Pocket Oscillator

[ruffrecords]

Any reason for using an NE5532 rather than a TL072?


Cheers

Ian

[mediatechnology]

Any reason for using an NE5532 rather than a TL072?

Cheers
Ian

No reason you couldn't use one.

Two reasons for using a 5532 are it's increased drive capability and, for purposes of test, to see what the topology is capable of doing in terms of THD.

I'll drop a TL072 into the socket and see, under the same conditions, how it measures.
For most things it's probably good enough.

[ruffrecords]

Excellent. I was just thinking I used one in the original design most likely from the power consumption point of view, but in all probability that is dominated by class B output current rather than quiescent? The annoying thing about the pocket oscillator as it stands is that it has no automatic switch off. You only have to accidentally leave it on overnight a few times to flatten the battery. And at my age, forgetfulness is more common than I would wish. So the Mark 2 will have this feature (plus a few more) but that is another thread.

Cheers

Ian

[mediatechnology]

Using a TL072 into an open circuit load (no pad) at +6dBu I'm reading 0.0005%. With a 632Ω balanced load the distortion rises to about 0.005%.
Adjustment below +6 dBu is pretty touchy.

The 5532 draws more current but has the drive capability making it worth a few extra mA.
The AGC rectifier is unbuffered and is sensitive to the output impedance of the op amp.
Since the TL072 has lower drive current and higher closed loop output impedance there may be some level dysregulation with complex loads.

Another advantage of an output pad is load isolation and, in the case of the mic-level output, phantom fault isolation.

I had thought about an automatic turn-off as well.
EDIT: I wonder if a couple of 4020 CMOS 14 bit counters could be used to count the 1 kHz "clock" and power it of after 2^n seconds.
A single 4020 provides a 16.83 second clock; two 4020 (up to 2^28) would provide up to 3 days.

[mediatechnology]

The more I think about it a 4093 configured as a 1 Hz oscillator and a 4017 could be used as a long duration timer.
16,384 seconds is just over 4.5 hours.

I'm considering using a BS170/BS250 as high-side/low-side switches for battery disconnection.

[ruffrecords]

Sounds like a plan.

Cheers

Ian

[mediatechnology]

24 hours later it's still running on the two 9V batteries. They're down to about 7V each and the THD is still 0.0005% with it under load.

I'm thinking about a solution that will have a three position power switch that will have Off, Timer, and Always-On.

I'm going to work on the high-side low-side MOSFET switches first.

[Tubetec]

These are a good way to allow battery access without the need for a screwdriver , very durable and robust , used in many broadcast applications where 9V power is required , ie body worn transmitter packs , preamps and phantom power supplies .

Space constraints(and jokes ) aside , an old fashioned clockwork timed off switch makes a lot of sense .
They come with off ,on(timed) and always on positions with upto 60 or 120 minute off times .
The kind typically used in cooling fans .

[mediatechnology]

The oscillator ran for about 37 hours before the batteries ran down.

The timer reminds me of the heating lamp timers in hotel rooms.
I could put in a crank generator (like the crank radios) or maybe a coin slot. LOL.

[Tubetec]

The more I think about it the more I like the mechanical solution over the digital .
Its big and clunky ,with a way over rated switching current , but the cooker types also come with a dial and skirt which means you dont have to worry about a front panel with decals ,downside is they usually have a bell too ,but a hacksaw would deal with that very effectively .
https://www.ebay.co.uk/itm/177323412929
Its only the 120 minute counter has the always on position .

A single gang metal clad electrical box and blank front panel could house the whole lot up ,nicely screened .