I have a computer, I need a simple oscilloscope

[Robert Atkinson 2Participant @robertatkinson2]

Maurice,

Any of the current picoscope will do the job. The lowest perfomance model is DC to 5MHz bandwidth, two channel and has a built in signal generator as a bonus, The one I have to lend is a 2203 with that specification. New they are around £100 (currently the lowest spec model for sale is the 2204 10MHz version due to the chip shortage. It's £115 including two probes. Beware ebay "resellers" who are charging up to 3 times this!
I've been using Pico Technology products for over 20 years both personally and professionally and have never had any issue with the product or their support.

Robert G8RPI.

[SillyOldDufferModerator @sillyoldduffer]

Posted by DC31k on 24/09/2022 18:09:39:

Posted by SillyOldDuffer on 24/09/2022 17:46:55:

Operation: set the wiper to ground…

And that is the flaw in the plan. It relies on the knob-twiddler performing his duty with 100% reliability every single time without fail after every change to the input conditions.

If you could show something with a kind of dead-man's control on it that absolutely cannot put anything into the sound card input without positive human action, that would be the beginning of a way forward.

If you can incorporate something in it that will prevent overvoltage altogether (perhaps caused by an involuntary jerk due to sneezing when adjusting the pot), that would increase the reliability of the system.

I'm all in favour of risk managed approaches, and thought for Tim's investigations a pot would be sufficient. However, as this was a good excuse to get out of urgent house-work, I tried the experiment.

Apparatus: A meccano lash-up with motor and wheel that spins a small super-magnet past a sawn-off nail with a few hundred turns of magnet wire coiled around it. The magnet is the shiny dot on the opposite side of the wheel to the nail and coil:

A Siglent SDS1102 CML oscilloscope

Method:

Oscilloscope was connected to the coil on the 5V / graduation range and the wheel turned slowly by hand. This showed a pulse of about ±0.15V so the oscilloscope' sensitivity was increased to 200mV per graduation and the motor powered up with a 1.2V AA cell.

Results.

The oscilloscope registers almost exactly 10Hz so the motor is revolving at 600rpm

The voltage spike is larger than I expected, at 760mV roughly double the turned by hand voltage

Conclusion

I knew but forgot that the voltage generated when a magnet passes a coil is proportional to the speed of transit. Thus, connecting this circuit directly to a sound-card is riskier than I expected! At 600rpm my experimental set-up wouldn't damage a sound-card, but higher rpm exceeds the limit. (At 4000rpm the set-up produces 2.75V pulses.)

My pot circuit is too simple to protect the sound-card if the rpm varies.

————-

I think Tim is looking for a display like this:

The voltage swings ever more negative as the magnet approaches the coil, then flips to positive as the magnet passes TDC, and and falls as the magnet moves out of range. (Reversing the magnet would cause the pulse to go positive then negative.) Anyway, if I've remembered correctly, the voltage is proportional to the rate of change when the coil flips from negative to positive, which depends on how fast the magnet is moving – RPM.

Dave

[duncan webster 1Participant @duncanwebster1]

Using a hall effect probe would make voltage in dependant of speed, and unidirectional.

If the OP's computer has a bidirectional parallel port I can provide an old pico scope unit, no use to me, but too good to chuck

[DC31kParticipant @dc31k]

Posted by SillyOldDuffer on 25/09/2022 12:14:58:

The voltage swings ever more negative as the magnet approaches the coil, then flips to positive as the magnet passes TDC

Thankyou for your efforts. It is interesting stuff.

On the trace, where exactly is TDC? Is it at the point of maximum negative voltage, at the zero crossing point or at the maximum positive voltage?

How many degrees of rotation does it represent from peak to peak? How many degrees from leaving the zero line to crossing it from below to meeting it again from above? The curve looks slightly asymmetric. Any speculation why?

Thanks again.

[Robert Atkinson 2Participant @robertatkinson2]

Posted by SillyOldDuffer on 25/09/2022 12:14:58:

Posted by DC31k on 24/09/2022 18:09:39:

Posted by SillyOldDuffer on 24/09/2022 17:46:55:

Operation: set the wiper to ground…

And that is the flaw in the plan. It relies on the knob-twiddler performing his duty with 100% reliability every single time without fail after every change to the input conditions.

If you could show something with a kind of dead-man's control on it that absolutely cannot put anything into the sound card input without positive human action, that would be the beginning of a way forward.

If you can incorporate something in it that will prevent overvoltage altogether (perhaps caused by an involuntary jerk due to sneezing when adjusting the pot), that would increase the reliability of the system.

I'm all in favour of risk managed approaches, and thought for Tim's investigations a pot would be sufficient. However, as this was a good excuse to get out of urgent house-work, I tried the experiment.

Apparatus: A meccano lash-up with motor and wheel that spins a small super-magnet past a sawn-off nail with a few hundred turns of magnet wire coiled around it. The magnet is the shiny dot on the opposite side of the wheel to the nail and coil:

A Siglent SDS1102 CML oscilloscope

Method:

Oscilloscope was connected to the coil on the 5V / graduation range and the wheel turned slowly by hand. This showed a pulse of about ±0.15V so the oscilloscope' sensitivity was increased to 200mV per graduation and the motor powered up with a 1.2V AA cell.

Results.

The oscilloscope registers almost exactly 10Hz so the motor is revolving at 600rpm

The voltage spike is larger than I expected, at 760mV roughly double the turned by hand voltage

Conclusion

I knew but forgot that the voltage generated when a magnet passes a coil is proportional to the speed of transit. Thus, connecting this circuit directly to a sound-card is riskier than I expected! At 600rpm my experimental set-up wouldn't damage a sound-card, but higher rpm exceeds the limit. (At 4000rpm the set-up produces 2.75V pulses.)

My pot circuit is too simple to protect the sound-card if the rpm varies.

————-

I think Tim is looking for a display like this:

The voltage swings ever more negative as the magnet approaches the coil, then flips to positive as the magnet passes TDC, and and falls as the magnet moves out of range. (Reversing the magnet would cause the pulse to go positive then negative.) Anyway, if I've remembered correctly, the voltage is proportional to the rate of change when the coil flips from negative to positive, which depends on how fast the magnet is moving – RPM.

Dave

It would be interesting to see what your test setup produces with a soundcard 'scop software you proposed. Have to tried it?

[Tim StevensParticipant @timstevens64731]

Hello again
Dave (s-o-d) has got it about right.
The engine needs a spark at a progessively earlier time before TDC as the speed rises. In the past this was done by a centrifugally advancing cam in the contact breaker.
The ignition device needs to be set to 'open the theoretical points' when the output from the trigger coil gets to a set voltage. As the rotation speeds up, the whole wave increases in size, and therefore reaches a given voltage earlier. The system does not rely on a slotted disc, or a Hall sensor – they were a later development in vehicle use.

What I hope to do is to be able to tweak the shape and size of the voltage curve so that the advance produced can be matched (by road testing) to a given engine and given settings. The system I have already made as a first try does seem to work fairly well, and this is a fortunate accident rather than cleverness. What I seek the to change, for example, the radius of the rotor, the distance to the magnet, the strength of the magnet, the shape of any pole piece on the magnet, etc, to vary the advance curve in known ways at low, or high, revs, so the the original set-up can be improved. Or by changing things, to show that the original guess was jolly nearly right anyway and best left as it is.

Hope this helps

Cheers, Tim

[Maurice TaylorParticipant @mauricetaylor82093]

Posted by Robert Atkinson 2 on 25/09/2022 10:25:35:

Maurice,

Any of the current picoscope will do the job. The lowest perfomance model is DC to 5MHz bandwidth, two channel and has a built in signal generator as a bonus, The one I have to lend is a 2203 with that specification. New they are around £100 (currently the lowest spec model for sale is the 2204 10MHz version due to the chip shortage. It's £115 including two probes. Beware ebay "resellers" who are charging up to 3 times this!
I've been using Pico Technology products for over 20 years both personally and professionally and have never had any issue with the product or their support.

Robert G8RPI.

Hi, Thank you for your reply

Maurice

[duncan webster 1Participant @duncanwebster1]

If you use a processor you could have the trigger way early, then create the spark after a time delay dependant on rpm and inlet vacuum. I seem to remember old fashioned distributors had a connection to the inlet manifold. You can also have any relationship between the various parameters just by altering the code.

[Robert Atkinson 2Participant @robertatkinson2]

One thing to consider is the location of the magnet. The "obvious" and common place to put it is on the rotating part. This can however make shaping the pole pieces to adjust the waveform difficult. Another location that works is to put the magnet on the end of the core the coil is wound on furthest from the rotating part. This assumes of course that the rotating part is made from magnetic material. This allows more freedom in shaping both fixed and moving pole pieces.

Robert.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Robert Atkinson 2 on 25/09/2022 16:28:38:

Posted by SillyOldDuffer on 25/09/2022 12:14:58:

Posted by DC31k on 24/09/2022 18:09:39:

Posted by SillyOldDuffer on 24/09/2022 17:46:55:

…

…

…:

Results.

The oscilloscope registers almost exactly 10Hz so the motor is revolving at 600rpm

….

Dave

It would be interesting to see what your test setup produces with a soundcard 'scop software you proposed. Have to tried it?

I have now, downloaded from here:

Slightly fussy getting it to recognise the input. My windows 10 PC has a Line-In (coded pink) socket on the front panel. Plugging in woke up the Realtek Audio mixer and I had to select Line-In. The scope still didn't work until I opened its settings tab (far right) and changed the input from Lookback to Line-In.

As I was certain my lash up powered by a 1.2V battery wouldn't over-volt the computer input, I connected directly. Might try a pair of back to back diodes later.

The waveform is in close agreement with the Siglent so concerns about soundcards only having an AC input are put to rest in this case.

Won't have a chance to think about Robert's questions until later this afternoon, but they're interesting too.

Dave

[Robert Atkinson 2Participant @robertatkinson2]

Thanks for that Dave,
Looks like good waveform fidelity. Amplitude is quite a bit off though. I wonder if the software has a calibration routine? One advantage of this application is the input signal is isolated so no ground loops.

Robert.

[Author]

Posts