Ok, I'll read more...but I think I'll forget about ordering a dipole antenna (since they're expensive relative to every other component combined) and just go with the wire
If anyone thinks I am being nitpicky in my comments to new questions or new answers please let me know. I may seem that way because I had the same comments posted to me about splitting blocks of text into paragraphs, use proper grammar, don't use confusing terms, etc. I have seen Olin tear someone to pieces who posted a simple, confusing but rude question, so there must be some gray area here.
Hi. Frustrated. My oscilloscope exploded last week. I replaced the exploded capacitor, but something else died so it still doesn't work. Just got back from buying more parts. I'll see if I can get it fixed after supper.
In case any one was wondering, the K8/50 diode has the 50 in its name because it is literally a stack of 50 selenium diodes inside a spring loaded tube
This explains my difficulties in testing the two diodes in the highvoltage supply of my scope. 50*0.6V is 30Volts. I tested with 22V, so they both looked dead. They actually are dead, since checked in circuit, in operation. Both failed open.
@KrishnShweta It's supposed to be a variable resistor controlled by the base current. But it's not really a great model. If you can understand "I_c = beta * I_b, unless that requires V_c < V_e + 0.2 V", you're better off using that.
Replacing the K8/50 diodes with 2 strings of 4 x 1N4007 did the trick. Well, that and the signal diode in the trigger circuit.
@PlasmaHH I normally test them with the diode test on my multimeter. That didn't do any good at all. The test with 22V was through a 10k resistor. One way should have read 22V, the other way should have been much lower. But, 22V wasn't high enough to get 50 diodes in series to conduct.
@KrishnShweta Very kind of you. I've got my old clunker running again, so I'm happy now.
@JRE My tactic with pre-80's: Chain as many voltage sources in series until at least 100uA starts dribbling through. At blocking voltage divided by 20 still no current: problems!
With the resistor and creating a smooth as possible ramp, of course
But, admittedly, my lab has (75+30+32+32+32)[adjustable to 0 each]+24+48 per standard + several to dozens of times 12V and 24V LED supplies, so it takes quite long before my lab runs out of voltage
If you have issues like these more often, before my hobby started making me money, I had one old 30V lab supply from Philips (cost of repairs 14 Dutch guilders - supply: free) and a bunch of 1A rectified transformers for fixed steps and rocker switches to switch over
Not the greatest set-up, but with a good multimeter a decent way to be able to test low power devices up to 100VDC or even more
I hope not to have these issues often, but I'll try to remember your tips in case it ever happens again. Fact is, I'm not well equipped here at home. Most the time I have spent working on this has been looking some aspect of the problem and thinking "Well, I could check X, but don't have the tools. Check something else and try to think of a way to check X using stone knives and bearskins."
I put off measuring the HV until I was sure there was nothing left that could cause the problem except for lack of HV. Then I put together a voltage divider out a couple of 1/2 watt resistord so that I could measure upto 4000V with a meter rated for 1000V.
Did I mention that I don't like high voltages? The 300 volt rsils powering the tubes made me nervousas hell. 4000 was just too much to mess with straight off.
reverse leakage may not be consistent, creating a differential, for that purpose, many cases, one would place 100's of kOhm or several to dozens of MOhm in parallel with each, to catch the reverse leakage differentials
Of course, rated for the proper reverse voltage
Especially since the original was very old fashioned, the extra leakages from the resistors will likely only make a closer approximation of the stack's flaws
@AdamUraynar That seems to me to be a 50pF tunable cap, tunable components (variable resistor, variable inductor, variable capacitor) normally are indicated at their maximum value
@Marla That was one of the things I found in the old D43. There was a pair of wires that were sticking up in the air. I panicked because I didn't know where they should go, then found a reference in the calibration notes to twiddling the wires to adjust something, and relaxed.
@Marla And what was connected to them? Why did it work without the missing part? Was the missing part hindring something? Is this a Philips Type Bodge?
@Asmyldof Now that may explain a lot. The first time it exploded 20 years ago, I just swapped out the whole board, and all was good. When I realized the diodes were bad this time, I tried swapping in the diodes from the previous board - no go.
Looks like I'll be ordering some 1M Resistors rated for a thousand volts.
@JRE So, what you do is look at the datasheet for the specific 4000 diodes you used, look at the reverse leakage graphs and determine their delta. Add a resistor that handles double that, done
@W5VO Yes, where one end of the switch is connected to a capacitor, which is connected to nothing, and the other end to the floating case and the middle to an unknown part on the board
Ha. "Strip of 100." I panicked when I got to Conrad, and the computer said they only had 5 on hand. The salesman checked anyway, and they actually had a strip of 10.
@W5VO If I'd be running a shop, I'd sure as hell only order from Mouser, Farnell, Digikey and China-Mfg-Direct, but then, I'm not in it to become a billionaire over the backs of penny-rich hobbyists and kids
Or in the most extreme case, the individual manufacturers, but if you have enough overhead for a shop, you have enough overhead to pay an accounted to sort out the invoice mess Farnell sends you
@W5VO I know what Farnell has and hasn't. But even given the time to wait for Mouser and the order size to get free shipping at Mouser, there's still plenty things that are easier, cheaper or more managable through Farnell. But you must always weigh the invoicing and overhead costs
@JRE There you go with Conrad's brilliant QA system. I really don't know how much harder the point needs to be made
Though the disadvantage of "the Maker Movement" and "IoT for everyone" is that people learn that ordering from Ching Ho Peng to get Huang Hung Loh brand boards works fine, and stuff is wrong means ask on EE.SE until someone tells you what transistor to replace. Rather than known-sourcing things
So I can imagine actual shops dropping left and right
But in all fairness, since the mid 90's or some such, Conrad has sure as balls not helped in establishing that known-good is better than probably-okay
The only other close by shop that I knew of closed up some time ago. It was never open when I needed something, and then it just disapeared one day. There's some other shop where it used to be now.
@PlasmaHH They've sure got more of that kind of stuff than they have parts. The one in Mainz is pretty large, but they've downscaled all things "build it yourself." I actually considered buying a new scope rather than fixing my broken one, but they don't even have them in stock. If you want a scope from Conrad, you have to order it.
@JRE I wouldn't buy that either. You'll be back soon enough to buy parts to repair that, which are not available, having you buy yet another poopy-McBleh-brand scope
With the signal to noise and measurement accuracy of having your youngest nephew or niece draw what they think the signal could be like
@Asmyldof That made me really laugh. I've got couple of nieces (and they've got daughters, too) and I can well imagine them (the daughters of hthe nieces) doing better than a Voltcraft.
@JRE their scopes haven't improved since the 90s and the bandwidth is useless. Almost all voltcraft stuff is rebranded, so figure out what they copied badly. Plus the parts they have in stock are usually from the 90s too, very useful for parts that have shelf life. Dry popcorn guaranteed
And I find the diode functionality of my meters rather useless I have to admit