-
I'm currently in the market for a new power supply. I was all set to purchase a Rigol DP832 when I was approached locally with an offer to purchase an Agilent E3631a for $330.
I'm not sure if the extra $120 is worth the differences or not. The Rigol would have a warranty as well.
Thoughts?
Here is a spec breakdown:
Text table below for copy pasters.Agilent E3631a Rigol DP832 Price $340 $449 Ch1 Output 0 to +25V, 1A 0 to +30V, 3A Ch2 Output 0 to -25V, 1A 0 to +30V, 3A Ch3 Output 0 to +6V, 5A 0 to +5V, 3A Resolution Ch1 10mV/1mA
Ch2 10mV/1mA
Ch3 1mV/1mACh1 10mV/1mA
Ch2 10mV/1mA
Ch3 10mV/1mACh1 Accuracy .05% + 20mV
.15% + 4mA.05% + 20mV
.20% + 5mACh2 Accuracy .05% + 20mV
.15% + 4mA.05% + 20mV
.20% + 5mACh3 Accuracy .10% + 5mV
.20% + 10mA.10% + 5mV
.20% + 5mACh1 Readback .05% + 10mV
.15% + 4mA.05% + 10mV
.15% + 5mACh 2 Readback .05% + 10mV
.15% + 4mA.05% + 10mV
.15% + 5mACh 3 Readback .10% + 5mV
.20% + 10mA.10% + 5mV
.15% + 5mALoad Regulation .01% + 2mV
.01% + 250uA.01% + 2mV
.01% + 250uALine Regulation .01% + 2mV
.01% + 250uA.01% + 2mV
.01% + 250uAOther GPIB Included USB
Lan
RS232
GPIB option
Programmable
-
I would go for the dp832 because all 3 channels info can be displayed at once which I find really useful during testing...
-
IMHO, this is a clear case
- the RIGOL has 3 times the power output,
- you can see all 3 channels on a larger graphical screen,
- it has USBTMC and LAN connectivity so it is remoteley controllable without the vintage GPIB bus,
- you can buy options for smaller voltage steps, control outputs, ... later
- plus you get 3 years warranty.
I have several of them (but the DP832A version), and they just work -
My head is leaning towards the Rigol.
The Agilent Brand name carries something with me. I remember using the HP version of these supplies in school. Maybe nostalgia is getting me. -
I have a lot of recent (1-2years) Keysight and Rigol equipment.
I tried to base my decision on requirements and specs, when I was young I was poor and could not afford neither brand, so no nostalgia with me ... -
- you can buy options for smaller voltage steps, control outputs, ... later
Or just use Riglol.. -
I am a big fan of Rigol's equipment, especially given its bargain basement prices, but I have to go with the Agilent here. I own an E3631A, its about 20 years old, and it is an absolute joy to use. It won't break on you, and if it does the schematics are online and almost all the parts are easy to find jellybean components that can be easily and readily replaced with little to no cost to yourself.
Obviously if you need more than 80 Watts you wont go for the Agilent, but if you can live with the 80 Watts I think the quality and reliability are unsurpassed in the industry. -
E3631A is considerably better quality than the Rigol if a little less fancy. Also to note, you can occasionally pick up an E3631A for very little money if you're prepared to wait around. Have seen them going for equivalent of $150 here in the UK (usually when I have no money!).
Don't forget the other supplies from that line as well E3632A-E3634A which can kick out 100W or so. -
E3631A is considerably better quality than the Rigol if a little less fancy.
If both were brand new, I maybe could agree with this statement (altough I am not aware of so many failed 832, apart from the initial issue with the regulator).
But we are comparing an older unit, with a brand new unit that has 3 years of waranty for almost the same price. If you can pick an E3631A up for 150$, that another story off course. But for the prices that the OP mentioned, I really think he should opt for the DP832. I personally would never go back to a power supply that does not display the info of all 3 channels at the same time... -
Regardless of the price I'd still go for the E3631 for these reasons:
1. Rigol known overshoot. Think that is resolved but you don't ship a power supply that overshoots in this century ever.
2. Rigol filed off chip identification limiting repair options.
3. Rigol noisy as hell.
4. Rigol spikes on power up
5. Rigol crappy banana sockets on some units
6. Rigol outstanding bugs
I was going to buy one recently but decided to stick with my TTi supplies for now until something like an E3633A with some wumph comes along cheap.
E3631 has been on the market for a long time. You can still buy new ones and parts now. A second hand unit is a very good investment. So you miss readout for all channels? Not bothered myself.
Note that an E3631, even an old one was a $1000+ supply when it was new. -
Regardless of the price I'd still go for the E3631 for these reasons:
1. Rigol known overshoot. Think that is resolved but you don't ship a power supply that overshoots in this century ever.
2. Rigol filed off chip identification limiting repair options.
3. Rigol noisy as hell.
4. Rigol spikes on power up
5. Rigol crappy banana sockets on some units
6. Rigol outstanding bugs
I was going to buy one recently but decided to stick with my TTi supplies for now until something like an E3633A with some wumph comes along cheap.
E3631 has been on the market for a long time. You can still buy new ones and parts now. A second hand unit is a very good investment. So you miss readout for all channels? Not bothered myself.
Note that an E3631, even an old one was a $1000+ supply when it was new.
Hi, I can understand some of your points, but for me, these are quite ok.
1) Never noticed any during my testing, and I have the scope quite often connected to test turn on behaviour of the DUT
2) Maybe, but sofar both of mine have been going good for I think now 4 years
3) Define what you call noisy and how you measured this. From my measurements they meet their specs of 2mVpp and <350µVrms, which is identical to the E3631
4) Ok, that can be a dealbreaker for some. Defenately not only Rigol, and I consider this not so critical because low energy and low voltage. But for very sensitive loads, I would always disconnect the load, even with the most high end power supplies available. In my case when I am testing, the DP832 is always power on, I only use the enable/disable buttons.
5)
Never seen these, that would indeed be very annoying...
6) Never ran accross a bug, but I only use it as a power supply, no automation or advanced features. Do you have a reference? -
"The Signal Path" did a teardown and repair of a Agilent E3631A that you may find interesting
-
E3631A is considerably better quality than the Rigol if a little less fancy. Also to note, you can occasionally pick up an E3631A for very little money if you're prepared to wait around. Have seen them going for equivalent of $150 here in the UK (usually when I have no money!).
Just PM me when it happens, I have always money for you.
I can't stand the DP832 because it has not 3 isolated channels. The CH2 negative is connected with the CH3 negative and knowing me I will burn something one day because of that.
E3631A You have also schematics. To me no brainer E3631A all the time. -
I can't stand the DP832 because it has not 3 isolated channels. The CH2 negative is connected with the CH3 negative and knowing me I will burn something one day because of that.
Is this really the case? How about using the outputs in series then to create 0 to 60V ? -
Is this really the case? How about using the outputs in series then to create 0 to 60V ?
No problem, CH3 is the 0-5V channel. -
There is a very irritating COM line on the front panel connecting the NEG CH2 with NEG CH3.
-
That was another thing that put me off. It really wasn't much more effort to provide 3 isolated outputs and they didn't bother.
Good points about service manual. E3631A is available from keysight web site.
Also the E3631A design is derived from the older HP supplies which is derived actually from Harrison supplies from the 1960s. The fundamental design has not changed, only the technology that implements it. That's merely 60 years of improvements gone into the designs. -
I'm not so sure about that. I wouldn't buy a PSU which has such a connection because I'm pretty sure it will become a nuisance at some point.Is this really the case? How about using the outputs in series then to create 0 to 60V ?
No problem, CH3 is the 0-5V channel. -
I'm currently in the market for a new power supply. I was all set to purchase a Rigol DP832 when I was approached locally with an offer to purchase an Agilent E3631a for $330.
For a bit less than a new 832 is a SPD3303X-E and while the X-E of the series has got 10mV/mA resolution it's $389 and can do all the Rigol can do.
I'm not sure if the extra $120 is worth the differences or not. The Rigol would have a warranty as well.
Thoughts?
Excepting it offers 3 properly isolated channels.
Seems dumb they would design and build like that when if you really needed it you could just add a link between binding posts.
I'm not so sure about that. I wouldn't buy a PSU which has such a connection because I'm pretty sure it will become a nuisance at some point.Is this really the case? How about using the outputs in series then to create 0 to 60V ?
No problem, CH3 is the 0-5V channel.
-
I'm not so sure about that. I wouldn't buy a PSU which has such a connection because I'm pretty sure it will become a nuisance at some point.Is this really the case? How about using the outputs in series then to create 0 to 60V ?
No problem, CH3 is the 0-5V channel.
I was just referring to your Q to create 0-60V. Since the link is not between the 2 30V channels that is not a problem.
-
For a bit less than a new 832 is a SPD3303X-E and while the X-E of the series has got 10mV/mA resolution it's $389 and can do all the Rigol can do.
It looks like you can only display the settings for maximum 2 channels on the display?
-
... another difference seems to be that the low voltage output on the Siglent is only fixed values(2.5,3.3,5) and not continuously adjustable (0-5) like on the RIGOL.
Why did they do that ? -
I agree that a fully isolated 6V channel would be better for the RIGOL, but for what I need (analog work, with +/-12-15V, and digital stuff at 3.3 to 5V) the digital ground normally sits at the 0V level. It can be different, but this is rare ...
-
I agree that a fully isolated 6V channel would be better for the RIGOL, but for what I need (analog work, with +/-12-15V, and digital stuff at 3.3 to 5V) the digital ground normally sits at the 0V level. It can be different, but this is rare ...
Same here, never run into an issue with this. Off couse, I have 2 DP832's, so I can have already many "grounds" before that becomes a problem...
The Siglent looks like a nice alternative also. Worth comparing both in detail to see which specs matter the most to you... -
I agree that a fully isolated 6V channel would be better for the RIGOL, but for what I need (analog work, with +/-12-15V, and digital stuff at 3.3 to 5V) the digital ground normally sits at the 0V level. It can be different, but this is rare ...
That is why i got DP831. It has +-30 and isolated low voltage ch. -
The Agilent has a VFD display, which would be a no-go for me. There are many on ebay with a faded display, some are completely unreadable.
Also, I think they are not as robust as people think, imho, there some broken on ebay. I agree with people who say it is over-engineered (https://gerrysweeney.com/agileent-e3631a-power-supply-teardown-repair/). It uses a multi-layer pcb, ridiculously thin traces (that tend to break after years of thermal cycling), custom ICs.... The mosfet (IRFP044) on +6V rail seems not specified for linear operation...
-
The Agilent has a VFD display, which would be a no-go for me. There are many on ebay with a faded display, some are completely unreadable.
Also, I think they are not as robust as people think, imho, there some broken on ebay. I agree with people who say it is over-engineered (https://gerrysweeney.com/agileent-e3631a-power-supply-teardown-repair/). It uses a multi-layer pcb, ridiculously thin traces (that tend to break after years of thermal cycling), custom ICs.... The mosfet (IRFP044) on +6V rail seems not specified for linear operation...
Robust is a funny word open to interpretation.
Are these devices robust on the scale of 50 years worth of rough use? Probably not. Are they robust to the point that they turn on and give stable, accurate output every time after two decades of constant use? You bet.
The alternatives to the HP/Agilent power supplies on the market today haven't been around for multiple decades, and thus their reliability cannot be assessed in the long term. These HP supplies have been around that long, and while some show up broken on eBay (mostly due to egregious error on behalf of the previous owner), I would wager that for every 1 that is broken after 20 years of use there are another 5 that will sit on a rack somewhere plugging away without a single complaint for another 20. -
The Agilent has a VFD display, which would be a no-go for me. There are many on ebay with a faded display, some are completely unreadable.
Try replacing a DP832 power supply LCD in ten years. -
Try replacing a DP832 power supply LCD in ten years.
TBH I can't find one even now (didn't try hard). But this wouldn't turn me away from it. It's not a part that fails often (unless abused) and it doesn't fade (okay, backlight can, but often be repaired). -
... and if you try to replace the VFD in the HP supply, the repair cost will be probably a new DP832. In ten years the eBay will be flooded with used DP832s because the sold so many already ... A short look today listed a used DP832 for less than 200$ in working condition.
VFDs as such are no selling argument in 2018. -
Yeah, the thing is, E3631a is a phasing out product, while DP832 is still on sale.
There are still VFDs on ebay for E3631a, but for how long? About a year ago I bought a display for my K2000, now they are unobtanium. I can't even buy "an old broken unit for parts" because chances are it will have worn-out display. -
Buy one now and put it away.
-
Buy one now and put it away.
Sounds like a plan, but adds ~$50-100 to the price. Although, people say if VFD in a good condition it will last for quite some time, just don't leave it always on. -
They last 8-10 years left on 24/7. A lot of the supplies you see were part of test systems etc so we’re on 24/7
-
Don't be fooled by a simple VFD...
https://www.eevblog.com/forum/repair/hp-34401a-dmm-with-leaking-segments/msg1745069/#msg1745069
A PSU needs to delievers good quality current and voltage... and be easy to use, the rest comes after...
-
A PSU needs to delievers good quality current and voltage... and be easy to use, the rest comes after...
Displaying multiple channels at the same time doesn't come after for me... -
Spend a few weeks with a cheap Mastech power supply with a sticky tap relay and you'll change your mind. Oh Mr Pass Transistor has blown up and gone C-E short. 50V into my 3.3V device-o-clock.
Rigol is somewhere between the Mastech and the E3631A.
If you look at the E3633A for example it has OCP and OVP that actually work. DP832 overshoots like fuck. Long enough to cost you a nice board full of AD and LT ICs.
Priority stack for me: Performance > Reliability > Interface > Price -
Ohhh Mastech, I was a young EE and I got a triple output PSU Mastech from Amazon. Arrived open the box, plugged in, turned on and did not work. DOA. Opened it up to check for a quick repair, and my liver contracted. Blind students can solder better than Mastech. I never returned a new item so fast, it lasted litteraly 10min.
Anyway between Agilent and Rigol, I go for Agilent. OP has enough material to decide, at the end it's a personal taste.
PS: I gave my old hy3005dx to my father
:
https://www.eevblog.com/forum/testgear/mastech-dc-power-supply-hy3005dx-cheap-and-very-impressing!/
now I am Mastech free, and life is beautyful. -
I never owned a mastech just fixed them for people . I’ve got TTi PL supplies here. Cheap, well engineered and reliable.
To be fair the design is pretty neat but the problems are usually the quality of the construction which is dire, the quality of the parts used, the preregulator implementation and the lack of safety features. An SCR preregulator as per HP supplies costs only slightly more than a wun hung low 20 cent relay and actually stays working. Also a tracking crowbar at the very least. Reckon I could add both for $5
-
Mastech
Okay, when I say interface is important to me it doesn't mean about other parameters at all. I play a lot with analog circuits and for me ability to easily change parameters and see them is crucial for productivity. Although, an external DMM can solve this...DP832 overshoots like fuck. Long enough to cost you a nice board full of AD and LT ICs.
Really? I was under impression it is quite good and doesn't do it, except may be when turned on. Where did you get this data? BTW, people report here that E36313A also overshoots at power on (and yes, they discuss issue with dp832 as well, but that's a single report, I'd like to have confirmation from other people): https://www.eevblog.com/forum/testgear/rigol-dp832-ovp-overshooting-with-huge-delay/ . -
Interesting; I have a 16 year-old Mastech clone that is very well built and is a workhorse. The only problem is reported below:
https://www.eevblog.com/forum/reviews/instrutherm-fa3005-(mastech-hy3005)-fix-and-teardown/
Between a HP and a Rigol, I think the HP is guaranteed to give you more years of service, while the Rigol still remains to be seen. IMHO the VFD is the biggest drawback for the HP, but perhaps it can be replaced with some engineering (just like other HP gear that people revived with alternative displays) -
... cant complain so much about the RIGOL DP832As. I got 4 of them, up to 4 years old, and they are all still alive and kicking.
I need to measure the overshoot issue (never saw it personally), however.
SCPI programmability is OK, it got USBTMC connectivity, the A version can play some tricks with external switch ports, cycling, graphs and other stuff,
... There is more bang than the RIGOLs, yes ... but not for the same buck.