How about a TRIAC? If it is only for cutting off 10 or 20 volts, easy going.

you only get assymetry problems if you cut all of the waveform with big motors.

 

5 1n4006 ? Don't think so. When I was in machine maintenance repaired a lot of "90vdc clutch power supplies" and "fixed time delay" and "wire break sensor" octal "relays" that weren't more than a couple diodes or a 555.

But now that I'm in field service and customers pay hundreds per hour for me to be on site (I don't get hundreds per hour, unfortunately), it is actually more cost effective for them to have me replace the octal device with a smart relay, since I can do that in 30min or less. Otherwise I would have to bring the busted gadget back to the shop, crack it open, troubleshoot it, find out if we have the parts, order parts if not, receive them in, solder, drive back out, you get the picture. The Smart relays that i carry (omron Zen) only cost $150 and I carry them with me. 120v and 24v. One perfect use I've found for them is DC motor field reversal.

The only components I carry are resistors and 1n4004 diodes

I see your point.

I've run into these "bridge in a box" at a half dozen places over the years. In telemetry tx/rx controls for the city. In lamintor machines for my biggest cx.

I don't know if it's a standard part or not. Never tried to buy one.

It's a wonder companies can afford to stay in business relying on outside service then. (other than me)

Last year while working on a 20 ton antique bridge crane, and having a "wound rotor" motor re-wound, I found that the motor shop insisted on checking the resistor bank with their hi-tech equipment. While I was not there.

The bill was for over $400 an hour on site.

When the new motor wouldn't run, I called their shop. I've met the owner a few times. He was defensive and told me I was an idiot (which may be true). Saying " If there are 3 phases to the motor and resistors are ok, it will run".

Well, duh! After the $2400 resistor check, won't you come back?

I went home and made up a 10V 10A transformer with a 120v plug.
Then used it to power the resistor banks.
One by one I manually pushed in the shorting contactors and recorded the current, on each of the three resistor banks.

In a half hour I found the breaks that caused the initial failure. For 1/24 the cost of their wasted trip.

 

Once upon a time (somewhere in the 1990's), I bought some little, button shaped objects, to suppress the initial surge in incandescent bulbs. (You put them in the socket before you screw the light bulb in.) I wish I knew where to buy some now.

Any clue what they are called or where to buy them?

I wish I did too, as I remember them. As I recall they were a thermistor. I started working on a bulb extender for home lights. This is how far I got. I first thought of turning them on at a 0 crossing, but soon realized the 4mS was not long enough. Then I went to a standard dimmer with a triac and set the value of the resistor for the bulb to light at reduced power. The next thing I was working was a way to short out the resistor in about 30-40mS. The idea was to use a hefty enough triac to handle the full range of bulbs normally handled bya wall switch, perhaps 10 amps or so.
When I made the bulb extender for RV lamps I used the CL series mentioned earlier in this thread.

 

Maybe an NTC thermistor will work?

 

Thought that I could get some readings with a peak reading ammeter.

Doesn't work, as the filament cools slowly. Have to wait 15Min between readings.

Also it's too slow. Peak readings were only 3-4 amps. Changing little with as much as 30 ohms series.

Expected peak being 12 amps rms from the 10ohm cold reading.

Think I'll try one of these.
http://www.cantherm.com/products/inrush_current_limiter/cantherm_mf72.pdf

MF72-20D 15

20ohm cold
4amp max
.288ohm @ 4A

From 12A down to 4A surge should be good.

 

I don't think it is the surge itself that makes bulbs fail; its the evaporation of the filament metal over the life of the bulb that develops weak spots in the filament.

My guess is that in a new bulb, the surge stress on the filament is entirely within the elastic limit of the filament metal, meaning that zero damage is caused by the surge.

It's only after enough filament metal has evaporated off to create weak sections that the surge breaks the filament at that section. But the damage that allowed that to happen was caused by evaporation, not by the surge. The surge causes failure a little earlier than it would for a surge-limited bulb but the bulb has already sustained life-threatening damage by that point.

So for the longest bulb life, run the bulb at reduced voltage AND suppress the surge.

 

Thank you davebee. I already run whole rooms of light fixtures through a dimmer before the on/off switch and have buck transformers on my outdoor area lights. That's the best I can do for the evaporation part. Right now, I'm working on the surge part.

 

........
So, here in Europe voltage was increased in small steps over some decades to burn out all the old equipment, and to shorten the life of new bulbs as well.

It was sold to customers with the explanation, advantage in cable materials allowed an increase in voltage while maintaining the current insulation thickness.

My Gfather, who worked in radion when Boston was switching from DC to AC, said the power company soon figured they could sell more power with higher voltages although the regulators were told it was to diminish occurrences of brownouts. < function of capacity!