|
Post by chinook9 on Feb 13, 2015 14:46:19 GMT
I would build up the power supply on a bench first. connect it to each other (so no PCB's other than power supply) and test it under a load. Verify the output voltages and ONLY then cnnect it to the PCB's. Thank you for the recommendation Frans. I'm not exactly sure how I will do that, but I spotted a DIY thread yesterday that mentions this so I'll do a read. I expect I'll be back with a question or two before I turn on any power.
|
|
|
Post by chinook9 on Feb 14, 2015 4:46:41 GMT
I was not able to find anything about providing a load specifically for the Placids, however what I did find indicates I should be able to use a resistor as a load in each case (not a light bulb as I had been thinking).
PLACID HD: (initial, no-load setup would be 5.25 volts and 490mA CCS) I calculated resistance in ohms is equal to the voltage in volts divided by the current in amps: 11.93 ohms = 5.25 volts/0.440 amps. If this is correct, using a 12 ohm resistor as test load would require adjusting for a CCS of 497mA (437mA load + 60mA shunt). Is it safer at full-load to adjust for slightly less shunt current (i.e. 50mA?)
PLACID HD BP: The load would be connected separately with one load for each rail (one from + to gnd and one from - to gnd.)
Calculating the load results in 98.33 ohms resistance = 14.75 volts/0.15mA. If using a 100 ohm resistor for each rail this would require adjusting each rail for 207mA CCS (147mA load + 60mA shunt).
Recommendations/corrections please.
EDIT: I just realized that the recommended voltage supply for the HD BP is 15 volts. I believe I used 14.75 volts for the warmed-up voltage so that the voltage does not exceed 15 volts when cold. Does this logic hold water? I would much rather be safe than sorry. I believe there is a little bit of latitude to the lower side in the voltage and current but going over can cause problems.
EDIT: I have learned a bit since originating this post and have modified the original numbers. I'm not sure I can find them in town (Phoenix) but I will look tomorrow for a couple of 100 ohm 10 watt resistors, and one 12ohm 5 watt resistor. I will do some reading and calculating tonight and confirm that the wattage rating of these resistors is adequate. If I can't find them in town I'll order from Mouser and just have to wait a few days.
|
|
|
Post by chinook9 on Feb 21, 2015 6:40:06 GMT
Well, the Placids are built and bench tested as Frans recommmended. They are installed with initial voltage and current settings per Javier's recommendation. Everything has gone well so far.... I was originally planning to start by using SPDIF input directly to the BIIISE, however, I discovered that none of my computers has SPDIF out. Consequently I must start directly with USB. Although I progress rather slowly, I expect I will be ready to plug the required pieces together in the next week and I would like to know if Javier, or any knowledgeable DIYer, can confirm the following wiring connections from the DIYINHK USB to I2S adapter to the BIIISE: DIYINHK BIIISE LRCK D1 BCK DCK DATA D2 D1 is the word clock D2 is Data DCK is the bit clock I found the previous information on a Greek AV forum along with the following graphic. The components pictured above are an earlier generation of the components I have but the connections are the same except for the additional GND on the BIIISE. I believe I should use the GND X2. I would appreciate any confirmations or corrections. Any mistake at this stage can be quite costly. Thank you.
|
|
|
Post by chinook9 on Feb 22, 2015 21:19:47 GMT
Hello Javier: I am getting to the final stages of putting the components together and I was surprised by the fit of the BIIISE over the IvyIII. As you can see the spacers, on the Ivy are shorter than the pins connecting the BIIISE to the IvyIII. Do you know if this is normal or have I missed something? If necessary, I will get larger spacers, or I may just add an additional spacer above those installed.
|
|
Javier
Administrator
Digital bytes
Posts: 987
|
Post by Javier on Feb 22, 2015 22:52:27 GMT
You got the wiring correctly though I'd strongly suggest to keep the I2S cables as short as possible, in the picture they cut be cut much, much shorter so don't take it as an example. Also I'd try to keep the GND wire twisted around each signal cable for as long as you can, in the picture they are separated, IMO, too early and I think it is possible in BIIISE to keep them together to the end?
Don't know about the pins/spacers issue but I'll pass the question on to a friend who has build some and I'll post his say.
|
|
|
Post by chinook9 on Feb 23, 2015 1:39:24 GMT
Thank you Javier.
My I2S signal wires will be around 7cm, which I believe is OK.
I plan to twist the I2S Signal/GND wires up to very close to the BIIISE. All the signal wires will be the same length and the GND wires slightly longer but the same length each.
I will look around locally and see if I can find some spacers of approximately 4mm length to use below the BIIISE. I could reduce the length of the 6 pins connecting the BIIISE and IvyIII by 4mm but I would rather not do that.
Looking forward to see what your friend has to say.
Brian
|
|
Javier
Administrator
Digital bytes
Posts: 987
|
Post by Javier on Feb 23, 2015 12:06:21 GMT
He said he measures before soldering and when the pins are too long he pushes them down and cuts them. So you can go either way. The longer spacers are the easiest path but there is no problem in cutting the excess, actually it provides a shorter signal path which is always a good thing.
Sent from my GT-I9100 using proboards
|
|
|
Post by chinook9 on Feb 23, 2015 17:31:26 GMT
Thanks Javier. I'll probably just go ahead and cut the pins, but I may try it as is first.
Do you know of any problem that might be caused by using shielded Canare microphone cable for each SE output channel. I'd use all four conductors for each channel with two positive and two ground wires.
I don't know if inductance might cause a problem. I know pretty much nothing about inductance except in some situations it can cause problems.
|
|
Javier
Administrator
Digital bytes
Posts: 987
|
Post by Javier on Feb 23, 2015 18:12:56 GMT
I don't think there will be any problem cosnidering the relatively short lengths used, 20cm at most? though to be safe that is a question I pass on to Frans (are you there?)
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Feb 23, 2015 18:43:30 GMT
yes... I am here...
keep wires as short as possible and for digital data the exact same length. Personally I would not wrap ground wires around digital wires. It won't do anything but increase capacitance. Capacitance is digital's worst enemy. I would use coax instead as it is designed for higher frequencies OR perhaps better .. run ground wires in parallel to the data lines (so ground wires between data lines, like a flatcable) The more ground wires in parallel the better it 'works'.
For audio you can use unshielded wiring but if the transformer is in the same enclosure I would use coax or shielded cable.
I would use one single shielded cable per channel and route separate wires. The shield must be connected to ground on both sides.
|
|
|
Post by chinook9 on Feb 24, 2015 4:24:06 GMT
Thank you for the recommendations.
I will use flatcable for the I2S digital signal (about 7-8cm) and shielded microphone cable for each audio out channel(about 25 cm).
I have connected the shielding of the audio cables to the ground connections on both ends... at the RCA jacks and at the IvyIII SE outputs. The RCA jacks are isolated.
The audio cables are longer than I like but there is not much I can do about that now. Maybe on the next enclosure if there is one.
Getting close....
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Feb 24, 2015 9:26:41 GMT
Long audio cables isn't a problem. Just keep them as far away from the trafo(s) as possible. Better to have a longer cable routed away from the transformer than a shorter wire running close to them.
|
|
|
Post by chinook9 on Feb 24, 2015 15:37:37 GMT
Long audio cables isn't a problem. Just keep them as far away from the trafo(s) as possible. Better to have a longer cable routed away from the transformer than a shorter wire running close to them. Excellent! The trafos are on the other side of the enclosure. I put the right RCA output jack much closer (12cm) to the trafos than it could be but I never thought about it at that time. If necessary I could install shielding.
|
|
|
Post by chinook9 on Feb 25, 2015 4:03:27 GMT
Well Javier, I connected all the components except the USB to I2S converter and turned on the power.
The HD BP supplying the Ivy seems to be OK. Connected to the Ivy, the side 1 readings are 14.77 volts, 217 mA CCS, and 102mA shunt current. Side 2 is 14.76 volts, 219 mA CCS and 105 mA shunt current. I did not adjust these as I expect they're probably OK for now. The heat sinks don't seem to get very warm, even with such a high shunt current.
The Placid HD, however, is a different story. No load the reading are 5.25 volts, 500mA CCS, and 292 shunt current. Under load, connected to the DAC, it is 4.50 volts, 500mA CCS, and no shunt current. Connected to the DAC I was unable to adjust any of these readings with either VR1 or VR2.
I disconnected the DAC and adjusted CCS to 600mA and connected the load again. Connected to the load the readings were CCS of 620mA and 8mA shunt current. Again, unable to adjust these with VR1 or VR2.
When the DAC is connected to the Placid, the lights on the tridents appear to be normal, however, the LED on the AVCC module is very dim. On this latest AVCC module there is only one LED. The "Mute" light on the DAC also lit.
TP_VREF+ to TP_GND2 is 2.48 volts under load.
I checked the board for incorrectly placed components, solder bridges, and colder solder joints. I resoldered one joint but didn't see any other suspects.
I do have what I believe is called a lifted trace at C8. I overheated the board when I was soldering C8. It is only slightly lifted and I figure it is probably OK but it may not be.
Tomorrow I plan to wire up the USB to I2S converter and try again.
EDIT 3-25-2015: The CCS and shunt current readings were abnormal because the multimeter I was using was incorrect. I later found that with an accurate multimeter the readings were close to what would have been expected. Everything has worked very well.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Feb 25, 2015 6:03:03 GMT
It appears as though the DAC board draws more current than expected. Would set the CCS to 700mA. Under load there should be 600 going to the DAC and 100mA in the shunt. TP states the current draw should be around 440mA so around 600mA is a bit high
check the power supply voltages before and after the shunt. After the shunt you should be having 5.25V, before the sunt slightly higher.
Check the DAC PCB for parts that get hot/very warm.
use a multimeter to check traces for continuity (C8)
|
|