A two cents question as usual please:
My understanding is Hifiduino blog shows how to feed the PS via J2 to avoid the embeded rectifier bridge and inject DC after this rectifier bridge to feed the embeded caps.
Despite the excellent noise rejection of the embeded LDO reg before the active components, could it change something to feed with a fast transcient regulator 10 to 12 V DC via J2 ? To rephrase it : do those embeded ldo reg are good enough in transcient ? I think of chip reg like the TSA74xxx ! Does the DC to DC converter a transcient eater ????
to rephrase, I would like to understand if the subjective lacke of dynamic and mid- bass impact (percussions) at home is a result of the embeded PS or the result of the EQ (flat output line between 20 hz to 20 K hz) or simply the actual digital filter ?
Another question please :Soren have been giving the Tools on the original thread and a srk file about filters : does this one come from sOX website and have to be uploaded at it is in the dam1021 (so is it a first upsampling revision?) ? Btw : is it possible for an experienced diyer of digital filters to play with the EQ text to be loaded to permitt so subjective EQ in FIR1 curve to avoid a flat EQ ?
thanks,
sorry in advance for my poor understanding !
My understanding is Hifiduino blog shows how to feed the PS via J2 to avoid the embeded rectifier bridge and inject DC after this rectifier bridge to feed the embeded caps.
Despite the excellent noise rejection of the embeded LDO reg before the active components, could it change something to feed with a fast transcient regulator 10 to 12 V DC via J2 ? To rephrase it : do those embeded ldo reg are good enough in transcient ? I think of chip reg like the TSA74xxx ! Does the DC to DC converter a transcient eater ????
to rephrase, I would like to understand if the subjective lacke of dynamic and mid- bass impact (percussions) at home is a result of the embeded PS or the result of the EQ (flat output line between 20 hz to 20 K hz) or simply the actual digital filter ?
Another question please :Soren have been giving the Tools on the original thread and a srk file about filters : does this one come from sOX website and have to be uploaded at it is in the dam1021 (so is it a first upsampling revision?) ? Btw : is it possible for an experienced diyer of digital filters to play with the EQ text to be loaded to permitt so subjective EQ in FIR1 curve to avoid a flat EQ ?
thanks,
sorry in advance for my poor understanding !
Last edited:
It is not that strange.
The multicomp transformers linked in the first post have a typical regulation of 18% for the 5VA and 19% for 7VA. The regulation % indicates the voltage rise off load. So for a 7VAC, 5VA Multicomp the unloaded voltage will typically rise by 1.26V giving an unloaded output of 8.26VA.
For 7.5VAC to rise to 9.5VA unloaded indicates the regulation % of the transformer is around 27%.
Looking at the other way - if you have the regulation % spec you can determine what the off load voltage will be.
Interestingly the regulation % is related to the source impedance of the transformer. Higher regulation % = higher source impedance.
So the "heavier rated" a transformer is the lower the percentage is (in general).
Thanks Nige2000,
So you 've got an improvement ? How you could describe it ?
Does the pcb around the cells manage a low output impedance as well like a shunt reg can do (have a measurement about the output impedance ?) or is it a voltage serie reg ?
In the offical thread I saw a photograph of a member with a big PS and he seems to have a good listing feeling. The PS would have an influence despite the sota noise rejection managed inside the dac (so not on the noise but something else : transcient ?)
Yes I'm waiting some improvement about the digital filter as unluckily my tastes made my library to be 95 % 16/44 ! But I didn't understand if there are any members who have an experience about that, it seems very difficult to design with a lot of trade offs for the low sample rate as Soren testimonied and some others as well !
Pos the rePhase developer, (the FIR soft you can see in MiniDSP e.g.) fellow would help for some advices and tips about its soft, but not sure it was fo redisign a FIR filtr from scratch !
Would it be simple to try it NOS without FIR1 for 44 or 48 k Hz material ? To have a comparaison point to understand the fundamental difference for the red book materials on this DAC between the sound upsampled by the FGPA vs the Sound NOS.
Could be a starting point with further tests with sox to try upsampled materials on the pc played on the dam1021 NOS ! Like some many does with the best computers players to stream towards multibit NOS dacs (TDA1541, AD1862, etc) or slighty oversampled dacs (x2 only) ! Notice I'm a noob and don't understand nothing about FIR and upsampling vs oversampling... just some basics !
@ Bikerboy,
I have not : checked. Have you some ref of song in the classic Pop/rock or jazz history we can talk about the same reccording please ? PM if you want to talk about that with more precision as it can be long to compare some materials !
It is not the interstellar vaccum between the speakers but more a huge width which give the impression the speakers are far from each others. The vertical soundstage is at the opposite the best I've ever had. When I plug an another dac on the same wires (to use the same cd input into my pre), I have a re fulled center with my other DAC and cd player...So I surmise it's maybe the materials I tested ?! (but know it very well, I use the same protocol since 10 or 15 years iirc)
So you 've got an improvement ? How you could describe it ?
Does the pcb around the cells manage a low output impedance as well like a shunt reg can do (have a measurement about the output impedance ?) or is it a voltage serie reg ?
In the offical thread I saw a photograph of a member with a big PS and he seems to have a good listing feeling. The PS would have an influence despite the sota noise rejection managed inside the dac (so not on the noise but something else : transcient ?)
Yes I'm waiting some improvement about the digital filter as unluckily my tastes made my library to be 95 % 16/44 ! But I didn't understand if there are any members who have an experience about that, it seems very difficult to design with a lot of trade offs for the low sample rate as Soren testimonied and some others as well !
Pos the rePhase developer, (the FIR soft you can see in MiniDSP e.g.) fellow would help for some advices and tips about its soft, but not sure it was fo redisign a FIR filtr from scratch !
Would it be simple to try it NOS without FIR1 for 44 or 48 k Hz material ? To have a comparaison point to understand the fundamental difference for the red book materials on this DAC between the sound upsampled by the FGPA vs the Sound NOS.
Could be a starting point with further tests with sox to try upsampled materials on the pc played on the dam1021 NOS ! Like some many does with the best computers players to stream towards multibit NOS dacs (TDA1541, AD1862, etc) or slighty oversampled dacs (x2 only) ! Notice I'm a noob and don't understand nothing about FIR and upsampling vs oversampling... just some basics !
@ Bikerboy,
I have not : checked. Have you some ref of song in the classic Pop/rock or jazz history we can talk about the same reccording please ? PM if you want to talk about that with more precision as it can be long to compare some materials !
It is not the interstellar vaccum between the speakers but more a huge width which give the impression the speakers are far from each others. The vertical soundstage is at the opposite the best I've ever had. When I plug an another dac on the same wires (to use the same cd input into my pre), I have a re fulled center with my other DAC and cd player...So I surmise it's maybe the materials I tested ?! (but know it very well, I use the same protocol since 10 or 15 years iirc)
Last edited:
True. A 15VA trafo has lower impedance than a 5VA trafo, but when comparing the same VA rating lower regulation % = lower impedance.
PSUD II allows you to calculate the source impedance and off load voltage. Click edit on the transformer, then the "..." beside Vrms.
clarity, more fluid/free flowing, more ease
44.1 still not fantastic
direct from cell is good enough for testing
Yes I could, but there is nothing hard in the wirering : on the non bufered SE the gnd are the two center vias and the two external are the one who go to the hot pin of the RCAs. For the bufered SE I used the shematic correction of Hifiduino ! And mainly I drink any wine today !
Could it be something else ? I checked the external speaker wires (hope I have not one of my internal passive filter out of phase ?)... but mainly it deseappers when i swap the DAC ! It coud be also psycho : not the habits to have huge width !
Bad PS traffo ? Problem on the DAC itself (I find the assembly quality very high ! )
If you really want i take a photograph of the output I can but the result is quite similar on the two output : buffered and not buffered ! So ultimatly it could be an non detected problem of phase onto the speakers : 3 ways... LR4 tweeter filter out of phase with the LR2 medium on the both speaker or just in one ?
Hummmm when two people says I'm wrong I'm certainly wrong here, I saw no reason why my speaker would gave a different result than yours guys !
@ Nige, yes they are good enough... and better enougfh if they are alkaline cells as the impedance is very low with this technology which doesn't need for test a shunt reg to lower the output impedance (my simple understanding of the TNT site reviews about regulators and output impedance measurement).
It's a good sign than your test give better result, it says than there is a reserv of possible improvement between the external layout and the digital filter... good and positiv sign.
@ Paul : I like a lot what you wrote about an output traffo : could add it some harmonics in the highs ?
It seems urgent to wait first an (or some) improved digital before any other harder tweaks (output buffer, PS, etc !).
Last edited:
The fail is on the user side i.e. for those who evaluate active components, from a specification that starts at 12khz, for audio.
//
Agreed, and the transformer spec is probably for a nominal 230VAC input so if your mains voltage is above that you can add a bit more on; at 240VAC you would get an unloaded output of 8.62VAC.
Ray
2 days of run in! Back to the sdtrans384 via i2s. Sounds great. Unbuffered se output. If the buffalo III with Salas ps paints in a well lit room, the DAM 1021 makes images with laser and a spotlight. Red book ripped to flac, then converted to wave for the sd card player. The Roots-Walk Alone is all new. A little bright, maybe need sunglasses! The bass is endless. No dropouts like the buffalo.
as the sd trans is known to be one of the less jittered apliance it could be fun you try a spidf input and compare if you hear a difference when this FIFO is working with less jitter to close the debate : time capacity of buffer seems to make a difference (or not as would say some others ! Is it just a marketing argument than the time capacity of buffers for FIFO ?) : proof of concept about the soundstage and spidf for a stupid french !
could you give me again please an example of a song you use for your test ?
could you give me again please an example of a song you use for your test ?
Last edited:
Amanero input board
Here is a draft of an updated input board for Amanero users. It goes directly onto the J3 header of the DAC. Amanero is plugged in below the input board same as the DAC. BNC and TOSLINK connectors are also mounted on the bottom. This arrangement was chosen to fit the whole setup into a 1U case.
No RCA connector is planned for SPDIF. If you need to connect an RCA source, use a BNC-RCA adapter such as this one (Farnell p/n 3706722). Or better yet, use an RG-59 BNC cable and an RCA adapter such as Farnell p/n 1169554 on the other end.
The board includes optional indicator LEDs for Amanero, and also an optional 3.3V regulator for TOSLINK and front panel connector. If internal 3.3V line from the DAC is used, the regulator must no be installed. And vice versa, if external supply is used for 3.3V, J6 jumper must not be connected.
The transformer can be mounted either on top or on bottom. All SMD parts are 0805 size, the regulator is SOT223 (same size as 3.3V regulator on the dac board).
Does anyone know where to find Eagle CAD library for U.FL connectors?
Here is a draft of an updated input board for Amanero users. It goes directly onto the J3 header of the DAC. Amanero is plugged in below the input board same as the DAC. BNC and TOSLINK connectors are also mounted on the bottom. This arrangement was chosen to fit the whole setup into a 1U case.
No RCA connector is planned for SPDIF. If you need to connect an RCA source, use a BNC-RCA adapter such as this one (Farnell p/n 3706722). Or better yet, use an RG-59 BNC cable and an RCA adapter such as Farnell p/n 1169554 on the other end.
The board includes optional indicator LEDs for Amanero, and also an optional 3.3V regulator for TOSLINK and front panel connector. If internal 3.3V line from the DAC is used, the regulator must no be installed. And vice versa, if external supply is used for 3.3V, J6 jumper must not be connected.
The transformer can be mounted either on top or on bottom. All SMD parts are 0805 size, the regulator is SOT223 (same size as 3.3V regulator on the dac board).
Does anyone know where to find Eagle CAD library for U.FL connectors?
Attachments
A couple of suggestion:
The dam1021 LED output do not have a current limiting resistor, might be a good idea to add it on that board.
Another good idea is to add two schottky diodes (or one double), to convert the pins from a standard 1x4 switch to the binary format needed by the dam1021.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.