I damaged my Agilent 6642A

[Adrianoo]

Hi guys,

I have been using my Agilent 6642A for many years with no issues, until one fateful day when I accidentally connected a Lipo battery charger (input side) to the sense circuit instead of the power output of the Agilent unit. Please note, the damage was caused when the output was turned on and no voltage from an external source was being applied to the sense circuit.

I heard some snap crackle and popping going on inside the unit and then it blew the 3A 240V input fuse. I put a new fuse in and ran the unit again with the lid off and I could see 2 resistors glowing red hot before the input fuse blew again.

I have disassembled the unit and identified the two resistors which were red hot. They are R216 and R326

You can see the schematics on pages 158, 160 and 162 of the pdf document below:
http://ridl.cfd.rit.edu/products/manuals/Agilent/power%20supplies/CD1/Service/654xAser.pdf

My question to you electrical genius' is, other than the resistors, what else should I replace? I'm assuming something was damaged and that component is dumping huge amount of current through the resistors.

I know the basics of electronics and I've had a look through the schematics but I can't make heads or tails of it.

If you had to make a bet, which components do you think would have been damaged?

TIA

[Mechatrommer]

Resistors are tough nut, they usually fail last and take whatever applied to them even if they have to glow red and in the end disintegrated. And usually the short is not caused by them, whats downstream is, i bet some power transistors or mosfets.

[Adrianoo]

So would that be component U202, Q208 and XQ207 for the left side and U301, Q304 and XQ303 for the right side?

Visually all of those components look fine. Is there a way to test them individually?

[Dr. Frank]

Hello,
at first I doubt that you connected your LiPO charger to the Sense+ / Sense- lines (only), neither to the ext. current sensing lines.
That would never create such a heavy current overload somewhere, because all sense lines go into some kOhm resistors.

Either you have connected the charger directly to the + and - output jacks, or there was a parallel wire connection between Sense+/ out+ and Sense-/out-.

Then I assume that the charger provided much more than 20V output, maybe on the order of 30..40V or so, please check. What's the current capability of your charger?
Why did you connect another power supply to your 6642A, anyhow? Have you done that before? Extremely dangerous.

The external and excess voltage to the output probably caused the Over-Voltage protection to fire the SCR CR700, which shorted the output, i.e. created a huge current draw from your charger through the SCR, which probably fused into a short.
Then I suppose that your overvoltage may have damaged the negative discharge circuit  Q205, made it a short; maybe Q206, U202 are damaged as well.

Very probably Q207 with its steering circuit Q208, U202 and Q303 with Q304, U301 are damaged. At least Q303  and Q207 might be toasted, so that you have a permanent current flow from +Rail to -OUT (via SCR) or to -Rail (via Q205).

All other 5 power output transistor might have been damaged as well, but to open, instead of short.

As R216 is red glowing, probably R228 is burnt (open), as well R306, if R326 is glowing red.

Well, there's no other way than to de-solder all eight power transistors Q201, 203, 205, 207, 301, 303, 305, 307, write their designators on the transistors, and to measure their BE and BC diodes.
The SCR can be measured in circuit afterwards, whether it's shorted.
When you have found the damaged part you have to check their driving circuitry, i.e. mostly the FETs. Maybe you'll find out that all power transistors are fine, then it could as well be their conjunct FETs which are damaged only.
Then check all power resistor pairs, e.g. R216, R228, whether they are fine, i.e. whether you still measure about 0.125 Ohm in parallel. I hope you can do 4W Ohm measurements. Probably it's necessary to at least replace both the  pairs, which were glowing red.

In the end, also check the 8 rectifier diodes D401-404 and 411-414.

Frank       

[Mechatrommer]

So would that be component U202, Q208 and XQ207 for the left side and U301, Q304 and XQ303 for the right side?

Visually all of those components look fine. Is there a way to test them individually?

check:

R216 path
XQ207, XQ205, Q208, Q206
all its parallel equivalent (XQ203, Q204 etc)

R326 path
XQ303, Q304
all its parallel equivalent (XQ305, Q306 etc)

remove from circuit and test with transistor tester, or else you can check continuity or diode mode with multimeter between pins if any short indicates bad components. marked U (opamp) is less likely damaged but it wont hurt checking just to make sure. but check the above critical path first.

while at it, you can check further upstream to +rail to see whats going on there.. and -rail too... if you gathered the parts that are damaged and figure out the total cost is more than $100 (if you are lucky they are not unobtanium), i suggest you get parts from this supplier and dont make the same mistake again

[bdunham7]

Have you checked fuses F205-207 and F304-307?   

What does the unit do now when you turn it on with no loads attached?

I'm not sure what happened, but your description of the events doesn't make a lot of sense given the damage.  You might as well start assuming that 'something' happened and test everything.  If the fuses are all good, then you need to inspect/test all of those 3W 0R25 output resistors and their related transistors.  You may need to either be quite clever in your testing or be ready to do some unsoldering--or both.

[Adrianoo]

UPDATE: So I desoldered half the power transistors off the board (left side) and tested them all. None are shorted and diodes seem to be working as intended. All 4 of the power transistors have the same Ohms reading across the pins.

So I moved on to testing the FET's and it seems like 2 of them are not consistent with the results I am getting when measuring the rest of them. These are FET Q206 & Q302. I am testing these while still soldered to the board... is this ok?

I want to order the new FET's but it seems like the original FET is no longer in stock. I can see that a newer version which is labeled with a ending L (i.e. MTP3055VL)... can I use this FET instead? The main difference I can see is that it has a much slower rise and fall time and slightly higher input capacitance... See differences here:
https://www.apogeeweb.net/compare/mtp3055v-vs-mtp3055v.html

I also checked all fuses and they all seem fine.

I'm not sure how to test the opamps... so I guess I'll leave them alone and after replacing the FETs if it still doesn't work, I can try replacing the opamps 

[Mechatrommer]

looking at schematics (attached), i suspect it will be quite confusing to test resistance between mosfet's pins in circuit. for example reading Q306's pin 1 and 3 will read 0 ohm due to R331, so we cant tell if that pins are ok or not. i'm not sure why there are some 0 ohm connection between gate and source pin of mosfet. XQ305's pin 1 and 3 will read 215 ohm due to R313. XQ305's pin 1 and 2 should read high resistance or forward diode (iirc, or reversed). pin 2 and 3 should read high resistance too, if i dont miss anything. should be the same high resistance reading on Q306's pin 1 to 2 and pin 2 to 3... ymmv..

for opamp, quick start test is check no short between every pins, opamp pins between output, +ve, and -ve (inverting) input may read some resistance but should be not so low. the sure fire test is to desolder them, put them on test circuit and run them in say voltage follower (gain 1) config, feed signal generator at +ve (non-inverting) input and output should follow the signal seen in oscilloscope. but that will be a lot of work depending on your guts. ymmv.

it is also possible all the power drive transistors and mosfets are ok, the damage is in control circuit, then thats a whole lot of can of worm to test.

[Adrianoo]

UPDATE: So I ordered all new Mosfets and sense resistors and replaced them all this afternoon.

I have a new issue, so it seems. Previously as soon as I plug the powersupply in, it will pop the 3A input fuse. This no longer happens and I can safely run the unit without it popping the fuse or making the sense resistors glow red hot.

My new issue is that when I turn the unit on, it is outputting 20V DC continuously. If I turn the output on, the cooling fan turns on to 100% speed and the output drops to 15V DC.

Does anyone have any ideas where I should look next?

I noticed a few random components are getting slightly warm to the touch. I'm not sure if these are problem parts or if they get warm under normal operation.

Any help would be appreciated!

[Swake]

No error messages on the screen? I'm surprised it passes power-on self test with the symptoms you describe.

Follow the troubleshooting flow page 37 in the manual you have linked. We'll see from there.

[buta]

Not sure have you checked the capacitors before and after the rectifiers (diode or bridge) e.g. D402 and D403.
One of our 20 years old HP spectrum analysers was broken this month because of the aging capacitors.