EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: xenocide702 on September 30, 2013, 03:35:40 pm
-
Link to manual (http://cp.literature.agilent.com/litweb/pdf/5957-6360.pdf)
I was wondering if anyone knew of a way to operate the supply in negative constant current mode. The manual claims it's only programmable from 0.02 - 5.1188 A, but notes on pg 12 that "The negative current tracks the programmed value with a nominal offset as shown..." and claims the offset for my supply is 250 mA. This appears to be accurate, it does indeed track the Iset value, except that the offset is measured as being 237 mA, not 250 mA even.
Does anyone know if this "nominal" offset is constant with time, temp, voltage, current, etc. ? Is anyone familiar enough with this supply to say where that figure comes from? I'll be doing some more experimenting and will report back what I find.
Thanks for any input!
-
Reverse the polarity of the output.
-
That doesn't accomplish the same thing though does it? That'd be putting a negative voltage on the output terminals, which isn't supported by the supply by my understanding. It's not a 4 quadrant supply, just a sink/source. My understanding was that it'll tolerate output voltages lower than the set voltage and regulate the current accordingly, but it will not tolerate negative voltages across the output terminals.
In my use case I need to discharge a 12 V AGM lead acid battery. The plan was to set the voltage to 10.5~11 V and current limit to -1 A (or more likely 0.1 C of the battery), and run the battery down until it hit that cutoff voltage of 10.5~11 V.
-
Your plan sounds about right, although you don't actually set -1A, but +1A. I've done exactly this with the 6632B (which I don't think has the small current offset when operating in quadrant 2).
The thing to watch is the internal crowbar SCR. If this fires for whatever reason, it will short out your lead-acid battery. They recommend a series diode to prevent this when charging batteries. Of course, discharging the battery prevents using this option. An external fuse would probably suffice.
-
Telling what the goal is helps a lot. I also recommend the fuse. I've seen what a battery can do to a similar power supply when the crowbar kicks in.
-
Out of curiosity, what did happen to that supply when the crowbar SCR kicked in?
I suspect I may have blowed up my supply :(
I wrote a quick program to start discharging the battery, but through some bug in my serial routine the OVP kicked on with that big honkin unfused lead acid battery across the terminals :palm: |O No smoke, no noise, but after removing the load and power cycling I get zip on the output terminals. It indicates that it's in constant current mode. Vset is set to 0.001 volts, but the current indicator reads -0.2561A. If I change the voltage setting to anything non-zero the current reading drops to -0.0024 A and the "UNR" unregulated indicator on the display shows up. If I don't touch the voltage, but increase the current limit it shows whatever current limit I put in + 0.256 A.
I'm going through the service manual now.
Link to service manual (http://cp.literature.agilent.com/litweb/pdf/5957-6365.pdf)
On page 75 the decision tree "Vb (Q113) > -2V" I'm measuring -1 V on that pin. I'll have some more time to put into this hopefully tonight.
Should I make a separate topic for discussing the repair?
Thanks!
-
Out of curiosity, what did happen to that supply when the crowbar SCR kicked in?
I suspect I may have blowed up my supply :(
Uh Oh... In my PSU several PCB traces got burned of the PCB. After repairing the traces the PSU worked fine ever since.
-
Assuming the 6632A is the same as the B, reversed polarity across the terminals isn't supported. You'll find there's a protection diode in place, so if you do connect a (current limited) supply to the terminals the wrong way round, you get the characteristic -0.7V but that's it. It's not a useful, supported mode of operation.
You're quite right about how to discharge a battery, though: set the voltage to be whatever you want the final battery voltage to be, and the current limit to the discharge rate. The PSU will sink that amount of current until the voltage falls to the set point, at which point it'll switch to CV mode and sink whatever current is needed to maintain that voltage. Eventually it'll fall to zero.
Just don't try and switch the output off with a battery connected or the SCR will... oh :(
-
Lol, yep. It looks like it may have just taken out the main output fuse (F2). I'm poking around the supply anyway though, great excuse to learn a little bit about it. I'll jumper the fuse to test it when I get the board back in.
I've cut jumpers W4 and W3, which is supposed to disable the crowbar SCR function, which (if I'm reading the manual right) will just allow the output to float when switched off.
EDIT: Yep, sure enough, I jumpered the fuse connection with 30 AWG wire wrap wire and the supply fired right back up. New fuse is on order. Usually when I do something stupid like this it's a much harder fix :P
-
Hi
You have blown the fuse of the crowbar as you have already discovered.
Be careful that the crowbar can fire also if you have a battery connected to the pwr supply when you switch it on, it happened to me.
You can eliminate the offset in sinking current reading back the current from the 6632 and sending back the corrected value like : 1A wanted current ; 1,250 readback ; ISET (1-0.250)
After that the current is stable in time and temp.
-
Just don't try and switch the output off with a battery connected or the SCR will... oh :(
AndyC,
May I ask you why that is?
Why would Agilent design it this way?
I have a E3632A and changed a battery, when the mains breaker went out for a completely different reason.
I reset the breaker and the E3632A started to smoke. (12V Battery was still hooked up)
When I opened up the Power Supply, one big track had burned off at the bottom.
Then I repaired the broken trace with a cable and the PowerSupply works perfectly, at least it seems
Is there anything else that is might be broken, that I am not aware of?
Thanks for any input on this one.
-
Hi
The HP 6632A is designed to be a two quadrant power supply. It has a positive output voltage, but it has the ability to source current (normal power supply behaviour) and sink current (unusual).
The negative current limit is 0.25A higher than the positive current limit. So if you program the output current to 1A maximum and connect a power supply that is higher voltage than the HP 6632A it will sink up to 1.25A. That is it behaves like an electronic load.
The HP 6632A (and B) are unusual in the sink / source capability is more or less symmetrical. Other HP power supplies can sink current, but the current is typically limited to around 1/8th of the sourcing capability. They call this a down programmer.
Why did they do this?
To speed up test time when adjusting to a lower voltage. A shorter test time means more test through put.
It is not meant to be used as an electronic load.
OVP
(https://www.eevblog.com/forum/testgear/hp-6632a-negative-constant-current/?action=dlattach;attach=84505;image)
You do not want the OVP circuit to trip if you are connected to a battery. Here is the schematic from the manual. The fuse will blow.
Jay_Diddy_
-
The 663xB works perfectly well as an electronic load, I'm not aware of any reason to avoid using it as such. I've found it particularly useful as a battery simulator, in products which are designed to be powered from a rechargeable battery pack.
There aren't too many bench PSUs around which can supply power to a product, and then switch into a mode where they can load up that product's battery charger circuit and show its voltage and charging current.
The current limits for sourcing vs sinking are the same on the 'B' series.
This ability is nothing to do with the crowbar circuit, which is there primarily as an overvoltage protection device, and I believe it's also used to rapidly bring the output voltage down to zero when the PSU is switched off. It's that action which causes problems with batteries.
-
The 663xB works perfectly well as an electronic load, I'm not aware of any reason to avoid using it as such. I've found it particularly useful as a battery simulator, in products which are designed to be powered from a rechargeable battery pack.
There aren't too many bench PSUs around which can supply power to a product, and then switch into a mode where they can load up that product's battery charger circuit and show its voltage and charging current.
The current limits for sourcing vs sinking are the same on the 'B' series.
This ability is nothing to do with the crowbar circuit, which is there primarily as an overvoltage protection device, and I believe it's also used to rapidly bring the output voltage down to zero when the PSU is switched off. It's that action which causes problems with batteries.
So how do I hook it up, if I want to use it as an electronic load?
Let's say I have another supply that I want to stress test a bit. I would like to load that supply with 5V at 100mA for an extended time.
Should be possible using the HP 6632A, if I understand things correctly? How do I connect the two supplies?
-
Assuming the 'A' works the same as the 'B'...
Connect the two supplies together, '+' to '+' and '-' to '-' as you'd expect.
Set the 6632's current limit to 100mA, and its voltage to zero.
Switch on, and it'll draw 100mA from the other supply, regardless of that supply's output voltage. (Note, you'll need to ensure that the other supply doesn't ever exceed 20V or you'll trip the OVP protection).
If you don't want the 6632 to be able to drag the other supply's voltage all the way down to zero, then set its voltage accordingly. For example, I sometimes use a 6632B to perform discharge testing on battery packs, so I set the current limit to be the desired discharge rate, and the voltage to be the final voltage I want to see on the battery when it's 'empty'. The battery then gets discharged at constant current, then maintained at constant voltage rather than being over-discharged.
-
Ok, pretty much as expected. Thanks for the quick and clear explanation though!
Edit:
I verified that it works the same way also for the HP 6632A.
-
I know this is a very old thread, but I have a couple of questions about Andy's steps (2 posts up) for using the 6632b as a load with a psu connected to it...
I've got a psu connected to my 6632b, the psu is set to deliver 1.1A @ 1V. The 6632b is set to 1A @0.5V... my cables are short and the sense lines are not connected. I enable the psu output and then the 6632b output and the 6632b sinks 1A as expected, but I just want to make sure I'm not making any procedural mistakes.
1. what, if anything, do you do with the sense wires on the psu and/or the 6632b?
2. do you have to turn on the psu first, before enabling the 6632b output? (and will you eventually blow anything up if you do it in the wrong order?)
3. does the order that the outputs are disabled matter? the above crowbar/burned traces/blown fuse discussion (with a battery connected) has me a little nervous
Thanks!
-
The sense wires on the 6632B should be connected to its own outputs, either at the connected device (ie. the 6632B's load) or, if you don't need to compensate for the voltage drop in the leads, directly at the screw terminals.
Without knowing anything about your PSU I can only guess what you'd do with its own sense leads. Normally joining them to the PSU output would be a correct option.
When the 6632B output is switched off, it presents a high impedance and can be safely driven externally up to 20V. Its display will show the voltage correctly, and the current will be close to zero.
The issue comes at the time you turn the 6632B output off. This activates the crowbar circuit in order to bring its output down to 0V rapidly, and in doing so, will sink considerably more current from the terminals. If your external PSU is current limited, it'll detect a short term overload from which it should recover, and no harm is done.
A lead-acid battery, of course, is not current limited, and it'll blow the SCR to bits.
I believe there's an internal modification which can be done to the 6632B to disable the SCR if it's undesirable - you may want to check the manual for this.
-
Andy,
Thanks so much for your explanations. The crowbar bit makes complete sense now, and the point about the 6632b presenting a high impedence should have been obvious because I've been using it in "meter" mode with the output off :palm:
When using the 6632b as a power supply, I get that I should connect the sense wires to whatever the power supply is driving - to compensate for the voltage drop from the 6632b to the load.
But I might still be confused about sense wires when the 6632 IS the load, and sinking current from my psu (a 6641a)... I should connect the 6641a sense wires to the 6632b outputs (so the 6641a compensates for the voltage drop from it's output to the load). But since the 6632b is connected directly to the 6641a - no battery or anything in the middle - I should connect the 6632b sense wires to its own outputs (and not the 6641a outputs) because there's no voltage drop for the 6632 to compensate for?