I didn't have that issue with my wooden enclosure, but it's probably at the analogue output.
That is, I would tend to connect all balanced in- and outputs according to AES48: XLR pin 1 straight to the chassis and one point of the chassis to the circuit ground. For that last connection I'd take a point close to the analogue outputs and connect it to the ground pins of the output connector of the filter board.
Do you have any unbalanced inputs, such as S/PDIF? I think I kept my S/PDIF inputs completely insulated with an insulated cinch connector and a transformer right behind the connector, but I can't check if I remember that correctly right now..
That is, I would tend to connect all balanced in- and outputs according to AES48: XLR pin 1 straight to the chassis and one point of the chassis to the circuit ground. For that last connection I'd take a point close to the analogue outputs and connect it to the ground pins of the output connector of the filter board.
Do you have any unbalanced inputs, such as S/PDIF? I think I kept my S/PDIF inputs completely insulated with an insulated cinch connector and a transformer right behind the connector, but I can't check if I remember that correctly right now..
Last edited:
My initial plan was to wire pin one of the output XLRs to the ground of the filter PCB, then from the filter PCB to the ground loop breaker. Can the jacks be connected to the chassis directly and not the star ground? I had a concern with this method because my case will likely be powder coated.
Last edited:
I think your initial plan should also work. How do you want to make contact with your powder-coated case?
The idea behind AES48 is to let the shields and chassis form one big Faraday's cage, to take for granted that there may be ground loops, but to ensure that the current in those ground loops doesn't flow through any internal circuitry (which is often unbalanced, even in professional equipment). It's an approach that only works with balanced inputs and outputs, but it results in better RF shielding than pure star grounding.
The idea behind AES48 is to let the shields and chassis form one big Faraday's cage, to take for granted that there may be ground loops, but to ensure that the current in those ground loops doesn't flow through any internal circuitry (which is often unbalanced, even in professional equipment). It's an approach that only works with balanced inputs and outputs, but it results in better RF shielding than pure star grounding.
Ok I think I'm almost ready to turn on the DAC for the first time. Right now I only have optical input setup. Before I flip the switch, I had a question about using a DC power supply for the 6.3v. Do I just connect the positive and negative to pins 3 and 4 of the power terminal? I was going to connect the positive to 3, where I jumped the THT.
Well, first test boot was unsuccessful so I quickly turned it off.
All PSUs, including the on-board 5V supply, are working properly. When I turn on the DAC, the tubes begin to light/heat up, but the clipping light is on and none of the other indicator lights work. I have switch 3/C on the FPGA set to on, so the lights should be on even if not using surprise. I also quickly tried surprise to see if that would change anything and it did not.
Could it be that I flashed the FPGA before shorting the resistor pad on it?
All PSUs, including the on-board 5V supply, are working properly. When I turn on the DAC, the tubes begin to light/heat up, but the clipping light is on and none of the other indicator lights work. I have switch 3/C on the FPGA set to on, so the lights should be on even if not using surprise. I also quickly tried surprise to see if that would change anything and it did not.
Could it be that I flashed the FPGA before shorting the resistor pad on it?
Last edited:
That's strange because the 85A2 should turn on after the FPGA has read its configuration memory, which takes two seconds or so. On top of that it has its own ignition delay, but that is usually less than 0.1 seconds unless you would happen to have an 85A2 without radioactive primer. Does it also turn on with the FPGA module removed?
They should both remain off when there is no FPGA module installed. To reduce the risk of damage, I think you have to debug this first before reinstalling the FPGA module.
I'll make a list with test points and expected voltages, but I'll probably do that this weekend because I have no access to my normal computer right now. In the meantime you could check again if there are any shorts in the small-pitch SMD things, like the FPGA module connectors.
I'll make a list with test points and expected voltages, but I'll probably do that this weekend because I have no access to my normal computer right now. In the meantime you could check again if there are any shorts in the small-pitch SMD things, like the FPGA module connectors.
The first thing I did was check whether I had valid DSD as an input to the Valve DAC by using a simple low pass filter arrangement to actually listen to it. Is that possible with this FPGA version?
If you have the gear, I would then only power up the 5V stuff and check that it is working, or at least not producing magic smoke, then proceed in stages. Marcel included some test points for that IIRC?
If you have the gear, I would then only power up the 5V stuff and check that it is working, or at least not producing magic smoke, then proceed in stages. Marcel included some test points for that IIRC?
Can I ask how you connected your 6.3V supply? Did you put the positive to pin 3 and negative to pin 4? Trying to think of anything I may have done wrong.
I disassembled everything and went over every SMD contact, reflowing a few. Turned on without the FPGA, no clipping light this time, but the 85A2 is still dimly lit.
I'm assuming it's abnormal to have voltage at the indicator lamp points when they shouldn't be on? I measured the voltage at one of the LEDs and it was ~50V, and it was one that shouldn't have been activated. Did it cook all my LEDs?
I'm assuming it's abnormal to have voltage at the indicator lamp points when they shouldn't be on? I measured the voltage at one of the LEDs and it was ~50V, and it was one that shouldn't have been activated. Did it cook all my LEDs?
I've downloaded my own schematics, so I could already make a part of the list of voltages I expect without FPGA module:
+3V3: somewhere between -0.7 V and +0.7 V without module (3.3 V with module)
nswitch85A2: same voltage as net +3V3 without module (first 3.3 V and then 0 V with module)
collector Q8: -300 V without module (first -300 V and then +2.5 V with module)
pins 1 and 5 of U25, the 85A2: -300 V without module (first -300 V and then -215 V with module)
collector Q10: 5 V whenever the 85A2 is off, 0.2 V when it is on and the FPGA board is present
ignited85A2: 0 V without module (first 0 V and then 3.1 V with module)
+3V3: somewhere between -0.7 V and +0.7 V without module (3.3 V with module)
nswitch85A2: same voltage as net +3V3 without module (first 3.3 V and then 0 V with module)
collector Q8: -300 V without module (first -300 V and then +2.5 V with module)
pins 1 and 5 of U25, the 85A2: -300 V without module (first -300 V and then -215 V with module)
collector Q10: 5 V whenever the 85A2 is off, 0.2 V when it is on and the FPGA board is present
ignited85A2: 0 V without module (first 0 V and then 3.1 V with module)