Well the tweezer component tester is useful for SMT components, but you can get tweezer adaptors for a regular multimeter too...
If you need a very portable 'scope the little ones like that are handy, but they are usually very low performance compared to a bench unit. If you need a 'scope to troubleshoot, you probably need a whole load of other stuff from the bench too...
As for these specific units, no idea at all, caveat emptor.
If you need a very portable 'scope the little ones like that are handy, but they are usually very low performance compared to a bench unit. If you need a 'scope to troubleshoot, you probably need a whole load of other stuff from the bench too...
As for these specific units, no idea at all, caveat emptor.
Small and cheap. Hmmm, let me think .... not worth anything. Seen plenty.
You need an analogue oscilloscope, 100 MHz dual trace shouldn't be too expensive. They will show you way more detail and won't lie to you. Oh yeah, digital scopes have many capture gotchas. Same for meters, cheap ones lie and are extremely inaccurate.
Inexperienced people actually do need the most expensive instruments. They survive mistakes better for one. Back when meter movements were used, you learned to say "about so many volts" or whatever. Cheap digital is less accurate than the average meter movement. So you need to learn to take the accuracy specs and apply them to see where the true voltage may be. The range might surprise you. With many 3 1/2 digit (and more digits) the LSD (least significant digit, right most) doesn't mean anything at all. That is if the meter is in tolerance. Many brand new cheap ones I was asked to certify were not in tolerance right out of the box. They could not be corrected for all functions, all ranges either.
It took many years of scrimping and saving to get decent instruments when I was young. I got good ones each time I bought and that paid off. There is no cheap, quick answer. I bought used good ones, and they were still expensive. A $25, single trace, 500 KHz tube based scope (a Stark) was the first scope. Early 1970's and army surplus. Most other folks didn't even have a scope of any description.
You need an analogue oscilloscope, 100 MHz dual trace shouldn't be too expensive. They will show you way more detail and won't lie to you. Oh yeah, digital scopes have many capture gotchas. Same for meters, cheap ones lie and are extremely inaccurate.
Inexperienced people actually do need the most expensive instruments. They survive mistakes better for one. Back when meter movements were used, you learned to say "about so many volts" or whatever. Cheap digital is less accurate than the average meter movement. So you need to learn to take the accuracy specs and apply them to see where the true voltage may be. The range might surprise you. With many 3 1/2 digit (and more digits) the LSD (least significant digit, right most) doesn't mean anything at all. That is if the meter is in tolerance. Many brand new cheap ones I was asked to certify were not in tolerance right out of the box. They could not be corrected for all functions, all ranges either.
It took many years of scrimping and saving to get decent instruments when I was young. I got good ones each time I bought and that paid off. There is no cheap, quick answer. I bought used good ones, and they were still expensive. A $25, single trace, 500 KHz tube based scope (a Stark) was the first scope. Early 1970's and army surplus. Most other folks didn't even have a scope of any description.
Several strong points of disagreement here:
You absolutely don't need an analog 'scope, they weigh a ton and are huge, unreliable and have many disadvantages over modern 'scopes that far outweigh the disadvantages to me. Most of them don't have automatic measurement functions such as mean and rms voltage... DSO's may be only 8 bit, but the human eye reading a graticule is less accurate than even 8 bits! Plus 'scope front ends and probes are not that great for linearity as you get up to higher frequencies anyway, so 8 bit is probably a good practical compromise (some DSOs are 12 or 14 bit anyway). But small handheld units are often very limited by comparison to either kind of bench 'scope, and especially lack usable UI's and have poor trigger circuits. Of course any 'scope is better than no 'scope...
Analog meters are not and have never been highly accurate (or even reasonably linear!), but they are very good for monitoring rapid changes in voltage and current which a digital read out is not. I suggest a bench power supply should use analog current meters for this reason.
A medium cost digital multimeter is more accurate than you need for most purposes and they use dual-slope integrating ADCs which are inherently extremely stable, linear and monotonic. The linearity and repeatability is tied to the short-term stability of a quartz timebase, which is down at the ppm level or better, so for practical purposes they are completely linear and repeatable (at least for DC measurements)
The least significant digit actuallt does means a lot even if the calibration is well out, as slow drift can be accurately seen as the least sig digit clocks over one by one. You could never see that on an analog meter unless you have a laser mirror galvonometer and a hand lens! If you are making ratiometric measurements you get the full 3 1/2 digit accuracy out of a 3 1/2 digital multimeter whatever its calibration, that's the beauty of dual slope integration.
Ever looked up the tempco of an analog meter movement compared to a precision voltage reference chip? No contest - the latter are available down to 2ppm/C. For an analog movement both the spring's modulus and magnetic field strength each vary orders of magnitude more with temperature than a decent DMM...
You absolutely don't need an analog 'scope, they weigh a ton and are huge, unreliable and have many disadvantages over modern 'scopes that far outweigh the disadvantages to me. Most of them don't have automatic measurement functions such as mean and rms voltage... DSO's may be only 8 bit, but the human eye reading a graticule is less accurate than even 8 bits! Plus 'scope front ends and probes are not that great for linearity as you get up to higher frequencies anyway, so 8 bit is probably a good practical compromise (some DSOs are 12 or 14 bit anyway). But small handheld units are often very limited by comparison to either kind of bench 'scope, and especially lack usable UI's and have poor trigger circuits. Of course any 'scope is better than no 'scope...
Analog meters are not and have never been highly accurate (or even reasonably linear!), but they are very good for monitoring rapid changes in voltage and current which a digital read out is not. I suggest a bench power supply should use analog current meters for this reason.
A medium cost digital multimeter is more accurate than you need for most purposes and they use dual-slope integrating ADCs which are inherently extremely stable, linear and monotonic. The linearity and repeatability is tied to the short-term stability of a quartz timebase, which is down at the ppm level or better, so for practical purposes they are completely linear and repeatable (at least for DC measurements)
The least significant digit actuallt does means a lot even if the calibration is well out, as slow drift can be accurately seen as the least sig digit clocks over one by one. You could never see that on an analog meter unless you have a laser mirror galvonometer and a hand lens! If you are making ratiometric measurements you get the full 3 1/2 digit accuracy out of a 3 1/2 digital multimeter whatever its calibration, that's the beauty of dual slope integration.
Ever looked up the tempco of an analog meter movement compared to a precision voltage reference chip? No contest - the latter are available down to 2ppm/C. For an analog movement both the spring's modulus and magnetic field strength each vary orders of magnitude more with temperature than a decent DMM...
Hi Mark,
My god - are you ever misinformed! It's fine to disagree, but you are dead wrong in many ways. Think about the user here, a beginning tech or hobbyist.
Analogue scopes have very robust inputs, and can accept much higher voltages without damage. Something these small DSO things absolutely cannot. Beginning techs are far more likely to exceed input levels unintentionally. Yeah, they are larger, heavier. Too bad, that is the nature of the beast. However you will get a deflection if you have a signal roughly in range on the input.
Digital scopes, even my $25K scope, may not show a deflection if the horizontal sampling is too far off. Notice I said sample rate, they have all the issues of a digital system and can easily "lie" to you by indicating no signal when in fact you can have a strong signal. Some may emit clock and noise signals. To use a digital scope, you actually need to be trained and familiar with oscilloscopes already to avoid some pitfalls. Then there is the aperture when the scope samples, it is "blind" at all other times. On cheap DSOs, the processor for display, sample control and everything else is the same thing, meaning one processor is busy doing other things as well as controlling the sampling. Those have longer blind times. WHat that means is some won't catch transients often enough to make the operator aware of them. Analogue scopes do not suffer from this at all.
I have many digital and analogue scopes, really good ones and tried many inexpensive ones. I also worked professionally in a calibration lab Mark. I know instrumentation.
Meters.
Analogue meters were variable in quality. Cheap ones 2.5% FS, the better ones were extremely accurate (HP for example). HP changed the scale at 5 points for each individual movement to compensate for any non-linearity. The movement is average responding, most VTVM or electronic meters are peak responding. Better ones became RMS responding to 10 MHz and beyond. The valuable point I was trying to make is that you learned meters were not exact, so you mentally said to yourself, it is approximately so many volts or whatever. People tend to read off a digital display mindlessly without ever figuring out the accuracy. Also as I mentioned, cheap meters were commonly out of tolerance new from the box. This makes a bad situation worse. Just because something is digital has no bearing on accuracy. They survive being dropped better and are easier to read. Now frequency response. Digital meters these days are better, but cheap ones may not be accurate by the time you reach 1 KHz. My HP, Keysight and Fluke handheld meters are good to 100 KHz, the bench models 300 KHz. The RF meters use different detection methods so are good to the sensor cut-off.
The meters being considered do not use precision references, so don't even try to bring those up. Believe it or not, analogue meter movements that are good are temperature compensated. So compare equivalents please or you are distorting the truth. The temperature drift isn't that bad compared to the accuracy of the meter, analogue or digital. In the lab, yes. We are talking about much more expensive products and much higher accuracies. That is why we record calibration temperature for all devices.
So even with my 6 1/2 and 7 1/2 meters, I calculate my error budget when making some measurements. Some measurements I don't, but I do not record all the displayed digits either. There is always some error in measurement. Cheap instruments commonly have large errors, and often that last digit is just decoration. People need to understand that. You cannot go brain dead and rhyme off whatever a display says.
You bring up dual slope integration (I used to design meters). There is error, there is drift depending on the circuit and converter. But you left out the biggest source of error completely. The input attenuator! It drifts under load (it is the load for the input signal) which can be significant. Different temperatures on the elements will increase error. Never mind capacitance for AC signals. The better meters use a thick film attenuator assembly with minimal capacitance and thermal coupling. You still can have drift. Then there is the reference for your converter. My meters use an ovenized reference which remains on. Earlier meters used references with a temperature sensor that is corrected. Earlier ones use heated buried zeners (even currently like the 3458A) use those. But this isn't what you'll find in the meters being discussed. Those use the iffy references on-chip. The manufacturers of some common meter chips are often really cheap off-shore concerns that do not compare to the original number of the chip. I can tell you some are what I would call wildly inaccurate.
Clock references. Come on Mark, in the examples given? Ceramic resonators or maybe even R-C, not even a real quartz oscillator.
So in the context of what the OP is discussing, you're off-base. You also are not considering a person not trained in metrology at all. People need to learn and understand where errors enter measurements. That LSD you seem to like? That is the biggest lie in a cheap meter. Learn to ignore it with low cost equipment, it is the most subject to drift errors as you well know.
The very best way to train anyone with measurements is to hand them obviously limited equipment. They gain an understanding of errors and error sources, and to regard the indicator as not absolutely accurate. A digital oscilloscope is no more accurate than an analogue one unless it is expensive, but then the expensive scope can completely misrepresent what is going on. My MSOX3104T often shows a false trace, and that is one of the better scopes. The cheaper ones have more of those problems and I keep my good analogue scope around for that reason. There are times when they are the better tool. For analogue (digital signals are analogue), and certainly for a technician, the analogue oscilloscope is the best tool. It just doesn't have the toys DSOs can have - but they show the signal better. They also survive oopsy's better. This from experience from training technicians for decades.
My god - are you ever misinformed! It's fine to disagree, but you are dead wrong in many ways. Think about the user here, a beginning tech or hobbyist.
Analogue scopes have very robust inputs, and can accept much higher voltages without damage. Something these small DSO things absolutely cannot. Beginning techs are far more likely to exceed input levels unintentionally. Yeah, they are larger, heavier. Too bad, that is the nature of the beast. However you will get a deflection if you have a signal roughly in range on the input.
Digital scopes, even my $25K scope, may not show a deflection if the horizontal sampling is too far off. Notice I said sample rate, they have all the issues of a digital system and can easily "lie" to you by indicating no signal when in fact you can have a strong signal. Some may emit clock and noise signals. To use a digital scope, you actually need to be trained and familiar with oscilloscopes already to avoid some pitfalls. Then there is the aperture when the scope samples, it is "blind" at all other times. On cheap DSOs, the processor for display, sample control and everything else is the same thing, meaning one processor is busy doing other things as well as controlling the sampling. Those have longer blind times. WHat that means is some won't catch transients often enough to make the operator aware of them. Analogue scopes do not suffer from this at all.
I have many digital and analogue scopes, really good ones and tried many inexpensive ones. I also worked professionally in a calibration lab Mark. I know instrumentation.
Meters.
Analogue meters were variable in quality. Cheap ones 2.5% FS, the better ones were extremely accurate (HP for example). HP changed the scale at 5 points for each individual movement to compensate for any non-linearity. The movement is average responding, most VTVM or electronic meters are peak responding. Better ones became RMS responding to 10 MHz and beyond. The valuable point I was trying to make is that you learned meters were not exact, so you mentally said to yourself, it is approximately so many volts or whatever. People tend to read off a digital display mindlessly without ever figuring out the accuracy. Also as I mentioned, cheap meters were commonly out of tolerance new from the box. This makes a bad situation worse. Just because something is digital has no bearing on accuracy. They survive being dropped better and are easier to read. Now frequency response. Digital meters these days are better, but cheap ones may not be accurate by the time you reach 1 KHz. My HP, Keysight and Fluke handheld meters are good to 100 KHz, the bench models 300 KHz. The RF meters use different detection methods so are good to the sensor cut-off.
The meters being considered do not use precision references, so don't even try to bring those up. Believe it or not, analogue meter movements that are good are temperature compensated. So compare equivalents please or you are distorting the truth. The temperature drift isn't that bad compared to the accuracy of the meter, analogue or digital. In the lab, yes. We are talking about much more expensive products and much higher accuracies. That is why we record calibration temperature for all devices.
So even with my 6 1/2 and 7 1/2 meters, I calculate my error budget when making some measurements. Some measurements I don't, but I do not record all the displayed digits either. There is always some error in measurement. Cheap instruments commonly have large errors, and often that last digit is just decoration. People need to understand that. You cannot go brain dead and rhyme off whatever a display says.
You bring up dual slope integration (I used to design meters). There is error, there is drift depending on the circuit and converter. But you left out the biggest source of error completely. The input attenuator! It drifts under load (it is the load for the input signal) which can be significant. Different temperatures on the elements will increase error. Never mind capacitance for AC signals. The better meters use a thick film attenuator assembly with minimal capacitance and thermal coupling. You still can have drift. Then there is the reference for your converter. My meters use an ovenized reference which remains on. Earlier meters used references with a temperature sensor that is corrected. Earlier ones use heated buried zeners (even currently like the 3458A) use those. But this isn't what you'll find in the meters being discussed. Those use the iffy references on-chip. The manufacturers of some common meter chips are often really cheap off-shore concerns that do not compare to the original number of the chip. I can tell you some are what I would call wildly inaccurate.
Clock references. Come on Mark, in the examples given? Ceramic resonators or maybe even R-C, not even a real quartz oscillator.
So in the context of what the OP is discussing, you're off-base. You also are not considering a person not trained in metrology at all. People need to learn and understand where errors enter measurements. That LSD you seem to like? That is the biggest lie in a cheap meter. Learn to ignore it with low cost equipment, it is the most subject to drift errors as you well know.
The very best way to train anyone with measurements is to hand them obviously limited equipment. They gain an understanding of errors and error sources, and to regard the indicator as not absolutely accurate. A digital oscilloscope is no more accurate than an analogue one unless it is expensive, but then the expensive scope can completely misrepresent what is going on. My MSOX3104T often shows a false trace, and that is one of the better scopes. The cheaper ones have more of those problems and I keep my good analogue scope around for that reason. There are times when they are the better tool. For analogue (digital signals are analogue), and certainly for a technician, the analogue oscilloscope is the best tool. It just doesn't have the toys DSOs can have - but they show the signal better. They also survive oopsy's better. This from experience from training technicians for decades.
Small Scope , meter and signal generator for around 89 dollars.
Still rather useful.
Repairing game machines, amps and cars back in the 90's
Stuff like this would have blown our minds.
Still rather useful.
Repairing game machines, amps and cars back in the 90's
Stuff like this would have blown our minds.
I agree useful in some cases, but not to rely on. But useful in the hands of someone who understands what they are.
Absolutely! This stuff would have totally blown our minds. We would still quickly grab the instruments that worked better and we could rely on.
Absolutely! This stuff would have totally blown our minds. We would still quickly grab the instruments that worked better and we could rely on.
Full disclosure, I'm a test equipment snob. So much good high end test equipment has been available for cheap that I don't see why anybody would use a cheap meter. I use both analog and digital. That said, I've never seen a meter so bad it wasn't useful for audio servicing. We just don't need very much accuracy. If all I could afford was a Harbor Freight meter, it wouldn't be a problem.
In the scope department it's a little more complicated. Until recently I'd only use good (ok, boat anchor) Tek analog scopes. You can see every bit of fuzz on a waveform and they don't lie to you. The controls are fairly simple and logical. IMO, everybody would benefit by starting out with such. The problem for the beginner is they often need service and not everybody can do that. The good analog stuff is getting fairly old by now. New high end is out of reach but we have a crop of very inexpensive imports at every quality level. That Fnirsi is crazy cheap and rated to +/- 400 volts, but I wouldn't be surprised if it failed below that. Or not; you just don't know. I do a lot of work with direct connections, not X10 probes, so I do want my inputs sturdy. I rarely work on tube stuff so the risks would be low if that was what I had. The Fnirsi also has a signal generator so the beginner doesn't have to buy yet another piece.
Remember that beginners who stick with this stuff for a bit aren't beginners anymore!
Now, I said "until recently." Eventually one has to downsize and I recently replaced all my scopes and signal generators with a Siglent 12-bit scope and 16-bit arbitrary waveform generator. Until the introduction of 12-bit scopes I wouldn't have touched a digital, but at the 12-bit level you can see what you need to see, the math functions are plenty accurate and the FFT is far more useful. 12-bit Siglent scopes start at about $339, so not at all out of reach. The only thing I miss for audio work are the high-gain Tek plug-ins, but few people would have those anyway. Note that the maximum input voltage of the Siglent is exactly the same as the Fnirsi. BTW, if you buy the new low end Tektronix the input is rated at 300 Vrms, about 425 Vpk, not a crazy difference, though it is CAT II, which gives more confidence in the design.
One more benefit of the digital world is documentation. If you want to record screen shots to save or to post, or if you want to save the actual data as a CSV file for analysis in Excel or another program, it's trivial.
In the scope department it's a little more complicated. Until recently I'd only use good (ok, boat anchor) Tek analog scopes. You can see every bit of fuzz on a waveform and they don't lie to you. The controls are fairly simple and logical. IMO, everybody would benefit by starting out with such. The problem for the beginner is they often need service and not everybody can do that. The good analog stuff is getting fairly old by now. New high end is out of reach but we have a crop of very inexpensive imports at every quality level. That Fnirsi is crazy cheap and rated to +/- 400 volts, but I wouldn't be surprised if it failed below that. Or not; you just don't know. I do a lot of work with direct connections, not X10 probes, so I do want my inputs sturdy. I rarely work on tube stuff so the risks would be low if that was what I had. The Fnirsi also has a signal generator so the beginner doesn't have to buy yet another piece.
Remember that beginners who stick with this stuff for a bit aren't beginners anymore!
Now, I said "until recently." Eventually one has to downsize and I recently replaced all my scopes and signal generators with a Siglent 12-bit scope and 16-bit arbitrary waveform generator. Until the introduction of 12-bit scopes I wouldn't have touched a digital, but at the 12-bit level you can see what you need to see, the math functions are plenty accurate and the FFT is far more useful. 12-bit Siglent scopes start at about $339, so not at all out of reach. The only thing I miss for audio work are the high-gain Tek plug-ins, but few people would have those anyway. Note that the maximum input voltage of the Siglent is exactly the same as the Fnirsi. BTW, if you buy the new low end Tektronix the input is rated at 300 Vrms, about 425 Vpk, not a crazy difference, though it is CAT II, which gives more confidence in the design.
One more benefit of the digital world is documentation. If you want to record screen shots to save or to post, or if you want to save the actual data as a CSV file for analysis in Excel or another program, it's trivial.
lol!
Hi Conrad,
I agree with you 100%. I will say that 12 BIT is much better, but my 100 MHz Philips scopes show more detail. They have finer traces than any Tek I have, especially the 2465B CT.
Another thing to consider the bandwidth of the analogue signal path. With Tek and other good scopes, it is not the limiting factor. So many other factors come into digital signal capture. I will say I love the toys (signal capture). With mine, I can not only capture the trace, but digitize the actual signal and play it back - even from the arb in the scope.
Most low cost digital scopes have inputs rated for 5V maximum. Most cheap probes are switchable x1 - x10, more damage occurs when the probes are switched accidently from x10 to x1. I highly recommend anyone with a switchable probe glue or fix it permanently in the x10 position. Use BNC cables and adapters for direct connections so there is never any question or late night stupid moments forcing a new scope purchase!
About audio servicing. Well, not entirely true. Many cheap DVMs are not suitable for even setting bias current. You're measuring in mV, sometimes just a couple mV and that means the LSD. So for such a basic measurement, you really need something half decent. Many cheap meters are 100 mV FS or higher.
Another thing I just thought of is how your DVM deals with noise. Does it average, peak detect - what? That's going to change the measurement and sometimes by a lot. Amplifier measurements typically are drifty, noisy measurements. You must know how your meter responds to these, and I often hang a scope on it to make sure I know what I am measuring.
All this boils down to a simple (but not so simple) concept. You must know and understand the equipment you use. This is extremely difficult for beginners or people even a few years in. Work in a calibration lab or similar, that will teach you in a hurry. Experimenting on your own even with "simple" measurements will help immensely.
Hi Conrad,
I agree with you 100%. I will say that 12 BIT is much better, but my 100 MHz Philips scopes show more detail. They have finer traces than any Tek I have, especially the 2465B CT.
Another thing to consider the bandwidth of the analogue signal path. With Tek and other good scopes, it is not the limiting factor. So many other factors come into digital signal capture. I will say I love the toys (signal capture). With mine, I can not only capture the trace, but digitize the actual signal and play it back - even from the arb in the scope.
Most low cost digital scopes have inputs rated for 5V maximum. Most cheap probes are switchable x1 - x10, more damage occurs when the probes are switched accidently from x10 to x1. I highly recommend anyone with a switchable probe glue or fix it permanently in the x10 position. Use BNC cables and adapters for direct connections so there is never any question or late night stupid moments forcing a new scope purchase!
About audio servicing. Well, not entirely true. Many cheap DVMs are not suitable for even setting bias current. You're measuring in mV, sometimes just a couple mV and that means the LSD. So for such a basic measurement, you really need something half decent. Many cheap meters are 100 mV FS or higher.
Another thing I just thought of is how your DVM deals with noise. Does it average, peak detect - what? That's going to change the measurement and sometimes by a lot. Amplifier measurements typically are drifty, noisy measurements. You must know how your meter responds to these, and I often hang a scope on it to make sure I know what I am measuring.
All this boils down to a simple (but not so simple) concept. You must know and understand the equipment you use. This is extremely difficult for beginners or people even a few years in. Work in a calibration lab or similar, that will teach you in a hurry. Experimenting on your own even with "simple" measurements will help immensely.
Yeah, that thought occurred to me- did they spec it with a X10 probe? Maybe I'm too trusting, but I don't think they did. Reading carefully probably won't help as they aren't very specific. BTW, I've never been a fan of PC instruments that use a sound card as you need some kind of input device to keep from smoking it. Nor do I trust them until verified by real actual test equipment and if you have real actual test equipment, it might be better to just use that!
Some cheaper devices I bought in the past (north of $200) did spec with 10:1 probe, but you had to really read. The comment was "with supplied probe", then later 5V maximum. These cheap instruments are normally used by students at audio line levels. Same as sound cards that get blown up on the bench.
Edit:
Hi Ray, I was thinking of one I got with a 50V input maximum, then the others were 5V. So you're right, 100:1 probe!
Edit:
Hi Ray, I was thinking of one I got with a 50V input maximum, then the others were 5V. So you're right, 100:1 probe!