One roll of duct tape

Eh? The CD4007 is MONOLITHIC, a single-chip. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

         To save time, Let's just assume I know everything.

Yeah, there's capacitance and bias restrictions because of that. But, you DO get wires for the source, drain, and gate available externally.

When a (chopper-input) lab instrument died on me (sometime in the eighties) and I needed a dual MOSFET for the switch, that little thirty-cent part was just as effective as the manufacturer's eighty-dollar module replacement. It didn't have all the gold plating, though. I forget what I did about the substrate bias; maybe a capacitor on the power pins (so the gate drive would charge it through the protection diodes)?

I envy the electric generator designers the problem reporting modes available to them. If something goes wrong, it usually goes through the roof in the form of spinning 150 tonnes of gear. It would have wiped the popular IT "methodologies" in a matter of days.

Best regards, Piotr

Yes, "discrete" as is separate pairs, not as in "separate dice. It's three P/N pairs of transistors.

I've seen many instances of a (perhaps planned) product expiration timer:

1) use an aluminum electrolytic in a timing-critical spot 2) put it up against a heat source, not TOO hot, maybe a 1-watt TO-220 regu lator lacking a heatsink.

A threshold is crossed after enough years, and either your product goes cra zy, or it turns off and won't turn on again.

In our big rf systems, the offending part is usually the energy storage in the boot circuit for a switching power supply. If the Al cap is dry and th e switcher can't wake up quickly enough, then the internal voltage dips too low, and it gives up. Symptom: your big amps work fine until you unfortun ately had to turn them off. Then they won't turn back on again.

After a building power-outage, when all sorts of things in our department s eem to go belly up, usually the cause wasn't surges and spikes. Instead, t heir "electrolytic death-timer" had fired long ago, and they were just wait ing for the next power-cycling to announce their message: "Time to buy a n ew $12,000 pulse amp."

Heh, there's opportunity for an eBay business where one finds just the righ t pieces of equipment at auction "not working parts-only," replaces the one special $0.15 capacitor, and re-sells as "tested works fine."

Well, there's always the electrochemical coulomb timer which would make a reasonable warranty timer. No need for subterfuge. Just install one of these timers in your product and inform the customer that they don't actually own the product, but instead are licensing it's use from the vendor. At the end of some pre-specified number of hours of operation, the device self-destructs and you're expected to renew your license by purchasing a replacement product. It's somewhat like the hardware version of Apple's 5 year planned obsolescence:

The life of most types of capacitors can be calculated if the operating conditions and temperatures are known:

Here's a marginal example of how to reduce the life of a product: This is from an LG L246WP LCD monitor. Notice that there are pads for

3 filter capacitors, but only two are populated. What this does is raise the ripple current in each or the 2 capacitors by about 17%. Apparently, the displays were not blowing up quickly enough to meet someones predictions. Removing one cap reduced the life, saved a few cents on the component cost, and probably provided someone with a nice performance bonus.

Yep. I have exactly the same problem with computahs, servers, network hardware, and such. After it cools down, it won't start. I've used a hot air gun to re-heat the power supply so that it would start, which would buy me some time to find a replacement.

It also happened to me recently. My home DSL modem (Efficient 4100) had been powered by a UPS for the last 10 years. We recently had a simultaneous all day power failure and DSL outage, which inspired me to try various replacement DSL modems. When I went back to my original DSL modem, it would not power on. Just for fun, I blasted it with hot air from a heat gun, and it started normally.

Or, time to fix the $12,000 pulse amp. All you need to do is replace every electrolytic.

That would be me: The Wavetek/SSI 3000 service monitors were repaired mostly by replacing every electrolytic and tantalum capacitor. The HP8620 sweep generators all have similar electrolytic capacitor problems. Same with lots of other test equipment I've purchased on eBay. I don't mind the "tested, works fine" baloney. As long as the price is right, I can fix it myself.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Yeah, or (Sony did this) an audio output with a current-biasing scheme that depends critically on the transistor's beta (beta rises with age of transistor, and Vce(max) drops).

Then there's the motor control chip that's probably just two or three power transistors in one special package, and when the laser printer scan head stops spinning, you find it has heat-blistered. Thing is, the package has a part number that's unobtainium, and the module it's on is out-of-stock.

Every CRT device in the world is one broken filament away from unrepairable, except electron microscopes (I have a box of spare filaments for those).

Well, except for parts mules and LCD replacement kits. ;)

Cheers

Phil Hobbs

(Who recently bought an LCD kit for a Tek TDS 784A--and will install it sometime when life calms down a bit.)

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

and filament breaks are very low on the list of faults in such equipment.

NT

Those 1990s HP network analyzers use VGA for their monochrome display. If you replace the dimmed CRT with an appropriate VGA flatscreen, you suddenly have multi colored animated Smith charts!

Or, just stand a cheap 14" LCD monitor on top of the cabinet (about $7 from Re-PC.)