|
kid sinister posted:So I got a brand new, shiny Rigol DS1054Z after such glowing recommendations here. I'm still trying to figure out the thing. I've only ever used my 1950s o scope. I do like how simplified the positioning controls are! That Auto button is pretty nice too. I'm still trying to figure out the purpose of Trigger though. Let me see I got the gist of it. Is Trigger the way to make the oscilloscope put the interesting part of the waveform on screen? Yes, the trigger is how you make the scope show the interesting part of the waveform. It's set as a voltage level and either a rising or falling edge (or both). When the signal hits that voltage in the correct direction, it resets the screen and starts displaying the trace from that point. On an old analog scope it would do this by holding the horizontal sweep in place and then literally triggering it at the proper moment, sweeping across the screen. On a digital scope the trigger simply causes the scope to start recording and displaying data (and also a little bit of the buffer before, usually). On digital scopes and fancier analog ones, the trigger level can also appear on the screen so you can visually line it up with the waveform. On the 1054 it's an orange line iirc. So for instance if you wanted to stabilize a 120v sine wave, you could simply set the trigger to 60v and rising and the scope will reset every time the waveform is rising and it passes 60v. The appearance will be a stable wave on the screen. Adjusting the trigger level will apparently move the waveform left and right, because the sweep will start a little sooner or later. If you wanted to look for noise spikes in a digital signal, like I was dealing with over the summer, you might watch the trace to see where your digital logic levels are, and then set the trigger to a rising edge somewhere above the expected levels. A spike past that point would reset the scope so you could see the effect of the spike on the logic. You can also use the single-sweep mode to freeze and record that single event for analysis. The attenuation multiplier on your probes serves two purposes. First, it just divides the input value by that much, so e.g. a 600v input signal will present 60v to the scope's front end. That's useful just for out-of-range signals. Second, switching in the multiplier increases the resistance and decreases the capacitance of the probe, which is helpful if you're trying to cleanly read a high-frequency signal. Note also that your probes have a capacitance adjustment screw on the bottom for calibration. Hook them up to the test signal on the front plate (has a symbol that looks like a square wave) and turn that screw until the edges of the signal are sharp and clean. e: quoted for new page Sagebrush fucked around with this message at 02:46 on Sep 17, 2019 |
#
?
Sep 17, 2019 02:42
|
|
|
|
| # ? May 9, 2024 16:49 |
|
|
Maybe one of you guys could clue me in on a search term to find a connector I need to do a simple project. It involves several RGB LED strips radiating from a central point, much like a clock or compass rose. I need a right-angle male header to connect every other pair of strips to each other so they are all in one series. Here's my crude MSPaint of what I"m talking about :  I've tried searching for "stacked right-angle headers" or "linear right angle headers headers" or "flat right angle headers" but I'm not finding them. Surely they must exist. There is probably also a waaay better and smarter way of accomplishing this. edit: Forgot to mention that I can't use right angle connectors meant for LED strips because the ends of the strips are nowhere near each other. Unperson_47 fucked around with this message at 03:39 on Sep 17, 2019 |
|
#
?
Sep 17, 2019 03:34
|
|
|
Unperson_47 posted:Maybe one of you guys could clue me in on a search term to find a connector I need to do a simple project. https://www.digikey.com/products/en...e=1&pageSize=25 Digikey > Connectors, Interconnects > Rectangular Connectors - Headers, Male Pins > set number of rows to 3 and specify right-angle through-hole pins.
|
|
#
?
Sep 17, 2019 03:38
|
|
|
Sagebrush posted:Yes, the trigger is how you make the scope show the interesting part of the waveform. "switching in the multiplier"? You're talking about the internal circuitry of the probe switch? Yeah the manual said something about calibrating all 4 probes with the internal square wave generator, making the leading edges flat. I did it.
|
|
#
?
Sep 17, 2019 03:39
|
|
|
Yeah, when you flip the switch on the probe from 1x to 10x it connects a little voltage divider circuit inside that gives a 10:1 ratio.
|
|
#
?
Sep 17, 2019 03:41
|
|
|
poeticoddity posted:https://www.digikey.com/products/en...e=1&pageSize=25 Wow, that was fast. Thanks! For anyone else that might need these: these Digikey results lead me to a model number to look up which lead me to the ebay keywords I needed which in this case is "triple row right angle headers." I use eBay because I'm cheap as hell and buy headers in 40 pin strips and snip them to the size I need. Unperson_47 fucked around with this message at 04:12 on Sep 17, 2019 |
|
#
?
Sep 17, 2019 03:41
|
|
|
Remember!! You must set the probe switch AND the oscilloscope setting to the same value (ie both 1x or both 10x). Otherwise it'll display wrong on the screen and you'll feel like an idiot when you're trying to work out why your circuit is generating 15 v instead of 1.5 V. Do not ask how many times I have done this... It's really easy to bump the switch on the probe
|
|
#
?
Sep 17, 2019 09:12
|
|
|
How are the probes that come with the rigol? A lot of those cheap scopes are garbage. I use my included probes for poking at stuff that might explode or short out, and use a set of Tektronix probes I got on Ebay for making actual measurements.
|
|
#
?
Sep 17, 2019 14:12
|
|
|
Sagebrush posted:Yeah, when you flip the switch on the probe from 1x to 10x it connects a little voltage divider circuit inside that gives a 10:1 ratio. Also it should be noted that you should leave them set to 10x unless you know a specific reason why what you're trying to do needs 1x, since as noted 10x puts less load on the circuit and provides at least a tiny amount of "protection" (though don't depend on it) against you whacking the thing across a high voltage and blowing up the front-end. babyeatingpsychopath posted:How are the probes that come with the rigol? A lot of those cheap scopes are garbage. I use my included probes for poking at stuff that might explode or short out, and use a set of Tektronix probes I got on Ebay for making actual measurements. They're, idk, fine? I'd describe them as purely "fine". They're not fantastic, not awful.
|
|
#
?
Sep 17, 2019 14:24
|
|
|
What do bad probes do to signals under measurement?
|
|
#
?
Sep 17, 2019 14:25
|
|
|
kid sinister posted:my 1950s o scope Don't talk about your grandmother that way. 
|
|
#
?
Sep 17, 2019 14:38
|
|
|
Are there any EE modelling tools out there where you can just import like Eagle PCB files to test how signals will behave and determine the kind of terminations needed (also reflections, insertion loss, phase balance etc)?
|
|
#
?
Sep 17, 2019 14:40
|
|
|
BattleMaster posted:What do bad probes do to signals under measurement? The bench scope that my former team let the software engineers use had lovely probes which added quite a bit of capacitance to the signal under test. It wasn't always a big issue but at one point I tried to probe a 26MHz clock going into a part which is hugely sensitive to the cleanliness of said clock, and the part would always just seize up and stop. We ended up taking it to a EE's bench scope with lower capacitance probes and then it worked fine 
|
|
#
?
Sep 17, 2019 15:09
|
|
|
Shame Boy posted:They're, idk, fine? I'd describe them as purely "fine". They're not great, not terrible.  But yeah they're totally fine for everyday use. I don't know how they stack up to high end probes but I've never had problems using them on analog or digital signals within the scope's bandwidth.
|
|
#
?
Sep 17, 2019 15:28
|
|
|
All my Rigol probes have unreliable 1x/10x switches so it usually takes a few tries to get it to actually stay in the mode it indicates. High end probes are usually only 10x, probably because the 1x mode is only useful for a handful of low speed measurements. Keysight probes at work have spring loaded probes, it's really great and I wish there were cheap options using something like the super cheap P75 pins. (We use modified Noga indicator stands for probe holding so the fragility isn't an issue)
|
|
#
?
Sep 17, 2019 16:45
|
|
|
Bedshaped posted:Are there any EE modelling tools out there where you can just import like Eagle PCB files to test how signals will behave and determine the kind of terminations needed (also reflections, insertion loss, phase balance etc)? Yes and no. You need models for each bit of the circuit, which usually means it is really hard to do the whole design at once. Theoretically Altium can do that type of sim in-program if you have electrical models for every part. Usually, I use a program like ltspice to analyze a simplified version of a circuit, and incorporate what I learn into the full design.
|
|
#
?
Sep 17, 2019 18:17
|
|
|
I've never used it myself but the Cadence SI tools use mostly IBIS models for device I/Os with some kind of 3d field solver for the PCB trace model. I think the IBIS model is a simplified model that gives the parameters for a specific I/O pin in a given mode (like for DDR3 outputs/inputs), this is specific for each pin since the bond wire length and actual routing on the die is unique. This is obviously a lot simpler to model and configure than say a full SPICE model of an FPGA (which would also be insanely slow to simulate). The end result is usually an eye diagram at a specific node for each trace you simulate, or a group of signals for timing analysis. This is way overkill for most applications. DRAM buses are kind of special since they're extremely high speed for a parallel single ended bus and most buses use multiple DIMMs on top of that so simulation can be a big help. In other cases it's usually sufficient to control PCB trace impedance, terminate properly, and length match traces reasonably well. For most CMOS level signals a source series termination with star-routing is almost always fine.
|
|
#
?
Sep 17, 2019 22:22
|
|
|
The Rigol probes are simple 10X probes. The do give you 4 for all 4 channels, which is nice. I just discovered that you can pull the tip off to get a point instead of the hook! My 1950s scope just used twisted pair. All the attenuation was done in the scope with a switch. Speaking of attenuation, how do I set that in my Rigol to match the probes? Fake edit: if I switch the scope to 10X and the probe to 1X, would I technically amplify the signal?
|
|
#
?
Sep 17, 2019 23:10
|
|
|
I just unlocked all options on a Rigol MSO5104, which is already nice out of the box. This one was also ordered from Amazon just like two DS1054Z scopes before. Those two came unlocked from the seller, though. The MSO5104 did not. The unlock enabled all sorts of goodies: onboard dual arbitrary waveform generator, 350 MHz BW, deep memory sampling, a bunch of protocol-specific stuff, and power analysis. Tiny complaint: While it is cool and good that Rigol ships their 4-channel scopes with 4 probes, it would be equally nice to get logic analyzer harness set with that scope instead of having to order it separately. Edited for picture. 
TotalLossBrain fucked around with this message at 00:54 on Sep 18, 2019 |
|
#
?
Sep 17, 2019 23:27
|
|
|
csammis posted:The bench scope that my former team let the software engineers use had lovely probes which added quite a bit of capacitance to the signal under test. It wasn't always a big issue but at one point I tried to probe a 26MHz clock going into a part which is hugely sensitive to the cleanliness of said clock, and the part would always just seize up and stop. We ended up taking it to a EE's bench scope with lower capacitance probes and then it worked fine I once listened to a talk given by a circuits professor who said that he studied/reverse engineered an oscillator circuit described in Jack Kilby's integrated circuit patent, and he claimed that the oscilloscope probe capacitance, not shown in the circuit diagram in the patent, completed the circuit, and actually made it function as advertised. Bedshaped posted:Are there any EE modelling tools out there where you can just import like Eagle PCB files to test how signals will behave and determine the kind of terminations needed (also reflections, insertion loss, phase balance etc)? I don't know of any free softwares that are oriented towards this kind of application. There is circuit simulator software you can buy which have integrated electro-magnetism (EM) simulators. You can import circuit artwork and a description and physical model of the circuit board stack-up into the EM simulator, and it will solve some form of the Maxwell Equations and provide you with a circuit model of the wiring, which you can then interrogate in the circuit simulator, and get physical quantities like insertion loss, return loss, amplitude & phase balance. You can also connect the circuit model for the wiring to circuit models for the components and get the simulated response of the entire circuit+wiring. Most circuit design doesn't warrant going to this level of detail, but there are a few applications where this is very useful. Keysight's Advanced Design System (ADS) & National Instruments' Microwave Office are competing softwares for this function, but you have to pay for them. You are in school though, right? If so, I think both companies offer free licenses to students to use their software. ANSYS has a really good EM simulator (High Frequency Structure Simulator (HFSS)) and I have heard that they have a circuit simulator, but the circuit simulator isn't very popular, I think. They also are more stingy about student licenses, I hear, although I haven't personally ever looked into requesting a student license from ANSYS.
|
|
#
?
Sep 19, 2019 11:17
|
|
|
You also wouldn't actually want to. This discussion comes up every so often here, and the truth is, if you're simulating, it will only be a tiny subsection of your circuit, and you will be adding/removing/replacing components to more accurately model the electrons, with stuff that isn't in your schematic. It's just not comparable. As an example, I've been doing some simulation for work this week to solve a ringing issue I'm having, and I've been spending hours adding little parasitic inductors everywhere to match what I've seeing, and only then using it to come up with solutions. Edit: here's the last time that discussion came up, where I wrote a lot more words https://forums.somethingawful.com/showthread.php?threadid=2734977&pagenumber=460&perpage=40#post494014802 ante fucked around with this message at 16:40 on Sep 19, 2019 |
|
#
?
Sep 19, 2019 16:35
|
|
|
What you are saying might be true for 99% of circuits, but there are circuits for certain applications where it is important to actually co-simulate the components with the circuit board wiring during the design process. These applications tend to be concerned with figures of merit like phase balance, insertion loss, & reflection, which were mentioned. There are also certain kinds of circuits, admittedly they are not very common, where basic circuit functions are are synthesized from the circuit board traces themselves. To not co-design the wiring with the components here would obviously be a a big mistake and would be designing in the dark.
|
|
#
?
Sep 19, 2019 17:36
|
|
|
silence_kit posted:What you are saying might be true for 99% of circuits, but there are circuits for certain applications where it is important to actually co-simulate the components with the circuit board wiring during the design process. These applications tend to be concerned with figures of merit like phase balance, insertion loss, & reflection, which were mentioned. I've been building something like a TMS device over the last few months when I've had time (I was bored), and it's been instructive how much attention I need to pay attention to the wiring and physical layout of the board. It needs to dump about a joule through a really precisely built inductor - which has been a pain. Those tiny little 6 pin boost converter chips are magic, can happily take 5v and put out 50v at a few hundred milliamps which'll happily drive a capacitor bank. I wanted to simulate the thing in SPICE but I don't think it's going to be worth building a model over just scoping the thing and checking against calculations, every single trace or parasitic inductor is as important as the thing I'm actually trying to power up.
|
|
#
?
Sep 19, 2019 19:21
|
|
|
sharkytm posted:Don't run the 5V from your mirror directly into your phone. If it's actually 5V, it's probably supplied from the BCM (or your car's equivalent), and it sure isn't going to like a 2.4A load when it expected to be driving a couple open collector inputs. Follow-up question: The BCM isn't going to like this because there's a logic circuit looking for a specific amperage to the window switch? This isn't an issue of a fuse capping the load on a particular circuit? A friend sent me this: https://www.amazon.com/Converter-Adjustable-Voltage-Regulator-Arduino/dp/B07PDGKBQN/ To regulate the voltage at 2.5 amps out to the power leads on the USB cable. I know the best answer is to run to a switched 12V source, but access through the door boot really sucks.
|
|
#
?
Sep 20, 2019 00:43
|
|
|
how practical/applicable is winding your own air-core inductors for general hobby-tier electronics projects? it's a hella tidy and cheap way to make any n all desired RF filters for SDR stuff, but I've barely worked w inductors aside from that and the passives that are called for w RF seem atypical across the board
|
|
#
?
Sep 20, 2019 01:28
|
|
|
I was winding wye coils in high school. You calculate roughly the size of the coils and length of wire, then you wind it. You attach it to a frequency generator on the input, oscilloscope on the output, do a few wipes up and down the waveform generator and find the peak on the oscilloscope. Voila, you know your resonance and efficiency. If nothing else, you learn a ton about induction design. And yes making a winding jig is worth the effort.
|
|
#
?
Sep 20, 2019 01:37
|
|
|
I'm a jeweller/metalsmith by training so I'm extensively equipped to wind whatever coil i want in whatever gauge i want to a professional standard, part of why i wanna lean on it as much as i can. so far i've been trusting the air-core inductor design calculators and physically tweaking the coils until they play nice(r), but i'll def do some proper diagnostics like you described soon also: decorative scrollwork/sinuous vine-mimicking layouts etc are a critically-underutilized feature of most circuit designs, smh Ambrose Burnside fucked around with this message at 01:53 on Sep 20, 2019 |
|
#
?
Sep 20, 2019 01:51
|
|
|
Ambrose Burnside posted:I'm a jeweller/metalsmith by training so I'm extensively equipped to wind whatever coil i want in whatever gauge i want to a professional standard, part of why i wanna lean on it as much as i can. so far i've been trusting the air-core inductor design calculators and physically tweaking the coils until they play nice(r), but i'll def do some proper diagnostics like you described soon That's really cool actually. My mom's been getting into metalworking to make jewelry and quite a lot of the stuff she does is similar to things I do for electronics (board etching / pattern etching, soldering, wire working, etc.) so we talk about it a bunch and give each other tips  Anyway yeah there's not anything wrong with winding your own coils, and it's generally the best way to go for radio stuff (as you've mentioned) since that requires small, finely-tuned inductance values, but for anything larger like a power inductor or transformer you're gonna need a core to wrap it around. You can buy them pretty cheap on amazon if you wanna get into that kinda thing, I find winding my own stuff relaxing so I do it now and then. Just search "ferrite core" and you find a bunch of stuff like this: Generic beads: https://www.amazon.com/Inductor-ferrite-10X6X5mm-Isolator-Filter/dp/B07N2MMDYP/ https://www.amazon.com/Toroid-Aperture-BN43-202-Ferrite-Torroid/dp/B01959LHN0/ Sticks you can cut to the length you need: https://www.amazon.com/CynKen-10mm140mm-Ferrite-Loopstick-Antenna/dp/B01MXQPP04/ Transformer cores: https://www.amazon.com/Transformer-ferrite-Isolater-Vertical-Bobbin/dp/B014LYMH4C/ https://www.amazon.com/PQ2620-Transformer-ferrite-Isolator-Inductor/dp/B07PV1BQW9/ etc. Shame Boy fucked around with this message at 14:24 on Sep 20, 2019 |
|
#
?
Sep 20, 2019 14:21
|
|
|
The inductors I was running were 1-2 MHz at a few thousand volts for power transfer. This was back when I made Tesla coils as was looking into solid state primaries. Nothing was more fun than the capacitor made of plastic plates and foil fly everywhere when it was discharged if you didn't tape it together.
|
|
#
?
Sep 20, 2019 14:34
|
|
|
I'm routing a differential pair through a bunch of parallel connectors, when I use the Altium differential pair tool it does this. Whereas I would normally do something like this  Which is preferable?
|
|
#
?
Sep 23, 2019 18:46
|
|
|
Your approach is better. Altium has just made two little antennae
|
|
#
?
Sep 23, 2019 21:35
|
|
|
While I might go with the second one as well, I doubt it makes much if any difference. They both look like they have good matching so the differential mode transmission should be largely identical. Routing the trace through the vias adds the slight mismatch of going through the via annular ring (which is smaller than the trace width), routing straight adds some tiny extra amount of stub length. When picking stub lines off a single diff trace to go to multiple connectors I doubt the extra "antenna" length matters, the entire bus is basically made of antennas in that case.
|
|
#
?
Sep 23, 2019 21:57
|
|
|
Splode posted:Remember!! You must set the probe switch AND the oscilloscope setting to the same value (ie both 1x or both 10x). Thank you so much for pointing this out. Suddenly, plugging other BNC equipment directly in makes SO much more sense on screen. It came from the factory set at 10X and I had no idea why measurements were off by a factor of 10.
|
|
#
?
Sep 28, 2019 02:39
|
|
|
kid sinister posted:Thank you so much for pointing this out. Suddenly, plugging other BNC equipment directly in makes SO much more sense on screen. It came from the factory set at 10X and I had no idea why measurements were off by a factor of 10. Hahaha yep gets everyone every time
|
|
#
?
Sep 28, 2019 22:49
|
|
|
I need a single normally-off, momentary-on push button switch to control two completely separate battery powered circuits. I've been looking at DPST and DPDT switches on mouser, but the schematics in the datasheets look like the circuits are crossed, even if maybe just sharing a ground, when the button is pressed. Does something exist like what I'm looking for? I also need it to be low profile, and I'm thinking maybe I'll just 3d print an enclosure for two SPST buttons and put a plate over the top of them so they are pressed together. This is for a costume, so, low order quantity and all that.
|
|
#
?
Oct 3, 2019 15:58
|
|
|
armorer posted:I need a single normally-off, momentary-on push button switch to control two completely separate battery powered circuits. I've been looking at DPST and DPDT switches on mouser, but the schematics in the datasheets look like the circuits are crossed, even if maybe just sharing a ground, when the button is pressed. Does something exist like what I'm looking for? I also need it to be low profile, and I'm thinking maybe I'll just 3d print an enclosure for two SPST buttons and put a plate over the top of them so they are pressed together. This is for a costume, so, low order quantity and all that. How much current are you dealing with? Maybe you could do a regular switch to flip the two circuits simultaneously with transistors or mosfets?
|
|
#
?
Oct 3, 2019 16:07
|
|
|
armorer posted:I need a single normally-off, momentary-on push button switch to control two completely separate battery powered circuits. I've been looking at DPST and DPDT switches on mouser, but the schematics in the datasheets look like the circuits are crossed, even if maybe just sharing a ground, when the button is pressed. Does something exist like what I'm looking for? I also need it to be low profile, and I'm thinking maybe I'll just 3d print an enclosure for two SPST buttons and put a plate over the top of them so they are pressed together. This is for a costume, so, low order quantity and all that.
|
|
#
?
Oct 3, 2019 16:15
|
|
|
armorer posted:I need a single normally-off, momentary-on push button switch to control two completely separate battery powered circuits. I've been looking at DPST and DPDT switches on mouser, but the schematics in the datasheets look like the circuits are crossed, even if maybe just sharing a ground, when the button is pressed. Does something exist like what I'm looking for? I also need it to be low profile, and I'm thinking maybe I'll just 3d print an enclosure for two SPST buttons and put a plate over the top of them so they are pressed together. This is for a costume, so, low order quantity and all that. What do you mean by it looks like they are crossed? I'm looking at DPST datasheets and I'm seeing stuff like this:  I would assume that the dashed line means they are physically connected by not electrically connected.
|
|
#
?
Oct 3, 2019 16:16
|
|
|
ickna posted:How much current are you dealing with? Maybe you could do a regular switch to flip the two circuits simultaneously with transistors or mosfets? I've hacked a few vapes that run on 18650s so I don't have to press the button on the unit itself. Each switch will activate one small aquarium pump running at 4.5v (on AAs) to pull air through the vape, and one vape. Basically I'm looking to have a few little mini fog machines strapped on a belt, with buttons hidden under my clothes elsewhere so I can make smoke come out of my cuffs, collar, or the bottom of my cloak. I could do what you suggest easily enough, I'm just kinda being lazy about it.
|
|
#
?
Oct 3, 2019 16:17
|
|
|
|
| # ? May 9, 2024 16:49 |
|
|
Cojawfee posted:What do you mean by it looks like they are crossed? I'm looking at DPST datasheets and I'm seeing stuff like this: Whichever switch that is, I didn't see it  I'll go look some more.
|
|
#
?
Oct 3, 2019 16:18
|
|