SPD3303X-E vs DP832

[Tom45]

I have both and one thing I haven't seen mentioned is that the Siglent is a lot shorter in depth and will fit on shelves that the Rigol wouldn't fit.

[BillB]

The OP never did provide a budget, so no real reason to hold back on recommendations. Although equating the recommendation of a very good and general $1.8k lab power supply to a ultra specialized +$10k power supply may be a bit over the top. 

Perhaps.  I was just having a bit of fun at nctnico's expense, whom I otherwise highly respect for making knowledgeable contributions to this forum.  I truly do understand and agree with his point as a professional that time is money, and quality tools matter.

I just didn't want the OP  (who has implied a price range in his post discussing two models) to feel like he's less than because he's not considering a price range four times over that in his post. 

Everything has its place, and the Siglent SPD3303X-E is a perfectly fine supply for the money.  In fact, some might be surprised that I use several in a professional engineering environment without issue.  Not every use case requires a $10K solution, nor even a $1.8K one.

[Gandalf_Sr]

BillB  Tell Jheeves that someone is making a fool of themselves but it's not you.

Plasmateur  I'd replace the Siglent AWG in your list with a Rigol DG1022Z that can be turned into a DG1062Z and have the 16 MPts/Chan option added using a 'magic' USB stick and a few simple SCPI commands. 

I'm guessing nctnico might suggest this Keysight E8267D PSG vector signal generator 250KHz - 44GHz though. /s

[tautech]

Plasmateur  I'd replace the Siglent AWG in your list with a Rigol DG1022Z that can be turned into a DG1062Z and have the 16 MPts/Chan option added using a 'magic' USB stick and a few simple SCPI commands. 

Cmon Ted, you're not done your homework, 14 bit vs 16 bit.  The one Plasmateur picks as first choice can be improved to 120 MHz and his second choice improved to 500 MHz.

Want something cheaper and 14 bit, well SDG1032X is just $319 and can be improved to 60 MHz.

We can play silly games all day. 

[nctnico]

I just didn't want the OP  (who has implied a price range in his post discussing two models) to feel like he's less than because he's not considering a price range four times over that in his post. 

Everything has its place, and the Siglent SPD3303X-E is a perfectly fine supply for the money.  In fact, some might be surprised that I use several in a professional engineering environment without issue.  Not every use case requires a $10K solution, nor even a $1.8K one.

I'm getting the feeling you feel less and need to justify your purchases. I don't think you are doing anyone a favour by insisting the cheap option is the best one for everybody. IMHO it is better to provide a wider range of options (even if they are more expensive or cheaper) to the OP to broaden the horizon and let the OP decide on what is the best fit for him/her.

[Gandalf_Sr]

I'm shocked! SHOCKED! That the Siglent vendor should suggest that the Siglent product is better.

I own a DG1022Z DG1062Z with the 16MByte/Channel option enabled.  It works, I have had zero issues with it, it has no missing capacitors, and the connectors don't drop out in normal use. OTOH, I have had bad experiences with both Siglent products I purchased (new) but when I raised these with the manufacturer, returns were refused and I was sent replacement components to fit myself in both cases! I was simply not supported and had to argue for a refund in one case and am still suffering in the other.  Given your outright cheer leading here, it's fair that I share such experiences on these forums Rob. From personal experience, I could not recommend Siglent regardless of whether they appear to be better than another brand/model on paper.  Others may have different opinions but lets hear them from actual purchasers rather than Siglent vendors.

Snide comments like "you're not done your homework" say so much more about you than they do about me Rob.

Indeed, we can play silly games all day but how about you stop cheer leading so hard and I'll go off and do my homework?

Peace be with you Rob.

[nctnico]

From hands-on experience the SDG2000 series is a decent signal generator (it has had enough time to mature).

[BillB]

I'm getting the feeling you feel less and need to justify your purchases. I don't think you are doing anyone a favour by insisting the cheap option is the best one for everybody. IMHO it is better to provide a wider range of options (even if they are more expensive or cheaper) to the OP to broaden the horizon and let the OP decide on what is the best fit for him/her.

Would I love to have a $150K workbench?  Absolutely!  Would I use a fraction of its capabilities? Probably not. 

I am actually rather proud of myself for getting as much capability that I have for as little cost as possible.  I'm also pleased that I convinced my company that A-brand equipment isn't always necessary.  I was able to outfit a half-dozen engineer's offices with a few thousand dollars worth of equipment that serves their needs 80% of the time.  Of course, the other 20% requires expensive, high-end stuff.

The alternative was we buy a couple of A-brand pieces that would have been shared.  The loss in productivity when people need to hunt down that 1 shared scope, then the probes, is substantial.  Many tasks go uncompleted as a result.  Now, productivity has greatly improved when the engineer can immediately reach across his desk and quickly diagnose a PCB bring-up issue with his $500 scope, $400 DMM, and $400 DC supply.

I don't believe I ever insisted that the cheapest option is always the best, but wanted the OP to know that many times it's good enough.
                       

[Plasmateur]

Plasmateur  I'd replace the Siglent AWG in your list with a Rigol DG1022Z that can be turned into a DG1062Z and have the 16 MPts/Chan option added using a 'magic' USB stick and a few simple SCPI commands. 

The SDG2042X is easy to hack up 120MHz....but at 120MHz the sine-wave looks a little gnarly.

Look great up to 100 MHz, which is alot more bandwidth than the Rigol option. I also prefer 16bit to 14bit. I can see where the long memory depth might be a weighing factor for someone to go with the Rigol option.

Now, if you wanna beat up on Siglent about something - talk about the DC power supply vs the Rigol option. For a long time I was going back and for on which one to pick. Ultimately I found the Rigol option to be better.

But the DC power supply I'm suggesting, which is neither Rigol or Siglent, is a little more in price, and crazy in value just like the DS1054Z and the SDG2042X.

IMO, These are just the best three devices someone can purchase in regards to value.

[nctnico]

But the DC power supply I'm suggesting, which is neither Rigol or Siglent, is a little more in price, and crazy in value just like the DS1054Z and the SDG2042X.

That GW Instek GPP-4323 looks interesting too as it can also serve as a 2x 50W DC load. OTOH the current readback accuracy isn't spectacular.

[TurboTom]

In my opinion this is the current hobbiest lab equipment Trifecta
....
And if you were to upgrade it would probably be something like this....

Code: [Select]

Debatable spot
Siglent  SDG6022X  2CH AWG
Instek   GPP-4323  4CH DC power supply
Same DC power supply.

It's actually quite funny where a discussion about power supplies may end up... 

At least regarding the SDG 6000X series of waveform generators (as suggested by @Plasmateur) with the current firmware, I can only recommend to wait a little longer until the flawy start-up initialization got sorted by Siglent so the insrument would output the signal that gets diplayed on the screen... (frst hand experience)

Otherwise, my suggstion regarding the PSU is to get a DP800 series unit (or a used "A" brand instrument with a decimal keypad). During the last week, I specifically kept an eye on what "programming" method I used when testing equipment: It was about 95% direct decimal entry of voltage /current limit before enabling the output. Any device with encoder input only would have taken me probably twenty times as long to get the parameters right.

Don't get me wrong, the encoder is still a valuable user interface, just not for initial basic testing. Moreover, the mechanical rigidity of the Rigol PSUs is something that shouldn't be neglected at a weight of 8+ kg (18lbs) of the instruments, and in my opinion Rigol is clearly in advantage regarding this subject. And above that, you can program cycles, ramps and what not for series cycling /testing of equipment. I'm not aware of any other instrument in this price range that would provide this feature (stand-alone that is). Admittedly, you can get similarly specified PSUs for less money, but if you only once need the advanced features that the DP800 series offers, the 100 USD/EUR or whatever currency you'll have to pay your instrument with, that this insrument is dearer than the next cheaper replacement, are neglible compared to the time saving the mentioned supply would provide.

Anyway, for a hobbyist starting off into electronics, any of the PSUs mentioned in this thread will do the job.

Cheers,
Thomas

[Plasmateur]

But the DC power supply I'm suggesting, which is neither Rigol or Siglent, is a little more in price, and crazy in value just like the DS1054Z and the SDG2042X.

That GW Instek GPP-4323 looks interesting too as it can also serve as a 2x 50W DC load. OTOH the current readback accuracy isn't spectacular.

Hey, can you elaborate on the current readback a little bit? You and TurboTom are making me want to re-evaluate my response.

The Rigol A series appear to have the same readback current resolution, but the programming resolution is 1ma whereas the GPP-4323's programming current resolution is 0.1ma.

Also, Turbo Tom brings up a great point about programmability. I fully automate whatever I'm doing, so a front panel interface isn't a big deal for me so long as I can initially get the hang of the instrument...it's all sending SCPI commands via python going forward from there.

Programming waveforms (1Hz or lower) can be done on the GPP-4323 as well. But I'm wondering....is the extra channel superior to what a DP832A can do?


Would be great to have a spreadsheet sticky like this one about power supplies and sig gens.

Or a poll sticked every one in awhile where people can vote on what the best value O_scope/Sig_Gen/DC_Supply -  which would complete what the undisputed trifecta is (for different price tiers). And we can all yell at each other and get really angry about our opinions.

[JxR]

And we can all yell at each other and get really angry about our opinions.

[JxR]

Hey, can you elaborate on the current readback a little bit? You and TurboTom are making me want to re-evaluate my response.

The Rigol A series appear to have the same readback current resolution, but the programming resolution is 1ma whereas the GPP-4323's programming current resolution is 0.1ma.

I'm guessing he is referring to it not having different current measuring ranges like: 10A/1A/100mV/10mV/etc.  Basically, you can read small currents directly on your power supply without having to use a DMM.  I think the Instek is about the same as the Siglent and Rigol in that regard.  It honestly isn't really something you get at that price point for new gear.  For me, it is a feature I prefer to have in a power supply like nctnico.

I'm also big fan of multi-quadrant power supplies. I remember seeing this Instek line a few months back, and they looked really nice for the price.  Honestly, I would much rather have one of those than the Rigol or Siglent.  I personally think you made a great choice.

[nctnico]

Or a poll sticked every one in awhile where people can vote on what the best value O_scope/Sig_Gen/DC_Supply -  which would complete what the undisputed

Such a poll is utterly useless because everyone's needs and budget are different. Forcing a 'best solution' on people is the worst thing to do. Just leave the options open and be clear about the pros and cons.

[nctnico]

But the DC power supply I'm suggesting, which is neither Rigol or Siglent, is a little more in price, and crazy in value just like the DS1054Z and the SDG2042X.

That GW Instek GPP-4323 looks interesting too as it can also serve as a 2x 50W DC load. OTOH the current readback accuracy isn't spectacular.

Hey, can you elaborate on the current readback a little bit? You and TurboTom are making me want to re-evaluate my response.

Look at the specs of the GW Instek. Current readback accuracy is (IIRC) 0.3% +10mA. I'd like to see that at 0.05% or so because in that case you have (about) the same accuracy as a 4.5 digit DMM. Part of the added value of having digital readouts on a PSU is not needing extra DMMs to measure voltage and current.

[BillB]

Also, Turbo Tom brings up a great point about programmability. I fully automate whatever I'm doing, so a front panel interface isn't a big deal for me so long as I can initially get the hang of the instrument...it's all sending SCPI commands via python going forward from there.

The keypad on the Rigol is a nice feature though, and being able to punch in a value and go is very handy.  But, that does takes up panel real estate, too.

[nctnico]

Also, Turbo Tom brings up a great point about programmability. I fully automate whatever I'm doing, so a front panel interface isn't a big deal for me so long as I can initially get the hang of the instrument...it's all sending SCPI commands via python going forward from there.

The keypad on the Rigol is a nice feature though, and being able to punch in a value and go is very handy.  But, that does takes up panel real estate, too.

I'd say a keypad is a must have on a more advanced PSU. Turning knobs is slow and tedious.

[tautech]

Also, Turbo Tom brings up a great point about programmability. I fully automate whatever I'm doing, so a front panel interface isn't a big deal for me so long as I can initially get the hang of the instrument...it's all sending SCPI commands via python going forward from there.

The keypad on the Rigol is a nice feature though, and being able to punch in a value and go is very handy.  But, that does takes up panel real estate, too.

I'd say a keypad is a must have on a more advanced PSU. Turning knobs is slow and tedious.

We can apply the same line of thought to multiplexed vertical controls on scopes yet they have general acceptance in todays world.
Much is now the UI works that make what might seem a daunting interface is actually quite simple, quick and effective to use.

[nctnico]

Sorry but that is a apples versus oranges comparison. On every PSU I own which has a keypad I use the keypad because it is quicker and more precise. Don't try to sell lemons for oranges.

Shared vertical controls on an oscilloscope are far less annoying compared to needing to turn an encoder endlessly. The root of the problem is that a digitally controlled PSU has a much finer grained control versus a PSU which has 10 turn pots (which are on the edge of being tedious to use). This translate on needing to rotate the knob on a digitally controlled PSU much more (unless ofcourse it has very good accelleration but that is something both Rigol and Siglent are not known for).

[tautech]

Sorry but that is a apple versus oranges comparison. On every PSU I own which has a keypad I use the keypad because it is quicker and more precise.

I'd bet some don't have detented encoders.

[nctnico]

Sorry but that is a apple versus oranges comparison. On every PSU I own which has a keypad I use the keypad because it is quicker and more precise.

I'd bet some don't have detented encoders.

Doesn't matter at all. For kicks I just tried using the knob on my E36313A to set it to 10V. I stopped at 4V because I already rotated the knob more than I'd need to get a 10 turn pot to the end. Pressing 3 buttons to get to 10V is quicker and easier. There is no amount of argueing which changes that.

[tautech]

Sorry but that is a apple versus oranges comparison. On every PSU I own which has a keypad I use the keypad because it is quicker and more precise.

I'd bet some don't have detented encoders.

Doesn't matter at all. For kicks I just tried using the knob on my E36313A to set it to 10V. I stopped at 4V because I already rotated the knob more than I'd need to get a 10 turn pot to the end.

Oh, so you can't select which digit you want to adjust ? 

Yep, that's something you need a numeric keypad fix for.

[nctnico]

Sorry but that is a apple versus oranges comparison. On every PSU I own which has a keypad I use the keypad because it is quicker and more precise.

I'd bet some don't have detented encoders.

Doesn't matter at all. For kicks I just tried using the knob on my E36313A to set it to 10V. I stopped at 4V because I already rotated the knob more than I'd need to get a 10 turn pot to the end.

Oh, so you can't select which digit you want to adjust ? 

That is even a worse solution. Now you need to push even more buttons  And it is potentially dangerous too. I have a DC-load which works that way and using the digit selection mode it is very easy to set it to a much higher current than required.

[JxR]

Look at the specs of the GW Instek. Current readback accuracy is (IIRC) 0.3% +10mA. I'd like to see that at 0.05% or so because in that case you have (about) the same accuracy as a 4.5 digit DMM. Part of the added value of having digital readouts on a PSU is not needing extra DMMs to measure voltage and current.

That is what I get for scanning the datasheet to quickly.  Numbers bled together and I read 0.03% (like the voltage accuracy).

I'll admit the accuracy is not the greatest on the Instek.  Although the fact that it can has programmable current sink for the same price as the Rigol would make me choose the Instek over it.  I still don't have a dedicated electronic load because I mostly deal with low power circuits.  The 50W sink capability like the Instek has would cover most of my use cases (if I didn't already have this capability).

Regardless, any of the power supplies in this price range means I'm going to have to eventually use a DMM to get a more accurate current measurement.  I would still rather buy a couple used Agilent "Mobile Comm DC Sources" for the same price just to have the lower current measurement ranges.