TioFrancotirador what type of Black Gates you want to use,the standard ?
2200uF 16V Black Gate Standard Type | Hifi Collective
2200uF 16V Black Gate Standard Type | Hifi Collective
Thank you Salas
The amp had no output coil, it is the L12-2 kit bought from ebay. It is discussed in a thread here and I received some suggestions. The speakers survived just fine, the outputs got shorted after the output transistors burned.
I will check all my soldering on the board soon, I have spare time now but I'll prepare the case for the heatsinks and regulators first.
For those looking for good cap in C2 position I found this:
High End Audio - Electrolytic capacitors
After reading it I am in doubt investing in BG.
Anyways, I was thinking where good C2 matters in terms of sound qality. I doubt it matters for powering DVDDs or clocks (however did not try) and I am wonder if it matters for AVDD in current outputs chips, where current is converted to voltage in further steps anyway. The chip I use is ak4495 (diff voltage output), seems like in this case it is also "first amplification" stage as well, which ... matters most in terms of sound quality, do not know, just thoughts. Maybe that is why Ref-D parts impacts sound quality in such application.
So my conclusion would be this:
For powering clocks and DVDDs user polymer cap in place of C2.
For powering AVDDs use Elna Silmic II/polymer (depending on taste) and definitely Fet transistor mod in R6 position.
High End Audio - Electrolytic capacitors
After reading it I am in doubt investing in BG.
Anyways, I was thinking where good C2 matters in terms of sound qality. I doubt it matters for powering DVDDs or clocks (however did not try) and I am wonder if it matters for AVDD in current outputs chips, where current is converted to voltage in further steps anyway. The chip I use is ak4495 (diff voltage output), seems like in this case it is also "first amplification" stage as well, which ... matters most in terms of sound quality, do not know, just thoughts. Maybe that is why Ref-D parts impacts sound quality in such application.
So my conclusion would be this:
For powering clocks and DVDDs user polymer cap in place of C2.
For powering AVDDs use Elna Silmic II/polymer (depending on taste) and definitely Fet transistor mod in R6 position.
Hi Salas,
I have question concerning 30VA transformer recommendation for Ref-D.
Is its power related to CSS and Max Vout. Meaing 30VA is rated for max CSS=600mA runing at Vmax=7V?
If I use CSS=200mA @ 7V and CSS=400mA @ 5V could I go bit lower with transformer power?
I know I can go with power as high as possible, but is there any reasonable power ratio formula?
I have question concerning 30VA transformer recommendation for Ref-D.
Is its power related to CSS and Max Vout. Meaing 30VA is rated for max CSS=600mA runing at Vmax=7V?
If I use CSS=200mA @ 7V and CSS=400mA @ 5V could I go bit lower with transformer power?
I know I can go with power as high as possible, but is there any reasonable power ratio formula?
Its related to Tx power consumption and very generous overrating. If it delivers 10VAC to a 600mA reg then its 6W constantly. So I rate at 3x =18W and then almost double because 30VA is a commonly available size in good makers of toroid & R-Core not that bigger to fit. So the impedance is lower, never hot, never sag.
In most cases you could use 15W Tx with comfort.
In most cases you could use 15W Tx with comfort.
Thanks!
What about AC voltage for Ref-D@7V. Manual says stick to 9VAC in any case. On the other hand there is recommendation that Vin should be 3-5V higher than Vout. So could I use e.g: 10VAC for Ref-D@7V then? Would it bring any improvement here? or maybe some ref-d parts values would have to be recalculated in such case ... Voltage after rectifcation will be higher anyway, but since I am going to order custom transformer of particular specs I would like to have it theortically most accurate/appriopriate.
9VAC meets the 3-5V Vi>Vo recommendation in the 7VDC Vo case too. Because it will bring 11VDC after rectification (when normal Vf Silicon diodes are used). 10VAC is still within that range.
That blanket recommendation was to can go up to 600mA CCS with the on board IRF9610's sink not much saturated. If you will never fix more than 400mA CCS for instance, the recommendation can be revised up to 12VAC for the 7VDC Vo setting. That will allow more VDS across the 9610 to where its parasitic capacitance figures are little better than with 5VDC Vi-Vo difference. Mainly about its output capacitance.
That blanket recommendation was to can go up to 600mA CCS with the on board IRF9610's sink not much saturated. If you will never fix more than 400mA CCS for instance, the recommendation can be revised up to 12VAC for the 7VDC Vo setting. That will allow more VDS across the 9610 to where its parasitic capacitance figures are little better than with 5VDC Vi-Vo difference. Mainly about its output capacitance.
Ok, so basically idea is to keep IRF9160 output capacitance at possibly low levels and not to exceed saturation levels. Which means that if do not exceed certain CSS levels at certain voltages I could go bit higher with AVC. So also for Ref-D 5V@400mA I could use around 10,5VAC I believe.
Hi guys,
I need your help (advaise).
I must admit that I know little about power design hence my question below.
I need a clean power for two DAC’s external clocks (DC +5V).
Which option would be better from SQ point of view:
1. Using only Reflektor D (set as 5V) or
2. Using reflector D (set as 6.4V) and very lov noise LDO e.g. ADM7151 (best I found) set as 5V
Would it be any sound benefit to combine very low noise LDO and Reflektor D?
TIA
I need your help (advaise).
I must admit that I know little about power design hence my question below.
I need a clean power for two DAC’s external clocks (DC +5V).
Which option would be better from SQ point of view:
1. Using only Reflektor D (set as 5V) or
2. Using reflector D (set as 6.4V) and very lov noise LDO e.g. ADM7151 (best I found) set as 5V
Would it be any sound benefit to combine very low noise LDO and Reflektor D?
TIA
VCXO clocks are sensitive to regulator quality. Differences between regs are easily audiobile. Clocks and AVDD are the most critical for good sq. I found best results powering them directelly with salas ref-d without local regs even shunts!
How much AC noise is expected on the output of the unit?
I am seeing considerable noise, up to 40 mV pk-pk at ~100 Mhz with my Rigol DSO. I believe most of this is environmental EMI, but when I probe with my Fluke multimeter, AC mode, I intermittently see 20 mV RMS being measured.
Any thoughts? Should I be concerned?
Note this is with the test resistor connected to the output terminals.
DSO captures of output terminals - AC coupled into DSO:
Edit: I have built two units, and they both exhibit the same output waveforms. So I wouldn't usually be concerned and would put it down to environmental EMI but as I said, multimeter is also measuring some AC...
I am seeing considerable noise, up to 40 mV pk-pk at ~100 Mhz with my Rigol DSO. I believe most of this is environmental EMI, but when I probe with my Fluke multimeter, AC mode, I intermittently see 20 mV RMS being measured.
Any thoughts? Should I be concerned?
Note this is with the test resistor connected to the output terminals.
DSO captures of output terminals - AC coupled into DSO:
Edit: I have built two units, and they both exhibit the same output waveforms. So I wouldn't usually be concerned and would put it down to environmental EMI but as I said, multimeter is also measuring some AC...
Last edited:
I have a year old oscilloscope picture vs a Texas evaluation board (TPS7A47 series). I hope this will help you. Light blue is the Reflektor, yellow is the Texas. Common DC input from a Korad cheap lab supply. To be same source since the evaluation board was DC input only. Although the Ref-D with its own bridge and filter cap it was showing the same.
And the close up screenshot. Vertical resolution 1mV. Always using the bandwidth limit scope's selection (B in bold) when measuring PSUs as the lab units manufacturers are doing.
It could be affected by the probes placement showing max the same thickness between them, bit thicker than those pic's yellow trace in worst case. I could not use the springy tiny loop area thingy that fits on a probe's nose because the connectors are wider. So I think its basically down to the measurement loop and scope's residual. They both certainly look comparably low wideband noise reg boards (20MHZ). Your measurement loop maybe picks up something noisy from the environment? Or an SMPS pollutes mains? Try with all other stuff like lights and non used gear powered down, various power adapters pulled out, or measure again in a different room.
And the close up screenshot. Vertical resolution 1mV. Always using the bandwidth limit scope's selection (B in bold) when measuring PSUs as the lab units manufacturers are doing.
It could be affected by the probes placement showing max the same thickness between them, bit thicker than those pic's yellow trace in worst case. I could not use the springy tiny loop area thingy that fits on a probe's nose because the connectors are wider. So I think its basically down to the measurement loop and scope's residual. They both certainly look comparably low wideband noise reg boards (20MHZ). Your measurement loop maybe picks up something noisy from the environment? Or an SMPS pollutes mains? Try with all other stuff like lights and non used gear powered down, various power adapters pulled out, or measure again in a different room.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.