HAMEG HM604 scope pre-fail?

[tpecar]

I was intending to put my old scope back into use, and, out of curiosity, did a quick check of its internals, where I surprisingly found this


Pictures of the whole board are enclosed at the end of the post.

The transistor (+ its socket) and the shielded inductor (blue package) look as if they had been exposed to excessive heat. They are part of the HV generation circuit. No other component seems to be affected.


Full schematic is available on R&S site, page 44

Since the last time I've used it worked ok, I went ahead and tested it, and sure enough, it's still operational.
I've checked the voltages provided by the LV section (+/-12V, +68 V, +140 V) and they are within spec. I didn't measure the voltages in the HV section, but the intensity control works ok, and one can get a bright trace at the highest time/div setting.
There are also no apparent issues with deflection, blanking, etc.

After I was running the scope for about an hour, I've checked that area and those two components didn't heat up at all.

So, one could consider this a cosmetic defect, but it is still unnerving.


Has anyone seen anything similar before? And should I replace those components before they go bad?

[Gyro]

As the bolt is Nylon, it wouldn't be able to put enough clamping force on the transistor to crack it like that. It looks as if it was careless handling when they were forming the leads, maybe held in a pair of pliers. I can't see if the crack extends over the top of the body or just scratches.

I suppose the transistor might fail long term, possibly taking out a few surrounding components (hopefully not the transformer winding), but it could just as easily last the life of the scope. At least, as it's socketed, it would be an easy replacement.

It looks as if it doesn't need very much heatsinking, but that thick silicone gasket is needed to cope with the high flyback voltage (watch your fingers in that area!)

[tpecar]

As the bolt is Nylon, it wouldn't be able to put enough clamping force on the transistor to crack it like that. It looks as if it was careless handling when they were forming the leads, maybe held in a pair of pliers. I can't see if the crack extends over the top of the body or just scratches.

Checked it out more closely, it seems that the whole top of the transistor package cracked.

At least, as it's socketed, it would be an easy replacement.

True, though it would probably require some convincing to get out, since the socket itself has melted.
BD681 seems to be a darlington transistor, however I don't know what to make of the Ph postfix.

It looks as if it doesn't need very much heatsinking, but that thick silicone gasket is needed to cope with the high flyback voltage (watch your fingers in that area!)

Hah, yeah, I know that much that poking into that section isn't recommended.

It does seem that it shouldn't heat up that much (and indeed it doesn't, as far as I've tested). I'm interested if anyone has any idea / explanation of what transient fault could cause it to misbehave like that.

[shakalnokturn]

A couple of ideas:

Check the 5 electrolytics shown in your schematic extract.

I'd bet on the cause being bad contacts on BD681's socket causing that to heat and inductor to suffer from the transients caused.

[Gyro]

Yes, I'd go for poor contact in the socket too, especially given the poor lead forming technique - your new photo appears to show a row of teeth marks. If the crack doesn't extend into the die/bond wire area then it's probably still ok, but if there is overheating in the socket, that sounds like a more pressing reason for replacement (maybe lose the socket completely and extend the leads slightly with flexible wire).

Yes, it's always checking electrolytics, but I think I'd put this one down carelessness in assembly. On the face of it, this seems to be the simplest explanation.


P.S. I suspect that the Ph! suffix on the schematic stands for Philips. I've attached a copy of the old Philips BD861 datasheet for reference. It would be best to see if any parameters vary from currently available parts - the only thing that I can see on a brief look is that the Philips datasheet shows Vcbo of 120V rather than 100V, but it would be best to compare parameter by parameter.

[tpecar]

Thanks to both of you, especially to Gyro for pointing out that the part was manufactured by Philips!

I had a look and 80's Philips parts had such markings + most other components in the scope seem to be sourced from them, so it makes sense.

As noted on the parameters, the Philips part does have a higher V_cbo (120 V instead of 100V of currently available BD681 parts), but given that other components are rated up to 50 V, it's probably within margin (correct me if I'm wrong).

Current parts (ON semi 681G) seem to have slightly different characteristics - Philips BD681 had |h_fe| (small signal current gain) > 10 (ON semi > 1), which might be due to smaller internal resistors (3k + 120, ON semi 8k + 120).
So they are probably not a drop-in replacement.

[Gyro]

Check out the ST  variant too, it has a rather more detailed datasheet (the ON looks to have been written as a generic second source). Although it has 15k + 100 internal resistors, the ST looks to have faster turn-on and turn-off times than the Philips.

It's not easy to know what parameter Hameg were selecting for - it may well have been speed as darlingtons are typically pretty slow. The ST, like the ON part specify minimum of 1 at 1MHz 1.5A. (I'm not sure if the >10 on the Philips was a typo, the 'typical' hfe vs frequency graph doesn't show anything like that good). It's hard to know the performance of the competitors were getting on for 40 years ago. I would expect a modern variant ought to be superior though.


P.S. It is the value of the second resistor that really determines the turn-off time of the main transistor. The circuit has direct control of the base of the signal (driver) transistor, so the first resistor has less effect.

[tpecar]

Ok, thank you Gyro for the insight, especially for the remark about the first resistor in the darlington pair, I should have known that :-/

Have to admit that I dove head-first into this project, but I guess that this is a great opportunity to finally delve into analog electronics.
I'll make the effort to understand the HV generation circuit, possibly do an LTspice sim, and figure out if the difference in characteristics affects the operation of the circuit.

I'll report back when I stumble over / figure something out.