Agilent 3631a

[Dandoni]

Hey there,

I am new to this forum and have been reading a few threads to get an idea of what equipment to buy. I am currently studying mechanical engineering at uni but after taking a couple electrical units I have taken quite a liking to the field.

I have found a used Agilent E3631a for around 1200 oz dollars. I cant seem to find a whole lot of info on this power supply. There seems to be a lot of talk on the Rigol dp832 or recommendations to get an older HP. For general messing around( vague, I know) , i have no idea on what specs I really need. I dont like buying things twice so i dont mind spending a little more on equipment that lasts and I dont mind buying used. Also recommendations for other test equipment for a decent setup. Help/advice is appreciated
Cheers

[Micke]

I have two of these, and I can really recommend it! Top quality, a power supply you can rely on.
I upgraded the firmware to v2.1 (EPROM 27C2001) in both supplies to fix the bugs with transients during some conditions  and not fully disconnected outputs.
See service notes https://servicenotes.literature.keysight.com/litapp/SearchSN.do?method=openExternalSNSearch&prodNum=E3631A
The only negative is the fan noise, and you can get unlucky getting a unit with dim VFD.
Very common fault on these supplies is that the rotary encoder gets bad, but easily fixed... I bought new ones cheap from Germany http://www.ebay.com/itm/10x-Inkremental-Encoder-12fach-ALPS-EC16-ohne-Rastung-/302025367910?hash=item46521d6166%3Ag%3A3KIAAOSwrx5UYhWb&nma=true&si=rqbe9mk7RPJ8Ci7jJKpmqax17kM%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557
This type of encoder is common in other HPAK equipment as well, so nice to have these in stock! 

[R005T3r]

Well,

something like the E3631a it's a very good unit. The only downside is that it's expensive and it has only 1A on the +/-25V output, and in my opinion is a bit strict.

If you are at the beginning of the journey in electronics, you may  want something less higher end (there's still time for improvmenent and for application specific ones), because you will find yourself with a lot of things to buy... And a big mistake is pretending to want everything and now: you will break the bank for sure!

Be sure to buy a decent brand: equipment have to be reliable, supported over time and trusty, othewise it's a waste of money (and time).

[Jester]

I keep an eye on ebay and the prices for the Agilent 6632, and 6633 seem like a relative bargain compared to most other quality power supplies, I prefer the B versions.  I picked up a 6633B 50V, 2A for $50 + shipping in non-working condition and fixed it for < 1$.  That being said there are plenty of fully working ones for < $200

I also like the Agilent E3615 (20V, 3A) and E3616 (35V, 1.7A)  for about the same price

[TAMHAN]

If you can live without the pots for turning, the HP 6624A also is worthy of attention. I love mine to bits - you see it, for example, in this video toward the end:
https://youtu.be/4zICevDOhr8?t=407

Plus, it also has GPIB. So you can automatize many tests...

[Dandoni]

Thanks all, excuse the short reply as I am in class, but opinions on 2x 6632b found a couple in good condition or 1x e3631a for capability and reliability ?
Cheers

[Dandoni]

Thanks everyone. Purchased a couple of 6632b's

http://www.ebay.com/itm/Used-HP-Agilent-6632B-100-Watt-System-Power-Supply-20V-5A-Fully-Tested-/112002830083?hash=item1a13e38f03:g:a8IAAOSwrnNXPs93


 . Hope they work ok from what i have read and suggestions here they seems to get a great rap. And it worked out quite a bit less than the local 3631a. There doesnt seem to be much used equipment floating around in australia that i could find online.

Now just got to learn how to use them(safely)! Now what else to buy?

[TAMHAN]

Definitely not a bad purchase. The biggest problem you might face is that these units can, and will, sink current - I fried a process computer like that once.

Other than that, the low voltage means that there should not be too much which can go wrong.

[Jester]

Agilent offered an option for front panel terminals. Most units don't have the front panel terminals fitted from the factory, however the front bezel and chassis have the cutouts so it is easy to retrofit them.

This power supply also has remote sense terminals that are handy when accuracy is important.  I seem to recall a thread (video I think) about adding the terminals, I did it in a slightly different way, see blue arrow below.

[Dandoni]

What is exactly meant by sinking current, isn't that a feature of the unit?

Also have started reading the manual, but what are the definite do's and don'ts for operating a power supply like this? Is there anything I should do first before operating the unit?

 It also seems like a good idea to run the posts on the front panel.
Thanks again for the replies

Dan

[TAMHAN]

Hello,
well, in my case, this happened. I powered an OPiPC with the 6624A, and then turned the voltage to 0 to turn the OpiPC off. Being stupid, I connected the normal power supply while the 6624A was still active with 0V and 1A. So, the supply happily started to drain 1A, which heated up the PCB of the OPiPC like shit...

[free_electron]

for the 662x power supplies : known defects are the 100nf capacitor on the front panel ( display board ). they short out.

[R005T3r]

Thanks everyone. Purchased a couple of 6632b's

http://www.ebay.com/itm/Used-HP-Agilent-6632B-100-Watt-System-Power-Supply-20V-5A-Fully-Tested-/112002830083?hash=item1a13e38f03:g:a8IAAOSwrnNXPs93


 . Hope they work ok from what i have read and suggestions here they seems to get a great rap. And it worked out quite a bit less than the local 3631a. There doesnt seem to be much used equipment floating around in australia that i could find online.

Now just got to learn how to use them(safely)! Now what else to buy?

  What else to buy? There's plenty of other things to consider:
A good oscilloscope(digital, >100Mhz), multimeter (take a dual display, bench one  it will be very useful for most of your needs), RLC tester(be sure to buy an RLC not only C ones even if it costs more), soldering station,  proper solder, microscope, function generator, DC load (it's very handy to have a programmable one, because you can have a variety of loads for testing anything in between), variac for regulating AC voltage(fundamental for homemade power supply), programmable power meter,and don't forget the for fundamental items: passive probes, current probes ( quite expensive), differental probes (for testing non ground referenced circuits) and the digital probe for your scope. Active probes are simply too expensive to consider.

But, immediately, my advice would be investing in some cables: if you don't have cables, how are you supposed to connect things up? Also, take some hooks, they are very handy for grabbing all sorts of pins.

Then, in case you want to do ham radio/signaling/communication stuff, there are also:
spectrum analyzers, signal generator, attenuators, signal amplifiers (to name a few).

For used equipment, remember that the instruments needs to be calibrated from time to time, or you won't be able to test things correctly, and it's something you can't do it yourself in some cases...

[chickenHeadKnob]

What is exactly meant by sinking current, isn't that a feature of the unit?

Also have started reading the manual, but what are the definite do's and don'ts for operating a power supply like this? Is there anything I should do first before operating the unit?

 It also seems like a good idea to run the posts on the front panel.
Thanks again for the replies

Dan

Yes!! before powering on open up your treasures and verify that the transformer primary side wiring is compatible with your mains standard ! Now these appear to be coming out of South Korea 220V@60 Hz    so you probably won't have to change anything but double check anyways. The wiring change is documented in the HP manual and is a bit persnickety compared to just flipping a switch on the back panel as you have pull spade terminated wires and switch them to the diagram.

An option to the front panel output jack mod is to put two (or 4 if you want to extend sense lines) jacks in a small project box and wire with 14 Gauge or similar silicone multi strand  to the back output terminals. That output box can then sit next to your project or be mounted anywhere you want.

Extending the sense lines requires some caution has it increases the propensity of the power supply control loop to oscillate.
I have a 6632B and like it a lot.

[Dandoni]

I spent the night reading the 6632b user manual and I will jumper it  from 220vac to 230/240.

I have quite a bit to gather. I pulled out my old electronics box as i used to work in a electronics manufacturing place here in Perth Western Australia some time ago. I was involved in assembly and rework, but never testing, so have very limited experience with test equipment.

In the box I uncovered a few rolls of multicore solder, solder wick, flux, cleaning pens , hand tools and some other bits and bobs.

i have starting filling the cart at digikey. I have a black and red roll of silicone test lead wire and 10 red and black alligator clips to make up connectors. Any recommendation on ponoma banana plugs and other connectors?

Now on to the output from the supply to binding posts  . Is this sufficient? One in red and one in black
http://www.digikey.com.au/product-detail/en/general-cable-carol-brand/76812.R8.01/C76812B-50-ND/5452557

If someone could also explain the sense connectors that would be helpful. From what i have read you leave s+ connected to +, and s- connected to -, and that senses the voltage at the source as oppose to the load, but if you want a more accurate voltage supply/measurement at the load you would disconnect the jumpers then connect another set of test leads?( is it just removing the voltage drop in the test leads?So would you have 4 leads connected to the load , sense leads being a twisted pair of what guage?  Also is says thats you cant sense at the front binding posts. Whats about the earth wire?

[Jester]

That wire will be fine.

This style Banana jack will fit in the front panel:  http://www.digikey.com/product-search/en/connectors-interconnects/banana-and-tip-connectors-jacks-plugs/1441927?k=%20%09%2072930-2%20

Either one red, one black if you parallel the sense and output terminals or 2 and 2 if you want the ability to sense at the load.


These items would be near the top of my list:
Banana plugs:
http://www.ebay.com/sch/i.html?_from=R40&_trksid=p4712.m570.l1313.TR0.TRC0.H0.X351646093156.TRS0&_nkw=351646093156&_sacat=0

Test leads:
http://www.ebay.com/sch/i.html?_from=R40&_trksid=p4712.m570.l1313.TR0.TRC0.H0.X171907570980.TRS0&_nkw=171907570980&_sacat=0

Good quality Tektronix grabbers, don't but the cheap ones, they are garbage and will only frustrate you when they constantly fall off.
https://www.google.ca/search?q=tektronix+grabbers&biw=1920&bih=960&source=lnms&tbm=isch&sa=X&ved=0ahUKEwi9s6-sv8vPAhVIQiYKHWnDAmoQ_AUIBigB#imgrc=JEPLWqqAukb-3M%3A

[Dandoni]

Jester what is the idea of having the banana jacks and the dual connectors? Is that just the sense connected at the output or is isolated? Why did you go for that setup over the posts? And is it not worth running the earth to the front also or is it rarely used?

[Dandoni]

Also CHK what precaution needs to be taken when extending the sense lines?

[Jester]

Jester what is the idea of having the banana jacks and the dual connectors? Is that just the sense connected at the output or is isolated? Why did you go for that setup over the posts? And is it not worth running the earth to the front also or is it rarely used?

Yes, one set of terminals for the output, and one set for the sense. Most of the time I just have the outputs shorted to the sense with the banana shorting bars as shown in the previous photo. However I have the option to run separate power and sense leads to the load if desired.

You can use what ever banana jacks or binding posts work best for you, I rarely use binding posts, and when I do I just use something like this:
http://www.mpja.com/Dual-Banana-Plug-to-Binding-Posts-Red/productinfo/14903+PL/

[chickenHeadKnob]

Also CHK what precaution needs to be taken when extending the sense lines?

The essential problem is that the parasitic inductance and capacitance of your sense line setup becomes part of the feedback control loop of the power supply regulator. That can push it into instability. Have a look at Agilent's Power Supply Handbook 90B, pay attention to pages 60-78. There is no simple preventive cure, and it has to be determined on an ad hoc basis. I find very little need for sense lines and prefer just to measure at the load with a multimeter and adjust the power supply accordingly if I am worried about voltage drop over the supply lines. A second hazard  is that the sense line can more easily become disconnected or shorted while you are futzing about your prototype, thereby throwing the control loop out of regulation.

link1: http://www.keysight.com/en/pc-1000000169%3Aepsg%3Apgr/dc-power-supplies?pm=LB&nid=-536902253.0&co=153731.i.1&no=100&sm=g&t=79831.g.1&cc=US&lc=eng
look to the bottom of the page for the pdf link.

alternate source: https://archive.org/details/DC_Power_Supply_Handbook_Agilent_Technologies_Application_Note_90B

I seem to remember Fluke or Keithley put out a similar document but can't find it right now.

[Jester]

For my front panel terminals, I seem to recall using #16 for the output and twisted shielded pair for the sense leads.

This seems to work fine with a 2A load, the voltage is about 50-60mV higher at the power supply end, so with the output set to 10V, I measure 10.000V at the load. With the sense at the power supply, I'm down to 9.908V, so about 100mV with 3' leads.

My electronic load can't switch super fast and at this speed no appreciable difference in transient response.