4ch scope?

[NateS]

Looking at scopes. $200-400. I need at least 2ch. I actually want a 4ch for a project I'm looking at (time difference of arrival. Honestly I could make it work with 1ch and then work out the timing part after I have my signal processing worked out, but I also do a bit of i2c/spi and I'm trying to decide if I should stretch for the Siglint/Rigol 4ch scopes or get a 2ch with a lot more fft data points than those.

[nickb]

What sort of times are you interested in? If you need small time intervals then you'll need a wide bandwidth scope and they don't some cheap. If not so small would a logic analyzer meet your needs?

[NateS]

Hmm good point -on bandwidth. The signals I'm looking for aren't that demanding (audio frequency) - but I do need to be able to resolve probably on the order of 5-10 us difference arrival. Eek - does that mean I need a 200Mhz scope? Hmmm that won't be a 4ch for sure. In all honesty - I've been wanting a scope for a looong time and the project said "Hey how about now?" All I really need out of this is to characterize the signal I need to amplify and feed that to my microcontrolller.

In that case - I guess the Siglint SDS1202X-E is starting to look a lot more attractive than the similarly priced Rigol DS1054Z. 200Mhz and 1000x bigger FFT. For protocol stuff, a protocol analyzer is cheaper than trying to do it with a scope.

[nickb]

It's the I2C / SPI rise tie that will determine the bandwidth: bandwidth = 0.45 / rise time. So for 10ns rise time (kinda slow), you'd need a 50Mhz scope. The resolving power is a property of the timebase and sampling rate. A 50MHz scope will sample at something like 200MSPS or more i.e a time resolution of 5ns. When looking at digital stuff you tend need lots of bandwidth as it's signal integrity that can be the killer. If you don't have enough bandwidth you could miss the nasty ringing that is causing your circuit to fail.

That said, a logic analyzer is my tool of choice for communications type stuff such as I2C and SPI. Get one with a built-in protocol decoder.

A mixed signal scope would be something to consider but they seem rather pricey. I suggest you go for a two channel 100Mhz scope and an 8 channel logic analyzer. You can get an Saleae 8ch analyzer clone for about $10 and they work great!


But the heart wants what the heart wants...

[NateS]

I did the math wrong. The rigol 4ch is gracious plenty to capture the entire window at more than enough resolution. Could I get by with a 100mhz 2ch scope? Probably. Could I get more fft points at the same price on a 2ch? Yep. Maybe if I sell my drums.

[Jazz P Bass]

Here is a fairly low priced 4 channel unit.
100Mhz.
https://www.ebay.com/itm/Agilent-HP-...4383.l4275.c10

[NateS]

Nice but it's analog. I'm trying to capture time difference of arrival on the order of a few of microseconds.

I thought I had found a super cheap 2ch PC based scope that would get me through the project, and could turn into my logic analyzer after getting a real scope, but alas. The Hantek 6022bl looks like a hard pass after a few reviews. I guess I'm back to saving up for a Rigol. That looks like the scope to beat for consumer DSOs.

[nickb]

The Hantek 6022bl is just 20Mhz and 48MS/s. It's not good enough for you needs.

[NateS]

Rigol DS1054z on the way. Sold a Roland D-50 and had enough. I seriously considered the 2ch Siglent 1102x-e for it's deeper memory and bigger fft, but I realized the Rigol has a very respectable memory compared to the $250-300 scopes and 1) it takes a while to get a million points at audio frequency, and 2) I can always do fft on the PC after capturing. The Rigol has a deeper memory than the Siglent's FFT, so as long as you don't mind the transfer step you can still have huge fft. And the 4ch will be nice for microcontroller stuff.

[NateS]

Man, it's been a long time since a machine made me feel this stupid. I finally got some triggering, but not on the things I expected. It did however show that my raspberry pi is woefully underpowered. If a glitch resets my encoders (ATTiny on each), the 3.3v rail dips to about half for a few hundred microseconds.

[NateS]

Success! First of all, I discovered that even though my Raspberry pi powers on and does stuff when powered off a hub - the 3.3v rail sags horribly when my i2c encoders glitch. (Bottom purple trace is an interrupt line).


Put a 2.5A PSU on the Pi and tonight I got i2c triggering working better, and got address triggering. (Still not happy with that noise on the interrupt line that seems to track SDA, but it doesn't seem to be hurting anything). Anyway - now for some real work.