-
The Rigol DS1052E
Posted by
Simon
on 17 Oct, 2009 14:14
-
Well like many I seized the moment when seeing a 50 MHz scope going for just £ 239 on ebay, I was unaware of the samples per second rate and its importance at the time but it seems that I got very good value for money for a very nice peice of kit..... well so I thought
It seems that there is a lot of noise on the trace, I have always assumed this to be because the power supply is not screened (naturally I tossed the warranty out of the window and opened it), however on screening the SMPS with tin foil stuck to cardboard not much changed, I've had a slight improvement after also taping a peice of tin foil on top of the input screening box so that it covers the analog support circuitry behind it (refer to daves video number 37 I think) but the interferance remains.
Its rather unusual because as I increase the V/div the Vpp of the interfearance also increases effectivly remaining about the same size on the screen as all V/div settings, now that I have added screening its improved slighty and at 100 mV/div and above its less than at 20 and 50 mV/div (X10 input setting) so bearable.
Has anyone else had this ? or have comments ? I can provide photos of my screening later
-
#1 Reply
Posted by
kc1980
on 18 Oct, 2009 09:30
-
Bump. I hope someone answers this. I have an itchy finger on a hair trigger for this Ebay purchase...if ya know what I mean. (pls take my credit cards away from me)
-
#2 Reply
Posted by
johnnyk126
on 18 Oct, 2009 15:57
-
-
#3 Reply
Posted by
Simon
on 18 Oct, 2009 16:08
-
right I'm reasured now, yes my interferance is the same as the pics dave put up (Thanks Dave), I'll take some pictures of my shielding and post them later,
I think your best bet is to put the sheet of foil taped to the input screen box over the circuitry behind it first and see if it makes a difference, amazinly screening the SMPS made no difference but I want to check that the foil is connected to the chasis
-
#4 Reply
Posted by
Simon
on 18 Oct, 2009 20:07
-
Right pics time:
http://www.simonsphotography.org.uk/ebay/scope1.JPGhttp://www.simonsphotography.org.uk/ebay/scope2.JPGThe most important part is the bit of foil stuck to the top of the input screened box, the bit of foil clad cardboard made little to no difference at all, the screws holding the SMPS down will need to be changed for longer ones (i used the ones holding the back on at the top) as the cardboard adds thikness, make sure the foil is connected to the chasis this should happen when screwed down to the mounting pillars but to make sure I nicked the nut and bolt holding the ac input socket to the case and taped a peice of wire to the foild and bolted the other end to the top of the case, again little to no effect. I think the foil around the power sire had little effect to.
what opinion would you guys have on removing the psu completely from the scope and making it external ? only problem is there is a second power wire going into the back of the screen and there are so many power wires, and no they are not all in parralel only some, its like somwe of the power supply is on the scope board RIGHT NEXT TO THE ANALOG SUPPORT CIRCUITRY, very clever rigol, you had the chance to make a perfect scope and flunked it !
-
#5 Reply
Posted by
Simon
on 18 Oct, 2009 20:17
-
another thing about the Rigol DS1052E, there is a "bandwidth limit" function (accessed from the channel menu) that limits the bandwidth to 20 MHz
now 10 ADCs at 100 MHz = 1 Gs/s but for 20 MHz and the same resolution thats 10 ADCs at 40 MHz..... to be or not to be overclocked
or with the liomit on 20 MHz is it still 1 Gs/s ? it does say something about the limit preventing artefacs....
-
#6 Reply
Posted by
EEVblog
on 19 Oct, 2009 01:24
-
If it's the same level as the noise on other units and other brands, what's wrong with that?
If you need ultra low noise performance then I'd recommend an analog scope.
I get similar noise on our $5000 300MHz Tektronix DSO here at work.
I've posted some more photos of a TDS3032B and TDS1012B
http://www.eevblog.com/images/Turning on the averaging function might help reduce it for repetitive signals.
Removing the supply would be crazy, you lose the nice compact portability of the unit. And it will likely not help at all. The noise is likely just inherent in the analog front end and/or associated signal path grounds and/or supplies and ADC sampling system.
Just learn to be at one with your scope sampling noise!
-
#7 Reply
Posted by
Simon
on 19 Oct, 2009 05:38
-
Just learn to be at one with your scope sampling noise!
Yea your right, I'm expecting the performance of a decent analog scope from a cheap digital scope, pity i sold on the analog one but then I needed that money
-
#8 Reply
Posted by
TrentO
on 22 Oct, 2009 05:26
-
Oh yeah?!! I got all you guys beat!
I'm the a-hole who bought a Rigol just to troubleshoot a hand-me-down Tektronix analog CRO, which I'm probably not going to use a whole lot anyway.
I know more about divider circuits and switching matrices now, than I ever cared to know!
Simon, you keep on truckin' -- let me know when you're done potting the whole thing.
-Trent
-
#9 Reply
Posted by
Simon
on 22 Oct, 2009 06:57
-
thats as much as I've done and as much as I can do i think ! do remember to cut a slot for the psu wires to pass through and BLOCK all case holes with tape with the exception of the ones over the psu otherwise there will be no air flow and you will overheat the psu, I didn't bother to block the holes at the base of the unit to make sure the scopes board gets some airflow too.
the other thing to remember is being 8 bit there will be a +-1 bit uncertainty, this will show up in anycase as a 3 pixel thick line, just the nature of thr beast, personally I can't understand why Rigol did not screen all of the analog circuitry
-
#10 Reply
Posted by
EEVblog
on 22 Oct, 2009 21:16
-
the other thing to remember is being 8 bit there will be a +-1 bit uncertainty, this will show up in any case as a 3 pixel thick line, just the nature of the beast, personally I can't understand why Rigol did not screen all of the analog circuitry
Because it most likely would have done nothing. They shielded what was important, the high impedance analog front end.
The support circuitry is much lower impedance so does not need the same shielding.
Remember also that noise can be conducted, not just electromagnetically and capacitively coupled.
That noise isn't coming from the switching regulator, you can shield it all you want it's not going to do a thing.
When the cheap Rigol has the same noise floor as a $5000+ Tektronix, you know something a lot more complicated than simple PSU noise and shielding is at play here.
If you wanted to stop all noise from that supply, you would have to not only magnetically shield it (al-foil does nothing in this respect, you need mu-metal), but also put suitable chokes on all the wires coming out of it. But you are chasing a red-herring with the supply.
Dave.
-
#11 Reply
Posted by
Simon
on 22 Oct, 2009 22:25
-
Dave, whats mu metal ?
I did notice that screening the supply made practically no difference at all, personally to anyone wanting to have a go I'd say just inser the foil above the analog support circuitry, this is quik and easy: just cut a sheet of foil of appropriate size and tape it to the analog input box, if you use decently thick foil (15 um) then it will support itself.
-
#12 Reply
Posted by
EEVblog
on 22 Oct, 2009 23:53
-
-
#13 Reply
Posted by
Simon
on 23 Oct, 2009 06:39
-
thats interesting, where do I get this stuff, what sort of thickness does it come in ? if this is the answer thats probably why the scope (being quite cheap) was not sheilded in the first place because of the exspense of setting it all up.
-
#14 Reply
Posted by
Dago
on 25 Oct, 2009 14:09
-
thats interesting, where do I get this stuff, what sort of thickness does it come in ? if this is the answer thats probably why the scope (being quite cheap) was not sheilded in the first place because of the exspense of setting it all up.
The reason why it was not done was because it will not help, like Dave said. You could prolly cast every component invidually in mu-metal and you'd still get noise, just get over it
Not all noise comes from external sources, it can be thermal noise, sampling noise etc.
-
#15 Reply
Posted by
Simon
on 25 Oct, 2009 15:53
-
yes I know, but the fact that putting foil over the analog support circuitry helped a little proves that I've spotted a little something, I'm quite happy to throw some metal around the psu and a little beyond if it cuts the noise in half, from what I gather they didn't bother because it was already "good enough" although could be made better, being a QC inspector I'm well aware of the fact that sometimes some quality or other factor is not made to be as good as it could be because in the scope of the thing its almost irrelevant but a big difference in cost, if rigol's aim was to make a scope as good as anyone elses then they achieved their objective, to make it better meant more money and not really making more sales because its already a very good competitor,
-
-
just browsed in here after missing out on a discounted uni-trend scope, did some looking around and found this rigol and the debacle about the overclocked adc's and the apparent potential to be modded for full 100mhz bandwidth.
i generally buy products that have been dissected, analysed and modded, so i love this kinda thing.
so i'm at ebay with my finger on the BIN button gambling that someone will work out the hack... do i click?
-
-
Click it, you won't regret it! I'm really happy with mine, and I'm pretty sure everyone here is of the same opinion.
(well, it all depends on your needs really...)
-
#18 Reply
Posted by
EEVblog
on 28 Oct, 2009 10:12
-
Yeah, I don't think you'll regret clicking either.
The Rigol isn't a mid or high end scope that's for sure, so don't expect miracles, but the value for money is second to none, even without a 100MHz mod.
Dave.
-
-
cool, thanks guys...
i have a few scopes allready, all inexpensive ones; instek grs-6052 (3 year old analog/digital), gould 1421 (mid 80s analog/digital) and a cheapy velleman hps-10 handheld
the gould is going to be sold off to part pay for the new rigol, still in two minds if i should sell the instek as well, but something tells me it's handy to have an analog scope to hand sometimes
-
#20 Reply
Posted by
Simon
on 28 Oct, 2009 17:47
-
Keep 1 analogue scope, i sold my Gould300 to part pay for the Rigol DS1052E and boy have I regreted it, if your in a situation where no noise is essential or its a very small signal you will need that analogue scope, my Rigol will probably not take me lower than a signal of 100 mV because at that point the inherent noise of the scope (like any other digital scope) will become part of your signal.
-
-
I joined the Rigol Bandwagon, in part from Dave's review. I'm still learning how to use the single-shot features for capturing digital signals. I'd love to see a tutorial on how to use it to troubleshoot things like SPI buses, etc. I'm learning, but Dave's experience (and all of yours) sure cuts the learning curve down. I've been in the software business for 15 years (still work for the evil empire) and I know code and the business side of it very well, but hardware and EE side is new (and VERY interesting) to me.
-
#22 Reply
Posted by
KuchateK
on 29 Oct, 2009 14:10
-
I'd love to see a tutorial on how to use it to troubleshoot things like SPI buses, etc. I'm learning, but Dave's experience (and all of yours) sure cuts the learning curve down.
I would also like to see this.
-
#23 Reply
Posted by
TrentO
on 29 Oct, 2009 20:00
-
I got my Rigol 1052E today, ordered it from the HappySeeds dudes eight days prior, with shipping from China to the Central Valley of California. Interestingly enough, [continuous] checking on the EMS tracking system showed that my parcel was stuck in customs at LAX for four days, and was only released about eight hours ago.
I know that the USPS is pretty good, but I doubt that they got it from LAX customs to my door in eight hours-- and I wasn't shipping flux-capacitors or dark matter (of which every pound weighs 10,000 pounds.)
I think they padded the delivery time and defaulted the blame on customs. But either way, eight-full days from button click to my door-- not bad at all.
It smells like crap-- like burnt plastic, but it looks, and so far, functions GREAT!
And you KNOW WHAT?-- I'm gonna take it apart right now! <--- what a nut-job, huh?
-
#24 Reply
Posted by
flolic
on 29 Oct, 2009 21:23
-
But either way, eight-full days from button click to my door-- not bad at all.
You lucky man, according to EMS shipping number, my DS1102E is stuck in GUANGZHOU China from October 23rd...