Pcb is FR4 standard with HASL and lead, the cheapest one option, 1oz on all 4 layers, around $48, ENIG with lead free + gold is much higher in cost, around double the price. Components around $68, my customs clearance + pdv for pcb and components around 30%, and my work = total 250eur for populated pcb. For enig pcb it cost higher than 250EUR, must recalculate proportionaly by cost of enig pcb+customs and pdv! Shipping don't know since never sent anything internationaly, I must check that. My country is not inside European Union and not iside that VAT system, I must pay around %20 for VAT and around +10% for customs, total 250EUR which I have calculated is around price!
Last edited:
Ok, I will prepare my budget for few pcb's and components soon, month-two it will be ready! I'm started revising our mini concept, idea is placing 3 alu caces on top of each other, one case will contain DDPD, seccond one DDPD frontend, and 3rd one power supply digitaly controled, with oled display for volume control level display. DDPD frontend will contain USB input, two optic inputs, bluetooth, fm radio, oled too. Remote control & opt encoder, or maybe buttons insead of opt encoder. It will be not that soon but its in my plan at least I want to have something like that in my home : )
Black front
Black front
Last edited:
What need to be tried next on our DDPD, I didn't measured DDPD but I'm believing it need an better LC filter, for example minimum seccond order filter with first pole at 50khz and seccond pole at 25khz, thats a first thing to try. One of the hard things to implement is post filter digital feedback, that reguire something very fast to be able to change dsd input bits on the fly by the error at post filter feedback. Hmm how that can be done? I'm believing an simple and ultra fast 2bit window comparator which need to be ultra fast and ultra simple, to contain two capacitors which will store sine wave voltage from the post LC as a ref value. So on every sample (which need to be synchronised with dsd clock) every next sample is a ref. Ref is compared with next sample. So we have 2 bits, one bit is ref and seccond bit is next sample. Comparator need to watch sine wave at post LC filter and look for voltage level, no need to know how there volts is, just simple function e.g. if voltage goes high than ref value is 1 , seccond bit become high or vice versa if voltage on next sample going high or low. So high or low bit is compared with dsd input bit and in case of bit not match with dsd input bit than dsd input bit get changed. Something like that and we have error correction. Hmmm this is just my idea but how this can be done I have no idea. : ) Things might be done with an ADC at post filter but its even not easy to implement, reguire an PGA and things to get the same level as dsd input and its very hard to implement, even ultra fast mcu is need to process all that. I'm still believing an simple analog thing is reguired here as like an simple window comparator and few more 74ahc at input to change bit on the fly.
Last edited:
"didn't measure", "believing", "try" and "no idea" are red flags. Nothing against your project (on the contrary!) but publishing may be better after actual building and measuring. Also concentrate on one project and not on two in the very same thread. This to avoid stress and lack of time for you and disappointment at the readers side.
Apart from that it is always OK to see new projects!
Apart from that it is always OK to see new projects!
Ok I understand. The projects are connected to each other, they use the same frontend. But this is in development stage and I told why I didn't done measurement, its in plan, curenly stale since pcb bug, but plan is to do it till after frontend reguired for measurement get ready. I'm just stating what's on my mind because maybe someone else will come forward with a good idea, if I don't mention what's on my mind, maybe someone else won't have an idea either, I think it's ok? Every idea is welcome.
Ok thanks, the initial project is finished and it is in the first post, everything else, every improvement, idea, post, is welcome! I mentioned post LC filter and also post filter feedback, that's next! Idea for post LC filter feedback is very welcome! Measurement is curently hard to do without digital frontend as everything here is digital, for example I can't use Rew to do audio loop since Rew doing PCM but we have here DSD, impossible to do without DSD frontend right now. Even idea for this is welcome!
I could recommend a Rigol MSO they have a built in I2S Decoder and some other features suitable for audio. Alternatively there are some Digilent Analog Discoverys out, since the released the V3 now.
Prices increased again, as I just have seen. Sometimes they have an offer with discount. I bought my 5074 at 795,- Euro + TAX last summer and had almost all SW options included! AFAIK it is possible to "open" the 5074 up to 53xx beeing capable of handling up to 350 MHz - but have not tried yet because of warranty. (all these HW and SW options are allready included and just enabled by SW codes).
Well I had expected that the prices for this series will go down, because Rigol recently announced their new series with better DACs. ..
Well I had expected that the prices for this series will go down, because Rigol recently announced their new series with better DACs. ..
The first preliminary measurement result from my colegue, I will do further measurements when ddpd frontend get finished, I will put aux0025 filter with atenuator between ddpd and Cosmos ADC for better result, this preliminary measurements is not that bad at all! : ) The problem here is pin headers connection, next DDPD revision will contain u-fl connection between DDPD and frontend since pin headers is like antena and it definitelly is not good for high frequency things, connection must be coaxial for better result! Even this looks very good, it promising! Minimum volume voltage must be allso revided, minimum voltage must be not < 1.65V as 74 registers regure that minimum. Also 1000uF condensers (x8) I will remove as it cause voltage regulator CCV not to function since acumulated energy inside those 8 capacitors give 6A but we need CCV limit 1A, it will be replaced with 64uF, also tantal capacitors will get replaced with ceramic capacitors because I need DDPD pcb very small to be able to use enclosure from post 25. I'm started revising DDPD! It will probably get digital supply in preparation to things from post 26, I will let you know the progress!
The result from my colegue:
"Measurements with Cosmos ADC and suggested settings, sine reproduced from HQPlayer and with two levels from the player, with 0dBFS and -18dBFS, the third recording is with short inputs at the end of the measuring cable. Cosmos is set to 1.7V input level, DDPD output voltage to minimum 1.24V"
The result from my colegue:
"Measurements with Cosmos ADC and suggested settings, sine reproduced from HQPlayer and with two levels from the player, with 0dBFS and -18dBFS, the third recording is with short inputs at the end of the measuring cable. Cosmos is set to 1.7V input level, DDPD output voltage to minimum 1.24V"
Attachments
Last edited:
Hi Savan,
Thanks for posting your preliminary results.
As to distortion I sort of was expecting this distortion level: I have been trying out quite some logic ICs in the last couple of years and in my experience none of them "like" to deliver current if the distortion is also to be kept at a low level. E.g. the Potato Semi ICs reached a distortion level of ~- 70 dB when delivering appr. 1 mA. With only slight variation I have had similar results for all the other ICs I have tried (NC7SZ, 74AC, 74AUP) ...
Just FYI if you are interested.
Cheers, Jesper
Thanks for posting your preliminary results.
As to distortion I sort of was expecting this distortion level: I have been trying out quite some logic ICs in the last couple of years and in my experience none of them "like" to deliver current if the distortion is also to be kept at a low level. E.g. the Potato Semi ICs reached a distortion level of ~- 70 dB when delivering appr. 1 mA. With only slight variation I have had similar results for all the other ICs I have tried (NC7SZ, 74AC, 74AUP) ...
Just FYI if you are interested.
Cheers, Jesper
My idea for post-filter feedback = an ultra-fast sample-and-hold window comparator, on the post filter, capable of processing the sample in <150ns (so we still have that much time left for post-processing on the dsd input gates while the sample is still running at 2.8MHz) and contains a sample & hold, where on the first sample it should read and hold the voltage value on the post filter in its hold register, and on the next sample compare the new value with the "hold" value and we simply don't have to to know how many volts there are on the post filter, but enough to know if the sine amplitude on the post filter decreases or jumps, that means if the amplitude jumps the output from the comparator would send HIGH pulse to the dsd input gate, or vice versa, if the amplitude decreases, the output from the comparator would send LOW pulse to the dsd input gate, at the dsd input it would have an additional gate that would convert high/low to low/high and vice versa and that would be an error correction technique? The FB circuit should be synchronized with the dsd clock. Is there such a fast window compator with sample and hold?
The first idea was in the direction of setting ADC on the post filter, but that is not a solution because at the same time it is necessary to bring ADC to the level of the input DSD in order to possibly compare the input and post filter signals, which means that a PGA would have to be added, as well as an adequate MCU , the whole thing is very complicated and seems impossible to do. So I'm looking for something simple. Any idea is welcome!
The first idea was in the direction of setting ADC on the post filter, but that is not a solution because at the same time it is necessary to bring ADC to the level of the input DSD in order to possibly compare the input and post filter signals, which means that a PGA would have to be added, as well as an adequate MCU , the whole thing is very complicated and seems impossible to do. So I'm looking for something simple. Any idea is welcome!
Last edited:
Some progress on revided DDPD, idea is from post 26, so in this stage 1 (first blue alu) must be room for oled display, mcu, 3 buttons, infra red, back side must contain 4 sma_to_ufl conectors (dsd input), dc socket and four banana chasis connectors for speakers, my hope everything will fit in the box : ) Whole pcb is a lot smaller than initial one. Infra red is for remote control. Oled will display volume level. One button for power and two buttons for volume control. Volume control is digitaly controled buck regulator, from 1.25V to 5V. Next is designing stage 2, digital frontend. Stage 3 no need, so only two alu enclosures instead of 3 as per post 26 : )
Attachments
Last edited:
I tried to register at Forum Primenjene Elektronike but the new registrations are currently disabled... bummer.
Small suggestion.... I see that you are planning to use USB Vbus -> U40 -> and then use that to power up the U7 and U1 and half of U38. I'd suggest that you at least consider a separate regulator for at least each of these ICs... and even better, a separate regulator for each Vbus pin.
I have no doubt that the end result will be extremely transparent, open, spacious... invigorating.
I applaud the initiative and I wish you the very best of luck.
Zemunac