Agilent E3610 Regulation Performance

[ian.ameline]

I recently purchased (at a pretty good price) An Agilent E3610A -- it is clean, appears to be in good shape - no evidence of any repairs having been done.
It was manufactured in 2000.

Everything appears to work -- it delivers the right voltage and current, even with differing loads. The noise, ripple and drift are all well within tolerance.

But it appears that the recovery of voltage regulation is somewhat slow. Allow me to describe my improvised test;

Load = tail light bulb @15V, 0.61A

When I go from no load to the tail bulb, The voltage drops from 15 to 5 volts, and takes almost a second to fully recover to 15v



Is this normal? If not, where to start with the E3610A -- inside it's clean as a whistle -- no evidence of any leaking caps etc...

Thanks!

(I tried to post / attach a sceen grab from my scope of the recovery -- not sure how to add/upload images here.)

[tombi]

I don't know about this supply in particular but could it be that there is a pre-regulator that is keeping the voltage at a bit above the output.

I am thinking that when it goes into current limit the pre-regulator drops the voltage to just above the output voltage (5V) but then when it comes out of limit it takes time to recover.

So it might be normal.

Tom

[ian.ameline]

I don't know about this supply in particular but could it be that there is a pre-regulator that is keeping the voltage at a bit above the output.

I am thinking that when it goes into current limit the pre-regulator drops the voltage to just above the output voltage (5V) but then when it comes out of limit it takes time to recover.

So it might be normal.

Tom

According to the schematics, the 3610 does not have a pre-regulator. I'll probe around the voltage error op amp tomorrow and see if I find anything unusual. I'll also check all the test point voltages, although I expect those will be fine -- the unit is mostly working - just that slow recovery. Going from .61A to 0 has almost no blips at all on the scope. It's the othet way - 0 to .6A at 15V that takes a second to recover fully.

[floobydust]

I say this is normal, an incandescent light bulb has inrush current 10x until it warms up. Try a resistive load instead.

[tombi]

Ahh sorry - no I had that the wrong way round.

I initially thought you mean that while under load it dropped to 5V and then when the load was taken away it took a while to recover.

Tom

[ian.ameline]

I say this is normal, an incandescent light bulb has inrush current 10x until it warms up. Try a resistive load instead.

That sounds reasonable -- but the supply does not appear to go into CC mode during that brief interval.

I notice from the schematic that there is a 470uf (50V) (C3 on the schematic) capacitor across the outputs (along with some resistors) A fluke MM testing capacitance across the outputs (unit unpowered) shows 0. But I suspect the resistors across the inputs are messing with that reading. I'll try putting a 470uf across the binding posts and see if there is any change.  Perhaps it has failed. (It is a nichicon PW series - rated for 5000 hours at 105c)

[ian.ameline]

I say this is normal, an incandescent light bulb has inrush current 10x until it warms up. Try a resistive load instead.

Winning answer -- scope on the output of the current error amp shows the unit is going into CC mode for much of the duration of that recovery. If observed carefully, the CC LED briefly lights too.