1
\$\begingroup\$

If you could please bear with me: I'm just starting out in electronics and I've had an idea for a nice experimental mod that will also be useful.

I've recently bought a pretty cool little 8 x 8 LED spectrum analyzer off eBay:

Similar to mine

And I'm currently building a container for it. I'm going to add a couple of mods. Among them, an ON/OFF switch, so I can get rid of the USB connection and have it permanently connected, but only ON whenever I really need it.

The thing is, this little device has a couple of configurable settings, none of which are, by default, the ones I prefer. So after installing the ON/OFF toggle, every time I toggle the device off, I'll lose my settings, and have to reconfigure them when toggling it back on... which is of course a hassle, because this is meant to be a little decorative plus when I turn on my music and relax, not a task to complete every single time.

So I've been researching what would be the best (slash-maybe-not-the-most-complex) way of adding the possibility for the device to remember its settings even when disconnected to power (or powered off).

I've read a couple of forums and articles and have come across the terms latching relay and slow-discharge capacitor... but I'm still mostly lost. I'm even wondering if just adding a coin cell at some point will do the trick but I'm almost sure it can't be that easy!

So if anyone could put me on the right track, in other words tell me, based on this specific project, what's my best bet, that'd be GREAT! Because right now, after a couple of days of research, I've turned out nothing, at least nothing I can make any sense of.

Thank you in advance!

P.S. Or... is there a way to reconfigure the WD1088S chip?

P.S.S. I've just had another idea... it might be a bad one, but I'm not sure if it won't actually work OK. This device works with 3-5V and has a "static working current [of] less than 3mA". So what if I attach it to a 3.6V 1240 mAh camera battery and have the charger connected to the battery with the ON/OFF toggle in between. That way, at least, I won't be using the electrical current while the device is not in use. Yes, the device won't really be OFF, but as long as it's kept idle, it should have enough power to not turn OFF during about 12 days. I can even use it with the battery (it should work about a day, with an working current of 30mA). I know it's not an elegant solution, more like a workaround... But it's the only thing I can think of. What I'm afraid is I'll be contributing to shortening the lifespan of the whole thing... Anyway, I'm out to learn, so what do you think of this idea?

Edit: I found the PCB schematic and the IC datasheet. Perhaps that can help...

PCB schematic and IC datasheet

\$\endgroup\$
9
  • \$\begingroup\$ Short of being able to modify the "firmware" mentioned in automatic translations of the only info I can find, there probably is not a good way to do this. Normally a modern circuit might be designed to drop into a very low power mode to only hold state, without any clocks running, but it is unclear if this supports that. You could try building a small supervisor circuit to trigger a reset and then press the right button combinations, but you might have to hope for the best, unless you can find a way for your supervisor MCU to detect the state by watching a few LEDs. \$\endgroup\$ Commented Apr 23, 2016 at 20:50
  • \$\begingroup\$ Or you could make your own custom equivalent from scratch with a modern MCU. \$\endgroup\$ Commented Apr 23, 2016 at 20:51
  • \$\begingroup\$ Chris Stratton is correct. Or you could go all Rube Goldberg on the project and add a small micro to do what ever it is you do to configure the thing to do what you want on every power up. BTW, super caps and coin cell battery back ups only work for processors that draw really small amounts of current. If the circuit wasn't designed for this from the start it is really hard to add & much easier (as Chris said) to start over. \$\endgroup\$
    – st2000
    Commented Apr 23, 2016 at 22:35
  • \$\begingroup\$ Hi, guys, thanks for your input! I've just had an idea... I've added it at the end of my orignal post. What do you think? \$\endgroup\$ Commented Apr 23, 2016 at 23:18
  • \$\begingroup\$ Do you really carre about how much current the device draws when you are not using it? If it doesn't display anything where then is no audio, then just have a SPDT switch (or DPDT if stereo) that either passes through the audio, or shorts out the input so nothing is displayed. \$\endgroup\$
    – tcrosley
    Commented Apr 24, 2016 at 2:40

2 Answers 2

1
\$\begingroup\$

Assuming it doesn't display anything when then is no audio input, then leave the circuitry on all the time and have a SPDT switch (or DPDT if stereo) that either passes through the audio, or shorts out the input so nothing is displayed. That becomes your on/off switch.

\$\endgroup\$
6
  • \$\begingroup\$ Thanks a lot, tcrosley! I would have to couple this solution with the one offered by @mkeith. I'll be looking into these possibilities. But still it looks like a lot of modifications for what will essentially be a workaround. So I want to keep on researching a little more, to see if I can find a proper/elegant solution. Maybe I'll have to redo the entire circuit... but because it's a learning experience, perhaps it's best than just applying a workaround, right? \$\endgroup\$ Commented Apr 24, 2016 at 15:20
  • \$\begingroup\$ This is also a very good answer. \$\endgroup\$
    – user57037
    Commented Apr 24, 2016 at 16:49
  • \$\begingroup\$ I'm still working on this project... or rather stuck. I've followed your advice and added a switch to cut off the audio input. In any case, even turning of the audio source should've been OK in this case. The problem is that every so often the device "picks up" some signal from somewhere and does a little twirl of lights. \$\endgroup\$ Commented Apr 28, 2016 at 17:23
  • \$\begingroup\$ @QuestionerNo27 Rather than just disconnecting the audio, did you take my advice and use a SPDT switch so the input to the spectrum analyzer is shorted to ground when the switch is "off"? That way it shouldn't pick up any noise. \$\endgroup\$
    – tcrosley
    Commented Apr 28, 2016 at 19:11
  • \$\begingroup\$ Hi@tcrosley, I haven't yet been able to do exactly that: I live in Paris and you don't know how fiendishly hard it is too get parts here! (To give you an idea: I was looking for a good 20k log pot to fit a small amplifier for 14 month before I finally got one in Spain!) So I've had to order the SPDT from China and am now waiting for it to arrive. I'll get back to this when it does. Thanks! \$\endgroup\$ Commented Apr 29, 2016 at 9:30
0
\$\begingroup\$

The basic idea of adding a battery to keep memory from being lost is an OK idea. But you have to look into various potential problems.

Is the full range of battery voltage acceptable (at first glance, based on your description, it sounds like it is).

Is the switch connected and of such a design that it disconnects the battery before it connects the external power source? I think the wiring is OK (based on your description) but is the switch a break before make type? You don't want to short the battery to the power supply momentarily while switching. Some switches DO temporarily short all three terminals together during switching. That would not be ideal. There would be some high currents flowing every time you switch. It could lead to long-term reliability problems with the switch or the power supply. It might even damage the battery.

Finally, if you do use a break-before-make type of switch, is there enough capacitance on the board to hold up the input voltage during the transition from one power source to the other? Since it only uses 3mA, you don't need a lot of capacitance, but it would be nice to have at least 100 uF on there. There is no way for me to know how much might already be on the board. You could just add 100uF to be safe. If that black cylindrical component is an electrolytic capacitor, then you probably are OK on capacitance. Otherwise I would advise adding a 100uF capacitor. Probably aluminum electrolytic will be most practical.

Have fun and keep experimenting!

\$\endgroup\$
1
  • \$\begingroup\$ Thanks a lot for the input, mkeith! I'm going to continue looking into this possibility. But still it looks like a lot of modifications for what will essentially be a workaround. So I want to keep on researching a little more, to see if I can find a proper/elegant solution. Maybe I'll have to redo the entire circuit... but because it's a learning experience, perhaps it's best than just applying a workaround, right? \$\endgroup\$ Commented Apr 24, 2016 at 15:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.