[Summon Volt Nuts] E3615A repair

[opa627bm]

RAWR,

I have scored two Agilent E3615A from a local auction for 40$. One working fine but the other does not have output. After careful investigation,the reference IC was bad. I don't have the replacement chip in hand but I do have a LM399 from an old equipment. So I modified the surrounding circuit accordingly and here is what I got so far:

The top trace is from the working powersupply and the bottom is the one with LM399.
The oven is running at 38V, 67C.
The output is slowly drifting in uV range....

It seems that this part can work without the oven. Due to the part has extremely short leads (de-soldered from other board) I suspect it is heating up the surrounding component and causing the drift. Maybe I should disable the oven or running on lower voltage?

Any input will be appreciated !

Regards,
Li

[Zucca]

I don't have the replacement chip in hand but I do have a LM399 from an old equipment.

Nice score, but why not buy a new replacement chip?

[sync]

The oven is running at 38V, 67C.

You are running the heater between the regulated +15V and the unregulated negative bias supply. With the ripple on it and varying mains voltage it can maybe exceed the 40V absolute max. rating of the LM399. Not good. Use the regulated -12V instead.

btw: you can't measure the temperature of the LM399 this way.

The output is slowly drifting in uV range....

It seems that this part can work without the oven. Due to the part has extremely short leads (de-soldered from other board) I suspect it is heating up the surrounding component and causing the drift. Maybe I should disable the oven or running on lower voltage?

What do you expect? The power supply is full of parts which will drift with temperature and other things. Even without the heat of the LM399 it will drift.

[opa627bm]

I don't have the replacement chip in hand but I do have a LM399 from an old equipment.

Nice score, but why not buy a new replacement chip?

I couldn't wait 

[opa627bm]

The oven is running at 38V, 67C.

You are running the heater between the regulated +15V and the unregulated negative bias supply. With the ripple on it and varying mains voltage it can maybe exceed the 40V absolute max. rating of the LM399. Not good. Use the regulated -12V instead.

btw: you can't measure the temperature of the LM399 this way.

The output is slowly drifting in uV range....

It seems that this part can work without the oven. Due to the part has extremely short leads (de-soldered from other board) I suspect it is heating up the surrounding component and causing the drift. Maybe I should disable the oven or running on lower voltage?

What do you expect? The power supply is full of parts which will drift with temperature and other things. Even without the heat of the LM399 it will drift.

Do you know how to measure LM399 temperature properly?
It has drifted 2mv today

[TiN]

RAWR, Volt nut here, who calls?

You can just use single supply to heater, +15V regulated and ground on other side, unless you running negative voltage reference.
Also measure reference voltage, just to make sure you have it stable.

If you suspect thermal issues, you can hook LM399 on wires just to check if it's solves your drift issue. It can be just some other parts in PSU drifting due warmup.
Worth to have PSU cover/chassis installed, to allow everything reach thermally steady state.

[sync]

Do you know how to measure LM399 temperature properly?
It has drifted 2mv today

I don't know a good way to measure the temperature. A destructive way is to open the metal can of the LM399 and measure the bare silicon chip with an IR camera. But it shouldn't necessary to measure the temperature. Just check that the voltage of the LM399 zener itself is stable.

What did you measured exactly? It shows 10V. So there are other components involved. For example the resistors R104, R105 which are used to amplify the reference voltage. They are only specified with a temperature coefficient of 100ppm/°C. That's two orders of magnitude worse than the LM399.

[TiN]

Internal die temperature of LM399 is around 90°C. It cannot be measured directly, but can be estimated from power consumption of heater element.
Usually if it's works, you don't really need to know actual temp