Tek 485 attenuator (FIXED) - advice wanted on substitute transistor

[Lorenzo_1]

My Tek 485 developed problems on both attenuator modules. Attenuator 2 would display only a flat line using 50 Ohm impedance. Attenuator 1 had an identical problem, but on the 1MOhm input.  The first module proved to have a simple microswitch problem. The AC/DC/Ground microswitch pair had something preventing closure - invisible to the naked eye but easily fixed with IPA and fine tissue.  Remarkable how little it takes to hold these switches open. The second unit had a shorted 2N2906 (Q1) feeding the  relay K1/K1S1 (see attached schematic - annotated in red). Q1 & K1, together with U80, switch input from the HiZ attenuator board to the 50 Ohm attenuator board and also handle input overloads.  I had a fairly close substitute transistor to hand (a BC557) and this brought the 1Meg display back to life immediately.  Seems to work fine, but wondering now if I can leave it in place or should substitute a 2N2906 in slower time? Not sure whether their slightly different specs are material here. Advice appreciated.

[SingedFingers]

bc557 is a pretty good substitute. I'd just leave it in there.

[David Hess]

Compared to a 2N2906, the BC557 has a much lower collector current which could be a problem.  A 2N4403, 2N2907, or BC327/BC328 would be more suitable.

If you measure the resistance of the relay's coil, then we can know.

[Lorenzo_1]

Thanks for the quick replies.  I haven't tried to measure coil resistance as I wasn't sure whether the relay is a conventional type or some sort of solid state device. However, It's described in the manual as 'relay, armature,  spdt, 21vdc, 450mW, 3.5A, 30V dc'. So it sure looks to be a conventional coil and contact design, albeit pretty small. These attenuators are so finely made and the switches so touchy that I've been very cautious fiddling with them more than necessary. Will have a go at measuring coil R.

[Lorenzo_1]

 K1 relay reads 1.2kOhm. Voltage drop across K1 is 14.4V when Q1 active and should be drawing about 12mA by my reckoning.  15V supply to Q1 reads ~14.8V. Not sure why Tek used the 2N2906 here - seems to have higher current capacity than needed. How important is it to have the same transistor in both attenuators? If it's sensible to match them, I'm inclined to replace the BC557 with the original spec item - there look to be some 2N2906s in TO18 packages available cheaply online.

[SingedFingers]

Tek used all sorts of random transistors from what I can tell, sometimes I think to showboat a bit or use up stock. The only important ones are in the signal path. If it's switching and 12mA just about any transistor will do the job if the beta isn't horrible. Best thing to check is switch it on at the base with a resistor and a bench supply, leave it a couple of minutes and feel it. If it's running hot then use a bigger one.

TBH transistor selection became a non issue really in the mid 1980s. Before that, characteristics were somewhat more spread and less reliable. Bar UHF+ you can probably get away with 5-6 different transistor types for most things.

[David Hess]

21 volts at 450mW is 21.4mA and 980 ohms so even at full voltage and 980 ohms instead of 1.2 kohms, the 100mA collector current rating of a BC557 should be sufficient and there is nothing wrong with the current gain of a BC557 at that collector current.

You could put a better transistor in like one of the ones I suggested but it is probably better to leave it alone at this point instead of risking further damage.  There is no need to match these transistors at all.

[Lorenzo_1]

Thanks for the advice. The scope is working really well - I gave it a good workout. I'll leave well alone and hope it continues to behave.