I was winding up one of my old HP spectrum analysers (my best one, as it happens, been sitting around unused for years) on the variac yesterday. Got to about 170VAC from 60 over several hours and just got impatient I admit it. Wanged it straight up to 240 and *POOF* the all too familiar magic smoke came out. Patience was never my strong point. So my Sunday is going to be spent trying to find and fix whatever it was that went bang. I kind of knew something was in the offing when the display started to twitch. You know how it is with a CRT tube when it starts winking at you. Any predictions as to the likely culprit? I'll post back later...
You mean the X2 type on the incoming power? Good tip. I shall do so. I've been kind of distracted on opening the top cover by the sight of soot over some of the EHT bits, in particular the 18KV line to the CRT anode. I'll try to post a pic if I can overcome Flickr's barriers.
My guess(tm) is the big electrolytic filter cazapitors are shorted, causing something between the line cord and the caps to smoke. Smoke usually means some resistive device burning up, although it can be a hot trace scorching the PCB. The color of the smoke would be helpful. Black is carbon comp or oil, white is plastic wire insulation, gray is phenolic, brown is epoxy or a dead mouse or insect infestation. There are smells associated with each one that is recognizable. I had a mouse take up residence inside an HP sweep generator last fall the demonstrated gray smoke and a horrible stench. If you can't find the source of the smoke or smell, shove one end of a 1/4" ID vinyl hose up your nose, and inhale while moving the other end of the hose all over the device. (Yes, I actually do this). It takes some practice and fortitude. For a mouse, have a vomit receptacle handy. You won't find any of this in books on electronic repair.
If you're dealing with very small amounts of smoke, that's either invisible, or pervasive (smoke everywhere), I suggest one of these airborne particulate detectors: My office was filled with smoke when I arrived one day and I had no idea where it was coming from. The Nikken "air quality" detector found it fairly quickly.
If you have money, consider buying an IR camera. By IR, I mean heat as in far-IR, not the common digital camera which does near-IR. Prices are coming down and you get something reasonable that plugs into an Android smartphone or iPhone. If you have a source of heat causing problems, that will find it. I borrow one from a friend occasionally. I've used one with great success isolating a bank of stepper motor controllers with conveniently exposed power transistors. It was easy to distinguish between warm normal, too hot, and stone cold.
Jeff Liebermann email@example.com
150 Felker St #D http://www.LearnByDestroying.com
The IR camera is one of those tools I rarely need, but it is REALLY helpful when I do. I bought the Flir for Android and took it back. Bought the SEEK and kept that one.
The Flir looks better because they combine the IR and regular cameras into one picture, but the actual IR resolution is less. The SEEK uses the camera in the smartphone and you have to swipe between visible and IR. Problem is that the IR and visible cameras are typically not along the same vertical or horizontal line and far apart. For close work, parallax is a killer. You can move the pictures relative to one another for a fixed location, but if you move, it can change a lot.
There are a couple of things to look out for: Smartphones must have USB OTG. For older phones, it's hard to tell whether it has OTG. There's an android app to tell if you have OTG, but I have phones that claim to be OTG, but don't work with the camera. I tried powering the camera externally without success. The microUSB sockets can face forward or backward. Obviously, IR selfies are not what you want. Ditto for side locations. At the time I looked, there was no way to rotate the image. I thought I could use it with the image rotated, but my brain couldn't handle it. I used one of the Best Buy $10 smart phones they had last year. Had the side connector. I built a clip into a phone protective case and mounted the camera as close as possible to the visible camera with the correct orientation. It worked, but the software didn't like my extension cable and locked up periodically. I also tried mounting the camera on a flexible extension so I could use it in an awkward location and still see the screen. Same cable problem. Some extension cables would not work at all. Others merely locked up a lot. That's not an IR problem, it's a USB problem. I yelled at customer support, but they were unsympathetic. They avoid the issue by claiming that only very few smartphones are supported. Maybe newer software fixes that...dunno.
If you get a SEEK from somewhere other than seek, you can't tell what you're getting. The original model was fixed focus, and worthless for close up work. You could buy a close up lens on ebay, but it was $55. Newer models have a variable focus lens that works very close up. Problem is that they did it without changing the model number. The only way to tell is to try it. There's a second model with variable focus, but narrower field of view for hunters. Looks exactly like the others.
One big problem with IR is the emissivity of the surface you're looking at. If you put a shiny metal pan on the stove, it looks cool to the IR sensor...just don't touch it, it's hot. Trying to determine the temperature of your shiny motorcycle head is a waste of time. That's a whole 'nother discussion topic...
Bottom line, if you have $200 to spend on a toy, the IR camera is a useful one. You'll be doing silly things like determining which of your dozen thermos bottles works best. Good times...
There was not really any visible smoke to speak of, Jeff. It was primarily the smell that indicated something had gone phut. Smelled like toasted PCB, but I can't see any burned areas at all. Might be a different story if I tore it right apart but so far as I can see, which is like 90% of it, nothing. One thing I can say is it's *definitely* not due to any infestation! Funny thing was, I couldn't detect the slightest smell at all after removing the covers; it was a very transient event with no noise either.
I shall look into those links; many thanks.
Yes, they're clearly very useful to have around. I'll see what's available here in the UK.
You were dialing the unit up in an effort to reform the electrolytic capacitors, and if you didn't know that before now you do (look up the term).
Electrolytic capacitors in the primary lower voltage power supply are the first suspects. I am not familiar with that unit, but there should be a section with B+ and possibly B- supplies - their filter caps are suspect. Also check if their bridge rectifiers were damaged, however that usually doesn't happen when a cap blows up.
Of course there are or should be fuses on the main power bus lines, check them and replace after finding the culprit.
The pictures suggest that the problem might be in the high voltage power supply. That's surely a switcher. Don't know if the main supply is a switcher. It's been suggested that you shouldn't try to reform caps by slowly ramping up the AC input voltages on swtichers. Disconnect the load side before you reform caps. Ditto for the switcher that is the HV supply. It probably doesn't like having it's input voltages all over the place.
Many circuits are designed to operate under specified conditions. Ramping up the AC is not a specified condition.
Depending on the competence of the engineer, she may have considered odd failure modes. But, if fixing it cost a penny, there would be resistance from the bean counters.
Point is that people will do stuff outside the spec, like ramping up the AC. You can have failure modes that you'd never think about if you considered only the specified worst-case conditions.
There may be no obvious cause for a failure caused by operator error trying to fix something that wasn't broke.
You raise some very good and interesting points here which I shall bear in mind in future. So far as the sooty EHT components are concerned, it would seem that these can arise from a blown component elsewhere within the case; the smoke particles thereby emitted are attracted to the EHT cabling due to electrostatics. So quite possibly it's a red herring and there may be nothing amiss with the CRT supply circuitry at all. It may simply be evidence of a previously-fixed failure elsewhere.
The problem is, this may be catastrophically wrong with modern gear with swtiching power supplies. Old tube gear could very often be brought back to like like this. But, switching supplies are constant power devices, and if they start up at exceedingly low voltage, they will draw excessive current. So, one may have to reform capacitors out of circuit on such gear.
Older gear with CRTs can have anazing amounts of black dust that has been attracted to the HV parts. It can come from local air pollution, just normal household dust, etc. not necessarily something that burned up. If you had a major component burn up, then the whole inside of the case would be black.
So, those black HV wires are not a sure sign of something burning.
It is often a bad idea to bring electronics up slow. Switching supplies do not like it. Some of the older tube equipment does not do well either unless most of the tubes are removed. You can run into what is called cathode stripping.
If the capacitors need the reforming, many times they will not last very long after. The equipment may seem to work ok, but it will often not be at its best.