exploding wirewound resistors

ickel

g a

't

all

ng,

sort

ation

ly quicker if you're sure you're only going to solder 1 or 2 joints, which is seldom the case. Their soldering power is the pits compared to a 60 or 1

Also good as a source of 60Hz B-fields when looking for ground loops, and such.

George H.

estruction prone.

That is interesting. How do you couple the field into the ground loop?

Say, for example, 50 ohm coax, or a 110 VAC power cord. Then what do you look for as the output? And where do you find it?

Oh I just wave it around. (orientation is important, obviously) It's rather a blunt tool. But I've sometimes found some loop I hadn't thought about, usually it's a loop you have to live with.

Typical case might be a system with two coax inputs. Sticking the gun between the coax cables give you a big signal... which I observe on the 'scope.

I use to worry a lot more about ground loops. (Star ground.) More recently I've come to the opinion that it's good to 'grasp your ground firmly' (like a nettle :^) ground early and often. (It's a complicated subject.)

George H.

Yes, I agree. I have switched over to sheet grounds. My desktop is a sheet of stainless steel, around 48 * 36 inches. My equipment racks are all steel shelving with straps to each piece of equipment. Still, in the winter when the humidity is low, I can draw an arc about 1 inch long to the metal. This drives my computers nuts. They go into complete lockup and require power down and reboot to recover. Fortunately it seems that none of the files are damaged.

My solution to the problem is to make a 80 megohm static discharge tree out of a woman's perfume bottle. The cap is a 1" washer, the base is a 3" washer that sits on the stainless desktop. In between is a stack of 10 megohm resistors in series, and a small NE-2 neon bulb to show the discharge.

Sometimes it takes a few seconds to drain. I am amazed at the number of Coulombs I can generate by simply taking off my coat.

My computers are happier now.

Roight. Stuff that's good for 60 Hz ground loops is horrible at RF.

Part of the issue is that the source impedance of a ground loop is so low--a 1-m loop of RG-58 shield is about 3.3 uH with a series resistance of about 20 milliohms. A milligauss of 60 Hz field produces

Cheers

Phil Hobbs

OK, I'm done. Here are some victims:

I've got two parts that look reliable to pulse at 15 joules or so. Vishay AC05W looks best, didn't blow up until hours at 21 joules per shot.

TE ER74 is a small 3 watt part, but very tough. Wouldn't die at 17 joules over 30 hours, 0.1 Hz. But it swings a lot in body temperature, which is a worry long-term, so the 5 watt Vishay, with a lot more mass, is probably the safer choice.

I think we can shut down at maybe 4 joules. I did test some big wirewounds that failed at 5J (like the big white rectangular one) in a couple of minutes of pulsing.

Drastic. Can you take some large smd resistors and stack them on top of each other? A bit of pain in production, but you have to spread the heat a bit better.

On a sunny day (Fri, 29 Sep 2017 11:28:59 -0700) it happened John Larkin wrote in :

From a statistical POV you should test more than one of the best you found, to see if there is any production variance?

I doubt that smds would stand millijoules, and production would attack me with sharp/hot objects. The surfmount wirewounds were pretty bad, blew up and squirted goo out the cracks.

The 5-watt Vishay axials looks fine. They will be up in the air stream, not heating the PCB surface.

What I learned here is that most of my expert opinions were wrong.

That's exactly what my wife said yesterday. Oh, all right.

Yes. Ground everything as often as possible, on a PC board or in a system.

Use differential signals when you have to.

Huh, yeah, when to go differential?* I get to do this DC lab amp soon, next week. I'll post a schematic for criticism. Measuring signals from down the bottom of a probe (~3' of SS, and 3' of cable) a wide range of source impedances.

George H.

Disappointing compared with your ex-fets picture. ;)

Interesting, thanks.

Cheers

Phil Hobbs

Or use transformer coupling to break ground loops.

Differential is no panacea. LVDS has about the same noise rejection as 5V CMOS.

One cannot escape their obligation to consider, and preserve, the voltage between signal and ground, no matter how many signals or polarities one uses!

(One of these days, I'm going to need a coax connector not firmly grounded, and it's going to have better CMRR and EMI than anything else, despite how terrible that sounds. Just takes a little logic.)

Although RS-422/485 receivers (and CAN, among others) do a good job of helping you not need to worry about it.

Tim

As acids go, HCL is fairly tame, it's is the stuff in your stomach for instance. personally I have an irrational fear of electric shock after a few close calls.

for small quantities you seem to be stuck with nickel ingots at nutty prices.

apparently nickel is about $10 per kilogram in bulk

It's not HCL but HCl - hydrogen (H) and chlorine Cl. There is no element with the symbol L - lithium is Li, lanthanum La, and lead Pb.

The trick is finding somebody who will supply a small sample.

We used them when I worked for Cambridge Instruments in the UK, but the current price list says that they want you to spend upwards of $100 on any one transaction.

has a smaller minimum spend.

HCL is also "Muriatic Acid", often used around pools, and in products like CLR. I don't "fear" electric shocks but they do get my attention. Try 100J through the chest from front to the back. Trust me, it hurts.

It all depends on the concentrations involved. For some reason it seems to be very difficult/expensive to source pure Hydrochloric (Muratic) Acid whereas 98% Sulphuric is commonly and cheaply available in the big DIY stores and plumber's merchants (although probably not for much longer thanks to some of London's very stupid/evil morons. The one acid that I won't go near is Hydroflouric. That is *seriously* nasty stuff indeed; real 'horrorshow' if I may borrow the term from Anthony Burgess.