XMOS-based Asynchronous USB to I2S interface

Hello Douwe, I'm interested how you got 3.3V with LifePo4-batteries: which brands of batteries, how many mA, which circuit for controlling charge/discharge you attached to them, which charger are you using, where you bought, etc.

Thanks for helping
Roberto


Hi Lorien, would it be possible to feed the board with 3,3V in stead of 5V? The specs of the xmos chip allow 3,3v. It would be nice when we could use LifePo4-batteries, works great on my current Hiface.
regards, Douwe
 
umm every lifepo4 battery ive used is 3.3v nominal as its normal voltage. at fully charged it will be 3.6 but this drops to 3.45 very quickly and spends the rest of its charge between 3.45 and 3.2 before falling off a cliff, which is where you should cut it anyway. so you dont really need a regulator as generally most parts will allow this much tolerance in the supply as long as its steady

A123 ANR26650A and the new updated ANR26650B is the best IMO but levinpower is supposed to be pretty much the same, dont buy from ebay its too risky. current you say? lol these things are insane, they will put out an extended pulse of 140A, but will run at 40A all day long. yes you read that right
 
@Lorien,

this is the "working area" and music comes out. ;)
More later ......
 

Attachments

  • IMG_0022_kl.jpg
    IMG_0022_kl.jpg
    153.3 KB · Views: 980
Hi Lorien, would it be possible to feed the board with 3,3V in stead of 5V? The specs of the xmos chip allow 3,3v. It would be nice when we could use LifePo4-batteries, works great on my current Hiface.
regards, Douwe
Unfortunately WaveIO card does not offer this feature! I know that XMOS I/O ports are working with 3.3V BUT you forgot that there are voltage regulators before XMOS chip (in fact there are all over the place). AFAIK minimum voltage applied to external PSU connector must be between 5V and 6V. There could be a chance to wire those batteries to this card by means of using few diodes between the two... just to drop the voltage to suggested levels. I remember that it was someone willing to try this path but no news so far ... so I don't know if it will work as expected.

@Lorien,

this is the "working area" and music comes out. ;)
More later ......
Nice setup you got there Carsten! One small thing I will do is to short the wires of PH regs - if possible. Thank you for sharing!
Kind regards,
L
 
I was about to say the same thing; using such expensive high performance regulators and leaving long leads and a connector between them and the load will destroy all benefit of having them. I would get rid of the connector and hard wire them as well.

i was meaning to supply directly, couldnt he plug the batteries directly in? they are pretty much perfect voltage and most 3v3 compliant parts will allow +/-5% not for the faint hearted though as these things will literally melt themselves and the load if shorted

and @ Lorien: I saw your reply, no problem about taking it to PM, i'm sorry i let it go on as long as I did. as for my application, i'll send you some info and pics privately rather than in the thread.
 
Last edited:
Hi,
thank you.
There are some things to do, like - - - changing to none isolated I2S, ....
Carsten
Why is this? I had the impression one would like optimum isolation between power domains. . as that is said to reduce jitter-induced distortion.
The isolator behaves pretty well I understand. (there was a post on that).

[That is why I would expect an isolator on the USB input, but that is just my rambling].
 
Thanks cusp for the info, the A123 is not easy to find in Europe but I have probably found an Internet store.

-Roberto

umm every lifepo4 battery ive used is 3.3v nominal as its normal voltage. at fully charged it will be 3.6 but this drops to 3.45 very quickly and spends the rest of its charge between 3.45 and 3.2 before falling off a cliff, which is where you should cut it anyway. so you dont really need a regulator as generally most parts will allow this much tolerance in the supply as long as its steady

A123 ANR26650A and the new updated ANR26650B is the best IMO but levinpower is supposed to be pretty much the same, dont buy from ebay its too risky. current you say? lol these things are insane, they will put out an extended pulse of 140A, but will run at 40A all day long. yes you read that right
 
if they are selling to the public, then i would not trust them to be real. more likely grey market or flat out fakes as these are a highly sought after product and the company normally only sell to OEMs. i do know where to get legit new product though. where are you talking about? i dont want to post the company's info in public as i dont know whether they are acting within their agreement and i want to continue to be able to get them
 
They can be legally found (sometimes rebranded) in stores selling Radio Control products.

-Roberto

if they are selling to the public, then i would not trust them to be real. more likely grey market or flat out fakes as these are a highly sought after product and the company normally only sell to OEMs. i do know where to get legit new product though. where are you talking about? i dont want to post the company's info in public as i dont know whether they are acting within their agreement and i want to continue to be able to get them
 
umm every lifepo4 battery ive used is 3.3v nominal as its normal voltage. at fully charged it will be 3.6 but this drops to 3.45 very quickly and spends the rest of its charge between 3.45 and 3.2 before falling off a cliff, which is where you should cut it anyway. so you dont really need a regulator as generally most parts will allow this much tolerance in the supply as long as its steady

In general, I thought most parts were +/-5% of 3.3v, for operating voltage range. IOW, voltage range min of 3.135V to max of 3.465V. They usually have a higher absolute max voltage, but I would not want to hit the absolute max voltages.

Randy
 
easy enough to deal with, ive never had any problem charging to full because under load the drop from 3.65 down to 3.45 takes a matter of a couple of minutes generally. sometimes they will fall down to this n their own. if you really worry just charge them to 3.45, as thats about 95% full minimum
 
no they cant; well not as loose cells as this is against the reseller agreement and they would lose their resellers rights, maybe old or rebadged, but certainly not new stock. ive been using these batteries for years and yes i frequent the RC forums (like RC groups which is probably where you are talking about and they sell the levinpower cells i mentioned) as thats where i found out about them, not audio and i use a cellpro 10XP 10 cell RC fast charger. you find rebranded, which are NOT legit, but rather grey market cells, sometimes rejects. but the ANR26650m1B has been out for quite a while now and nobody sells them online still because they havent been faked or hit the used/rejected state yet.

believe me, i buy from a legit reseller, there are no legally rebadged, or legit single cells for sale. you can buy the A123 racing packs which are welded but that is the only way apart from through industry that you can buy them legit. many claim all sorts of things, but they are stretching the truth. the power tools packs and RC packs are still the old ANR26650A as well afaik. i'm been saving a bit of coin to buy some of the prismatic cells, even lower impedance, MUCH higher power capability (1200W) as if that was needed and much higher capacity 19.6Ah vs 2.3 (anr26650m1A (old cells)) or 2.5Ah (anr26650M1B new cells) about $100 a pop for the prismatic AMP20M1HD-A
 
Last edited:
Go L!

By posting a 'ready for production' product on a diy forum Lorien has received much help that may be useful during a design review of a basicly diy product. WaveIO is extremely well 'positioned' as a stereo audiophile product whose allowance for external power drives jitter to extremely low levels, chiefly because GMRs are evil.

In addition, his product can be used by every popular personal computer supporting USB and having sufficiently short DPC latency to avoid buffer under runs. The effect of an under run is, in any case, only a short pause in the music, not a system crash.

Given that WaveIO sits near the juncture of digital and analog it brings forth an unusual mixture of readers, some with in-depth expertise on one side or the other but rarely on both. My sympathy goes out to Lorian (a.k.a. Lucien) partly because I once developed a product that was purchased by thousands of users, some of whom were not sophisticated enough to utilize it properly. My satisfaction guarantee and personal tech support kept the return rate way below 1% and required less than an hour a week of my time. That was in 1985.

The situation is much more dificult today, in part because ease of use has amplified the Dunning-Kruger effect with personal computer market penetration approaching ubiquity. I take my hat off to 'L' for his product positioning, excellence in production, and calm demeanor in this thread.
 
I see gusp your point, I'm not in the market.

My question: where can I buy top-notch rechargeable batteries at 3.3V?

You suggest me batteries that only NASA can buy :)

-Roberto

P.S. Maybe this has become OT, maybe it is useful also for the WaveIO (5V)

no they cant; well not as loose cells as this is against the reseller agreement and they would lose their resellers rights, maybe old or rebadged, but certainly not new stock. ive been using these batteries for years and yes i frequent the RC forums (like RC groups which is probably where you are talking about and they sell the levinpower cells i mentioned) as thats where i found out about them, not audio and i use a cellpro 10XP 10 cell RC fast charger. you find rebranded, which are NOT legit, but rather grey market cells, sometimes rejects. but the ANR26650m1B has been out for quite a while now and nobody sells them online still because they havent been faked or hit the used/rejected state yet.

believe me, i buy from a legit reseller, there are no legally rebadged, or legit single cells for sale. you can buy the A123 racing packs which are welded but that is the only way apart from through industry that you can buy them legit. many claim all sorts of things, but they are stretching the truth. the power tools packs and RC packs are still the old ANR26650A as well afaik. i'm been saving a bit of coin to buy some of the prismatic cells, even lower impedance, MUCH higher power capability (1200W) as if that was needed and much higher capacity 19.6Ah vs 2.3 (anr26650m1A (old cells)) or 2.5Ah (anr26650M1B new cells) about $100 a pop for the prismatic AMP20M1HD-A
 
Last edited:
well you can buy the official A123 racing RC packs that are 2 of the old type welded together in series and either make a matching connector that uses them in parallel, or cut the welds. or you can buy some of those levinpower ones at RC groups (cant remember the actual shops name ATM) which do appear to be either old, stolen or diverted last gen A123 and just charge to 3.45. they'll work and from reports the levinpower ones are actually quite good, but they arent officiel rebadged A123 as claimed; thats all I was saying and i would stay away from buying things that claim to be official A123 on ebay. also i can probably put you in contact with my source, i just cant post here and you'll probably have to buy 10.

its borderline OT, but i dont really think this post is, we are discussing alternate power sources for WAVE IO.
 
Thank you qusp
-Roberto

well you can buy the official A123 racing RC packs that are 2 of the old type welded together in series and either make a matching connector that uses them in parallel, or cut the welds. or you can buy some of those levinpower ones at RC groups (cant remember the actual shops name ATM) which do appear to be either old, stolen or diverted last gen A123 and just charge to 3.45. they'll work and from reports the levinpower ones are actually quite good, but they arent officiel rebadged A123 as claimed; thats all I was saying and i would stay away from buying things that claim to be official A123 on ebay. also i can probably put you in contact with my source, i just cant post here and you'll probably have to buy 10.

its borderline OT, but i dont really think this post is, we are discussing alternate power sources for WAVE IO.
 
Hello,
@qusp, thanks for your remark ...
I was about to say the same thing; using such expensive high performance regulators and leaving long leads and a connector between them and the load will destroy all benefit of having them. I would get rid of the connector and hard wire them as well.
Yes, it comes in the "final version", this is only my playground.

@triode_al,
thanks for these hint
Carsten
Why is this? I had the impression one would like optimum isolation between power domains. . as that is said to reduce jitter-induced distortion.
The isolator behaves pretty well I understand. (there was a post on that).
Oh, that i misunderstood these post (my english is not the best), i will search again for these post.

[That is why I would expect an isolator on the USB input, but that is just my rambling].
I have tried the "ultravox" from diyparadise.com (USB Isolator based on Analog Devices' ADUM4160.), but then the mac down't recognized the WavIO anymore.
yeo think that the ADUM4160 can only 16/44.

@all,
does anybody knows another chip that can handle more?