Advice needed on replacing power transistors HP 467A

[PM3295]

I need to replace some of the power transistors in this old HP amp/power supply.
The problem is that they are germanium power T0-3 parts.

What precautions or modification will I need to make if I replace them with something like a 2N2955?

All help appreciated.

[vern]

There are still TO3 germanium transistors around, like 2N1559.
You could also just try and put silicon transistors in, the output voltage might be a little different because of the different base - emitter voltages (silicon about 0.7 and germanium 0.3V) and because this circuit has no regulated feedback, it's just straightforward zener diode as reference and transistor for regulation. Same for the current limit adjust.
But if you can live with that it will work, 2N2955 is a good choice.

[PM3295]

I wonder if I should just throw out the old regulation scheme and put a modern regulator in for both rails? The maximum current draw per rail is less than one amp. It will destroy the authenticity of this classic product of course.

[vern]

If you want to keep the authenticity then you have to use germanium transistors of course.
If not, you could put in modern regulators like the LM337. It's not TO3, but TO220 which you can easily mount on the heatsink.
Should work fine!

[metal]

you might need to play with CR135 on Q6 base

[titi]

Hello,
the use of the LM337 is not possible because the input voltage is 50v and the maximum for LM337 is only 40v.

If you don't care about current limit adj.

As the 2 parts (Positive and Negative Power Supply) has they own transformer, it is possible to use the LM317HV, to do 2 insulated positive Power supply and connect the positive from the first one to the common (ground) of the second one (Keep only the 2 transformers and insulate the 2 common of the transformers (pin 5 J2 up) (pin 2 J2 down)).
You obtain a symmetric power supply with 2 Positive Power Supply.

Why to use LM317HV ? because this model accept up to 60v in input and in differential and it can give 1.5A.

I think the use of TO-3 is important because if you have 50v in Input and 34v in Output with 1A (as the schematic show) you will have 16W to dissipate, so not easy with a TO-220 to dissipate this power without over-heating.

This can be a solution is the 2N2955 give not a good results.

Best regards.

[vern]

titi,
the LM317HV is a positive regulator, to use it in this application you would have to change a lot of the circuitry.
You can still use the LM337 in the circuit, the 40V rating is for input - output difference, not for absolute input voltage.
There are only three voltages at the LM337 in this case: input -50V, output -30V, adjust -28.75V.
The input - output difference is only 20V, which is ok.
However you have to take precautions for fault cases where the voltage can exceed the 40V limit, i.e when turning on the system, because the output caps are zero V and the input is 50V - not good. The same problem with a short, 50V in and 0V out.
A power zener diode with about 35V from input to output helps in these cases. Since the LM337 has internal temperature shutdown it won't blow up if you overload it.

[metal]

Look at page 9,10

http://www.ti.com/lit/ds/symlink/tl783.pdf

Also this is interesting: http://www.diyaudio.com/forums/power-supplies/225737-dx-supply-output-adjustable-stabilized-power-supply.html

[OscarH]

Hi,
I have done repairs in the past - my HP time - on similar power-supplies using Ge transistors.
467A was introduced in Nov 1964. Germanium Transistors were used as commonly available, there were no technical reason other than availability. 
Replacing with Si transistors is working fine, and the drop in voltage between Ge & Si is usually within power-supply output specifications, who at that time, were not so accurate than today.
If you want to have exact same output voltage, as Metal said, you can play with CR135 and CR143, but this is again probably not needed.
OH

[PM3295]

I appreciate all the good suggestions. Will the supply run cooler with silicon devices and have better regulation?

[metal]

Not necessarily true..

[vern]

Will the supply run cooler with silicon devices and have better regulation?

There would only be a tiny difference because the output voltage would be slightly lower, regulation will be the about the same. Might be a little better, depends on the transistors.
If their current gain is higher than the germanium types it could be a little better.

[flyback]

Hi PM3295
in my opinion, SI transistors are okay, but Q6 & Q8 need some more tweaking, else the current limit functionality will not work. (see attached schematic)
You may try to add 2 germanium or Schottky diodes for an additional voltage drop of around 0.4V .. 0.5V
Just 1 additional silicium diode may also work, but I'm not confident that the adjustment range of the potentiometer will be enough.
Good luck.
fb

[zab]

There is no problem in replacing these with silicon type as the circuit is emitter follower or common collector used for current amplification. The output voltages will be little bit less but no issue in operation.