HP 3325A function generator's High voltage option (option 002)

[lyonsk]

Hello guys,

  I got my hands on HP 3325A function generator with Option 001 (owenized oscilator) and Option 002 (High voltage output). It was sold as having power supply issues but is fully functional except of high voltage option. This option is simply 2x amplifier that amplyfies the output. It is separate board with own +/- 30V power supply. Disadvantage is that both owen and high voltage board is permanently turned on causing the caps to die very soon.

  Except for the dead caps, diode bridge and +15V stabilisator I found one dead transistor. It is Q6 which should be 2N3904 both according to what is marked on the part and the part list and should not be a problem to buy. Since the schematic schows that the amplifier is symetrical I asume that the complementary transistor Q7 must match the new transistor. I was not able to identify what standard type of transistor is Q7 based on the part number 1853-0020 but 2N3904 is typicaly matched with 2N3906. Is it possible that it is indeed 2N3906 but selected to tighter tolerances and therefore it has own part number?

So what now?
  1.) Should I try to find a match for original Q7 or replace both?
  2.) How should I compare the properties of the two complmentary transistors? hFE? Vbe?

[RobK_NL]

There is nothing in the manual to indicate that these two transistors are / should be matched, so I wouldn't worry too much about it.

[lyonsk]

Hello guys,

  I pulled some transistors out and measured their hFE using this circuit:
http://www.geofex.com/Article_Folders/ffselect.htm

You can check attached schematic where orange numbers represents hFE.

To me it looks like each side of the circuit should have the same overall hFE so basicaly hFE of Q6 should be equal as Q13...

I replaced the transistor with the closest one I had (hFE = 228) and the circuit works. Output is within tolerance when checked for DC voltage using generator's offset function. I don't have spectrum analyser. Maybe the offset could be a little bit better. Now it is -35mV but as one is supposed to use this option with amplitude over 10Vp-p I say it is good enough.

[Kleinstein]

hFE matching should not be critical in that circuit. It is more or less like a normal audio amplifier. The main part to set the offset is the input transistor pair (Q1). If at all it would be about Vbe matching in a way that  U(Q6) - U(Q4) = U(Q7) - U(Q5). However the setting is not that critical. If really needed one could adjust the offset (many options). i won't expect good DC precision from this circuit anyway.

Q6 and Q7 are just for setting auxiliary voltages and compensating some temperature drift. So they are not critical.