RIGOL DP832 still a good Bench PS?

[2N3055]

i‘m also looking for a bench power supply and was looking for the dp832. my main concern was that is is quite deep. but yesterday i found these two videos: ------- and i‘m quite disappointed. his device is 5 years old, but has anything changed since then?

Why wouldn't things be changed? If that was a problem in all supplies there would be thousands of reports of them dying out there...
There were some problems of first revisions, overheating, this PSU problem and display problem (My DP831 had that). They gave me brand new one that was new revision and it's been solid for years now...

I would be wary about buying older used one (because I might get one of those first revisions with problems), but buying a new one shouldn't be a problem.
But as many pointed out here, there are other options that might be better for your need, be it smaller, more functions etc etc..

I won't get into which one to buy, but I will tell you this: being able to readback voltage, current and power of DUT directly, with good precision,  is something I use all the time.
Also, being able to control it with PC (or Raspberry PI) is another thing that is useful.  Smart PSU are very useful if you take time to use advanced capabilities.

[nctnico]

A SPD3303X-E is reasonable value at $ 389 but the 3rd channel is only 2.5/3.3/5.0 V @ 3.2A.
At least it has 3 properly isolated channels and good power ON behavior.

But no numeric keypad which is essential on a digitally controlled PSU. Without a numeric keypad you keep twisting the knob ad nauseam. It is just too tedious to operate.

[Berni]

The DP832 is a bit expensive but it does bring a fair bit of bang per buck.

It is a nice PSU when you need multiple channels. Much more convenient than using separate PSUs. The common ground between Ch2 and Ch3 is not really that bad once you get used to it (You often need the digital rail to be commoned up with something else). Unlike most 3 channel PSUs the Ch3 is a proper channel with adjustable everything, this is very nice to have in my opinion. You can also hack all the options on it easily. It does have some quirks but its not too bad. The fans can also be replaced with quieter ones, it makes it much less annoying to me.

These Korad supplies and similar are fairly poor build quality. Dave blew one live on video, i blew up one myself within a few days of getting it. I was not impressed

[tautech]

A SPD3303X-E is reasonable value at $ 389 but the 3rd channel is only 2.5/3.3/5.0 V @ 3.2A.
At least it has 3 properly isolated channels and good power ON behavior.

But no numeric keypad which is essential on a digitally controlled PSU. Without a numeric keypad you keep twisting the knob ad nauseam. It is just too tedious to operate.

Not so, select the unit to adjust and turn the encoder.
The fact that we're always out of stock of SPD3303X-E might suggest otherwise.
More are on the way and just about to order even more.

[nctnico]

You know what they say: there is a sucker born every minute  Once you have used a digitally controlled PSU with a keypad you'll never want to use one with only a twisty knob.

[Berni]

But no numeric keypad which is essential on a digitally controlled PSU. Without a numeric keypad you keep twisting the knob ad nauseam. It is just too tedious to operate.

Not so, select the unit to adjust and turn the encoder.
The fact that we're always out of stock of SPD3303X-E might suggest otherwise.
More are on the way and just about to order even more.

Also have digital PSUs that do that encoder knob thing, it is better than a analog knob but still pretty annoying to use.

On the DP832 if i want to have 5V at 200mA i just key in    5    Volts    2    0    0    miliAmps. No need to even select what i am going to type, it just knows that its voltage when you press the button for volts. Only thing to watch out is to make sure the correct channel is selected. Its faster to type it in, also requires no looking at the display (as you would have to look what digit you are turning). If i do want to set it with a knob (Like fine turning the voltage up and down live) i can still select the appropriate parameter and spin the digits with the knob. Having both options on hand i always pick the numeric keys for setting a fixed nice round voltage number. Even with the rigols weird circular numeric keypad (you kind of get used to it eventually).

[rsjsouza]

My US$0.02

Yeah, I do need to round it out with some other equipment as well, that's the main reason I want to stay under $450 with a PS. I have thought about going with a single PS and adding more as time goes on, which is probably the smartest way, but at some point I would probably want three channels anyway and bench space will be more of a premium than money.
(...)
Anyway, long story short, I thought I would re-educate myself as I built up the lab, by building simple circuits as I read books and follow along videos. My thoughts were to start off with a decent PS and a Scope. I have some lesser gear and tools already, couple DVM's, soldering stations, etc. but no bench type equipment.

Have you considered building your own? Power supplies are relatively simple pieces of equipment and provide a great learning experience, especially when suited to the power levels of Arduino and Raspberry Pi. The reason I mention this is because you seem to have passed that level of knowledge where everything is daunting and seems insurmountable.

Another alternative would be to try and find an used but older model in the local market (Craigslist, Offerup, etc.) - although the risk of getting something that has a failure is higher, the classic models (from HP or Tektronix for example) have full schematics and service information and can be a good platform to learn a bit more. Obviously that the selection becomes more restricted to what is available in your market, but it would be certainly a cheap alternative with a lower risk of regret.

I, just like many on this forum, spent most of my hobbyist (and then later professional) life using single channel power supplies. Sure, in the 1980s/1990s lab power supplies were mostly out of reach for the common man, so we had to make lots of compromises.

A SPD3303X-E is reasonable value at $ 389 but the 3rd channel is only 2.5/3.3/5.0 V @ 3.2A.
At least it has 3 properly isolated channels and good power ON behavior.

But no numeric keypad which is essential on a digitally controlled PSU. Without a numeric keypad you keep twisting the knob ad nauseam. It is just too tedious to operate.

Not so, select the unit to adjust and turn the encoder.
The fact that we're always out of stock of SPD3303X-E might suggest otherwise.
More are on the way and just about to order even more.

Although the keypad is not 100% mandatory, it is indeed quite helpful. A bad twist on the knob and your circuit may go up in flames. Also, if the voltage/current is far from the current set point AND the firmware is not well implemented (without encoding accelleration, for example), the twisting becomes tiring quite fast. Another aspect is the sheer quality of the encoder used: in typical low cost equipment, the wear and tear of encoders is quite severe.

Does that mean the Siglent is a bad contender? Not at all, but if you can choose and is within your budget, this is an important factor to consider.

The popularity usually means the power supply performs decently and has a very good price point, but it has just limited correlation in a comparison between models/manufacturers.

[nctnico]

BTW a nice PSU on the used market is an Agilent 66311 (15V 3A IIRC). It needs a different Noctua fan to reduce the noise a bit but other than that it ticks a lot of boxes. The dual range current display (with uA resolution for the low range) is very handy.

[RoGeorge]

i‘m also looking for a bench power supply and was looking for the dp832. my main concern was that is is quite deep. but yesterday i found these two videos: and i‘m quite disappointed. his device is 5 years old, but has anything changed since then?

Rigol fixed that design mistake many years ago.

The very first design was like the one shown in the video.  Dave caught the reset problem in a video, too, many years ago, and as a result Rigol changed their design adding a radiator to that voltage stabilizer.  They do not produce the malfunctioning design any more.  The additional radiator can be seen from outside, without opening the unit, so it's easy to identify if by accident you receive an old stock with the design bug.



LATER EDIT:
Found it, that was in 2013:
https://www.eevblog.com/forum/blog/eevblog-512-rigol-dp832-bad-design-investigation/msg284232/#msg284232

[justme1968]

thanks for the heads up! so it will probably be a dp832 after all. if it just were a little less deep...

[RoGeorge]

It is very deep and very heavy.  Not recommended for a normal office desk.

My workbench had a shelf extension in the back, making it 85cm deep, so I can align the front panel of the DP832 with other instruments.  A casual desk is only 60cm deep, that will left almost no room in front of a DP832.

The good news is it can stand eventually on the floor, like a desktop.  DP832 has deep corners with rubber pads and a 90 degree power cord plug, so it won't damage your cables when sitting on its back panel, on the floor.

[thm_w]

Have you considered building your own? Power supplies are relatively simple pieces of equipment and provide a great learning experience, especially when suited to the power levels of Arduino and Raspberry Pi. The reason I mention this is because you seem to have passed that level of knowledge where everything is daunting and seems insurmountable.

At that point I would just get a RD6006 and build my own case around it. But you need access to salvage PSUs, craigslist or second hand store.
https://hackaday.com/2020/03/16/review-the-riden-rd6006w-dc-power-supply-module/

[rsjsouza]

Have you considered building your own? Power supplies are relatively simple pieces of equipment and provide a great learning experience, especially when suited to the power levels of Arduino and Raspberry Pi. The reason I mention this is because you seem to have passed that level of knowledge where everything is daunting and seems insurmountable.

At that point I would just get a RD6006 and build my own case around it. But you need access to salvage PSUs, craigslist or second hand store.
https://hackaday.com/2020/03/16/review-the-riden-rd6006w-dc-power-supply-module/

That or a big transformer, four diodes, a large capacitor and a suitable box.

[DriftlessNet]

Have you considered building your own? Power supplies are relatively simple pieces of equipment and provide a great learning experience, especially when suited to the power levels of Arduino and Raspberry Pi. The reason I mention this is because you seem to have passed that level of knowledge where everything is daunting and seems insurmountable.

Another alternative would be to try and find an used but older model in the local market (Craigslist, Offerup, etc.) - although the risk of getting something that has a failure is higher, the classic models (from HP or Tektronix for example) have full schematics and service information and can be a good platform to learn a bit more. Obviously that the selection becomes more restricted to what is available in your market, but it would be certainly a cheap alternative with a lower risk of regret.

I, just like many on this forum, spent most of my hobbyist (and then later professional) life using single channel power supplies. Sure, in the 1980s/1990s lab power supplies were mostly out of reach for the common man, so we had to make lots of compromises.

Actually I have build a couple of PS's over the years, the last one being some time ago but it has since got lost somewhere in a move after I lost interest and was neck deep in work related worries. It was when you could still walk into a local RatShack and pick up the all the parts you needed, to give you an idea, not counting their last few years they spent as a glorified Cell Phone store.

Anyway, as you said I am a bit past that, but still want to hit the ground running as I refresh my knowledge. My first go around was mainly practical knowledge and the hands on learning, with very little (just enough) of the theory and math behind what was happening. While that was plenty for my job at the time, it only really seemed to be good for repairs and constructing simple circuits such as power supplies. When trying to design more complex projects I seemed to get hung up and I think it was due to the lack of the math/theory knowledge (and decent lab equipment). I could easily follow a schematic, or a plan in an electronics magazine, to build something,  but I never could sit down and design a more complex project on my own from scratch. That is the direction I want to head now that I have more time on my hands.

As mentioned earlier, I have build several things now using Arduinos and associated shields, but again these are almost "finished" projects from the perspective I am looking at. I don't claim I will reinvent a Arduino, but maybe designing/building a few custom shields sounds like fun. I also want to design/build my own Audio Power Amp, just two channel and maybe a separate sub amp as a more immediate challenge.

Anyway, maybe I will just get a single PS and work on a home built 3-channel to start after I bone up on the topic a bit more. Thanks for your input btw.

[Berni]

It is very deep and very heavy.  Not recommended for a normal office desk.

My workbench had a shelf extension in the back, making it 85cm deep, so I can align the front panel of the DP832 with other instruments.  A casual desk is only 60cm deep, that will left almost no room in front of a DP832.

The good news is it can stand eventually on the floor, like a desktop.  DP832 has deep corners with rubber pads and a 90 degree power cord plug, so it won't damage your cables when sitting on its back panel, on the floor.

The DP832 is a bit on the deep side but not too bad. Most test equipment for bench use (apart from the most modern ones that moved to the new vertical standing style like scopes) need about 30 to 40 cm of depth. On top of that this form factor has a great benefit of being nicely stackable, sure it might use 2 or 3 times the depth compared to a modern slim standing scope, but can you stack 3 of those scopes on top of each other?

Not only is the DP832 in a reasonable depth range for bench gear, it is also 1/2 rack width, so this makes it the same width as a lot of bench gear from HP Keysight, Keithley, Fluke, BK...etc Making it stack on top of it perfectly. I do have a 40cm deep set of shelves behind my bench for this reason.

But yeah compared to a cheep little single channel bench PSU the DP832 is not very portable and you would not want to move it too often with how heavy it is. Then again so are all the other 3 channel linear PSUs of a similar power range. So yeah if you are after something you can just easily grab and plonk down on any random bench then you either need a smaller single channel linear PSU or a compact switchmode one.

[hpw]

My fish...

. It is heavy, enormous and costs some to ship!
. fan as load as a hover  and is constant even on low load 
. cannot set both as 10.0V one 10.02V, the other always 20mV off even on tracking mode
. be careful about the common GND if you have +/- 15V and + 3.3V on third! While made not logically to connection +/- V GND
. LAN and one digit more extra bucks 
. this as DIY project, for this price hard to build

Nice tracking mode for the first two as +/- in common. Current is still separated 

So locking how to set the first two with equal output voltage!

Overall OK for bucks, consider the GND connection behavior if 3 requires a common ground

Hp

[edtyler]

I have had the DP832 for over 2 years and it has been quite useful and reliable. It is used to source 48V at 2.5A and also 14.6V at 6A.

It is much more useful than single output supplies, of which I have at least 4. The fan noise is noticeable at full output, but less so at lower loads. Overall, a good value for the money. I would buy it again as a general bench supply.

The power on spike issue exists with many linear supplies, including those from HP and Lambda. I was trained to never have any sensitive loads connected when the power supply is turned on or off.

So far, I've only run across one bug that is annoying enough to mention: If one has CH1 and CH2 operating in parallel with a monitor condition on both channels set to turn off that channel when the current drops below a specified value. The logging for that channel shows the drop to the cut off current value and then, instead of showing zero amps for the rest of the log, it shows a ramp up to about 0.7A and remains there indefinitely. This is not a big problem because the deflection point is easily spotted. It does interfere with getting statistics (watts) on that channel unless one sets the ending sample number at the deflection point, which is easy to do.

[bson]

The Rigol is a fine PSU, if a little big for what it does.  But any smaller comparable 3-channel supply like the Keithley 2231A-30-3 is going to cost more.  And the DP832 is still feature-rich, for example it has LAN and a sense input.  (Though annoyingly the sense input in the back, using a special connector that is guaranteed to get lost.)  If you have a decent remote controlled Agilent 34401 compatible DMM it's also easy to calibrate yourself.

[RoGeorge]

(Though annoyingly the sense input in the back, using a special connector that is guaranteed to get lost.)

DP832 doesn't have sense inputs.

The green connector in the back (its pair is delivered together with the DP832) has only digital inputs and outputs.  Those digital I/O pins are programmable and can do a couple of things described in the manual.  They are especially useful for automating testing, to trigger the source externally or alike.

[bson]

Hmm.  I remember wrong then.

Edit: confused it with my Keysight 66312A.

[Berni]

Yeah DP832 has no external sense inputs. They are internally wired to the banana terminals.

But then again i never really used sense inputs on a general purpose bench PSU. These 3A that it can supply is still such a low current that simply using cables with a decent amount of copper in them (lots of mine are 1.5mm2) makes the voltage drop almost non existent.

Most of the times i actually used remote voltage sensing is for stuff involving very large currents >20A where even a big fat hose of a cable can have a considerable voltage drop. Almost all gear that works with those sort of currents has a remote sense in some shape or form. Quite often seen in the form of one of those green screw terminal connectors.

Especially on PSUs with a lot of channels having separate sense terminals can really clutter up the front panel, tho it admittedly also makes the PSU also look more impressive (even if you rarely even use them)

[hpw]

So I went to Manual Calibration and 

VDAC and channel 1 / 2 / 3 mysteries!

Just enter into the util, page 2 for manual calibration menu and please may tell about similar behaviors.

While you may go to each level steps and save it, but will never show up your entered values in case on a second checkup!

VDAC Channel 1:

[1] 0.45V and measured 152mV

[5] 2.0V and  measured 1.685V


VDAC Channel 2:

[1] 0.53V and measured -254mV

[2] 0.73V  and measured -0.0522mV

[3] 1.55V  and measured 0.779mV

[4] 2.7V  and measured 1.931V

[5] 3.68V and  measured 2.909V


VDAC Channel 3:

[1] 0.12V and measured 104mV

[5] 1.05V and  measured 1.033V

This means Channel 2 is ways off to the given references  even negative  !!!

How to proceed with calibration download on RIGOL & calibration the related manual (to enter default PW:2012)

Was just a trial to get ride 20mV offset of the tracked CH1 & Ch2 channels

Hp

[hpw]

Here more about other already known issues:




https://www.eevblog.com/forum/blog/eevblog-549-rigol-dp832-lab-power-supply-followup/

[thm_w]

So I went to Manual Calibration and 
...
How to proceed with calibration download on RIGOL & calibration the related manual (to enter default PW:2012)

Was just a trial to get ride 20mV offset of the tracked CH1 & Ch2 channels

Never do manual cal. There are automated python cal scripts here that will: be more accurate in the end, and less hassle to complete. Best if you have a USB/networked DMM.
https://www.eevblog.com/forum/testgear/dp832-calibration-using-python-pycharm-running-on-windows/

If its close enough already just leave it though.

[RoGeorge]

The one in the video is caused by the voltage drop on the internal wires.  Can be fixed by a thick bridge between the GND of the channels 2 and 3.  Ideally the bridge would be a thick copper wire soldered between the two GND connectors internally, but can be as well an external bridge between the two GND banana plug.

That's more like a theoretical corner case.  In real life you would see more errors caused by real wires going to real circuits, so yeah, could have been better but doesn't really matter.



About calibration, I think that is supposed to be made in an automated way, it would take a lot of time and patience, and be very prone to errors to calibrate manually.  AFAIK there are over 400 points to measure, read, then introduce manually
- 48 measurement points for CH1 VDAC
- 48 measurement points for CH1 VADC
- 32 measurement points for CH1 IDAC
- 32 measurement points for CH1 IADC

- same for CH2
- same for CH3 except VDAC and VADC are only 18 measurements each instead of 48

On the CD that came with DP832 there is a "DP80_CaligrationGuide_EN.pdf" that says it is recommended to calibrate once a year but they also wrote:

RIGOL does not recommend manual calibration by users. If calibration is required,
please contact RIGOL customer service department or the local distributor.

As a side note, my DP832 have some small errors, but I never calibrated it because it doesn't matter for my needs, and wires will introduce more errors anyway.  Also I'm not very sure about the correct calibration procedure, and how big exactly is the calibration table (how many calibration points are), I've seen contradictory info floating around.  In the programming manual there are no calibration commands, but from the disassembled firmware by tv84 there are hidden SCPI commands for DP800 calibration:

Code: [Select]

:CALibration:SET
:CALibration:MEAS
:CALibration:STORe
:CALibration:CLEar
:CALibration:STARt
:CALibration:END
:CALibration:SENSE

:PROJect:CALIbration:DATA:VOLTage:WRITe
:PROJect:CALIbration:DATA:VOLTage:READ?
:PROJect:CALIbration:DATA:CURRent:WRITe
:PROJect:CALIbration:DATA:CURRent:READ?
:PROJect:CALIbration:DATA:CURRent:ADDRess?
:PROJect:CALIbration:INFO:WRITe
:PROJect:CALIbration:INFO:READ?
:PROJect:CALIbration:INFO:ADDRess?