.

If all else fails, I make up a board and fab it myself in-house. Making the board itself is quick once you have figured out your process (figuring out the process takes a lot of failure though!), half an hour from opening the cupboard with the nasty chemicals to having a board and tidied up. For anything other than a really basic board, laying out the board typically takes quite a bit longer, as I capture the schematic at the same time (most of my designs live through their prototyping life on paper).

If I do make up the board myself, I can go down to 0.5mm pitch reliably, but I am unable to drill vias under 0.7mm without breaking drill bits, that may be my crappy drill press though, but it was advertised as being specifically for PCB drilling. Thankfully I do little through-hole on my own board layouts and usually maintain a solid ground plane on the underside so drilling is typically limited to ground vias.

Some links:
http://www.busboard.com/surfacemountpcbs
http://www.proto-advantage.com/store/index.php?cPath=4000
http://schmartboard.com/surface-mount-boards/

[BenKenobi]

I was wondering about how to handle surface mount packages. Wouldn't it be so cool if you could just print your PCB on something akin to a 3D printer - gee should I patent that idea ...

[Howardlong]

I was wondering about how to handle surface mount packages. Wouldn't it be so cool if you could just print your PCB on something akin to a 3D printer - gee should I patent that idea ...

Breakout boards.

Someone did a PCB printer, or at least talked about it. Not sure it this was the one. $3.5k, eeeek.

https://www.voltera.io/

[joeqsmith]

I was wondering about how to handle surface mount packages. Wouldn't it be so cool if you could just print your PCB on something akin to a 3D printer - gee should I patent that idea ...

I've made generic PCBs for prototyping surface mount ICs.  The problem is I never get them generic enough to be really useful. 

[BenKenobi]

Is it that the PCB's aren't generic enough or that packages aren't made with any 'consistency', as for the printer I don't think at 3.5k I'll be buying one any time soon.

[kalel]

Conductive ink printer could be an interesting thing at a low price (if the technology is useful, prices might come down eventually).

Still, it mentions:

~5 USD
Board cost.

I'm not sure how this compares with the cheapest Chinese fab house (for those that can wait up to a few weeks), but I understand that even more expensive - but rapid - could be worth a lot for some people in some cases. I don't even know much much the current home etching methods cost per board.

[tggzzz]

- some of the probe leads cost as much as the scope !!!!

A quick scan of a mass market distie shows £5230+VAT, and that's far from the most expensive! http://uk.farnell.com/tektronix/p6248/probe-oscilloscope/dp/4352610

Concentrate on bog-standard 10* "high" impedance probes (preferably with a short ground connection), *1 probes for low-level signals, *10 resistive divider "low" impedance Z0 probes, ignore *1/*10 switchable (because you will accidentally move the switch, sooner rather than later). For high voltages (i.e. >50V) get a decent HV differential probe. And watch the max voltage vs frequency rating very carefully.

[BenKenobi]

  £5230+VAT 

Where do you keep them in a safe - I'm forever losing stuff like this without even leaving my man cave ... now where did I put that wire coat hanger - that'll do the job ...

[Avacee]

I was wondering about how to handle surface mount packages.

I find these SMD to DIP adapters are a godsend for prototyping onto either vero/strip/bread/etc board.
http://www.ebay.co.uk/itm/20PCS-SOP14-SSOP14-TSSOP14-to-DIP-PCB-SMD-DIP-Adapter-plate-Pitch-0-65-1-27mm-AS-/291979406901

There are adapters for each of the pin counts - this link shows the 14 pin version.
One side is 1.27mm and the other is 0.65mm and if you don't suck at de-soldering you can reuse your expensive IC on your final PCB.

[tggzzz]

£5230+VAT 

Where do you keep them in a safe - I'm forever losing stuff like this without even leaving my man cave ... now where did I put that wire coat hanger - that'll do the job ...

Pah. That's low cost.

Consider "P7700 series TriMode probes are available for order now worldwide. Probe prices start at $8,500.  A set of 5 TekFlex solder tips is available for $250". Why do they come in 5-packs? ISTR each tip has a guaranteed lifetime of 10 touches. Yup $5 every time you touch a node.
http://news.tektronix.com/2016-03-15-New-Performance-Oscilloscope-Probes-Tackle-Todays-Smaller-Faster-Designs

[Howardlong]

Conductive ink printer could be an interesting thing at a low price (if the technology is useful, prices might come down eventually).

Still, it mentions:

~5 USD
Board cost.

I'm not sure how this compares with the cheapest Chinese fab house (for those that can wait up to a few weeks), but I understand that even more expensive - but rapid - could be worth a lot for some people in some cases. I don't even know much much the current home etching methods cost per board.

I am not sure how the process is better than "home etching" from an ease-of-use or productivity perspective.

Home etching costs peanuts, but you do have to make the effort to get your process to work, and expect plenty of failures to get to that process, just make sure you document what works, and possibly more importantly what doesn't. Keeping to the same recipe is an extraordinarily good idea. Using different stock, in particular changing you pre-sensitised boards, or chemical mix/dilution, is not a good idea. I do my own etching sporadically, I can go weeks without needing to do a quick board then suddenly I might do a dozen in a couple of days, but it took quite some weeks if not months to get my process to work.

For the avoidance of doubt, my process very closely follows Mike's (http://www.electricstuff.co.uk/pcbs.html), but I "discovered" it without prior knowledge, I just did self-learning. What is nonsense and a waste of money for kitchen table etching are bubble tanks. Use your eyes and learn, bubble tanks will just get in the way. Plastic take-away containers are designed for the purpose ;-)

As a tool, what is very useful for any prototyping is a decent PCB guillotine, and while they're not cheap, my they're handy, way more useful than any bubble tank.

[BenKenobi]

Used to do a similar thing when I had a darkroom - digital is so much less hassle.

[cdev]

I'm using first a breadboard, then a piece of PCB inside a cardboard box to prototype.  I can then write all the pertinent notes on the outside or inside. I dont route the wires in from the top, I cut holes for them and use dip sockets and plugs which I already have.

This is new but for a long time Ive used home-made cardboard boxes to enclose projects. With RF sometimes you have to have a snug case, this is easier to get right with copper coated cardboard or foam core than anything else.

I've often tried to use metal cases but just couldn't do what I wanted to do so ended up back with cardboard and copper and foam core..

I've used foam core for antennas too.

[kalel]

I'm using first a breadboard, then a piece of PCB inside a cardboard box to prototype.  I can then write all the pertinent notes on the outside or inside. I dont route the wires in from the top, I cut holes for them and use dip sockets and plugs which I already have.

This is new but for a long time Ive used home-made cardboard boxes to enclose projects. With RF sometimes you have to have a snug case, this is easier to get right with copper coated cardboard or foam core than anything else.

I've often tried to use metal cases but just couldn't do what I wanted to do so ended up back with cardboard and copper and foam core..

I've used foam core for antennas too.

Cardboard boxes sounds great for prototyping. Have you posted an image of such a project?

[ChrisLX200]

Copper clad board and Veroboard for me. I have a few breadboards but don't get on with them too well - always seem to get bad connections, perhaps I should invest in better quality ones :-)  I use a 3D printer for most temporary boxes because it's so easy to design and print something that fits near exact.

ChrisH

[Red Squirrel]

Mostly bread boards for small stuff and crocodile clips for bigger stuff.  For more permanent stuff I'll use perf board.  I have not gotten into any complex projects or SMD stuff yet, but once I do I will probably build my own break out boards to combine certain things together into some break out boards.   

[John B]

I don't do a lot of high frequency stuff but I am experiencing the shortcomings in breadboard use, I find the soldering up of stuff a hassle - and I always muck it up.

I marvel at how tidy some people seem able to work but I have problems transferring schematic to reality so I tend to breadboard a lot.

I have used Vero but I find the copper tracks very limiting and end up cutting tracks and putting jumpers all over the place and it is useless for anything with an IC or two, I've had some that wouldn't take solder readily.

I use Express Schematic and Express PCB to help me lay out prototypes on 0.1 inch through hole board with individual plated holes, not strip board. Once you reach a certain level of complexity and number of components, I can't imagine trying to lay it out as you solder.

The up-down left-right square grid does eventually become limiting for space, density and routing options, but it should suffice for many projects which can be put into use as is.

[Howardlong]

Used to do a similar thing when I had a darkroom - digital is so much less hassle.

You don’t need a darkroom. Use decent pre-sensitised boards and peel off the protective backing just before you place them in the UV box. As long as you’re not leaving the boards out a long time unprotected or outside in direct sunlight, normal office or domestic lighting isn’t a problem.

While I understand the retiscence around the chemicals, once you have your process it’s quick and reasonably simple. Just don’t let the ferric chloride near stainless steel kitchen sink, or conduct operations in your formal wear, or you’ll have some explaining to do.

[bd139]

Good advice. My wife hasn’t forgiven me for wrecking the nice porcelain kitchen sink with ferric chloride yet. 

[tggzzz]

Good advice. My wife hasn’t forgiven me for wrecking the nice porcelain kitchen sink with ferric chloride yet. 

Old timers remember why most people swapped to stainless steel sinks as fast as possible. And it wasn't just because of our type of activity

[bd139]

Yes she dropped a glass in it a week after she mandated that it was installed and they installed it. Big crack. Looks like shit now.

Plastic trays for this stuff. I’m a fan of the Poundland Tupperware boxes. They also won’t melt if you’re doing an IPA wash.

[woody]

Using a breadboard for the first tinkering, then a cnc mill to make a working prototype. The advantage is speed; the disadvantage is that you often have to redesign the board to make the final pcb. This due to the fact that thru-holing, vias and very fine tracks are unfeasable. But for simple one-offs a mill is quite usable.

[tggzzz]

Plastic trays for this stuff. I’m a fan of the Poundland Tupperware boxes. They also won’t melt if you’re doing an IPA wash.

Oh yes indeed. They come in so many sizes that they can be used for just about anything.

But I do find the cheap ones are more brittle than desirable.

[BenKenobi]

What is kind of clear here is that there really isn't what I'd describe as an 'elegant' solution for this, that the more complex 'packaged IC' components get, the higher the frequencies get that the prototyping methods become really challenging.

I'm not dead against chemicals but I can be a real clumsy doofus with such stuff, I know the IPA story well having had the bottom drop out of a container whilst washing something - on the dining room table.

I like the idea of breakout boards for the SMD packages, I think I'll make sure to add an extra package for the making of these as I go.

I think some of the techniques used by you guys are a little tricky for one like me who lacks all the experience in build and layout strategies that you probably take for granted, I really need to force myself to learn some PCB layout stuff but I'm just not the artistic type so I struggle a lot with it.

Please keep the ideas coming, perhaps a eureka moment will occur ...

[tggzzz]

What is kind of clear here is that there really isn't what I'd describe as an 'elegant' solution for this, that the more complex 'packaged IC' components get, the higher the frequencies get that the prototyping methods become really challenging.

For theory, concentrate on the physics of the inductance of a wire (start with 1nH/mm) including the wires "inside" SMD capacitors (>1mm long!), the current needed to charge a load capacitance during the switching/transition time, and hence the voltage induced across a wireinductor. Then you will realise why short connections are essential!

For practical construction of medium frequencies (up to UHF), see how radio hams do it. Note particularly the use of ground planes.

I like the idea of breakout boards for the SMD packages, I think I'll make sure to add an extra package for the making of these as I go.

I think some of the techniques used by you guys are a little tricky for one like me who lacks all the experience in build and layout strategies that you probably take for granted, I really need to force myself to learn some PCB layout stuff but I'm just not the artistic type so I struggle a lot with it.

Bugger artistry; physics is all that counts in this context! Short leads. Ground planes.

[bd139]

Nice PDF on it here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT1.pdf and here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT2.pdf

Quick excerpt:

[BenKenobi]

Happy days - that put a light on - that pad idea looks really usable .. off to buy some stuff - I'll experiment with that method for sure. You could even layer that with standoffs.

Rqd 1 Nibbler, 1 PCB shear, materials, new soldering station ... update - ouch - how much is a PCB shear  anybody got a secondhand one taking up too much space ..

[A Hellene]

I usually am too lazy not to be using solderless breadboards and stripboards.

But, where signal routing matters (because there is hundreds of amps of current flowing or there are nasty dU/dt and/or dI/dt signal change rates --especially when dt is of the order of tens of nanoseconds) I do not hesitate to make prototyping PCBs --even at home, by messing with fine transparency printed films, Positive photoresist sprays and FeCl3 on FR4 copper clads.

For example, this is my latest design of a voltage-doubler lead-acid batteries desulfator:


PCB layout.jpg


Runtime.jpg


About.jpg

And, since I am not satisfied with the real life performance of the voltage-doubler prototype, I'm now back to the drawing board for a voltage-tripler version of it, or even for one with an uCU controlled >50V voltage-booster on board...


-George

[bd139]

Happy days - that put a light on - that pad idea looks really usable .. off to buy some stuff - I'll experiment with that method for sure. You could even layer that with standoffs.

Rqd 1 Nibbler, 1 PCB shear, materials, new soldering station ... update - ouch - how much is a PCB shear  anybody got a secondhand one taking up too much space ..

I use a Stanley knife to cut the main board. Score both sides for a bit then it’ll snap cleanly. The 0.8mm strips you can cut with a pair of aviation shears pretty easily. I usually cut them into strips then skip off bits with aviation shears as and when I need them.

[tggzzz]

Nice PDF on it here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT1.pdf and here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT2.pdf
Quick excerpt:
(omitted for brevity)

I don't like the "gnd bus" in that particular example: it is only attached (inductively) to the plane at a single point. Why not use the ground plane directly, and completely omit the gnd bus? Fig 13 looks much cleaner in that respect.

Nonetheless, those articles are informative and a good starting point.

[bd139]

Mainly because you have to tombstone the SMD parts to get them on the ground plane and conductors.

[tggzzz]

Mainly because you have to tombstone the SMD parts to get them on the ground plane and conductors.

Prettier or electrically better: your choice. Fig 11 would be better if the gnd bus was soldered to the plane along the bus' length, not at a single point.

But there are alternatives: double sided PCB with one side a plane, and the other side "milled with a dremel".

[BenKenobi]

What about matrix board ?

I didn't pick that up on the circuit shown re the ground but yes since it is on a copper sheet aka ground plane why not go direct - I thought this was the idea.

On the desulphation project take care - excessive constant voltages on 'gel' type or glass mat batteries will not have the effect you expect, the battery will be effectively trashed if you're not careful. Such methods only work on 'liquid' lead acids because the liquid can carry stuff away from the plate more readily. For Gels and AGM's you need to consider pulsing, raise the voltage yes but do not apply it constantly to such a battery or the gas builds faster than it can dissipate / the cells overheat - the result is never good.

[A Hellene]

[...]
On the desulphation project take care - excessive constant voltages on 'gel' type or glass mat batteries will not have the effect you expect, the battery will be effectively trashed if you're not careful. Such methods only work on 'liquid' lead acids because the liquid can carry stuff away from the plate more readily. For Gels and AGM's you need to consider pulsing, raise the voltage yes but do not apply it constantly to such a battery or the gas builds faster than it can dissipate / the cells overheat - the result is never good.

That's correct!

On the flooded and glass-mat type lead-acid batteries you can do whatever you want to do, if and only if you are also monitoring their temperature changing rate. On the other hand, on the gel types you must be very careful with each pulse energy amount because you do not want to create (permanent) gas bubbles into the gel electrolyte polymer films that exist between their electrode plates. Thus the instantaneous (<0.5-10us max.) strikes of ~500A intended pulses, strictly depended on the procedure statistics by the uCU that creates them (meaning the length and the frequency of the pulses).


-George

[BenKenobi]

Well I've just added another project to my winter list ...

A 3 Axis DIY CNC Router / Mill / Engraving machine - can cut out tracks and such from copper coated boards.



Several interesting projects in one - building, coding, PCB design / production, and my favourite puzzle - debugging ....

[ChrisLX200]

Well I've just added another project to my winter list ...

A 3 Axis DIY CNC Router / Mill / Engraving machine - can cut out tracks and such from copper coated boards.

Several interesting projects in one - building, coding, PCB design / production, and my favourite puzzle - debugging ....

Yes, thinking about that one myself! Could be a very useful tool for smaller PCBs but I'm not confident of attempting to tackle anything above about 6" square. Needs a self-levelling system to work well.

ChrisH

[Red Squirrel]

Wow that little machine does look pretty cool.  Could be used to etch stuff like front panel plates for projects and stuff too.   I kinda like the idea of making PCBs at home but don't really want to invest in a chemical based setup, it looks quite involved if all I want to do is a few one ofs.

[BenKenobi]

Well that one in the video is pretty expensive but there is plenty of DIY stuff around for a lot less, worst case you can build up from your own steppers etc, the research adds to the fun much like 3D printing.

I don't see that it can be so different from a 3D printer, tons of stuff out for those in the open source domain. Could either buy a dedicated kit of parts or I've found some plans from a Swedish firm. The setting of levels, origin and boundaries are standard CNC procedures, just need to decide the detection method, but it looks interesting.

Thinking about this you could make one with an interchangeable head and have it be a printer too - so you can print your enclosures, engrave your labels, build your PCB's, print your knobs ..... the plot thickens.

[Red Squirrel]

Yeah could make a fun project. I probably would try for the DIY route myself.  The issue with premade stuff is that it's rarely compatible with Linux without jumping through so many hoops, so if I do it myself I can make sure it is.

[bd139]

Looks like too much work for me.

[Yansi]

Nice PDF on it here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT1.pdf and here: http://www.aoc.nrao.edu/~pharden/hobby/HG-MANHAT2.pdf

Quick excerpt:

I have some strong doubts this will work properly. Maybe at 35MHz, not 350. 
I think the grounding is not good enough in there and the trace impedances will be one big mess. Not even close to 50ohm.

[bd139]

I've built similar. Certainly good to 145MHz.  Haven't had to go higher yet myself but I have seen this approach working well even up to 1GHz or so. Don't expect a repeatable build though.

[tggzzz]

I've built similar. Certainly good to 145MHz.  Haven't had to go higher yet myself but I have seen this approach working well even up to 1GHz or so. Don't expect a repeatable build though.

Yes, but the ERA-1 in that cuircuit is DC-8GHz, and as we all know "amplifiers oscillate, oscillators don't".

[bd139]

That’s why you buy a 100Mhz scope - don’t have to worry about those oscillations if you can’t see them then 

Edit: actually you can with an ammeter

[chris_leyson]

I don't know it might work at 350MHz doesn't look too bad, I would have used an extra ground pin though.
Below is a picture of Manhattan style construction, it's the main 320-650 MHz VCO in an Adret 7100 signal generator, Circa 1980. The transistors with the red paint blobs are BFR91 and the metal can transistors are BFQ22. Would be nice to have a go at replicating this with SMT devices, nice generators BTW with low phase noise.

[bd139]

Similar construction to Marconi 2019 of the same era. That went to over a GHz.