I had a 1.5A solar charger built into the original shiznit design, but it was solar only (not wall wart) and you had to set the MPPT voltage of the panel by hooking up a discharged battery + solar panel and turning a pot until you measured the panel MPPT voltage across a couple test points with a DMM. Not exactly friendly.
Anyway, the shiznit 2.0 prototype is currently just a prototype, intended mainly to prove the DC/DC + amplifier concept rather than being an exact representation of the final product. If everything goes well with it thermally, I can probably shrink down the design - use 0402 components, QFNs, DC/DC FETs in more exotic packages, etc - and make room for a 5V output, solar charger with wall wart capability, or other features. It'd be nice to have the card be the only device you need in a Boominator, which was really the point of the original shiznit card.
Mind isn't made up yet. We'll see how testing goes with this card, what I can come up with for a solar charger, etc. and go from there. Still waiting on the 3787s as I type this.
Anyway, the shiznit 2.0 prototype is currently just a prototype, intended mainly to prove the DC/DC + amplifier concept rather than being an exact representation of the final product. If everything goes well with it thermally, I can probably shrink down the design - use 0402 components, QFNs, DC/DC FETs in more exotic packages, etc - and make room for a 5V output, solar charger with wall wart capability, or other features. It'd be nice to have the card be the only device you need in a Boominator, which was really the point of the original shiznit card.
Mind isn't made up yet. We'll see how testing goes with this card, what I can come up with for a solar charger, etc. and go from there. Still waiting on the 3787s as I type this.
Have you thought about a 4-ch Class-D chip for a future design? Like the TI TAS5414C or similar 
I'd really dig something like a MiniDSP, 4-ch Class-D amp and the Linear LT3787 integrated on a compact card (Mainly because I use real tweeters, the P.Audio PHT-407N).
That way, the audio could get transfered via I2S from the MiniDSP (or NanoDIGI) to the TI TAS5414C
I'd really dig something like a MiniDSP, 4-ch Class-D amp and the Linear LT3787 integrated on a compact card
(Mainly because I use real tweeters, the P.Audio PHT-407N).That way, the audio could get transfered via I2S from the MiniDSP (or NanoDIGI) to the TI TAS5414C
Have you thought about a 4-ch Class-D chip for a future design? Like the TI TAS5414C or similar
The TAS5414C is by some accounts inferior in audio quality. It also has 6 times the quiescent current and dramatically lower efficiency.
In general principle miniDSP or other DSP function is not always desirable as it introduces delay which rules out it being used as a monitor for live music. I would rather not see a built-in DSP function but instead use external DSP function from for example a bluetooth reciever (reprogrammed Kalimba DSP in BlueCore5 or later) or miniDSP where there is already delay introduced.
Also, 4 channel output is not always desirable.
In the same way that 2-ch output is not always desirable?
I was purely thinking about my own user needs there
Do you know any other 4-ch amps that receives I2S signals, with better audio quality and lower quiescent current? Or would it just be better to implement two TPA3118 chips?
The latter but that would also require a total redesign of the dc-dc converter. It's by far better to just have 2 of these amps instead. If one could be set in slave mode that would be optimal.
Last edited:
Decision made.
I'm stuffing the shiznit V1 charger back into the upcoming shiznit design. I'll set it up by default for the CIS panels that canopysound.dk sells, but I'll provide an adjustment pot, couple of test points and a procedure to adjust the MPPT voltage for a different panel type if you use something different.
I'll provide a wall wart input also - you'll need a 3 wire, switched DC jack (I think, haven't thought this through 100%) and a 19.5V laptop brick or 24V wall wart as a power source.
I'm stuffing the shiznit V1 charger back into the upcoming shiznit design. I'll set it up by default for the CIS panels that canopysound.dk sells, but I'll provide an adjustment pot, couple of test points and a procedure to adjust the MPPT voltage for a different panel type if you use something different.
I'll provide a wall wart input also - you'll need a 3 wire, switched DC jack (I think, haven't thought this through 100%) and a 19.5V laptop brick or 24V wall wart as a power source.
IT'S ALIVE!
Threw together a video which goes over the design and demonstrates it in action. Still got a fair bit of development work left to do but so far, things are going well.
Shiznit V2 card update - YouTube
Threw together a video which goes over the design and demonstrates it in action. Still got a fair bit of development work left to do but so far, things are going well.
Shiznit V2 card update - YouTube
I have a slightly off topic question: I've been trying in vain to add a Raspberry Pi to an old boombox, the problem being a nasty ground loop and/or impedance mismatch. It's using a buck step down converter to deliver the 5v from the 12v supply and has a terrible hum & digital noise in the system.
If I were building a boominator with your Shiznit amp, would there be any way to integrate something like a Pi or even a charging iPhone, from a single power source, without running into this problem? I temporarily fixed my issue by using two separate battery sources in the boombox but this is complicated and not ideal.
Also, if you have any tips on fixing my current problem I'd love to hear it. I tried a ground loop isolator but ended up losing over half my output volume. Thanks!
If I were building a boominator with your Shiznit amp, would there be any way to integrate something like a Pi or even a charging iPhone, from a single power source, without running into this problem? I temporarily fixed my issue by using two separate battery sources in the boombox but this is complicated and not ideal.
Also, if you have any tips on fixing my current problem I'd love to hear it. I tried a ground loop isolator but ended up losing over half my output volume. Thanks!
Current status:So how is the process advancing? When can we expect the 'Shiznit amp' ready for consumers?
- Prototype's fully built/verified.
- DC/DC is working, amplifier works and doesn't give off much heat as all, and idle power consumption is pretty reasonable.
- Audio presence detection using the AVR's ADC didn't work as planned, I can't reliably tell the difference between quiet audio and no audio, or reliably capture the envelope of the music (and reliably set the DC/DC level)
- I've decided to throw the DSP back on.
The DSP has much better dynamic range and it'll do a much better job of capturing the audio signal level. It'll also do nice things like soft clipping of the audio input, if the input level suddenly increases and the DC/DC voltage isn't sufficiently ramped up.
The DSP also allows a few extra fun things, like an "Halifinator" mode, which removes stereo below a certain frequency.
The input ground isn't technically isolated, it's connected to the main ground via a 100 ohm resistor, and a differential amp is used. Basic concept is #3 in this PDF:
http://www.ti.com/lit/an/sloa143/sloa143.pdf
None, unfortunately; haven't had time to even start the new DSP design, too many other things going on. It's planned out, I just need to get a few hard days at it to pick a couple of parts and bang out the schematic/PCB.
I absolutely promise that it'll be available for purchase before the next boominator season
I absolutely promise that it'll be available for purchase before the next boominator season
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.