Home lab parts

Regular Al electros are OK up to their rated voltage IME. They die very fast if you go above that, and last somewhat longer if you derate them a bit. Mostly they die from overheating due to excess ripple current.

Cheers

Phil Hobbs

What is blob board? Squares that are pre-tinned?

Then maybe make extras and see how yours sell on EBay. If they become a hot commodity rename your company to "ElectoOptical Innovations and Blob Board LLC" :-)

Yup. You make solder bridges to connect nearby stuff, and use wires for further away. It makes it possible to do good hand-wired protos with TSSOPs and SC70s and DFNs and so on (QFNs are more of a problem). I did my first pHEMT/SiGe:C NPN cascode that way--it worked well up to about 6 mA drain current, and then started oscillating around 12 GHz.

If I do my own, they'll be 0.8 mm thick with a ground plane on the back, with blue solder mask and the EOI logo in ENIG (bling bling).

Cheers

Phil Hobbs

If you're going to do that, include some 4007. (The CMOS chips, not the diodes, though you want those as well.) 4007 is more versatile when you want to do something unusual.

Also, if you are getting 4066, better get some 4016 as well (or instead). If you are using either of those, you must not be very interested in ON resistance, and the 4016 has the advantage of not having the complicated back-gate switching circuit, that can inject charge from the supply rail each time it switches, especially if they did not take adequate care of break-before-make switching of the backgate. Look at the internal circuits in the datasheet, though that is quite likely wrong too.

That sounds nice. It's basically what I use the perf board for, but has a ground plane.

Don't you have issues with copper lifting though? Perf board prevents that.

CH

You need a light touch with the iron, but that's easy with a Metcal.

Cheers

Phil Hobbs

Tants are nice when you want a bit of ESR. Just derate them 3:1 on voltage across a power rail.

Regulators and opamps also tend to be stable if you put a lot of very low ESR polymer cap across their outputs.

It's the Jim Williams approach to control loops. A big enough cap will stabilize anything.

Sometimes they only look stable, because the oscillation amplitude is too small to see easily. A 1-ohm resistor in series with VCC will tell the tale--the supply current goes through the roof.

Cheers

Phil Hobbs

Sure, if you want to do something like this, it's sensible to try it first. I inject a current square wave into the output and look at transient response vs amount of load and big cap value.

I don't know that I entirely trust Spice models here, even if there is one.

RRO amps tend to be happy with low ESR loads because the comp cap has one end on the output pin. Older voltage regs have low loop gain.

Poly alum caps are great, so we need to figure out how to use them.

Here's the VH regulator for my tiny pulse generator, output near zero to 45 volts. C5 is a 56 uF polymer.

Looks like the optimum value for R3 is zero.

Nothing is as bad as white proto boards :) Except the odd variants with peculiar layouts.

Perf board's sole upside is that it's there on the shelf ready to use. For little circuits & doodads that's great. For anything demanding, make a pcb.

NT

com/ They have a retail store in NE Ocala as well. The address in on their website. One kit for everyone would cost more and have a lot of stuff that most people would never use. Maybe Analog, Digital and Power supply compone nt kits would work.

to get a few that I needed. Over 1000 Akro-mills plastic drawers of oddball components ad hardware cover a couple carts and one eight foot workbench t hat could be put to better use.

bsolete part in my face "I usually had something they could use. One guy be t me $20, that there was n way that I had any of those old, stud mounted 8m F, 450 VDC military capacitors. I took his $20 and sold him one cap for $6. I had about 20. ;-)

Wish I still had the very first grab-bag I bought. Lots of 1950s semicons o n long obsolete packages.

Dead bug rules.

Cheers

Phil Hobbs

They made sense decades ago when removing parts for re-use was necessary. Now having to take your products apart makes no sense.

NT

Tants have had too many failures. Al electros fail too but don't normally short or explode. Max 1/3 rated v on tants is the rule now. If healed after construction they can take more, but that's mostly not done.

NT

That depends. For example, an 8-stage crystal filter that cost me severla hundred wasn't left idle in an older box when I needed it in a newer one. Same with pricey sampling diodes and such.

If you feel the urge to experiment with Ge seminconductors you can get Russky versions on EBay:

Amerikanski as well:

Ge diodes like the 1N60 might still be in production. Though one has to be careful, some also come as Schottky with the same part number.

Of course, for the real vintage feel a diode has to be in a black painted glass case:

that I can stick the chip down face up. I sometimes cringe thinking about busted off pins, but in over 40 years of doing this, I've yet to have one b reak.

Since I have something of an addiction, I have piles of Popsicle sticks tha t I use for all sorts of things. You can cut, file or sand them into whatev er shape you need, too.

I did, very early on as a kid. It was an RF transistor, the most expensive component I owned and I had only that one. Almost five bucks or several weeks of allowance back then which really peeved me.

Dentists are probably benefiting the most from the products of popsicle manufacturers. Next up would be insulin manufacturers ...

I've also got a sweet tooth put I try hard to suppress it. One wee piece of not very sweet chocolate after dinner, that's it.