XM PCR Digital outs

[Sawzall]

I know a couple of people emailed me interest in making a digital output for XM recievers (a satellite based radio broadcaster here in the states). Up to this point all of the units had analog outputs only. Given that they have focused on cars, this makes sense. Plus, digital outs gives the RIAA and the rest the shakes - even though the quality of those bits is pretty shaky if you consider they are compressed to within an inch of thier lives...

Well it seems the XMPCR unit, which was designed to connect up to use your computer as a control surface, using analog outputs into your soundcard, has the digital signals present on the board... a fine fellow went hunting and found the digital data and the clocks. An investigation of the usefullness was made and he produced a working version using the CS8402. But its a dead chip, so he moved on to the CS8405. There he has had issues, especially in getting the TTL and a PLL to work right.

So I went looking for other alternatives. Found that the TI DIT4096 seems to work perfectly.. so much so that in hardware mode, it requires no other active devices to be made work. Its a TSSOP chip, so while I wait to get one of the chip adapters. I will wire it up point to point - since it is just connecting up high and low, using ground and the power from the USB.

attached is schematic - anyone with suggestions are welcome to chime in.

[tiroth]

Deadbug-ing a TSSOP is going to be fun...

[Sawzall]

I miswrote that... I am waiting on the chip adapter to wire it up point to point. You would have to be either nuts or really into pain to try it otherwise. S***e, I can barely see those pins.

[Guido Tent]

Originally posted by Sawzall

attached is schematic - anyone with suggestions are welcome to chime in.

Hi

I'd suggest reclocking of the SPDIF, if low jitter is required

Furthermore I'd suggest putting the 110 resistor at the output of the transformer and readjust it because the driver is not zero ohm

best regards

[Sawzall]

Thanks for the suggestion on placing of the 110 resistor. I just copied from CS and TI's application notes...

as for reclocking for jitter, I suspect that it would be guilding the turd, so to speak but I am open to thoughts on it if it really might help. I took the really simple approach to this figuring "how much should one spend screwing around with a $70 receiver that is pushing bits."

It has actually turned out more interesting than one might think at first glance. They implemented this using 44.1 - which makes sense since they did not want to ever change rate from the CDs that they are playing. So it turns out their entire plant is at 44.1. They also used I2S (which made the job simpler). The amazing thing is that they have managed to get decent sound. For those that do not know this service, they basically have 12.5 Mhz of bandwith in the SBand. They deliver about 100 streams to your radio over this. The 40 or so talks channels sound a little better than a good radio, so the 60 music channels are getting most of the bandwidth. I know more than most about how satellite signals are created so it amazes me how will they have been able to deliver on the "CD" quality. Or perhaps it points out how really bad CD's are... take your pick.

(as an aside to those who might want to do something like this - XM has just started showing thier next generation recievers. Basically, its a Walkman. So they are going to be even harder to work as they get smaller).

[Da5id4Vz]

Do you have any suggestions for a source on XM hardware? I haven’t found the flee-bay auctions or froogle searches too encouraging.

$150+ for a Delphi w/ a cassette adapter seems a bit X-tream after I add in the $120 annual service fee.

I’ve got a 2 day trip in the car coming up and I was thinking it would be fun to try the service.

Perhaps I can fake a GUI or spoof the AUI's from a Palm Pilot to work with the PCR, then for under $200 (plus service subscription) I could have a PDA too.

-Dave

[Guido Tent]

Originally posted by Sawzall
Thanks for the suggestion on placing of the 110 resistor. I just copied from CS and TI's application notes...

as for reclocking for jitter, I suspect that it would be guilding the turd, so to speak but I am open to thoughts on it if it really might help. I took the really simple approach to this figuring "how much should one spend screwing around with a $70 receiver that is pushing bits."

It has actually turned out more interesting than one might think at first glance. They implemented this using 44.1 - which makes sense since they did not want to ever change rate from the CDs that they are playing. So it turns out their entire plant is at 44.1. They also used I2S (which made the job simpler). The amazing thing is that they have managed to get decent sound. For those that do not know this service, they basically have 12.5 Mhz of bandwith in the SBand. They deliver about 100 streams to your radio over this. The 40 or so talks channels sound a little better than a good radio, so the 60 music channels are getting most of the bandwidth. I know more than most about how satellite signals are created so it amazes me how will they have been able to deliver on the "CD" quality. Or perhaps it points out how really bad CD's are... take your pick.

(as an aside to those who might want to do something like this - XM has just started showing thier next generation recievers. Basically, its a Walkman. So they are going to be even harder to work as they get smaller).

Hi

It appears CS and TI do not understand transformers

Jitter: It depends on the sound quality you want to achieve

enjoy

[Sawzall]

You live close enough that I can turn you on to the dark side... I have a Sony plug and play laying around if you want to borrow it for your trip. Uses a magnetic antenna so you can just slap it on the roof of the car.

As for the price of the units, I suggest you keep an eye on the promotions. I have seen the Skyfi units on sale with rebates at Best Buy and the like for about $50 to 75. There is no market yet on Ebay, and it will be a while before there is since there are just not that many out there, and so far, they have had less than 3% churn on them. (which is an amazing number - shows you how addictive the service is. You know broadcasting. Their average listen time is *hours*, not minutes. ) The other option is wait till the next generation come out - but given the shortages that happened last Xmas on them, I don't know if that will work.

Secondly, they don't want them too cheap. They are subsidising these by some number - it was about $100 last time they reported - new numbers come out August 7th.

And a good part is that they have zero market value stolen - the guy buying it cannot get the service billing changed unless they can come up with the password (or credit card) used to register it the first time. Having had my stuff stolen more times than I want to think about, this makes me feel good.

Drop me a line if you are interested on borrowing the P&P.

[Da5id4Vz]

HEY!

That’s very generous! “turn you on to” why do I feel like Ken Kesey meeting Neil Cassidy for the first time?

I will likely take you up on the offer assuming my employer doesn’t have me sequestered while my wife needs to head back to New Orleans. (Complicated huh?)

I’ll know within the next week what’s going on with at least a small part of my life.

I think that stalking a sale w/ rebates from one of the big retailers is a great idea. Since placing the post I did a bunch of reading about the XM PCR. It looks like it has all kinds of potential for some fun. The XM fan site has a bunch of protocol stuff controlling the PCR via the USB port. I tried sending an email to you with a bunch of ideas I was having in this area. Looks like good territory for a PIC application.

My employer has similar issues with electronics for our product line. DSL modems still cost a small fortune.

[Sawzall]

The PCR is a great opportunity to play with it. They built thier reciever in pretty much a "one" chip package with industry standard parts around it to control the thing. The FTDI usb chip is pretty open with drivers for everthing. They used a client server scheme with a "TCP/IP" type of communication - and did not attempt to hide the signals in anyway. So folks have already put together alternative control software that lets them have the receiver at home and run and listen to it at work even. Although, it makes a lot more sense to run it over the home network - have the reciever sitting someplace, and then using a client to play it where ever you might be in the house. Running it over the net probably breaks the user agreement, but they have so far had a pretty hands off attitude - as long as you are not trying to rip them off from paying the subscription fee, its ok with them.

The digital outs is another winner. Lots of folks have wondered why they didn't just make them come with the outs in the first place. Well, running them down the USB would just make them sound worse since that is a crappy way to send sound. And in the alternative, I don't think that most soundcards have digital inputs. I use a pro level soundcard, so it has them. Others are just connecting it up to thier usual sound equipment. No doubt the Digigram card I use is a hell of a lot better than the DAC in the PCR.

The other issue is of course the RIAA - they just get scared of digital anything. But the other side of it is the consumer. They may think that a digtal signal means that they would be able to record them in a MP3 or the like. NOT. Transcoding is just not viable - so having the digital outs would just create consumer disappointment when it sounded like crap on replay.

But to your point on a PIC application - for sure. Simple matter of programming as they say. The other area of exploitation is for business users - they offer a rate of $25 a month for commerical use. This market is big and no one has developed a player focued on this niche yet. Although the number of business' that could play out the 66 Raw stream is probably limited.... (being a satellite service, they are not censored, so some of the channels, like gangsta', can be rough to say the least).