Found a TDS460 with jittery trigger - worth getting as a first decent scope?

[NRoach44]

Yeah, I already got a HP 1662C for $200 shipped, which I'm quite happy about and it did help with an assignment.

As it is I don't see the likely hood of a 4channel reasonable bandwidth scope that's in good condition appearing for a price within my range. I just felt I needed to demonstrate this to the people who keep saying "you aren't looking hard enough"

Thanks guido, this is the sort of thing I was looking for when I asked on the forum. I think I'll leave buying old scopes until I have the space, time and willingness to repair them.

[rs20]

Yes it does. Have you ever used a scope which used ETS? Glitches will show as an unstable part of a signal and because ETS is usually slow it stays there long enough to get noticed. ETS is not some kind of glitch filter! The only really important thing when using ETS is that you need a stable trigger. Once you know there is a glitch you can adjust the trigger accordingly to get a better picture of the glitch.

Mmm, if you define "show" as "display in the form of glitchy, apparently-high-frequency corruption that is completely confusing to a beginner, completely irrelevant to a discussion prompted by a self-styled beginner, and when discovered, would necessitate that the user in question either figure out an advanced triggering mode* to isolate the glitch or switch to an oscilloscope with non-awful realtime performance like a ds1054", then yeah, I guess you are pedantically, technically correct.

* Yes, pulse width triggering mode is advanced for a beginner.

[vk6zgo]

Scopes may not be as readily available in Oz as in the USA or Europe,but they are not that scarce.
It's Australia,for Pete's sake!------Not some coral atoll in the middle of the Pacific.

Perhaps the OP should look around a bit more.
If he is poor,by the time he has spent the money & time fixing the TDS 460.he could buy something more appropriate.

I'm going to try and be as less confrontational as possible, but please don't keep saying I "should look around more".

I'm only really going to spend up to about ~$300 total on the scope if it's used (and if it's approximately comparable the 1054z, because I'd probably have to buy probes, won't have the warranty, etc.) and unless it's miles better than the 1054z that isn't going to change.

Take into consideration that I probably will use the extra inputs on the Rigol.

Just to get it out of the way, let's look at element14: http://au.element14.com/webapp/wcs/stores/servlet/Search?catalogId=15001&langId=43&storeId=10184&categoryId=800000019504&sort=P_PRICE&pageSize=25&beginIndex=1&showResults=true
Their cheapest scope is a Tenma (their generic brand) 2ch 25MHz 250MSa/s for $527 pre tax

That also appears to be the same basic unit that Altronics sell.

Let's look at craigslist Perth's electronics section: http://perth.craigslist.com.au/search/ela
It's not even worth searching for oscilloscope.

Gumtree perth: http://www.gumtree.com.au/s-perth/oscilloscope/k0l3008303
Occaisionally they do pop up here, but it's either terrible or I don't have the cash while it's available.

Ebay (filtering for used and worldwide):

Kikusui: Freight. It's going to be bloody expensive.
Triumph: A nice relic but I'm not collecting.
Tek 7633: Produced from 1974. I'll pass.
Tek 2440: Actually doesn't look too bad, but holy shit that power on test.
Tek 485: 2 chan, 350MHz. I feel I'd use the channels more than the frequency. $360
Tek TDS320: Over cost without freight.
Tek 2465: not bad, but over cost without shipping again
Fluke 92B: Over cost, not worth it IMO
Protek p2020: 20MHz. Pass
Tek 2465: About same cost as the Rigol, marked as "FOR PARTS"
Tek 465B: Half as many channels as the Rigol, for 120% of the price
HP 1715A: $200, might be worth considering if it wasn't for Parts/Repair, with no probes.
Tek T935A: That frequency and that cost? Not even with the "Great! Cool!" in the title.
Tek 485: Almost at the budget but is sold as-is.
HP 1722B: As previous but ancient and only dual input.
Hitachi V355: Dual input, 35MHz. Looks fairly simple.
Tek 2440: Again, looks OK but the seller mentions "The device has intermittent start-up errors, fails the self test, and has a weak display... adjustments and repairs will be required and should be expected."
Tek 2465A: Over budget but not bad. Has been red stickered but apparently the seller can't find any fault.
Tek 2430A: No shipping, not touching it. (Probably expensive)
Tek 222: Looks like a nifty little machine, but doesn't use standard BNC probes, need an external power brick (probably actually a good thing) and is only dual input.

This is where I gave up (I'm not going to bother writing more about them). There is another Tek 222 later on that I may consider but it still comes down to the following:

Do I spend $200-$300 on an Okay-ish scope with 2 channels, and probably end up switching to the Rigol later anyway? These scopes also don't have serial decoding.

I am not touching a PC scope as I would like my PC to not be tied to old hardware or software.


Edit: Clarifications.

Gumtree & Craigslist are a waste of time,so you are left with eBay.
Again,if it isn't in Oz,it's a waste of time & money.

Another possibility is the Perth Hamfest,which is in the beginning of August,& often has cheapie Oscilloscopes.

If you object to all & every one of the 'scopes on your list,it looks like you may as well save up for the Rigol.
I don't have one,or any DSO,for that matter,but the Rigol brand has a good reputation on this forum,so it would be my choice.

[edavid]

Tek 2465: Probalby needs new caps in the power supply (mine did). The A version has a big chance of leaking caps.

I disagree with the probably; I have several, and none of them have bad power supply caps.  Only the 2465B has the leaking cap problem.

Tek TDS320. Same as the 460; leaky caps, schematics, performance.

It's a lot different, since it's 500MSPS.  It's a nice scope, except for the lack of memory.  Schematics are available for the TDS340A/TDS360/TDS380, and are usable for the TDS320.

[NRoach44]

Gumtree & Craigslist are a waste of time,so you are left with eBay.
Again,if it isn't in Oz,it's a waste of time & money.

Another possibility is the Perth Hamfest,which is in the beginning of August,& often has cheapie Oscilloscopes.

If you object to all & every one of the 'scopes on your list,it looks like you may as well save up for the Rigol.
I don't have one,or any DSO,for that matter,but the Rigol brand has a good reputation on this forum,so it would be my choice.

If I search ebay with "up to $300, australia only" on top of the previous search, there's 28 results. I'm not going to go through them all, but long story short there's only mostly 2chan low-bandwidth scopes, and vintage machines.

I went to the Perth Hamfest last year, there was the 10MHz DSE scope I have now, and a vintage dual channel there too. I went with the DSE as it would actually fit in my bag. I'll probably end up going this year as well, but I doubt there'd be a decent scope there this time either.

I am definitely going to purchase the Rigol now, as a few of the posts here (when they aren't arguing about ETS) go on to say something along the lines of "If it's old it's not worth it for a beginner*, doubly so if it's broken"

Edit: *at least for my use cases any way.

[vk6zgo]

I remember now,you're the bloke who uses public transport.

Unfortunately everything is geared to cars,so anything that isn't light becomes a nightmare.
That said,I took an old Desktop,complete with CRT monitor & keyboard, home on the bus back in the 1990s-----never again!

I Googled for the TDS460,& lots of people seem to have the "jittery" problem.

I was surprised that it was quite a compact unit--some of the  really old DSOs from the late '80s,early '90s
had massive great raster displays & were nearly as heavy as a 7613,plus were an awkward shape.

Most of the early DSOs were pretty horrible for realtime use due to lack of memory depth
To get around this,they reduced the sample rate for long times/div,which is OK if you aren't trying to look at analog video at field rate,or 50Hz hum on a pulse train.
Aliasing makes a total mess of things.

That put me off DSOs for ages,but modern ones are light years better!

[NRoach44]

Heh, I've got a friend who is on his Ps and has a car now  I'm on the "do 25 hours of driving and wait 6 months" stage myself, but due to living on campus at Uni I can't just ask the parents to supervise me.

The other thing is I'd rather not get another big machine like the HP 1662C as it's already enough of a pain in the arse to move my stuff around before I got the HP.

Emona's Perth branch isn't too far away. I'm going to try and call them to see if they have one in stock that I can pick up.

[Wuerstchenhund]

Yes it does. Have you ever used a scope which used ETS?

Yes, I have, actually quite a lot back in the days when scopes had very poor real-time sampling rates.

Glitches will show as an unstable part of a signal and because ETS is usually slow it stays there long enough to get noticed. ETS is not some kind of glitch filter!

Wrong! *Some* glitches (i.e. those that have a high occurrence rate) are somewhat visible on ETS, and if they're large enough you might even recognize that there's a problem. That doesn't change the fact that most common glitches and problems simply very likely won't show up in ETS.

The only really important thing when using ETS is that you need a stable trigger. Once you know there is a glitch you can adjust the trigger accordingly to get a better picture of the glitch.

Not in ETS, because even with advanced triggers ETS won't miraculously show signal deviations it hasn't shown before. And without ETS you're back to the bandwidth limitations of the poor real-time sample rate.

Fact is that you can't use ETS to probe unknown signals (or signals with unknown components), because for ETS to remain valid you have to establish that certain signal parameters are valid (i.e. repetitiveness) even before you start your measurements. You may or may not see something which you believe is true but unless the stringent requirements of ETS are met it very likely isn't. So ETS completely worthless for that, which probably rules out a majority of debugging tasks already.


Your reply shows very well the typical fallacy that people who don't appreciate the severe limitations of ETS often fall into. Sure, you can find situations where ETS shows you signal deviations, and you can certainly construe situations (which others have in this thread) where a signal sampled in ETS looks somewhat similar to the same signal sampled in real-time mode, but these are very limited border cases which don't change the physical realities, and it would be stupid to assume that just because it seems to you as ETS is perfectly useable that this is generally the case, because it isn't.

And even if you don't want to accept the maths/physics behind ETS then just ask yourself this: if ETS was as sufficient for common debugging work as you and others in this thread claim it is, why don't we see more new sample scopes? I guess Rigol and Siglent could drop their prices even further by using slower ADCs and using ETS instead. But reality is that sampling scopes have mostly disappeared aside from some cheap-ass USB "scopes" and the upper high end (i.e. bandwidths in excess of 50GHz to 100GHz) and even in the latter area they are dying out in favor of real-time scopes. Even more so, ETS mode is also slowly disappearing in newer scopes as well. And the fact that they are means that there's probably a good reason for it. Go figure.

You (and all the others that made this thread to one of the most embarrassing ones I've seen on this forum) can continue to keep your head in the sand of course but reality is that, as a general debugging tool, ETS is downright useless. You might actually have a better chance of finding random glitches in your signal by reading fish guts than with ETS.

[T3sl4co1l]

Mmm, if you define "show" as "display in the form of glitchy, apparently-high-frequency corruption that is completely confusing to a beginner, completely irrelevant to a discussion prompted by a self-styled beginner, and when discovered, would necessitate that the user in question either figure out an advanced triggering mode* to isolate the glitch or switch to an oscilloscope with non-awful realtime performance like a ds1054", then yeah, I guess you are pedantically, technically correct.

* Yes, pulse width triggering mode is advanced for a beginner.

One only needs to think in terms of what digital sampling means, in order to figure this out.  None of the detractors in this thread seem to appreciate DSOs as DSOs; if one wills that a cat should act like a dog, one will forever have the problem that it is a fundamentally different creature...

Not for a beginner?  Probable.  But a beginner will have trouble getting a stable display on an analog or RTS DSO all the same, so it's a pointless argument.

Wrong! *Some* glitches (i.e. those that have a high occurrence rate) are somewhat visible on ETS, and if they're large enough you might even recognize that there's a problem. That doesn't change the fact that most common glitches and problems simply very likely won't show up in ETS.

Just what kind of signals are you watching that have such infrequent glitches?  And at a bandwidth much greater than typical?

To use your very own example of SMPS startup conditions, consider that SMPS rarely peak above 1MHz, and those are the small ones.  That's 1000ns per cycle.  500ns per on-period.  At 100MS/s, you have a whopping 50 samples to observe the pulse width variation in real time.

If you're looking at glitchy high speed data, you simply have the wrong instrument, period.  You need a logic analyzer with that type of capability.  Preferably with a glitch trigger setting, or something like that.

So again, your arguments are invalid.

Fact is that you can't use ETS to probe unknown signals (or signals with unknown components)

Have you an example of such an alien signal?

So far in all of my work, I have not seen a single situation that was obfuscated by ETS.  (The prior case of runt square waves is easily solved with advanced trigger, or a single capacitor to the trigger input.)

And even if you don't want to accept the maths/physics behind ETS then just ask yourself this: if ETS was as sufficient for common debugging work as you and others in this thread claim it is, why don't we see more new sample scopes?

Duh.  Marketing.  ADCs are cheap enough that the analog signal path is the limitation now.  It would cost quite a bit more to design and build a preamp that's capable of driving the full ADC's analog bandwidth (which is usually a few times its max. Nyquist frequency).

They can also charge more for the scopes that do offer extreme bandwidth.

It's also a cost saving feature to avoid ET hardware and software.  Although with everything going through an ASIC, it's up to the functionality in there.  But even having that in the first place, Rigol (and Keysight) get tremendous value in performance and production cost.

and the upper high end (i.e. bandwidths in excess of 50GHz to 100GHz) and even in the latter area they are dying out in favor of real-time scopes.

When cost is no object, you might as well get the best.

I'm not concerned about the design of scopes that cost more than a small house.

Tim

[guido]

Tek 2465: Probalby needs new caps in the power supply (mine did). The A version has a big chance of leaking caps.

I disagree with the probably; I have several, and none of them have bad power supply caps.  Only the 2465B has the leaking cap problem.

Tek TDS320. Same as the 460; leaky caps, schematics, performance.

It's a lot different, since it's 500MSPS.  It's a nice scope, except for the lack of memory.  Schematics are available for the TDS340A/TDS360/TDS380, and are usable for the TDS320.

Guess your lucky with the PS caps but others are not (i wasn't for starters and you can find more examples if you google around). Those caps start to bulge. With regards to the leaky caps on the A5 board, you are right.

As to the TDS320, i missed the sampling rate spec. The schematics are there indeed, i use the doc for the monitor pcb in my 430A. But the schematic of the monitor is only partly there (if you have the full pdf, i'm very interested ). When i looked around on the net i think i saw two TDS manuals with schematics in total.

As for the topic starter; if you would buy a second hand scope it is a good thing to check first if you can find the schematics on the net. And if it is easy to get parts.