Yes, that was plan A. But again, i only need relative stability within the group. Nothing more simple just than three zeners.
The goal is to catch and alarm serious imbalance and swap them out. I had mismatch with 10V and 12.75V in series. I don't know how that's possible, with internal current between them much higher than the external current.
I burnt out a few cheap CenTech meters, some with input fuse intact. Again, don't know how 12V battery can kill meter in 20V range. Fortunately, they still work in HA (High Amp) mode. So, i got plenty of HA HA HA.
But probably not because it is good. I suspect that it is all vanity electronics - custom made solutions for the problem the customer thinks that they have.
Most transformers today used printed windings. Hand-winding work for prototyping, but there are a lot of cores made these days which are designed to work with printed windings - which need thin substrates and thick copper.
The semiconductor parts you use don't offer internal simplicity.
You'd be better off looking at what a broad-line distributor has in stock , so you can see how much they cost, and how many they have in stock.
What you ought to do is think harder about what you are trying to - which seems to be to have a particular LED light up when an input voltage is in a particular range.
The usual solution for that is a single chip processor with a built in A/D converter and a program that lights up one of a number of LEDs depending on the input voltage.
But they tend to end up happier if you go to the trouble of working out what they actually need - which isn't always what they think they are asking for - and give them that. This may strike you as being intellectually arrogant if you haven't actually done, but people who meed fancy electronics frequently don't know a lot about what it can do, and a bit of helpful clarification was work wonders.
Sounds like a SPAD -
formatting link
An uncontrolled avalanche can destroy them very quickly, but if you limit the energy that a single avalanche breakdown can dump in the avalanche region they can work reliably for years. Phil Hobbs knows all about them.
how many thousands you have to sell before you catch up with the added cost design and finding a supplier compared to just picking something off the shelf and be done with today?
On a sunny day (Mon, 26 Dec 2022 08:05:28 -0800) it happened John Larkin snipped-for-privacy@highlandSNIPMEtechnology.com wrote in snipped-for-privacy@4ax.com:
False argumant, Snake oil sells too.
formatting link
is a test setup of my bench power supply. Note the small current transformer Been working now almost 24/7 since 2009 (charging batteries, testing and powering stuff)
formatting link
note the Itrip sense line to the micro in the circuit diagram.
Do you actually know what happens when a broadcaster goes 'black'? Never happened to me in all those years.
The masses will revolt!
Or you are in a studio with so many artists and something does not work? Costs millions to get it all contracted together again!
The 'show must go on' was always the word, and on time to the second.
Nobody gives a,, about the 12 cents you saved per unit on the < 100 you sold while spending 10 hours tinkering and babbling about it on Usenet at your wages... If it is millions then custom made parts are no issue either, China loves to make it, OK I know US tries to destroy the competition no chips to China, bad for bossiness, they will now make their own. US just an evil Mafia club, IQ dropping, full of fear for the competition from the rest of the world, sanctions, it is its end.
On a sunny day (Mon, 26 Dec 2022 17:41:32 -0800 (PST)) it happened Anthony William Sloman snipped-for-privacy@ieee.org wrote in snipped-for-privacy@googlegroups.com:
Yes PIC 18F14K22 plenty here, has build in voltage reference, ADC, outputs can drive LEDs. If that is all it needs to do it seems simple code to me. You could even flash the LEDs and have it do other things at the same time. Nothing much goes without micros these days....
Hey, if all you want is to know when something's off from its neighbor, probe the standard and twiddle a knob until the sum of input voltage and triangle-wave at a fraction of a volt lights two LEDs (on two comparators) equally. Then mismatch comparable to tri-wave amplitude makes one LED go dim, then dark.
Quad op amp and any old zener, resistors two LEDs and a capacitor.
John Larkin's median design and documentation time is about four days - or so he told us a long time ago.
There 's not a lot of new design time in most of his products, and he clearly resent adding to it.
He's not got a lot of product where he has sold 40,000 units - he doesn't seem to operate in high volume markets - and designing for that kind of volume really does involve shaving costs to a minimum and having a transformer custom designed to do a particular job at the minimum price.
Some of the stuff I've worked on sold a thousand or so units a year, and there we knew the price of every last resistor.
so if he was making a different powersupply, for a different market, in a different volume, the importance of the transformer cost would be different....
sure, the price in the component database from when it was bought, but at those volumes it makes very little difference if it is 1 cent or 5 cent
Our parts average about 1/6 of our selling prices, so we don't sweat the small stuff. We're actually selling intellectual property, but embodied in hardware. It is low volume, a couple of thousands of boxes lifetime max.
We expect to make about 2500 more of the pulse digitizer thing I'm redesigning now, and some of the power supply parts costs have gone way up lately. I think we can save about $20 a unit, which is worth a few days of engineering.
I used an LTM8023 as the +10 to +5 switcher, very nice gadget, but it's around $13 now. TPS562208 is 19 cents by the reel!
The Coilcraft Q3903-AL is over $6, so the DRQ127 looks good. I used an ISDN transformer in the HV supply, and they got expensive too. Is ISDN dead yet?
The box also has an LM8261 c-load opamp that's got expensive. I'm experimenting with c-loading cheap RRIO CMOS opamps. The lower island is stability is documented, but many amps have another stable region with really big ceramic caps. The Jim Williams technique is to add capacitance until it quits oscillating.
To do serious electronic design, you need to be independently wealthy, or have paying customers.
An emotional factor is that you never know if your designs are any good unless serious people are willing to buy them for serious money.
Different style:
formatting link
formatting link
This style of breadboard can be kept, in case we ever want to fire it up again. We formally document breadboards and keep them. That turns out to be valuable long-term.
My current case is bigger numbers. And juggling circuits for parts cost or BOM minimization or multi-sourcing is another interesting game.
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.