Measurements of the Linn Utopik power supply
Posted: 2024-01-03 12:08
Linn has apparently lowered the switching frequency a lot compared to the former dynamic power supply.
Read more here:
https://www.aktives-hoeren.de/viewtopic ... &start=210
”This is a ripple of 34mV for the dynamic and 20mV for the Utopik. Well, not exactly the finest DC voltage imaginable, but you keep in mind that there are a lot of filters and regulators behind it, that'll be fine. But something completely different only opens up at second glance: at the bottom left. At what frequency does the fidgeting occur? And lo and behold, the dynamic range is 831kHz. This is right in the middle of the medium wave band. That's a lot of HF, and you can create as many filters as you want, the relatively high frequency is transported everywhere through the air. Also in the circuits behind the filters, where you can find the HF again. And now let's look at the value for the Utopik: 33.9kHz. This is just above audible and is in the ultrasonic range. Nothing is being sent yet.
So, the main difference between dynamic and utopian lies in the frequencies used. The joke with a switching power supply is that you don't need such huge transformers as with the 50Hz power supply and therefore you rectify the voltage first. Then it is transmitted via the transformer at a higher frequency in an electronic inverter. The higher the frequency, the smaller the transformer required. But if you go towards MHz, like with dynamics, you have built a juicy HF transmitter at the same time. With Utopik the specification for the frequency was obviously as high as necessary and as low as possible.
Now I want to put the Utopik in my G-ADS3 and I'm curious to see how it compares to the dynamic. I don't have a direct comparison, but I know the G-ADS3 Dynamic vs. G-ADS3 G-LNT comparison well from the Selekt Edition Hub tests. I immediately notice whether the difference is larger, smaller or roughly the same. Although I hardly believe that things will get worse with utopianism ”
————
”So the Utopik is screwed into the G-ADS3. First boot test on the laboratory bench: The G-ADS3 insulted goes into error mode. There are four small green LEDs south of the processor and normally the boot sequence goes like this: first three light up and shortly afterwards the fourth comes on, namely the third from the left. It does, but immediately afterwards the second one from the left goes off again and the display remains dark. What's going on now? Undervoltage.
The Utopik has a slightly lower voltage at the output than the Dynamic. That's enough, but now the G-LNT makes things a little more difficult: I have to put a diode in along the digital voltage of 6.5V to prevent the G-LNT from flowing back. For this I use a so-called Schottky diode, which only has 0.3V forward voltage instead of the usual 0.7V for silicon. But in this case that is already too much. I find a diode in the almost infinite number of my drawers that only needs 0.15V, and now it boots. At least when you switch on-off-on. There's something else about the G-ADS3: I have distributed so many voltage-calming electrolytic capacitors with the lowest internal resistance on the circuit board that they pull the voltage down the first moment they are switched on. If you switch on-off-on, they are already a bit pre-charged the second time.
A bit of a bumpy start. But normally there is no G-LNT in it when you retrofit the Utopik. Something else shows up: If you have plugged in both power supplies, you can normally easily do the switching test with the G-LNT; this can be done on the fly, even while music is playing. With the Utopik, this also works from the Utopik to the G-LNT, but the other way around, the Akurate first reboots because the tension has dropped a little too low before the Utopik has fully taken over. Cosmetic defects now in the test, but as I said, if you put the Utopian into it, you actually don't have a G-LNT.
Hearing test. I'm exhausted. This plays almost as well as with my super complex linear power supply. This can score points with a slightly better authority in the bass range, but the main shortcoming of the previous switching power supply ("dynamics") is completely gone: this blurred, somewhat lame and opaque sound in a direct comparison with the G-LNT. That wipes away everything utopian. Cheers!
If the G-ADS3 with dynamics is at 100 (see the Selekt test) and the G-ADS3 with G-LNT is at 200, I would use 180 for that. The arbitrary definition of 100 and 200 was made here.
In principle, you can retrofit the Utopik yourself. If someone wants to retrofit the Utopik in their G-Linn, I would prefer that they send me the device for conversion, then I could take countermeasures if there are any problems when starting up.”
Read more here:
https://www.aktives-hoeren.de/viewtopic ... &start=210
”This is a ripple of 34mV for the dynamic and 20mV for the Utopik. Well, not exactly the finest DC voltage imaginable, but you keep in mind that there are a lot of filters and regulators behind it, that'll be fine. But something completely different only opens up at second glance: at the bottom left. At what frequency does the fidgeting occur? And lo and behold, the dynamic range is 831kHz. This is right in the middle of the medium wave band. That's a lot of HF, and you can create as many filters as you want, the relatively high frequency is transported everywhere through the air. Also in the circuits behind the filters, where you can find the HF again. And now let's look at the value for the Utopik: 33.9kHz. This is just above audible and is in the ultrasonic range. Nothing is being sent yet.
So, the main difference between dynamic and utopian lies in the frequencies used. The joke with a switching power supply is that you don't need such huge transformers as with the 50Hz power supply and therefore you rectify the voltage first. Then it is transmitted via the transformer at a higher frequency in an electronic inverter. The higher the frequency, the smaller the transformer required. But if you go towards MHz, like with dynamics, you have built a juicy HF transmitter at the same time. With Utopik the specification for the frequency was obviously as high as necessary and as low as possible.
Now I want to put the Utopik in my G-ADS3 and I'm curious to see how it compares to the dynamic. I don't have a direct comparison, but I know the G-ADS3 Dynamic vs. G-ADS3 G-LNT comparison well from the Selekt Edition Hub tests. I immediately notice whether the difference is larger, smaller or roughly the same. Although I hardly believe that things will get worse with utopianism ”
————
”So the Utopik is screwed into the G-ADS3. First boot test on the laboratory bench: The G-ADS3 insulted goes into error mode. There are four small green LEDs south of the processor and normally the boot sequence goes like this: first three light up and shortly afterwards the fourth comes on, namely the third from the left. It does, but immediately afterwards the second one from the left goes off again and the display remains dark. What's going on now? Undervoltage.
The Utopik has a slightly lower voltage at the output than the Dynamic. That's enough, but now the G-LNT makes things a little more difficult: I have to put a diode in along the digital voltage of 6.5V to prevent the G-LNT from flowing back. For this I use a so-called Schottky diode, which only has 0.3V forward voltage instead of the usual 0.7V for silicon. But in this case that is already too much. I find a diode in the almost infinite number of my drawers that only needs 0.15V, and now it boots. At least when you switch on-off-on. There's something else about the G-ADS3: I have distributed so many voltage-calming electrolytic capacitors with the lowest internal resistance on the circuit board that they pull the voltage down the first moment they are switched on. If you switch on-off-on, they are already a bit pre-charged the second time.
A bit of a bumpy start. But normally there is no G-LNT in it when you retrofit the Utopik. Something else shows up: If you have plugged in both power supplies, you can normally easily do the switching test with the G-LNT; this can be done on the fly, even while music is playing. With the Utopik, this also works from the Utopik to the G-LNT, but the other way around, the Akurate first reboots because the tension has dropped a little too low before the Utopik has fully taken over. Cosmetic defects now in the test, but as I said, if you put the Utopian into it, you actually don't have a G-LNT.
Hearing test. I'm exhausted. This plays almost as well as with my super complex linear power supply. This can score points with a slightly better authority in the bass range, but the main shortcoming of the previous switching power supply ("dynamics") is completely gone: this blurred, somewhat lame and opaque sound in a direct comparison with the G-LNT. That wipes away everything utopian. Cheers!
If the G-ADS3 with dynamics is at 100 (see the Selekt test) and the G-ADS3 with G-LNT is at 200, I would use 180 for that. The arbitrary definition of 100 and 200 was made here.
In principle, you can retrofit the Utopik yourself. If someone wants to retrofit the Utopik in their G-Linn, I would prefer that they send me the device for conversion, then I could take countermeasures if there are any problems when starting up.”