I wrote about an old oscilloscope which was giving me trouble. After some poking and prodding, I managed to get things tweaked back into operation. Then I was hacking on some software, pursuant to driving the 'scope as an X-Y display off a soundcard output, when I became aware of the 'scope making small popping noises. Small popping noises are not a normal operating sound for that 'scope. I used it enough to know _that_. Then, looking over at it, I saw that (at least some of) the tube filaments were dark, and there was nothing on the display. So I turned it off. The popping sounds continued, and I saw a wisp of smoke. I unplugged it and turned it over in search of the part that was upset. Turns out it is the power transformer (not surprising, in view of the dark filaments). It continued making popping noises well over a minute after I unplugged it, and for a few seconds I was even mentally making sure I knew where the nearest fire extinguisher was, since it gave the impression it might actually catch fire. It's now been at least ten minutes since I unplugged it, and the transformer is still too hot to rest my hand on for longer than about a quarter-second. And, it is giving off odours which I am also quite certain are not part of normal operation. Tomorrow, I'm going to try turning it on (very briefly, with a current-limiter in series with the primary) to see if it's healthy enough to give me useful voltage measurements. The only winding whose voltage I am confident of is the one that drives the tube heaters. Based solely on wire count I am sure there are at least two others - and based on the application I expect at least one for the final anode voltage and one for power to most of the circuits. Anyone know a good supplier for ten-kg power transformers? :-( /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

der Mouse wrote: YES! Well since this has to be a custom design,try here -- http://www.electra-print.com/

That was my own impression too. If it had been more accessible to air I suspect it *would* have caught fire, but it was enclosed in a fairly solid steel case.... I actually don't suspect that. First, because that would drastically reduce supply output voltage, and the thing did work for half an hour or so. (I found the trace; after poking around the circuit enough I found the right pot to adjust.) Second, because - after the transformer overheated - I felt all the caps in sight, and none of them were even warm to the touch, much less as hot as I'd expect if they were responsible for overloading the power transformer. No semiconductors at all as far as I've been able to find yet. While I haven't yet put much effort into searching, there are at least three rectifier tubes, so I see no reason to expect semiconductor diodes. Perhaps - though, as I say, shorted caps are low on my suspicion scale at the moment. Or do caps that fail shorted do so with a low enough resistance that they don't dissipate enough power to get warm? If so, it coulda been one of the caps failing in service. As for a replacement transformer, I'm actually considering using multiple different transformers, one per function, instead of a single core with lots of windings. I might be able to use stock transformers that way. But first, I'm going to poke through it enough to draw a schematic of the power supply portions of the circuit. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Possibly, though I find Chuck Guzis's guess that it's boiling varnish somewhat more plausible. Just a thought - does anyone know what happens to a steel-core transformer if its core goes above its Curie temperature? While I don't think it got anywhere near the Curie temperature of plain iron (about 1000?K - and paper catches fire at something like 500?K), I don't know the Curie temperature of transformer-core metal.... That's what I fear too. :( I haven't done anything with it, yet; I've been at work all day. When I get back home I'm going to trace out a schematic of everything the transformer feeds, mostly meaning the power supply, to tell whether the suggestion of pulling all tubes really will remove all load from the transformer. (I expect it to, given the apparent absence of semiconductor diodes, but am not about to count on it without checking.) I may do that, but first I'm going to see if I can get voltage measurements off mine my removing all load. If necessary I'll cut wires; even if this transformer works in isolation, I am not about to try to run the 'scope off it for more than moments, so I definitely will be cutting wires to remove it at some point. The thought is much appreciated! /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Yes, but it also changes the magnetic behaviour of the material in other ways too. (For example, the Wikipedia page on the Curie point says that the magnetic susceptibility theoretically becomes infinite.) I was wondering if these would affect its performance as a transformer. Oh, I wasn't worried about it. Clearly it doesn't get that hot in *normal* operation. I just got curious if anyone knew whether a hot core would make its behaviour as a transformer significantly different fixing the thing. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Welcome to the world of old tube equipment. :/ That popping sound is usually the sound of boiling tar in the transformer (and of course the latent heat keeps it boiling after power is turned off). The symptoms you describe are not unusual and it's most likely a short in the transformer. If, as you say, only some of the tube filaments were dark, it may be that there are multiple filament windings and one of them developed a short. In equipment of that sort, shorted caps don't usually present such a degree of problem to the transformer. The vacuum-tube rectifiers have some hundreds-of-ohms of plate resistance, as well as maximum emission (current) limits, and act as a limiting impedance for shorted/leaky caps. If it's any consolation, there probably wasn't anything you could do to avoid the failure. I pulled out the schematic for a Stark OSK-4, a very basic vacuum-tube 'scope, probably similar in category to the DuMont. Here are some observations which may (or may not) be of assistance in your situation, if you are RE'ing the power supply: - The CRT accelerating potential does not come from a high positive supply for the anode, rather it is a high negative supply for the cathode. It is obtained through a half-wave rectifier, supplied by a transformer winding internally stacked on one side of the main B+ winding. The filament for the half-wave rectifier is obtained from an additional tap on the high-voltage winding. - The CRT has it's own filament winding, as the cathode is operating at a high potential and the cathode-filament insulation potential is limited. (In other words, the transformer is providing the high-potential insulation.) - The B- (center-tap) does not go directly to ground, rather it goes through some resistance to ground so as to develop a negative supply (for cathodes of the V input amp.) If there's no rush and you're willing to wait on a possibility I can poke around the radio museum here in the next couple of weeks to see if there's a junkered/junkable DuMont (let me know the model number). We have piles of old vac-tube scopes that we should get rid of and I know I've seen a DuMont or two. Shipping (cost) would be from Vancouver.

Uh, thank you. I guess. :-) (I actually started hacking electronics with vacuum tubes, well before I did anything significant with semiconductors. Back in the days when you could find large numbers of vacuum tubes thrown out in the guts of consumer electronics.) That's not quite what I meant. I looked over, and all the tube filaments I saw were dark, but that was only about three of them, and I didn't look to see whether the rest were or not before diving for the power switch. So I can be certain of only some of them, but my observations are consistent with it actually having been all of them. That's what I'd expect. I'd expect to see red-hot plates on the rectifiers before I'd expect the transformer to go - especially a transformer as big, and, by implication, with as much excess capacity, as this one. That's reassuring, yes. Thank you. If it helps, the tube list for this thing (aside from the CRT itself) is: 12AU7 x8 6J6 x2 2X2A x2 6AQ5 x2 6Q50 x1 5Y3GT x1 0B2 x1 The 6Q50 might be 6Q5O or 6Q5Q instead; there's a dark mark at the lower right of the last character, but it's not clear whether it's an intended part of the printing or a bit of dirt. The tube actually in that socket is an 884; "or 884" is penciled next to the socket. Noted. Thanks. I don't know whether they'll be useful, but I'll keep them in mind. "TYPE 304 H". That's the closest thing to a model number I could find. Well, a lot of the work I'll need to do I'll still need to do even if you can dig up a spare for me. And it's unlikely to take less than a week, so.... If you do find a junk one, I'm tempted to ask if you'd be willing to send the whole thing, as a parts unit (most notably the tubes, at least after the power transformer). /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

For those of you following my Dumont oscilloscope saga (based on an offlist email, that set of people, while perhaps small, is definitely nonempty :-).... I just now tried hooking power up to it with all the tubes pulled (including removing the CRT's socket from the tube) and a (40W) light bulb in series with the mains. The light bulb came on, to the eye about as bright as full power. A voltmeter across the transformer primary showed about 30V (vs about 110-115 if I have the 'scope's power switch off). The clip-on ammeter shows nothing at all until I wind the wire around to give it a x5 multiplier, at which point it shows 1.2A, meaning actually about 1/4A. So either the primary is very sick, or some secondary winding is shorted and is loading the primary. I then tried something else. I have a filament transformer which produces (nominal) 12.6V centre-tapped at 1.2A (according to the markings on it). So I hooked half its secondary across the heater terminals of a convenient tube (of the 6J6s) and applied power (with that light bulb in series with the mains, to be sure). The bulb burnt dimly. (It is completely invisible if I leave the secondary open; this filament transformer itself seems to be OK.) Measuring the voltage across the primary, it had sunk to about half mains voltage; a clip-on ammeter on the secondary showed it pulling about 3.5A. (This is with all the tubes completely removed from their sockets, CRT included.) A dead short across the secondary carries about 4A. The voltage across the secondary - nominally 6.3V, and that is what I get with the oscilloscope disconnected - is about 1V; a lot is being lost in the transformer (not surprising, as it's carrying some two and a half times its rated current). The winding that provides the plate supply for most of the 'scope measures about 12V. Since it is rectified by a 5Y3GT and then fed to a 0B2, it must be designed to produce at least 75V or so, probably more like 90 or 100. If we assume the fried transformer is acting normally as a transformer for the two windings in question (admittedly a large assumption), this is right at the margin of reasonable: it's getting 1V instead of 6.3V on the heater winding and is producing 12V on the medium-voltage winding, for 12V*6.3V/1V=75.6V nominal. This is not quite enough to make the 0B2 regulate, even ignoring the drop in the 5Y3GT, but is in the right ballpark. However, with the same setup I am unable to measure any voltage whatsoever across what I think must be the high-voltage winding, the one that's rectified with two 2X2As to produce the HV for the CRT. Perhaps I'm wrong about which wires are the ends of that winding, or perhaps it's failed open.... I shall be investigating more, but it's now late enough I need to call it quits for the night. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

THANK YOU! (For not writing "It sounds like your...".) Ooh, yes, I forgot I was measuring plate-to-plate voltage, not plate-to-centre-tap voltage. So the voltage I was seeing is only about half what it'd have to be, even ignoring losses in the 5Y3GT and the "substantially higher" you mention. Yes, I think I agree. It is sufficiently dead in place that even if it's something fixable I'll still have to take it out to deal with it. There is. I wouldn't unsolder it either, but my principal reason is that most (all?) of the connections have the wire wrapped around before soldering until it's reasonably secure mechanically *without* solder. Good construction technique in most respects, but a pain when it comkes to desoldering. (Especially when the connection is not very accessible, which is the case for at least half of them.) Actually, I would more likely just tag each wire from the transformer (probably with a simple serial number), and write down where each tag goes (things like "V9 pin 5" or "C12 positive side" or "chassis") - and cut the wires as close to their connection points as I can easily manage. None of the insulation looks degraded, so I consider that, indeed, a very very slim chance. I sha'n't overlook it, though, and will check for evidence of such a thing when I crack the transformer case. (Given how electrically broken the thing is, I'll probably have to open it up just to tell which wires go together - there are a few I'm not sure of.) I'm not sure what a "clamshell" is, here. This transformer has metal (presumably steel) sheet laminate for the core, metal sides enclosing the windings, and bolts through the sides and the lamination layers to hold it all together. It's a common design for vacuum-tube-era power transformers, in my (admittedly limited) experience. And it's big; my "ten-kg" upthread was a somewhat flippant estimate, but it wouldn't surprise me if it's about right. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

der Mouse wrote: (Thats's a lot of 12AU7s) Yes, the tube complement may help with correlation if there isn't a perfect model # match. 6Q5G: http://frank.pocnet.net/sheets/041/2/2B4.pdf Thats a funny tube: "..intended for use in DuMont Cathode-Ray Oscillographs for sweep oscillator service.." http://frank.pocnet.net/sheets/049/8/884.pdf contains a small schematic of a sweep oscillator using the 6Q5/884. (http://frank.pocnet.net/ is a good ref for tubes including oddballs) OK, ..I'll see what might be found..

Possible insulation breakdown somewhere, still discharging a charged capacitor? I would disconnect all the secondary wiring and power it up with a mains light bulb in series with the primary (a well known way to do this is to connect the bulb across the fuseholder with the fuse pulled). The bulb should stay dark (if not, you've got breakdown or shorted turns in the transformer), you can then measure the output voltages. I would expect some of the windings to be centre-tapped (it's a fair bet this thing uses the 'biphase' full wave rectifier circuit). The 2 halves should be the same voltage, of course. It's not unheard-of for 'scopes to have series-strings of valve heaters run off fairly high voltage stabilised DC lines. Well, at least Tekky did this sometimes. Most 'scopes run the gun electrodes at a high (a few kV) -ve with respsect to the chassis. That means the deflection plates can be run at somewhere near chassis potential (a few hundred V above it), so they can be easily connected to the deflection amplifiers. There may be a post-deflection-acceleration electrode, which will run at several kV +ve wrt the chassis, although this may not be present in your 'scope. It's also not unheard-of for the EHT to come from an oscillator/transformer circuit run off the normal HT (B+) line, not straight from the mains transformer. You said some of the valve heaters were out. If that means some were stil alight, then there may well be 2 or more heater windings on this transformer, one of which has failed (or has been shorted out by a fault elsewhere in the instrument). Rewind it yourself? -tony

On Tuesday 04 July 2006 12:53 am, der Mouse wrote: It is? :-) Right. I'd forgotten about that. Those 4.7M resistors are to maintain the DC balance of what's normally driving the plates. You might check and see if they are indeed 4.7M. That's a heck of an imbalance, there. Probably leaking capacitors coupling at the amplifier output? -- Member of the toughest, meanest, deadliest, most unrelenting -- and ablest -- form of life in this section of space, a critter that can be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters" - Information is more dangerous than cannon to a society ruled by lies. --James M Dakin

On Monday 03 July 2006 11:39 pm, der Mouse wrote: Before the days of Tek and HP, Dumont scopes were some of the best available. At various times, I've had Eico and RCA scopes, and both brands had serious warm-up drift. Supposedly, the Dumonts were among the most stable. Cheers, Chuck

der Mouse wrote: .. generally the result of an imbalance between the symmetric halves of the push-pull amplifier for the axis. There should (may) be an internal pot ("DC BALANCE") for the appropriate axis to bring the amp into balance. (not that it matters unless you can get the scope working again.)

On Tuesday 04 July 2006 04:29 pm, der Mouse wrote: I think in my case it's the battleship gray paint and the little metal plate with various ID numbers on it that probably gives it away. :-) I gather from later stuff that you wrote that it's probably somewhat bigger than mine is too. Mine also has a handle on the top. And some bits on the bottom to wind the cord around. And the most unusual AC plug I've ever seen, it was three-prong but you could pivot the third prong out of the way if you wanted to use a 2-prong outlet... Ok. Yeah. Good luck with that transformer situation. -- Member of the toughest, meanest, deadliest, most unrelenting -- and ablest -- form of life in this section of space, a critter that can be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters" - Information is more dangerous than cannon to a society ruled by lies. --James M Dakin