Electronics-Related.com
Forums

Best Oscilloscope for $3k

Started by ChesterW August 26, 2014
Dear Group,

      I can finally afford a new oscilloscope to supplement my trusty 
TEK 465M. My budget is around $3k. I like the newer Agilent scopes, 
which I have used when someone else was paying. I do mostly low 
frequency high-precision instruments (analog signals less than about 1 
MHz).

      The pretty lower-end Agilent scopes are in reach, but I'm tempted 
to use one of these:

http://tinyurl.com/k88x74r

which is driven internally with an FPGA and one of these ADCs:

http://www.mouser.com/catalog/specsheets/DataSheetASD5020HS_v2.0.pdf

I make out the ENOB to be 11.6 bits at 125 MSPS with 14 bit samples 
using 4 channels, and 13.2 bits with 16 bit samples at 62.5 MSPS using 
one channel. It will also go out to 1 GSPS with 8 bit samples. Analog BW 
is 200 MHz.

I like the serial decoding for free and the ability to stream data to an 
external application for creating custom processing. I'm thinking this 
would be good for prototyping AFEs with signal processing via the PC.

So my thinking is that using this rather obscure configurable ADC that 
these guys improve performance over the big manufacturers ASICs for my 
particular applications niche.

What do you think?

ChesterW

On Tue, 26 Aug 2014 11:21:57 -0500, ChesterW <iamsnoozin@yahoo.com>
wrote:

>Dear Group, > > I can finally afford a new oscilloscope to supplement my trusty >TEK 465M. My budget is around $3k. I like the newer Agilent scopes, >which I have used when someone else was paying. I do mostly low >frequency high-precision instruments (analog signals less than about 1 >MHz). > > The pretty lower-end Agilent scopes are in reach, but I'm tempted >to use one of these: > >http://tinyurl.com/k88x74r > >which is driven internally with an FPGA and one of these ADCs: > >http://www.mouser.com/catalog/specsheets/DataSheetASD5020HS_v2.0.pdf > >I make out the ENOB to be 11.6 bits at 125 MSPS with 14 bit samples >using 4 channels, and 13.2 bits with 16 bit samples at 62.5 MSPS using >one channel. It will also go out to 1 GSPS with 8 bit samples. Analog BW >is 200 MHz. > >I like the serial decoding for free and the ability to stream data to an >external application for creating custom processing. I'm thinking this >would be good for prototyping AFEs with signal processing via the PC. > >So my thinking is that using this rather obscure configurable ADC that >these guys improve performance over the big manufacturers ASICs for my >particular applications niche. > >What do you think? > >ChesterW
Looks like Hittite acquired Arctic and Keragis, then ADI acquired Hittite. Personally, I like scopes with knobs and screens and such. It is an interesting idea, a scope with a fast 8-bit ADC and a slower, true 12 to 16 bit ADC.
On Tue, 26 Aug 2014 11:21:57 -0500, ChesterW <iamsnoozin@yahoo.com>
wrote:


>What do you think? > >ChesterW
I think that you ought to buy a cheaper PC-based scope first to see if you can stand the lack of knobs. Several years ago I bought a Link combination scope and logic analyzer. About $2k at the time. I grew to hate that thing with a passion. Almost everything required one to use both hands. And unlike a conventional scope where you know by feel which knob and switch does what, you have to take your eyes off what you're doing to make any adjustments. I ended up dumping it for a pittance on sleazebay. Then there's the matter of making room on your bench for a display, keyboard and mouse. In my lab, no way. OK, so I keep a sloppy bench. Even a laptop is in the way and the keyboard fairly quickly fills up with debris - stripped wire ends and such. My main scope nowadays is a Rigol. Tremendous bang for the buck. I like it but even it has too few knobs. It needs a knob per channel for vertical sensitivity and position. I don't know if it'll meet your accuracy requirements or not but it sure is a nice scope for what I do (primarily designing induction heaters). John John DeArmond http://www.neon-john.com http://www.fluxeon.com Tellico Plains, Occupied TN See website for email address
On 8/26/14, 1:00 PM, Neon John wrote:
> On Tue, 26 Aug 2014 11:21:57 -0500, ChesterW <iamsnoozin@yahoo.com> > wrote: > > >> What do you think? >> >> ChesterW > > I think that you ought to buy a cheaper PC-based scope first to see if > you can stand the lack of knobs. Several years ago I bought a Link > combination scope and logic analyzer. About $2k at the time. I grew > to hate that thing with a passion. Almost everything required one to > use both hands. And unlike a conventional scope where you know by > feel which knob and switch does what, you have to take your eyes off > what you're doing to make any adjustments. I ended up dumping it for > a pittance on sleazebay. > > Then there's the matter of making room on your bench for a display, > keyboard and mouse. In my lab, no way. OK, so I keep a sloppy bench. > Even a laptop is in the way and the keyboard fairly quickly fills up > with debris - stripped wire ends and such. > > My main scope nowadays is a Rigol. Tremendous bang for the buck. I > like it but even it has too few knobs. It needs a knob per channel > for vertical sensitivity and position. I don't know if it'll meet > your accuracy requirements or not but it sure is a nice scope for what > I do (primarily designing induction heaters). > > John > > > John DeArmond > http://www.neon-john.com > http://www.fluxeon.com > Tellico Plains, Occupied TN > See website for email address >
You guys are right. From using a (cheap) USB scope before, the lack of knobs is a real PITA. For this USB thing I'd make a box with some encoders and buttons to emulate a USB keyboard. The supplied software allows mapping keystrokes to the important functions, so at least I'll be able to have basic controls without using the mouse. My lab bench and computer bench are one in the same, so I'm OK there. Also usually messy, BTW. I wanted to get some opinions since I can't quite believe I'm picking this USB scope over a nice Agilent. Once I buy it I'll be stuck with it for a while, so I want to be sure. ChesterW
On Tuesday, August 26, 2014 9:21:13 PM UTC+2, ChesterW wrote:
> On 8/26/14, 1:00 PM, Neon John wrote: >=20 > > On Tue, 26 Aug 2014 11:21:57 -0500, ChesterW <iamsnoozin@yahoo.com> >=20 > > wrote: >=20 > > >=20 > > >=20 > >> What do you think? >=20 > >> >=20 > >> ChesterW >=20 > > >=20 > > I think that you ought to buy a cheaper PC-based scope first to see if >=20 > > you can stand the lack of knobs. Several years ago I bought a Link >=20 > > combination scope and logic analyzer. About $2k at the time. I grew >=20 > > to hate that thing with a passion. Almost everything required one to >=20 > > use both hands. And unlike a conventional scope where you know by >=20 > > feel which knob and switch does what, you have to take your eyes off >=20 > > what you're doing to make any adjustments. I ended up dumping it for >=20 > > a pittance on sleazebay. >=20 > > >=20 > > Then there's the matter of making room on your bench for a display, >=20 > > keyboard and mouse. In my lab, no way. OK, so I keep a sloppy bench. >=20 > > Even a laptop is in the way and the keyboard fairly quickly fills up >=20 > > with debris - stripped wire ends and such. >=20 > > >=20 > > My main scope nowadays is a Rigol. Tremendous bang for the buck. I >=20 > > like it but even it has too few knobs. It needs a knob per channel >=20 > > for vertical sensitivity and position. I don't know if it'll meet >=20 > > your accuracy requirements or not but it sure is a nice scope for what >=20 > > I do (primarily designing induction heaters). >=20 > > >=20 > > John >=20 > > >=20 > > >=20 > > John DeArmond >=20 > > http://www.neon-john.com >=20 > > http://www.fluxeon.com >=20 > > Tellico Plains, Occupied TN >=20 > > See website for email address >=20 > > >=20 > You guys are right. From using a (cheap) USB scope before, the lack of=20 >=20 > knobs is a real PITA. For this USB thing I'd make a box with some=20 >=20 > encoders and buttons to emulate a USB keyboard. The supplied software=20 >=20 > allows mapping keystrokes to the important functions, so at least I'll=20 >=20 > be able to have basic controls without using the mouse. >=20 >=20 >=20 > My lab bench and computer bench are one in the same, so I'm OK there.=20 >=20 > Also usually messy, BTW. >=20 >=20 >=20 > I wanted to get some opinions since I can't quite believe I'm picking=20 >=20 > this USB scope over a nice Agilent. Once I buy it I'll be stuck with it=
=20
>=20 > for a while, so I want to be sure. >=20 >=20
I have a Picoscope 4 channel, 12 bit version. I sometimes bring it to work = for measurements. Often I find myself using the Pico instead of the self-su= stained scopes. It has a FFT that is quatum steps better than the ones in t= he regular scopes and it's a breeze to document stuff, since you can just c= opy from the app to Word etc. The bus analysis comes for free and that's a big advantage Regards Klaus
ChesterW <iamsnoozin@yahoo.com> wrote:
> What do you think?
Short term, if the PC you are using it with craps out, you need to come up with another PC. This is usually not too hard to do, either at work or at home; it might be a good idea to keep your n-1 PC around a little longer as a backup. Long term, you have to be prepared to keep a PC running, probably on XP or 7, for however long you own the scope. (Or a virtual machine of XP or 7, as long as the VM's USB support is fast enough for the scope.) The fact that they claim to support XP through 8 is a good sign. You can probably count on getting updated drivers/software for 9 in a few years. After that, you may or may not get further software support. If they give you source code for the drivers, that is also a good sign for long-term support. However, even their Cadillac model at US$2800 is only about a week or two, plus or minus, of one programmer's time. If it won't work with Windows 10, you may end up ahead to just buy another scope, rather than paying someone to port the drivers for you. The newer "regular" scopes have kind of the same problem, since many of them are just PCs in a funny case with a nice A/D board. On the other hand, you would hope Agilent or Tek would pick a better "motherboard" and other components, so they would last longer. By the way, if you do have a newer Tek that runs Windows, and you haven't made the backup CDs/DVDs for it, do so. If you get a replacement hard drive from Tek, it comes blank - no software. At a previous employer, the nice Tek scopes (and spectrum analyzers, and other stuff) lived in sort of an IT limbo - they were on an isolated network, which was good, but they didn't get any backup or antivirus support, which was not so good. Matt Roberds
On 8/26/2014 4:35 PM, mroberds@att.net wrote:
> ChesterW <iamsnoozin@yahoo.com> wrote: >> What do you think? > > Short term, if the PC you are using it with craps out, you need to come > up with another PC. This is usually not too hard to do, either at work > or at home; it might be a good idea to keep your n-1 PC around a little > longer as a backup. > > Long term, you have to be prepared to keep a PC running, probably on XP > or 7, for however long you own the scope. (Or a virtual machine of XP > or 7, as long as the VM's USB support is fast enough for the scope.) > > The fact that they claim to support XP through 8 is a good sign. You > can probably count on getting updated drivers/software for 9 in a few > years. After that, you may or may not get further software support.
I have been wanting to switch to Linux for some time now but never made the jump. I may be mistaken but I believe a Linux compatible device is much less likely to be broken by new versions of the OS, no? I won't buy a LogicPort from Intronix because they don't support Linux or Android. -- Rick
Hi Rick,

On 8/26/2014 2:38 PM, rickman wrote:

> I have been wanting to switch to Linux for some time now but never made > the jump. I may be mistaken but I believe a Linux compatible device is > much less likely to be broken by new versions of the OS, no? I won't buy > a LogicPort from Intronix because they don't support Linux or Android.
Well, it depends on which OS features/support libraries the device depends. If it is handled entirely in userland -- and only uses the OS for the hardware interface(s) to the device -- then it is most readily continuously supported. E.g., (backward) compatibility libraries can provide all the hooks for the OS's API that *may* change, over time (this is how NetBSD/FreeBSD handle "old binaries" without requiring re-compilation). If, OTOH, support is tightly integrated into the kernel, then all bets are off. E.g., I have never ported my 9-track tape driver "forward" as the NetBSD kernel evolved -- simply because it was easier to just keep an old kernel running for those times when I needed to read/write half inch tape! Similarly, FreeBSD at one time (perhaps still!) had some low level hacks to support a special purpose LORAN-C receiver. Expecting these things to be supported moving forward assumes someone will have a "vested interest" in making that happen.
Den onsdag den 27. august 2014 00.06.44 UTC+2 skrev Don Y:
> Hi Rick, > > > > On 8/26/2014 2:38 PM, rickman wrote: > > > > > I have been wanting to switch to Linux for some time now but never made > > > the jump. I may be mistaken but I believe a Linux compatible device is > > > much less likely to be broken by new versions of the OS, no? I won't buy > > > a LogicPort from Intronix because they don't support Linux or Android. > > > > Well, it depends on which OS features/support libraries the device > > depends. If it is handled entirely in userland -- and only uses > > the OS for the hardware interface(s) to the device -- then it is > > most readily continuously supported. E.g., (backward) compatibility > > libraries can provide all the hooks for the OS's API that *may* > > change, over time (this is how NetBSD/FreeBSD handle "old binaries" > > without requiring re-compilation). > > > > If, OTOH, support is tightly integrated into the kernel, then all > > bets are off. E.g., I have never ported my 9-track tape driver > > "forward" as the NetBSD kernel evolved -- simply because it was > > easier to just keep an old kernel running for those times when I > > needed to read/write half inch tape! > > > > Similarly, FreeBSD at one time (perhaps still!) had some low level > > hacks to support a special purpose LORAN-C receiver. Expecting > > these things to be supported moving forward assumes someone will > > have a "vested interest" in making that happen.
I think a big thing is that you'll be able to reinstall on any hardware you find that is compatible without having to mess around with getting permission from a Microsoft server -Lasse
On Tue, 26 Aug 2014 17:38:10 -0400, rickman wrote:

> On 8/26/2014 4:35 PM, mroberds@att.net wrote: >> ChesterW <iamsnoozin@yahoo.com> wrote: >>> What do you think? >> >> Short term, if the PC you are using it with craps out, you need to come >> up with another PC. This is usually not too hard to do, either at work >> or at home; it might be a good idea to keep your n-1 PC around a little >> longer as a backup. >> >> Long term, you have to be prepared to keep a PC running, probably on XP >> or 7, for however long you own the scope. (Or a virtual machine of XP >> or 7, as long as the VM's USB support is fast enough for the scope.) >> >> The fact that they claim to support XP through 8 is a good sign. You >> can probably count on getting updated drivers/software for 9 in a few >> years. After that, you may or may not get further software support. > > I have been wanting to switch to Linux for some time now but never made > the jump. I may be mistaken but I believe a Linux compatible device is > much less likely to be broken by new versions of the OS, no? I won't > buy a LogicPort from Intronix because they don't support Linux or Android.
It depends on the nature of the interface. If it requires a special driver, you really need the source code so you can compile for new kernel versions. If it's just a USB or TCP/IP or other standard interface, it's relatively robust to version changes. Not perfectly immune, just relatively. We got bit on a special PLC controller with a linux driver for a CNC machine - the supplying company wouldn't give us the source, just the binary driver that was locked to a particular (now antiquated) version of Linux. We took it off the network given the risk, which is a nuisance.