Latest DAC/AMP combo from Soekris at the recent RMAF show CANJAM AT RMAF 2015 – BRIAN MURPHY | Headphone.Guru
An externally hosted image should be here but it was not working when we last tested it.
The interesting take away is that the DAC/AMP board supports DSD.
The standalone DAC board on the side also looks interesting and different from the DAM, it looks smaller and the connector laayout is also different.
The is a mention of a DAM2021 on the soekris.eu site noting concrete, maybe a typo
Looks like VRef has been significantly changed on both boards
I just bought rev.2.
Look at all the vref caps on the dam1121 (right board), feedback well taken from the looks. Curious if the smaller package shift registers have any differing impedance characteristics... And those blue resistors, what type would they be? No doubt high-precision. Too bad its OEM-only.
You're welcome, Soren.
--
The Community.
You're welcome, Soren.
--
The Community.
Have someone tried selecting the I2S MCLK OUT frequency with the FSEL IN pin?
With no Inputsource the dam outputs 24.5764MHz. Switching FSEL pin to GND should change the Frequency but it didn´t.
Firmware is 0.99
Found an answer at post 1119:
So I assume its not supported in the current firmware release.
Soekris, when will this be enabled?
Most interesting. I was thinking of ordering the DAM1021. Now need more info on the Dac1101. Good that it will be out in Nov2015. Need to wait some time to hear user impression.
I can also verify that it does NOT work with 0.99 firmware. Connecting a beaglebone black to the iso i2s interface with the botic driver works fine with 48KHz and relatives but on 44KHz and its relatives play with a higher pitch. Quite frustrating that although not yet working its not mentioned in the manual.
Is there a filter issue?
Hi Soekris,
I´ve read at moredamfilters:
NOTE: There is currently a bug in the firmware that prevents IIR filters loading correctly. Until this is fixed only the first IIR filter appearing in the .skr file will load.
So you can do basic testing of single filters but it's very difficult to do anything meaningful with 1 * 2nd order filter.
Is this correct and when it will fixed?
Christian
Hi Soekris,
I´ve read at moredamfilters:
NOTE: There is currently a bug in the firmware that prevents IIR filters loading correctly. Until this is fixed only the first IIR filter appearing in the .skr file will load.
So you can do basic testing of single filters but it's very difficult to do anything meaningful with 1 * 2nd order filter.
Is this correct and when it will fixed?
Christian
The Vref mods try to improve the load transient response of the DAM ladder supply. So it might be interesting to know the current draw of the DAM ladder.
Here a table of the current drawn (in mA) if bit i (of the positive ladder) is set, i.e. if the output voltage is 2^(-i) volts (DC).
For any sum of that voltages the total current draw is the sum of the corresponding currents.
One property to be observed is that, except for the 3-4 most MSB (an effect of the "thermometer bit section" of the ladder), the current draw is constant ~0.26mA independent of the bit position.
So the current draw is in local scale not proportional to the output voltage. It is more correlated to the number of set bits.
Here a graph current draw over output voltage for 20 bits only, due to the amount of computing (i.e. the most LSB sift register bits are all zero).
The step pattern arrises from the most MSB bits.
The heavy oscillation is a result of that with increasing voltage the LSB, and so locally also the number of bits set, change fast. With the full 27 bits of resolution the oscilation would be even stronger resulting in a about 2mA higher draw.
The current draw for a positive half of a FS sine wave would thus be a coarse stepped approximation of a sine wave with a heavy HF noise (HF = normally ~3MHz, or with NOS: f_sample).
The load of a sine wave consists of the above followed by zero load for the same time (when the other ladder is busy).
Here a table of the current drawn (in mA) if bit i (of the positive ladder) is set, i.e. if the output voltage is 2^(-i) volts (DC).
Code:
i : current
0 : 1.200000000000000000000000000000 mA
1 : 0.700000000000000000000000000000 mA
2 : 0.375000000000000000000000000000 mA
3 : 0.293750000000000000000000000000 mA
4 : 0.273437500000000000000000000000 mA
5 : 0.268359375000000000000000000000 mA
6 : 0.267089843750000000000000000000 mA
7 : 0.266772460937500000000000000000 mA
8 : 0.266693115234375000000000000000 mA
9 : 0.266673278808593750000000000000 mA
10 : 0.266668319702148437500000000000 mA
11 : 0.266667079925537109375000000000 mA
12 : 0.266666769981384277343750000000 mA
13 : 0.266666692495346069335937500000 mA
14 : 0.266666673123836517333984375000 mA
15 : 0.266666668280959129333496093750 mA
16 : 0.266666667070239782333374023438 mA
17 : 0.266666666767559945583343505860 mA
18 : 0.266666666691889986395835876465 mA
19 : 0.266666666672972496598958969116 mA
20 : 0.266666666668243124149739742279 mA
21 : 0.266666666667060781037434935570 mA
22 : 0.266666666666765195259358733892 mA
23 : 0.266666666666691298814839683473 mA
24 : 0.266666666666672824703709920868 mA
25 : 0.266666666666668206175927480217 mA
26 : 0.266666666666667051543981870054 mAOne property to be observed is that, except for the 3-4 most MSB (an effect of the "thermometer bit section" of the ladder), the current draw is constant ~0.26mA independent of the bit position.
So the current draw is in local scale not proportional to the output voltage. It is more correlated to the number of set bits.
Here a graph current draw over output voltage for 20 bits only, due to the amount of computing (i.e. the most LSB sift register bits are all zero).
The step pattern arrises from the most MSB bits.
The heavy oscillation is a result of that with increasing voltage the LSB, and so locally also the number of bits set, change fast. With the full 27 bits of resolution the oscilation would be even stronger resulting in a about 2mA higher draw.
The current draw for a positive half of a FS sine wave would thus be a coarse stepped approximation of a sine wave with a heavy HF noise (HF = normally ~3MHz, or with NOS: f_sample).
The load of a sine wave consists of the above followed by zero load for the same time (when the other ladder is busy).
Last edited:
I made an error in the last post. For the total load current for an arbitrary voltage the currents need to be added on vectorial level (and not the (correct) currents for one bit set) and then the currents corresponding to the bits added to the final value. The correct current draw/output voltage graph is (now hopefully ) this:
The main observation that the current is locally more correlated to the number of bits set, remains valid.
The main observation that the current is locally more correlated to the number of bits set, remains valid.
Problems....
I got my new rev 2 DAM,and connected everything and have been playing for a couple of hours and changing some filters too,so that I know everything is working as it should.
Then I thought of trying the Crap filter,but when I have loaded the filter and restarted the DAC it writes the usual numbers on putty that it does when it starts,but it keeps writing:
LO44
LO44
LO44
LO44 and so on and the Led keps blinking what is the fix for this??
I got my new rev 2 DAM,and connected everything and have been playing for a couple of hours and changing some filters too,so that I know everything is working as it should.
Then I thought of trying the Crap filter,but when I have loaded the filter and restarted the DAC it writes the usual numbers on putty that it does when it starts,but it keeps writing:
LO44
LO44
LO44
LO44 and so on and the Led keps blinking what is the fix for this??
