I skipped the gnd and power. Both the computer (wandboard) and WaveIO are sharing GND via the power supply. Works fine.
MArco
I have a basic question, but I must provide boring details for it to be answerable.
I have ordered a WaveIO from Lucian and plan to build it into a metal chassis. This is to sell off a Berkeley Alpha USB between my PC and a W4S DAC-2 after favourably comparing my brother's WaveIO which was just floating there with power from a cheap 5V switch mode plug pack. Prior to this comparison the Berkeley was always a clear winner when compared to many other USB/spdif converters with the Halide Bridge being by far the sloppiest. This time they sounded the same (waveIO = Berkeley).
The metal chassis will be Earthed to the mains and house a 9V totoidal tranny with just a single secondary winding. It will also house a third party power supply PCB to provide a full wave rectified and LM317 regulated 5Vdc output with the 0V reference isolated from the chassis. This will provide the two DC leads to the WaveIO screw terminals. i.e. the PSU will not tie the WaveIO to mains Earth.
The chassis will have a hole to receive the thread of a plastic-bodied BNC right angle PCB connector mounted alongside H4. i.e. no direct connection between shell of the BNC and the chassis.
For the USB input I plan to run a short cable from the USB header pins to a plastic bodied panel mount B type receptacle. Again no contact between the USB and the mains earthed chassis.
The question: Should I retain this all-over chassis isolation of the WaveIO card by using a plastic stand-off at H4, or should the ground plane of the WaveIO be single-point earthed using a metal standoff at H4?
My intuition says is to use plastic and rely on any Earthing via the coax shield to the DAC (since that's how it would have been in the test with my brother), but I'd be interested in any ideas.
Thanks.
I have ordered a WaveIO from Lucian and plan to build it into a metal chassis. This is to sell off a Berkeley Alpha USB between my PC and a W4S DAC-2 after favourably comparing my brother's WaveIO which was just floating there with power from a cheap 5V switch mode plug pack. Prior to this comparison the Berkeley was always a clear winner when compared to many other USB/spdif converters with the Halide Bridge being by far the sloppiest. This time they sounded the same (waveIO = Berkeley).
The metal chassis will be Earthed to the mains and house a 9V totoidal tranny with just a single secondary winding. It will also house a third party power supply PCB to provide a full wave rectified and LM317 regulated 5Vdc output with the 0V reference isolated from the chassis. This will provide the two DC leads to the WaveIO screw terminals. i.e. the PSU will not tie the WaveIO to mains Earth.
The chassis will have a hole to receive the thread of a plastic-bodied BNC right angle PCB connector mounted alongside H4. i.e. no direct connection between shell of the BNC and the chassis.
For the USB input I plan to run a short cable from the USB header pins to a plastic bodied panel mount B type receptacle. Again no contact between the USB and the mains earthed chassis.
The question: Should I retain this all-over chassis isolation of the WaveIO card by using a plastic stand-off at H4, or should the ground plane of the WaveIO be single-point earthed using a metal standoff at H4?
My intuition says is to use plastic and rely on any Earthing via the coax shield to the DAC (since that's how it would have been in the test with my brother), but I'd be interested in any ideas.
Thanks.
Hi Lorien:
Still trying to figure out whether I should use a conducting (metal) standoff to attach the Wave board to my metal Hammond chassis through that grounded H4 point. (I'm going to use an insulated USB B jack, with a connector to the header on the board, setting the appropriate jumper.)
Is it better to have the board shielded by the chassis through this ground connection, or not?
Thanks for your help in advance.
Larry W.
Seattle area
Still trying to figure out whether I should use a conducting (metal) standoff to attach the Wave board to my metal Hammond chassis through that grounded H4 point. (I'm going to use an insulated USB B jack, with a connector to the header on the board, setting the appropriate jumper.)
Is it better to have the board shielded by the chassis through this ground connection, or not?
Thanks for your help in advance.
Larry W.
Seattle area
Zapped it seems I missed your question or you missed mine (post just prior to your last), but maybe we're asking the same question (you a little more concisely perhaps). I don't think it's a matter of shielding so much though since the metal chassis will do that anyway assuming it's earthed to the mains - more a matter of avoiding ground loops. Looking forward to the replies.
@ Ian and Larry: to be honest I didn't had the chance to test any chassis configuration with WaveIO, the main reason is the fact that I don't have the time nor proper equipment to "see" the differences. Anyway, please do remember that the ground connection from the screw-type power connector is directly tied to the H4 mounting hole through the internal ground planes of WaveIO. So, powering the card by using thin wires while connecting the board to chassis through H4 will create a ground loop from the start. Also, I do not like to rely on connections that are not too "safe" for me so, if I have to choose to use or not H4 then I would NOT use it. I like to keep things 'simple' meaning that WaveIO will be powered through twisted/thick/short power cables from a dedicated PSU and then, that PSU will be connected to a common ground (preferably in a star configuration). In this case please do try to isolate the connectors directly tied to WaveIO from the case otherwise ground loops will appear again.
Maybe for small chassis grounding the WaveIO through H4 directly could become 'appealing' but for large cases I would try to avoid it.
I hope it helps,
Lucian
Maybe for small chassis grounding the WaveIO through H4 directly could become 'appealing' but for large cases I would try to avoid it.
I hope it helps,
Lucian
OK, I've assembled the necessary bits & pieces (including the 1uH choke as suggested by Triode_al. Thanks!)
Now I'm wondering about how to ground this. I'm currently using the isolated spdif output for the co-ax output but now I'll have the +5V return to consider. Can I just run this back to the 5V supply into the board, or into the spdif return?
All and any help gratefully received!
OK, I've assembled the necessary bits & pieces (including the 1uH choke as suggested by Triode_al. Thanks!)
Now I'm wondering about how to ground this. I'm currently using the isolated spdif output for the co-ax output but now I'll have the +5V return to consider. Can I just run this back to the 5V supply into the board, or into the spdif return?
All and any help gratefully received!
If you will connect the return to the WaveIO's ground plane then you will defeat the trafo's purpose: isolation... but since you are going to use fiber optics then I guess it will not count too much! Anyway, the isolation is good to keep noise at distance so I suggest to take this advantage. For the sake of DIY please do use the SPDIF return with your 5V ground and keep WaveIO's GND away from it. Please do let us know about your improvements.
Hi Stefan,
the WaveIO's SPDIF output voltage is 0.5Vpp so from what I can see you cannot drive the miniDSP directly but you have to use a voltage translator for that. @ J11 connector, the "hot" wire is the squared-shaped one while the return is the rounded pad.
Kind regards,
L
Thanks Lorien. So as my original intention was to have both a co-ax and an optical spdif connection - any reason why I shouldn't use the isolated spdif of the Wave board for co-ax and the non-isolated connections for the optical?
Be careful with connecting a TTL output to a 3.3V input. Vcc for TTL should be 5V +- 10%, which might be too much for an input that expects a maximum input of 3.3-3.6V. It might survive, but then again it might not. As Lorien noted this doesn't apply in this case (as the WaveIO has a lower voltage output), but it's still an important note.
EDIT: Also, you might run into data errors related to differing high and low thresholds. I wouldn't expect this to be likely, but it could happen I guess. If you connect a 3.3V output to a TTL input this is much more likely to happen of course.
Last edited:
Good note
the minisharc expects a minimum input of 3.3V but from what I've read it is a 5V level input. Unfortunately the documentation is not very good and the response through the forum is slow at best. Great products but like other companies in the DIY arena they can be let down by support. Compared to Loriens excellent support on this list they are lackingI've ordered one of the twisted pear level converters so hopefully this'll get me sorted.
The optical side will be hard to drive. That's because WaveIO outputs SPDIF signals with 0.5Vpp in 75R load. In single connection the coax will work but if you are going to use the non-isolated output then things will be messed up! With this voltage swing the optical transmitter will sleep forever and you will not get any signals from it. I suppose you will need a converter too, wired directly to Wave's SPDIF output and from then try to drive both coax (with appropriate resistor dividers) and optical in the same time. Hmm... converter idea seems to be a reasonable one because you will not get any phase delays between the two outputs... but I'm thinking out loud even though there's something that I don't like regarding this path
Thank you! I'm doing my best.Good note
I've ordered one of the twisted pear level converters so hopefully this'll get me sorted.
Have a look at the Ultra High Speed Buffer module by dvb Projekt ( 10 USD) should work. Here is a picture.
It is a finished board - but [imho] however make a local earth on the board (In Bernards setting he does not need it, his lines are very short. . .
Last edited:
Lucian, may I ask - what is the actual "source voltage" across each LED take-off? In a previous post in this thread a member suggested to another that he place a 5K resistor in series with a standard LED to limit current to your specified 8mA, but 5K seems very high. If the "source voltage" is (say a guess) 3.3V, and your LED has a 2V drop, then the resistor ought to be around 180R in order to limit current to 8mA. However we need the actual "source voltage" to calculate this.
I'd just measure it, but I'm yet to receive my WaveIO card.
Cheers,
I'd just measure it, but I'm yet to receive my WaveIO card.
Cheers,
Quick question as I just start the assembly:
I have located the jumper "PSU select" on this card, used to choose between the USB power or an external power, but there is another jumper named "EXT BRD" (look at the yellow arrow on the attached picture) : what is the purpose of this jumper ?
tks
I have located the jumper "PSU select" on this card, used to choose between the USB power or an external power, but there is another jumper named "EXT BRD" (look at the yellow arrow on the attached picture) : what is the purpose of this jumper ?
An externally hosted image should be here but it was not working when we last tested it.
tks
