Building a DIY Music Server

[JayM]

Hello everyone! 

 

My real name is Jelle, but all my American/English friends call me Jay. Somehow my Dutch name keeps – unintentionally - being pronounced as Jelly or Yellow.

 

So, hello everyone, I am Jay! ;)

 

For a - relatively short – while I’ve been working closely with Emile, Edward, Ton, Ted and Wout at Taiko audio. We’ve recently moved from our old, shared facility in Hengelo to our own new headquarters in Oldenzaal. Our new location and set-up will allow us to do a lot more development and testing.

 

To manage the ever-increasing complexity of our operation, Emile has asked me to work with him as General Manager for Taiko Audio. In the coming months we will try to streamline our production and logistics process to expedite the time between order placements and actual shipping. This will enable us to fulfil orders faster and faster in the coming months.

 

That brings me to the reason why I’ve decided to introduce myself in this thread.  Until now we’ve only shipped a very limited amount of DC DC-ATX’s for testing.  If our supplier delivers some of the last needed components for the DC DC-ATX as scheduled, we will be able to ship all existing (paid) pre-orders on the 30th of April. 

 

16 hours ago, bit01 said:

@Exocer - I thought I had signed up for the Taiko DC DC-ATX but have not heard anything about it since - are they taking orders now and how does that work - or was yours a special case?

 

Thanks.

b.

Everyone who pre-ordered a DC DC-ATX can expect an email in the upcoming days:)

[JayM]

2 hours ago, magnium said:

Welcome @JayM

 

For those who pre-order the Taiko DC DC-ATX and interested in the Taiko DIY case, is there a possibility of consolidating the orders to save shipping costs?

 

Hard to say if I will order the case without knowing the design and cost but just throwing it out there!

 

Unfortunately for now it is not possible to consolidate shipping of these two items due to our internal logistics. The DC DC-ATX's only require a few more components before we can fully assemble and ship them. The DIY chassis however requires a lot more production steps to be completed before they can be shipped, this might take a while.

 

I have high hopes that I will be able to share some drawings or 'renders' of the DIY chassis with everyone on this forum tomorrow when I'm back at the office. Exact costs are not yet clear, this will be communicated in due time.

 

Hopefully I've answered all your questions.

Jay

[JayM]

21 hours ago, magnium said:

Welcome @JayM

 

For those who pre-order the Taiko DC DC-ATX and interested in the Taiko DIY case, is there a possibility of consolidating the orders to save shipping costs?

 

Hard to say if I will order the case without knowing the design and cost but just throwing it out there!

 

As I promised yesterday, here are some pictures to give an indication how the DIY chassis will look.

The external size of the chassis is 483mm wide, 496 mm long and 177 mm high (including the feet).

The internal size of the chassis is 416mm wide, 483 mm long and 150 mm high.

The whole chassis will be CNC-machined out of high grade aluminium.

We still have to add the recess for IEC on the back plate, the rest of the chassis is pretty veracious to the final product.

The chassis will come in black, this will be the exact same colour as the black version of SGM Extreme servers, unfortunately other colours are not an option so far. 

 

We hope to send the definite plans into production somewhere near the end of next week.

 

 

 

[JayM]

2 minutes ago, ASRMichael said:

Thanks for the information. Do you have information;

 

- passive CPU cooling capabilities? What is TDP rating?
- Will CPU heatsink & heat pipes be included? 
- will it be compatible with other motherboards & CPU? 

- would also be good to see a render of full size motherboard & atx inside the case. 
 

Thanks

 

Yes, the chassis will be set-up for passive CPU cooling, the TDP rating will be 2 x 85W at normal usage

Yes, CPU heatsinks and heat pipes will be included but only for the LGA3647 square.

Yes, it is compatible with motherboards mini ITX up to SSI EEB and all compatible CPU's under 2 x 85W at normal usage.

 

Unfortunately we are not be able to provide a render of the chassis with a full size motherboard & ATX as of now. When we receive our prototype chassis we might be able to provide some pictures with a full size motherboard & ATX in the case during test fitting.

 

Hopefully that answers everything!

 

[JayM]

Here is a specification sheet of our GaN based DC DC-ATX. Hopefully this provides some more insight! :)

 

This is a new development, not implemented in the SGM Extreme music servers. This was originally designed to facilitate a cheaper server model, without compromising performance. As the demand for our music servers is keeping us rather busy, and so delaying the potential launch of a smaller server. As we will not get around to this anytime soon, we’ve decided to make this available for DIY projects. This design allows you to have a high-performance power supply which can be powered from an unregulated power supply as low as 100 USD costs in parts (transformer, rectifier, filter capacitance). Everyone should feel free to experiment with the unregulated DC-rail, its make-up will largely determine total system performance and sound. We are eagerly looking forward to your experiences!

 

 

 

 

[JayM]

10 hours ago, ASRMichael said:

On 4/13/2021 at 9:20 PM, ASRMichael said:

Thanks can you just clarify; So the case is compatible with other motherboards & CPU’s but the Taiko cpu heatsink & heat pipes are not? 

 

Yes, CPU heatsinks and heat pipes will be included but only for the LGA3647 square.

@JayM

 

Hi can you please clarify? 

 

Hee ASRMichael, sorry did not see your question.

 

The chassis is compatible with pretty much all standard size motherboards (mini ITX up to SSI EEB).

For now we only have a LGA3647 square heatsink planned, this will fit on any LGA3647 square socket. You cannot use this 'LGA3647 square' heatsink on a 'LGA3647 narrow' socket.

 

The CPU heatsink will have  four recesses for 6mm heat pipes. 

 

Hopefully this answers your remaining questions, if not just let me know! :)

 

[JayM]

On 7/6/2022 at 4:44 PM, basillus said:

HDPLEX H5 does support Asus Sage C621E Dual CPU, when they are able to make the CPU cooler... First it was December last year, then May this year and now it is....?

Nenon do You know, if it is just the heatsink near the C621E MB there is used to both cpues at the Taiko Diy, like in the Taiko Extreme?

 

Is the Diy Taiko chassis already gone?

I got this massage: 

• Sorry, we do not have enough "DIY Audio Server Chassis - Option 2: 2x LGA3647 square cooler + heat-pipe bending pliers + 9x heat-pipe (includes one spare heat-pipe) + heat-paste" in stock to fulfill your order (0 available). We apologize for any inconvenience caused.

 

Unfortunately the first batch of the Taiko Audio DIY Chassis got sold out way faster than expected.

 

Some people already contacted me, but to prevent this happening at the second batch as well please send me a private message if you are seriously interested in buying a DIY chassis.

 

If more people than expected contact me I still have the ability - right now - to acquire more raw materials, so that this won't happen again. I will need to know during this weekend though. Since the parts with the longest delivery time (about 1,5 month) will be ordered on Tuesday. As soon as these parts come in we can ship out the orders which means we can ship everything in 2 months from now (3 months maximum, in case of set-backs). So it won't be as long a waiting time as last time...:) 

[JayM]

1 hour ago, MarcelNL said:

Interesting to see the rectifier arrived in the US earlier than at my house not a hundred miles or so away, same happened when the ATX shipped.

 

Some people who had additional orders with us received their Caps/rectifiers earlier (shipped last week) to combine the shipping. All other orders will be picked up by the shipping carrier tomorrow morning👌

[JayM]

21 hours ago, baconbrain said:

Next challenge with my Taiko DIY Case appears to be that the pre-drilled and threaded holes in the base plate for mounting the Taiko DC-ATX are to close to the Asus WS C621e Sage Mainboard. Here a picture showing the overlap of the DC-ATX when aligned with the mounting holes:

 

 

Given that I my ARC 6 Power Supply is external, I will not need all the space within the case so I can mount the DC-ATX flush with the baseplate using just 2 of the 4 mounting holes. Which now looks like this:

 

 

Needless to say, I am wondering how Taiko can oversee something like this?

 

I realize that all new products have teething problems but given that this is now the third issue with the chassis there seems to be a certain lack of quality control on the Taiko side. Certainly not ideal for a product at this price point.

 

Would be curious to know if others encounter this issue or if I was just unlucky enough to get the bad apple of the bunch.  

 

21 hours ago, Johnseye said:

 

I'm going through the same exercise as you. All the same components however I'd like to put my supply components inside the case. We already shared our similar experience with the L bracket. I'll let you know after this weekend how the mobo and atx converter fit. 

 

First of all our apologies on behalf of the entire team here at Taiko Audio!

 

Regarding the L-brackets, so far we've in total been contacted by 4 people who had issues fitting the L-brackets. By far, most brackets were produced correctly. However due to human error, some brackets had been placed in the wrong position in the CNC mill which caused the wrong holes to be threaded. In the next DIY batch we've changed the design so that this can't happen again they will only fit one way. The people who were affected by an incorrect L bracket will be provided with a correct one. One person offered to fix it himself and asked us not to ship a bracket.

 

On the bottom-plate issue; As soon as Baconbrain contacted me by email (same issue as post above), I spoke with our design and production team what could have happened, because this is obviously substandard in terms of quality and the service we want to offer everyone. Here is a quick recap what happened:

 

We did physically test a DIY chassis here at the factory, with the same motherboard and our own ATX. It all fitted together brilliantly, however one of the the motherboards we used came with a shield back plate on which foam was attached (normally we don't have that). Because of the extra thickness of the foam the backplate of the DIY chassis would come under a little stress and bend ever so slightly. With a shield back plate (no foam) everything fitted perfectly. To make sure that all motherboards - including the ones with a foam covered shield back plate - would fit. Our design team moved all threads for the bolts on which the components - such as the different kind of motherboards - are connected 2mm to the front. This is where our mistake was made; our design team thought SolidWorks was moving all threads for all the bolts 2mm to the front but the ones for the ATX were apparently not selected properly. Originally there was 1mm between the sides of the ATX and the motherboard, and because only the motherboard has been moved 2mm, there is exactly 1mm overlap.

 

Today, as soon as we figured out what went wrong - especially since we test fitted everything here in the factory months ago - we started working on a solution. We already started production of a copper adaptor plate to be fitted between the ATX and bottom plate of the chassis. We will continue to work on these adapter plates tomorrow, and are hopefully able to ship these to everyone who ordered a DIY chassis - in the first batch - tomorrow afternoon. At the very latest these will be shipped out on Monday!

 

We once more sincerely apologise for the inconvenience caused by the situation as described above.

 

[JayM]

1 hour ago, drjimwillie said:

@JayM

Will you please post a picture of the L bracket, either correct or incorrect, so I can compare it to mine?

I will not be building the Taiko DIY Case for a few months and I would like to be able to identify if my L bracket has been fabricated correctly.

Thank you

 

8 minutes ago, Johnseye said:

 

Here's a picture of how the bracket should NOT look. Let me know if you want specifics. 

Exactly what Johnseye says. So you'd want the holes that are NOT threaded in this picture to be threaded and the ones that are threaded should not have been. Hopefully that clears it up 

 

For general info, I promised to ship all DC-DC ATX adapter plates out today. So I drove to the DHL hub of East Holland which used to be open until 20:00, however they had changed opening hours and closed at 19:00, I was there at 19:15... :(

 

I will schedule a regular pick up for tomorrow morning.

[JayM]

On 8/28/2022 at 2:40 PM, MarcelNL said:

Just dropped in the regular Saligny, same thing...stable 36 V , slow discharge withoud load but once I connect the MB and switch it on the voltage drops to 19V...so I am thinking the ATX is faulty...

Main power was off in the 2 weeks, and I did indeed disconnect all cables between attempts trying to start from scratch each time.

 

Guess I'm in for a visit to Taiko, lucky it's in driving distance..

 

 

On 8/28/2022 at 2:57 PM, MarcelNL said:

Think I found the probable cause....a hairline fracture in the PCB right next to the power in connector. I never liked the way that part of the PCB sticking out in thin air, it's where some force is to be expected...where the other connectors are located the PCB is better supported.

 

invisible to the naked eye but using a stereo microscope it sticks (literally) out...(issue is more clear using two eyes on the scope)

 

 

 

On 8/29/2022 at 5:17 PM, Nenon said:

 

My guess is this is a treacled down design from the Taiko Extreme where it's plugged in once and not touched again, rather than a heavy duty solution designed from the ground up with durability as the main goal. And of course those Molex connectors are not designed for 14AWG solid wire. 

 

The R&D cost is quite significant on these things, and I don't think our small DIY market can cover the cost of designing this technology and product from the ground up. Hypothetically, if they sell 100 units and the profit is 50% of the retail cost, that will be 62500 Euro before the hefty Netherland taxes. As far as I know the R&D cost on these things is a six digit number, not five. 

 

While some people see this as "I paid so much money, I want everything for this amount", I see it a little differently. This DC to DC ATX is quite reasonably priced and provided to the DIY community out of passion. Emile was a passionate DIY-er just like us and opened his own business to be able to manufacture products he envisioned that no one else did and was nearly impossible to do as DIY. If he gets any useful feedback from the DIY community, that of course will be a bonus for him. But as far as I know, he is not offering these products for money making purposes. Taiko is operating at a much higher level than those DIY products and busier than ever at the moment. I personally see these DIY components as a favor and am very thankful for that. I probably have some influence convincing Emile to launch these products as he did not really have the time to support the DIY community. Documenting my experiments and posting recipes for ULPS helped convincing Emile to sell the DC to DC ATX (followed by the other DIY products).

 

It's really sad if your PCB cracked. Talk to Taiko and hopefully they will find a good solution for you. We are DIY-ers, and so knowing this, we can come up with some way to reinforce the Molex connectors, especially if we use very stiff cables. 

That might be a good idea for us to do if we stiff cables are used. If someone comes with an idea please post. 

 

@MarcelNL - keep us posted about your trip to Taiko and how they handled the DC to DC ATX fix/replacement. 

 

Just read up on this issue, unfortunate situation.

 

We looked at the picture of the hairline fracture on your ATX pcb and based on the location of the fracture this is definitely not the cause of the issue. You've also contacted me by email regarding this issue, since I am not too familiar with the design and limited in the technical knowhow, I've asked Emile to answer your questions he got around to it last night but it is in Dutch:

 

"Lijkt erop dat de spanning over de maximale toegestane waarde gaat wat de protectie trigger om de FET's te beschermen, als dat meerdere malen gebeurd dan kunnen de FET's kapot gaan, die zijn heel erg gevoelig voor “uitschieters”. Dit gebeurd als b.v. de impedantie te laag is versus de gebruikte capaciteit, te laag-Ohmige chokes, te laag-Ohmige gelijkrichter of trafo, derhalve belangrijk om de “geteste” waarden te gebruiken."

 

In short: Probably the current has been to high (multiple times) which triggers the protection, however if this happens too often the FET's can get damaged. 

 

See attached picture for an additional schematic, with English explanation below.

 

Example of what can happen with a CLC schematic with a choke substitute suggested in this thread. Please use tested / simulated examples as posted by Nenon if you are not familiar with such procedures. Duncan Amps PSU designer is an easy to use application which is accurate enough to model your system, please use that at the very least before changing anything in your design. Substituting lower DCR and/or higher Inductance chokes without adjusting capacitor values increases the likeliness of this to happen resulting in a damaged DC-DC ATX and in extreme cases a damaged motherboard, as the DC-DC ATX protection will always act a little bit “too late”. If a system is set-up this way, it may work a few times but in the end may (and probably will) damage your system because during its lifetime it will be powered on and off and get exposed to this oscillation effect every time. Active rectification, no soft-start and lower secondary impedance transformers can also exaggerate this effect.

 

As confirmation to Nenon's post regarding R&D this is definitely a big cost driving factor, this is the reason we have to design our products with different markets in mind since there is no way R&D can be supported by only the DIY branche. This is the reason the ATX is designed that way, the holes and cutout corner of the copper plate make the DC-DC ATX mountable in te SGM Extreme under all the capacitors otherwise this would not not have fit. Emile really wants to support the DIY market purely out of joy, because he started that way himself. I can vouch for the fact that we have little to no margin on the DIY products as is, so having to PCB's made one that would have fit in the SGM Extreme and one purely for the DIY market would have increased the price many times over since the market is just to small. We do not want to do that to the DIY community.

 

Hopefully I've touched on all subjects and clarifies the issue at hand :)

 

 

[JayM]

2 hours ago, MarcelNL said:

@JayM

Thanks for the response, I'm using components pretty close to the V3 design by Nenon using a softstart (resistor). My LPSU is designed for 36V output (delivering 35-36VDC) but I will battle some more with the PSU design tool to verify. Component specs such as impedance of the active rectifier are unknowns. Toroidy 500VA audio transformer, Lundahl LL2773 chokes in common mode is a 5A, 65mH, 0.22Ohm per coil, three 47000uF caps (two RIFA PEH169 (which kemet, which likely is Mundorf HC), and one Mundorf HC 47000uF)

 

Meanwhile I'm not really any closer to a solution or a confirmed root cause/problem isolation that might lead to a solution, is there a way to replace FET's should they be toast, how does one troubleshoot the ATX, can Taiko troubleshoot the ATX? Don't get me wrong, I appreciate the feedback in the forum it's just that it's a bit awkward to do troubleshooting/customer service through a forum.

The GaN fets are BGA (Ball Grid Array) and it’s a complex design, if it’s broken a replacement is the only practical workable solution.

[JayM]

19 minutes ago, MarcelNL said:

Lacking a response from Taiko I had some silent time to consider my next step(s) for the LPSU, I was close to building a final (if such a thing does exist in DIY😇). I'm slightly concerned about the custom 220.000 Mundorf HC which is rated at 40V, Now the sim shows that with that capacity Voltage upswing is non existent so I'll take my chance with that.

What I'm going to for DC ATX is open, pending the resolution/solution Taiko may offer yet at thius point I am leaning towards the HDPLex 800W which is a fair bit cheaper and seems to have a broader input voltage bandwidth.

The Bquiet PSU makes the server sound as if it's held back, some generic hi-fi piece.

 

Unfortunately we are not always available to answer straight away. In case of technical questions like yours I really have to talk with people here at Taiko (in this case Emile), because I don't want to provide some generic answer but actually give you a response that might be useful. Also to be 100% honest Emile simply does not have the time (he wants to, believe me...) to answer every technical question regarding DIY components. This is because there are many questions related to Taiko Audio products like the SGM Extreme and other new developments that simply have to be given priority. 

 

Emile looked at the graphs and components you provided and he'd suggest "in your case to use the parallel connection option on your chokes and perhaps get a higher secondary resistance transformer like the Noratel we‘ve purchased for this project with 22/22/440 capacitance values. Contact us and we can work something out on exchanging your dc-dc atx combined with a Noratel transformer  and/or additional caps."

 

Hopefully this helps you out:)