|
Post by stanleyb on Aug 13, 2018 17:59:20 GMT
I can't help noticing how many Pi based DAC boards there are out there. I have no idea how good any of them are, and whether there is a market for yet another DAC board. Not that I know anything about the Pi interface at this moment in time. But I do wonder if I am missing out on a sales stream.
|
|
|
Post by sondale on Aug 13, 2018 18:43:26 GMT
Stan,
If you look at the people on this forum, and others, quite a few people are fronting their Beresford DACs (especially the SEG) with Raspberry Pis fitted with a board to generate SPDIF. If you could take the essentials of the SEG or other DACs and create a board to fit the RPI with its HAT connector then you could be on to winner.
There are of course a lot of people now making DACs for the RPI and they are starting to creep upwards in sophistication and price.
A lot of the manufacturers are sticking to the same small footprint and therefore their DACs become 2 or 3 boards high which would include a power-supply plus other bits.
It might make sense to see how small you could make the SEG without having to worry about inputs (you get that from the Pi I2S pins), outputs would be RCAs and you would probably need 5volts input.
I do not know whether people would want headphone output - in my opinion you would need a fair amount of power to do that properly.
So a lot to think about.
|
|
|
Post by John on Aug 13, 2018 18:44:38 GMT
Probably Stan. I think you would also do it better
|
|
|
Post by stanleyb on Aug 13, 2018 19:01:44 GMT
I shall do a case study. What I can say is that looking at what is on the market right now, most of these companies have just slapped a bunch of components on a PCB and hoped for the best.
|
|
|
Post by MikeMusic on Aug 13, 2018 19:30:34 GMT
This is a growing market you ought to be in Stan Best of luck
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 14, 2018 9:00:18 GMT
There are one or two which are pretty passable and a couple more due in the not too distant future, the question is in real terms what is really possible in that type of space and sat on top of a processor which is not emi RFI optimised . Again power supplies to the units etc
A suggestion would be to mount the two sepertaly in a custom enclosure that is designed to compliment the design, oh wait could that be a streamer?
The SEG for the money is good Stan it can be improved quite markedly as well by a few updates and circuit layout changes but it is still two boxes, 3 if you take an off board psu
My suggestion would be rather than looking for a sticking plaster solution fix the root problem
Then you gave to weigh up cost of R&D v's potential RRP the pi route is really designed for diy guys and young kids looking to start electronics projects for the first time
The market will settle down and the next latest fad will take over as forums promote what tickles a particular fancy imho
intetesting times uber mass market with ultra thin margins in relation to the cost means thousands of units gave to be sold to make a reasonable living and the hassle that entails with no tech guts playing with rhem
just a few thoughts, look at how many people who gave actually purchased these things and reverse the figures then assertsin whether it is worth the effirt
|
|
|
Post by Slinger on Aug 14, 2018 11:35:31 GMT
As an outsider, looking in, I'd have to agree with Tony. Firstly it's a question of market, and secondly, It's a question of margins. The only current "hi-fi" solutions seem to be complete streamer units at horrendous prices, bearing in mind we know roughly what goes into them. Personally, I went the route of using an interface HAT and an external DAC. That seemed to me to be the highest quality/bang-for-bucks D.I.Y. approach because many of us already have excellent DACs (many people already have one of your excellent DACs in fact), so buying another excellent DAC is a bit redundant? It adds to the price but not, necessarily, the SQ. I know I'm in "teaching-my-grandmother-to-suck-eggs" mode, but you need to identify not only a market but a sub-section of that market before anything else. There are so many variables: HAT, two cards, one card, streamer, server, power, etc.? A cheap(er) single-box pi-based streamer than is currently available, with the Beresford name, (which already carries a certain cachet) would probably be the most marketable solution to my non-businessmanlike mind. This is an example of the " competition" at around £1300.00
|
|
|
Post by MartinT on Aug 14, 2018 13:44:00 GMT
I agree with the above views.
Another way of cracking it could be to make a SEG-streamer. By that, I mean integrate a Pi (or even Pi Zero) into a slightly larger case, take the I2S output from the Pi and use your reclocking skills to make it as rock stable as you can. Beef up the PSU to feed the Pi with isolated 5V, add some supercappery, licence some software like Volumio and you have a streamer-in-a-box.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Aug 14, 2018 17:58:09 GMT
There are one or two which are pretty passable and a couple more due in the not too distant future, the question is in real terms what is really possible in that type of space and sat on top of a processor which is not emi RFI optimised . Again power supplies to the units etc A suggestion would be to mount the two sepertaly in a custom enclosure that is designed to compliment the design, oh wait could that be a streamer? The SEG for the money is good Stan it can be improved quite markedly as well by a few updates and circuit layout changes but it is still two boxes, 3 if you take an off board psu My suggestion would be rather than looking for a sticking plaster solution fix the root problem Then you gave to weigh up cost of R&D v's potential RRP the pi route is really designed for diy guys and young kids looking to start electronics projects for the first time The market will settle down and the next latest fad will take over as forums promote what tickles a particular fancy imho intetesting times uber mass market with ultra thin margins in relation to the cost means thousands of units gave to be sold to make a reasonable living and the hassle that entails with no tech guts playing with rhem just a few thoughts, look at how many people who gave actually purchased these things and reverse the figures then assertsin whether it is worth the effort
|
|
|
Post by stanleyb on Aug 14, 2018 22:28:59 GMT
Those are all very good ideas. But what interests me in the idea of a Pi board is that the DAC chip that I use has several advanced functions that can only be accessed via I2C. In the case of expansion, the Pi can accept touch screen displays etc. But to get the best out of a DAC PCB mounted on a Pi would require a separate PSU. Using the same PSU to power everything is adding to the noise. On top of that the Pi PSU is 5V. To have a discrete output stage would require at least 12V for the best dynamic range.
|
|
|
Post by Slinger on Aug 14, 2018 23:04:33 GMT
You'd probably still have to consider providing the option of a case if you're thinking of designing a proprietary HAT. Would both power supplies in the same box be do-able?
|
|
|
Post by stanleyb on Aug 14, 2018 23:27:23 GMT
It's not impossible.
|
|
|
Post by MartinT on Aug 15, 2018 5:24:24 GMT
But to get the best out of a DAC PCB mounted on a Pi would require a separate PSU. Agreed - this is what I have done with my Allo DigiOne HAT board, feeding it separate (non-dirty) 5V.
|
|
|
Post by MartinT on Aug 15, 2018 5:26:35 GMT
I wonder if a Pi Zero would give you more flexibility and lower the integration costs? It could well fit into a SEG sized case.
|
|