Recently I got a Roksan Black CD player on my bench, for some modifications/improvements. So, I have found out about a serious fault, either a design fault or a production fault... In my opinion it should be a production fault, which it should definitely be cached by the final tests of these products out of the production lines. But it was not... It seems to me that these CD players are sold out so, as the one I got, and this is not good...
The post DAC circuits on Black CD player it are designed and build to be powered by standard (LM7815/7915) fixed regulators. The I/V stage it should be powered by +/- 15v regulators, as the buffer opamps as well. I appreciate this design as a good one, being able to ensure a high quality output signal. Well, maybe not just the best choice for these fixed regulator devices, but anyway, good enough.
The problem here is that all these regulators mounted on board (see picture), it get on their inputs 14,5v. On the output pin of these regulators I measured +12v and -13v... This is very wrong! All these regulator devices it not regulate anything!
The toroidal transformer used to power these devices it is not properly made to deliver the right voltage or power so that the 15v regulators get a higher voltage on their inputs (at least +/-17v unregulated DC), so to be able to do their regulating function. The consequence of this fault is the all post DAC opamps it are powered by fully unregulated DC.
Here is not about a possible power fault on board, which it may cause an over current, lowering so the voltage delivered by the transformer. The toroidal transformer it is just not made/designed to deliver more voltage on this winding output. Why Roksan have used a such faulty transformer? How it happened that nobody have measured the voltages in this area on the production lines (voltage test points are provided on board)? For sure this fault is not singular, being specific to a singel product. This is a production line or product design fault! I just wonder myself how Roksan it can sell out such devices...
I am thinking to warn here the owners of a such CD player about this fault.
The post DAC circuits on Black CD player it are designed and build to be powered by standard (LM7815/7915) fixed regulators. The I/V stage it should be powered by +/- 15v regulators, as the buffer opamps as well. I appreciate this design as a good one, being able to ensure a high quality output signal. Well, maybe not just the best choice for these fixed regulator devices, but anyway, good enough.
The problem here is that all these regulators mounted on board (see picture), it get on their inputs 14,5v. On the output pin of these regulators I measured +12v and -13v... This is very wrong! All these regulator devices it not regulate anything!
The toroidal transformer used to power these devices it is not properly made to deliver the right voltage or power so that the 15v regulators get a higher voltage on their inputs (at least +/-17v unregulated DC), so to be able to do their regulating function. The consequence of this fault is the all post DAC opamps it are powered by fully unregulated DC.
Here is not about a possible power fault on board, which it may cause an over current, lowering so the voltage delivered by the transformer. The toroidal transformer it is just not made/designed to deliver more voltage on this winding output. Why Roksan have used a such faulty transformer? How it happened that nobody have measured the voltages in this area on the production lines (voltage test points are provided on board)? For sure this fault is not singular, being specific to a singel product. This is a production line or product design fault! I just wonder myself how Roksan it can sell out such devices...
I am thinking to warn here the owners of a such CD player about this fault.
Attachments
I am very sure. Input voltage is in this case, 230v. The device is well functional all point of views. The only issue, which it seems to me, not being specific to only this device, is that the original power toroid transformer it output (on load) 2x11vAC, which it result (on load too) +/-14,5vDC on the inputs of the 7815/7915 regulators (analogue section - post DAC). Please note that the currents needed by the circuits loading these lines, are quite low (into few teens mA). All other power lines voltages of the toroid, it are as it should (the same for the rest of power DC lines).
In this reported case it is either the toroid transformer wrong for this winding, and this it should be cached by the production line testing procedures. Or there is a faulty population of the PCB, with wrong voltage regulators, which also this it should be cached by the device testing procedures...
The fact is that the +/- 15v power lines of the analogue circuitry (post DAC), in Roksan Balck CD player, are first not 15v at all (measured +12v and -13v), and second, these lines it are not regulated, but only filtered by the planted capacitors... All the regulators mounted on board for this section it no do their function, because it are wrong powered.
It could be interesting if another owners of this device may open the box, then doing a simple measurement on the specified test points.
In this reported case it is either the toroid transformer wrong for this winding, and this it should be cached by the production line testing procedures. Or there is a faulty population of the PCB, with wrong voltage regulators, which also this it should be cached by the device testing procedures...
The fact is that the +/- 15v power lines of the analogue circuitry (post DAC), in Roksan Balck CD player, are first not 15v at all (measured +12v and -13v), and second, these lines it are not regulated, but only filtered by the planted capacitors... All the regulators mounted on board for this section it no do their function, because it are wrong powered.
It could be interesting if another owners of this device may open the box, then doing a simple measurement on the specified test points.
Last edited:
Some more details...
Please note that the S1 marked winding on toroid, it is going to the front panel power system.
Question: it is right having a main common power line for both noisy drive motors and its controllers, optical servos, and the DAC chip both digital and analogue sections (all together)?
Please note that the S1 marked winding on toroid, it is going to the front panel power system.
Question: it is right having a main common power line for both noisy drive motors and its controllers, optical servos, and the DAC chip both digital and analogue sections (all together)?
Attachments
Last edited:
Some words about Roksan design of the post DAC analogue section.
The differential outputs of the DAC chip (PCM1798) it goes into differential I/V circuits (NE5534). Then the differential audio signals are converted to unbalanced for RCA outputs (OPA134). Then the unbalanced audio signals goes into an unbalanced to balanced converter, to be outputted to XLR (SSM2142).
Each analogue processing section it have its own 15v regulator (+/-). The +/-15v regulators are powered (input) with 14,5vDC... Measured on TP+/-15v on PCB: +12v, -13v... Filtering caps on input/output of the regulators - 330µF.
Up to you to judge such approach...
The differential outputs of the DAC chip (PCM1798) it goes into differential I/V circuits (NE5534). Then the differential audio signals are converted to unbalanced for RCA outputs (OPA134). Then the unbalanced audio signals goes into an unbalanced to balanced converter, to be outputted to XLR (SSM2142).
Each analogue processing section it have its own 15v regulator (+/-). The +/-15v regulators are powered (input) with 14,5vDC... Measured on TP+/-15v on PCB: +12v, -13v... Filtering caps on input/output of the regulators - 330µF.
Up to you to judge such approach...
Last edited:
Hi,
This is my cd player. I'm a bit surprised (and pissed off) over finding out this. I have Roksan Blak amp and cd player and Roksan Tr5s2 speakers. I really like the amp and the speakers, but always thought there was something off with the cd player. I got first the amp, then I tried the speakers and made a package deal on speakers and cd player almost a year ago.
The norwegian dealer didnt have the cd player, so it was ordered directly from Roksan UK last summer. I had to wait a few weeks for that reasons.
I sent an email to Roksan, looking forward to read their reply....
This is my cd player. I'm a bit surprised (and pissed off) over finding out this. I have Roksan Blak amp and cd player and Roksan Tr5s2 speakers. I really like the amp and the speakers, but always thought there was something off with the cd player. I got first the amp, then I tried the speakers and made a package deal on speakers and cd player almost a year ago.
The norwegian dealer didnt have the cd player, so it was ordered directly from Roksan UK last summer. I had to wait a few weeks for that reasons.
I sent an email to Roksan, looking forward to read their reply....
Hi, difficult to judge but can it be that S2 and S4 are (partly) interchanged or that a thinking error was made in that section?
You can solve this clear design error by adding a 2 x 15V toroid or even better an R-core transformer (if there is space). Or, as you already remarked, by replacing all 7815/7915 for 7810/7910 but make sure you order those from Onsemi (lower noise). There is way more than enough headroom with +/- 10V. I don't think Roksan will respond as this is embarrassing but not dangerous.
* Please note that I wrote 7810/7910.... It is not wise to want to regulate to + and - 12V with 2 x 10 V AC so in this case + and - 14.5V after rectification/filtering... This is only safely possible with 7812/7912 LDO replacements but that will be costly. Always use 3V dropout in calculations with 78xx/79xx regulators and don't forget to take ripple voltage into account as well. If you would use 7812/7912 you would more or less make the same thinking error as Roksan did
We don't know if there are any other implications by lowering the voltage (normally not at all). If you want to play safe the best bet is to add a transformer. The output stage is overcomplicated and can be simplified for balanced use. If the owner does not use the balanced outputs better leave it like it is. This device is a Far East design with quite a few quirks.
Last edited:
Companies that seemingly are European or have European brand names often have stuff produced in the Far East since decades as labour was cheap over there. In time the design level also got better so the design also was done in the Far East and all knowledge was exported. In 2020 you may notice that the knowledge to design and produce is almost completely in the Far East and only management/sales/marketing is in your own country. This is not exactly a secret. Some brands were bought by Far East companies and the only thing British, German etc. is the brand name and nothing more than that.
I don't think such a design error would be made by a competent design team. This error was done in design, it was not seen in testing and it has passed quality control. Something unlikely in this ISO day and age. Many a DIYer would don't even make such a mistake
BTW the exact name of this device is Roksan blak .......
I don't think such a design error would be made by a competent design team. This error was done in design, it was not seen in testing and it has passed quality control. Something unlikely in this ISO day and age. Many a DIYer would don't even make such a mistake
BTW the exact name of this device is Roksan blak .......
Last edited:
I have mostly used this player as a transport, and the strange thing is that it sounds very good as that. I have an Oppo 105 and various cd and dvd/bd players around the house, and I tried all of them as transport feeding my vintage dac. But none of them sound as good as the Roksan as transport. Strange thing.
Some do like distortion or "warm" sound. Harmonics and such. Personal choice. However, this does not take away that this specific device is seriously hampered by a non functional PSU design. So obvious that judging by ear is better left to absolute layman. Many a typical consumer can be fooled by listening to a price tag. It is expensive so it will sound good and such. As a transport it does not make use of its analog circuits that are likely to be most influenced by this design error. Even though the user does only use half of its capabilities it still is an error best solved.
This device has an incredibly stupid and large design error (or better: several design errors as they were replicated) and it can only be seriously judged after correction of the error. Any review boasting its qualities the way it is now can be ported to dev/null.
* this device should (when the error is corrected) sound way better than a vintage DAC. In fact for that price (over 3000 Euro) it should outperform it and make it obsolete. No I am not going to ask which brand/type vintage DAC. Users that want help/info should deliver to receive.
This device has an incredibly stupid and large design error (or better: several design errors as they were replicated) and it can only be seriously judged after correction of the error. Any review boasting its qualities the way it is now can be ported to dev/null.
* this device should (when the error is corrected) sound way better than a vintage DAC. In fact for that price (over 3000 Euro
) it should outperform it and make it obsolete. No I am not going to ask which brand/type vintage DAC. Users that want help/info should deliver to receive.
Last edited:
Yes, your observation is right. Using the CD player as a transport, the design/production error for this device model it not affect that circuits, but only the final analogue stage. The (supposed) +/-15v it is meant to power only the final analogue section.
My supposition about such embarrassing mistake of an enough known producer (Roksan), is that the original electronic design it is right done. For some reasons, after the board design it was tested and approved, it was changed the toroidal transformer. Either they wired wrong the new transformer, or the transformer itself it was wrong designed for this purpose. They could very easy route the unused winding (2x15v) to power the final opamp stage (not using the 2x10v one). A such solution (swapping the toroid 2x10v winding with the 2x15v one), it can very well fix the problem for the existent faulty CD players out there.
Well, the explanations are no so important here and now. The most important is that Roksan final test approach on production line, it have not cached a such serious fault. It seems that nobody measure on test points provided on board, and so far the CD is spinning inside, the product is declared as "Passed"...
Then the Roksan Balck CD player is sold on market with this serious error in it, which it affect its functionality (analogue outputs), and it alter its datasheet specifications.
Well, the explanations are no so important here and now. The most important is that Roksan final test approach on production line, it have not cached a such serious fault. It seems that nobody measure on test points provided on board, and so far the CD is spinning inside, the product is declared as "Passed"...
Then the Roksan Balck CD player is sold on market with this serious error in it, which it affect its functionality (analogue outputs), and it alter its datasheet specifications.
Last edited:
It is called "blak". Roksan blak in small capitals even. Not "Black", "Balck" or "Blak" but blak. Its color is black and we should not judge it by its color. Oh oh...
Please note that the 2 x 10V and 2 x 15V windings are very different in current! I don't think the 0.2A of the 2 x 15V windings are enough. Maybe this has even led to the choice to use the 2 x 10V....
Only you can judge the exact situation but ... you said 2 x 15V is completely unused... So you can remove the contacts from the plug and put them in the connector of the 2 x 10V windings? This can easily be done by pressing the contacts with a needle or very small screwdriver. Then it can be tested. It would be best to measure drawn current to see if the windings are not overloaded. I hope the violet-yellow-blue wires are long enough to put them in the other connector for safe assembly and tidy look. As they supposedly would not be connected to anything the situation would be solved.
The point is clear but the choice for 7815/7915 in a 3000+ Euro device is not.
Please note that the 2 x 10V and 2 x 15V windings are very different in current! I don't think the 0.2A of the 2 x 15V windings are enough. Maybe this has even led to the choice to use the 2 x 10V....
Only you can judge the exact situation but ... you said 2 x 15V is completely unused... So you can remove the contacts from the plug and put them in the connector of the 2 x 10V windings? This can easily be done by pressing the contacts with a needle or very small screwdriver. Then it can be tested. It would be best to measure drawn current to see if the windings are not overloaded. I hope the violet-yellow-blue wires are long enough to put them in the other connector for safe assembly and tidy look. As they supposedly would not be connected to anything the situation would be solved.
The point is clear but the choice for 7815/7915 in a 3000+ Euro device is not.
Last edited:
The main improvement approach for this device it was adding an complete chain DAC and post DAC circuits, and dedicate it to the XLR outputs. The original DAC/post DAC design it remain functional only for RCA out (with corrected +/- power rails issue as well).
The modified device it have now two DACs working in parallell, sharing the same digital interface (one for RCA and another for XLR output). Well, some more other improvements are done as well.
In my opinion, the power system for this DC player model is designed quite unprofessional.
I can only say that the improvement for the outputted audio signals for this upgraded device, is just dramatic.
The modified device it have now two DACs working in parallell, sharing the same digital interface (one for RCA and another for XLR output). Well, some more other improvements are done as well.
In my opinion, the power system for this DC player model is designed quite unprofessional.
I can only say that the improvement for the outputted audio signals for this upgraded device, is just dramatic.
Attachments
Well, I see... "blak" indeed. Sorry for my wrong "baptising"...
No, there is no any current issue about swapping that torid windings. The analogue final section it use a current into 100mA. Some of the torid windigs specified currents are much larger than the needs of the electronics on board. It is very possible that Roksan have replaced later the original specified transformer for the designed board, with another one, which it fitted enough well, but it was made for something else. For production costs reasons maybe...
Yes, I fully agree that for a 3000+ euro device, the designers should chose some better regulators...
Last edited:
No. The original power system on board it is improved as well... Not my intention to make now an detailed list of the all improvements made...
To do it just right, the existent board it should be redesigned/repopulated entirely. Not applicable....
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.