Finally, we decided to order 500pcs Cosmos Scaler production, perhaps, at the same time when we'll order Cosmos ADCiso i.e. next few weeks.
Cosmos Scaler THD performance test. The input-to-output H3 difference is 1.4db at -161db i.e. Cosmos Scaler(gain=0db) residual H3 is below -175db@1kHz 1.7Vrms and 640ohm of Cosmos ADC input. Hence, at +26db H3 is expected at -150db. At the higher Cosmos ADC range(>1.7V) H3 should be even lower.
For more detail, Archimago test is here: https://archimago.blogspot.com/search?q=Scaler
also, I told about that "Cosmos ADC buffer" in that thread year ago.
Cosmos Scaler THD performance test. The input-to-output H3 difference is 1.4db at -161db i.e. Cosmos Scaler(gain=0db) residual H3 is below -175db@1kHz 1.7Vrms and 640ohm of Cosmos ADC input. Hence, at +26db H3 is expected at -150db. At the higher Cosmos ADC range(>1.7V) H3 should be even lower.
For more detail, Archimago test is here: https://archimago.blogspot.com/search?q=Scaler
also, I told about that "Cosmos ADC buffer" in that thread year ago.
As I see, the demand is about 200pcs but it is too small a number to order(especially for mechanic parts, 200pcs are just ridiculous, and I'll order 1K at least), so I expect a little batch of 500pcs will be available >1 year. Also, I need ASAP to refine the software algorithm of the autoranger. I noticed some glitches there, the reason is gain differences-tolerance, I guess.
20Vrms for balanced and 10Vrms for unbalanced sources. Relays, that's quite obvious according to so low distortions.What is the max voltage it can handle? Are you using relays or fet switches for ranging? I'm in for one whenever its ready.
Nop, but I added some silk-guidelines to avoid mistakes with the thermostat board installation.Will the V2 version of the ADC come with the temperature controller built-in? Not necessary?
I made some unification for the original Cosmos ADC and Cosmos ADCiso FW to match digital frequency responses. Both ADC versions will share the same FW. However, due to hardware differences, the analog frequency responses are different, hence, we need two compensation files, one for Cosmos ADC and another for Cosmos ADCiso. Both files flatten the response down to +/-.1db from .8Hz to 383.98kHz, or let's say +/-.3db for unlucky tolerance.
Attachments
Hello,
I am looking for a test solution to measure the noise of power supply rails up-to 36V DC.
This little Gizmo shows the best specs and performance, but I can only find information where the Cosmos is used for testing the performance of audio equipment. So I am wondering if this unit is suitable for the purpose that I have in mind?
If yes what do I need further besides the Cosmos ADC? I would also like to know which interface board/circuit I need to connect the output of the power supply rail to the input(s) of the Cosmos?
I hope that one of you can help me further.
Thnx, Ruthless
I am looking for a test solution to measure the noise of power supply rails up-to 36V DC.
This little Gizmo shows the best specs and performance, but I can only find information where the Cosmos is used for testing the performance of audio equipment. So I am wondering if this unit is suitable for the purpose that I have in mind?
If yes what do I need further besides the Cosmos ADC? I would also like to know which interface board/circuit I need to connect the output of the power supply rail to the input(s) of the Cosmos?
I hope that one of you can help me further.
Thnx, Ruthless
I use a +60dB LNA feeding a Focusrite Solo ADC. You do need to make sure you feed the rail through a capacitor before going into the LNA. My LNA has a 220uF bipolar at the input. You should be able to measure an unregulated rail without the LNA, but you still need the capacitor to block the DC.
The LNA inputs are fuse-like protected, each one contains 10ohm 0603 in serial, and 5V TVS to GND.
The short story why it is finally like that: the first LNA chip was AD SS2019, and I noticed that H3 has tend to degrade if an input reaches >5-8V.
Next, I decided to use THT1510 and I see no such distortion degradation, however, I kept the protection there just in case. In most cases overvoltage will burn the resistor 10ohm, TVS and THAT1510 are safe. A 36Vdc is quite high and a fast transient may burn the resistor 10ohm, a slow transient could be safe.
Anyway, better to add a couple of external capacitors and one resistor between them to cut DC safely.
The short story why it is finally like that: the first LNA chip was AD SS2019, and I noticed that H3 has tend to degrade if an input reaches >5-8V.
Next, I decided to use THT1510 and I see no such distortion degradation, however, I kept the protection there just in case. In most cases overvoltage will burn the resistor 10ohm, TVS and THAT1510 are safe. A 36Vdc is quite high and a fast transient may burn the resistor 10ohm, a slow transient could be safe.
Anyway, better to add a couple of external capacitors and one resistor between them to cut DC safely.
Take a look at Texas Instruments SLYY076 app note.
They are discussing there some interesting points about low noise voltage regulator measurements - coupling cap, shielding, etc.
They are discussing there some interesting points about low noise voltage regulator measurements - coupling cap, shielding, etc.
I think so high PS as 36V doesn't really need a super precise shielded rig as SLYY076 shows.
I would take simply 2 aluminum NP caps 22u/50V and 1k resistor between them to get AC coupling down to 20Hz, solder that rig to the PS and connect the APU after power On. 3.3-5V PS for some ADC or DAC's reference is the most critical case, and it is safe with a bare APU.
I would take simply 2 aluminum NP caps 22u/50V and 1k resistor between them to get AC coupling down to 20Hz, solder that rig to the PS and connect the APU after power On. 3.3-5V PS for some ADC or DAC's reference is the most critical case, and it is safe with a bare APU.
- Home
- Vendor's Bazaar
- Cosmos APU a notch+LNA $70 to outperform APx555b for $30000