Desktop DAC with a cheap ES9038-board…

Last year I mentioned that I had bought a cheap ES9038 DAC-board and now that’s starting to come together. This will be used as a controller to feed a pair of active monitors on my desk (that’s the “need” I mentioned in the original post) where it will save me a box as I currently have a separate USB DAC and passive pre). To make the box a bit more versatile (and to use some of that empty space inside…) I am thinking about including a Raspberry Pi Zero W in the box and connecting it to the I2S-input of the DAC plus add a potentiometer for volume control (digital on the DAC via an onboard uC). The ES9038Q2M DAC is asynchronous with its own clock so perfect for use with the RPi and this combination would give a Volumio-based source with two additional digital inputs and a volume control. That means that can be connected directly to a power amplifier input or active speakers – perfect for the desktop or a second system!.

The slightly unusual power arrangement with 5V input to a DC-DC converter is to support this setup. It will allow the RPi to be powered on constantly while the DAC power can then be switched on and off. I bought an early (”first generation”) of the ES9038 board with a slightly different power setup than the current version and crucially no DC input connector that I did not need (of course the newer versions of the DAC board can be used as well). Check the diyaudio-thread for the board for loads more examples of how people have been hacking this by the way – some of them are close to unrecognisable.

The original idea was basically to allow the power source to be a cheap 2A USB-charger, but reality seems to have thrown a spanner in the works here – none of the chargers I have tried will actually let the DC-converter start properly because of the high inrush-current it draws. Using a slightly more beefy 5V supply fixes the problem though, so the concept is more or less intact anyway 😀

Part of what drew me to this DAC-board originally were the on-board connectors, but as I couldn’t find any mechanical drawings having to make the back panel by hand seemed a bit of a risk. Using my own measurements and the datasheets of the connectors I managed to trace it out and with judicious use of paper printouts from Frontpanel Express to check the dimensions against the physical board it actually fits well for a first time attempts – although this is definitely an occasion where I would wish for a 3D-printer or a laser cutter to make prototypes before betting my money on ordering “the real thing” from Schaeffer/FPX (they are becoming quite expensive I think).

Soundwise this seems quite close to the other ESS-based DACs that I have tried. I am not sure that is such a good thing to be honest, but I am prepared to compromise a little bit on my desk where convenience is important and the differences aren’t night and day anyway 🙂

Advertisements

A gem for Christmas…

The other of Richard Murdeys “gem” designs (the first one was here) that I have worked on is the Sapphire headphone/line amplifier. Now in version 4, I’ve looked at it before but I didn’t have a compelling reason to start building anything. However, when Richard “upgraded” it to version 4 before the summer it caught my eye again (through here) and so I finally decided to start my own version. This was a while ago now, but due to some problems with my board orders (the first one didn’t show up at all and the second one just took nearly two months to arrive…) it’s been a lot longer than I expected.

The Sapphire is a current feedback design which can be built to drive low- or high-impedance loads, meaning it can be used as both a line preamplifier and a headphone amplifier. It’s fully-discrete circuit but uses standard parts that are easy to find and it’s actually a fairly simple build. Once again, for “my” version I set out from the published Eagle-files (v4.1) with a view to make minor changes but just like the Emerald it didn’t really hold. My board is smaller than the original and some of the routing is different because I used a “splayed-pin” footprint for all the transistors (I try to buy all my TO92s on tape, so the inline pins footprint makes everything much easier).

The only real changes I’ve made to the electrical design is the addition of a couple of indicator LEDs to show that the board is powered on. I find this quite useful for troubleshooting, it looks nice and there was space on the board for them 🙂 While I’ve been faffing around with PCB orders from China, Richard has made small tweaks and released v4.2, but that’s of course fine – if you want guaranteed boards with support etc. you should be buying his anyway 🙂

My prototypes are one of the “high-bias” configurations because I mainly have low-impedance headphones. That’s probably fine, but the small heatsinks get seriously hot (app. 65 degrees C) and the offset is also higher than I would have liked. However, I think that’s more to do with my iffy transistor-matching that any serious issues with the board layout itself. I’ll probably give it one more go in January and see if I can get the amp to behave as I want them to and then post the board files in case anyone wants to do more tweaking 🙂

An early Christmas present…

I generally make a point of buying myself a Christmas present every year and this year it came a little early 🙂 While I was looking for the Hypex Ncore module I wrote about a couple of weeks ago, a stereo ICEpower700AS2 popped up as well – so I bought that 🙂 This is going to be a “little brother” to my 700ASC-monoblocks but whereas the monoblocks (which I also hope to finish over this Xmas break) have added buffers and dual-input switching, this is just going to be made into a simple and no-frills power amp.

In contrast to the Ncore module the ICEpower amp has onboard heatsinks so mounting in a small(ish) enclosure should be fine – at least for home duties. As I already had a basic layout for both a bottom plate and a back panel, drawing them up was quite easy and the back panel order is already placed. The support PCB I did for the monos also works here which should mean that once I receive the back panel there should hopefully be very few blockers to wiring up the amp and getting it tested within the next weeks.

I have seen one comparison of the Ncore and the ICEpower module (although I can’t find the link at the moment) with the Ncore coming away as the clear winner, but I am looking forward to seeing if my own conclusions match that 🙂

(apologies for the poorly lit pictures, but winter in Scandinavia means no daylight when I get home from work :D)

Testing the F3 amplifier…

Earlier this year Nelson Pass graciously started distributing batches of the Lovoltech LU1014 Power-JFETs for free to diy’ers. You could get 4 pcs. and only pay for shipping and of course that offer was hard to turn down – all the more so since I had been looking at trying an F3 amplifier at some point. There are no “official” group buy boards for the F3 at the moment and the few redesign/group-buy initiatives I have seen have been false starts, so I picked up a set of amplifier boards from ebay instead.

The F3 amplifier is one of Nelson’s “unusual” First Watt amplifiers, in that it uses a Power JFET as the gain device. Power JFETs are rare, and as a result for those of us who were a bit slow on the uptake the LU1014 came and went without me buying any. Otherwise the F3 isn’t a very complicated design, but it’s got normal First Watt class A heat levels and even-lesss-than-First Watt levels of output and gain. As a result i am not sure whether I actually have any practical use for the finished amplifier, but as a listening experiment I am still going to give it a try 🙂

The F3 is also a single-rail amplifier, which quickly led me to the realisation that I didn’t really have a PSU board suitable for a single-rail power amp. When you have a good “back catalogue” of designs then that’s quite helpful and so taking one of my existing class A power supply boards and chopping it up to create a single-rail version wasn’t that hard. The end result should hopefully – at some point – end up as a (nearly) dual-mono F3 amplifier (meaning a single transformer is used).

Another complication is that for reasons I can’t really remember I decided to try using 3U heatsinks for this build which may end up being a mistake – they are going to get very hot I guess. Anyway, for now it is an experiment and hopefully I will have it (electrically) completed by the Christmas break so I can hear what it sounds like before I put time and money into finishing the mechanical design 🙂

Encore?

A quiet last few weeks here – at least on the surface. Two reasons for that really: 1) With an Xmas-break looming on the horizon the pace at work is picking up a bit and 2) for quite a lot of projects I am in the annoying phase where lots of important work is done, but it doesn’t really look like you are getting closer to a finished product and so it’s not really worth showing here. If nothing else though, it’s nice to have a good pipeline for next year 😉

However, one thing there is always time for is to buy new projects for the shelves 😀 As regular readers will know I have made lots of ICEpower-based projects, but practically nothing with the various Hypex-modules. However, recently one of the new Ncore NC502MP modules came up on ebay and so I pounced on that. The module looks very nice but I need to test it for a while to check the sound (waiting for proper cables at the moment) and then I’ll decide on a suitable enclosure for it. The original goal was to built a custom high-power integrated amp, but I may end up going in a different direction and do a pure power amp instead. One deciding factor will definitely be whether mounting the module on a simple aluminium bottom plate proves to be enough heat sinking, because if the module has to be on a “real” heat sink, then all my current enclosure ideas are definitely out the window!

Soundwise, I still expect that the benchmark for the Ncore to beat (at least in class D) is going to be my trusty 125ASX-based stereo amp and the 700ASC-monos (which incidentally are also among the designs that are I am currently inching closer to completion…)

Experimenting with the BBA3FE…

Not the most obvious acronym to decode, but it stands for “Balanced Burning Amp 3 Front-End” and it’s the first half of a power amp design that Nelson Pass launched for the ”Burning Amp” festival several years ago. It’s very closely related to the F5 circuit and without the accompanying power stages it’s also well-suited for pre-amp duty. The diyaudio store has been selling PCBs for years, but I’ve never been sufficiently interested to grab one (and the shipping cost and import duty for buying PCBs from the US makes it really expensive to get anyway).

Now, fast forward to a few months ago: For a while I’ve looked for a “real” balanced preamplifier circuit. I have several balanced designs already, but some of them are “cheating” by not being developed for balanced use and that obviously won’t do 😉 I’ve looked at the BA-3 before, but some months ago I did a double-take when I noticed a schematic for a balanced version in this thread (and yes, I know the thread is seven years old…). That circuit was more or less what I was looking for, namely a simple discrete circuit optimised for balanced use, and so I decided to try to make one.

My version is more or less the same as the original, but I decided to save some board space (and create some flexibility) by moving the output capacitors off-board and so they are not shown (but still very much required). The major downside of the circuit is that the input is based on 2SK170/2SJ74 JFETs that are obsolete and near-impossible to get. To add insult to injury, they should ideally be matched to around 8mA Idss which is more or less the most commonly required value – and therefore even harder to find!. However, the article also states that while matching is preferred, it is not essential, and so I managed to find some suitable pairs in my parts drawers.

I’ve only briefly tested the boards and they power up fine but the DC-offset is unstable so the output capacitors are definitely needed. Thermal stability and equilibrium with the bias-current is also something I need to work on (it’s going to require leaving the boards on for a while as far as I can see), but so far it is looking promising. As regular blog-followers will know I am a big fan of the B1 design and I don’t normally need gain, but as there are several reports of the BBA3FE sounding significantly better than the B1 I am obviously quite excited to make progress on this build.

PCB files will be coming eventually, but I made a couple of stupid mistakes in this layout that need to be corrected, and since I was so focused on this being a balanced pre I forgot to make it easy to do the SE-version as well. Translation: I really need to get a v1.1 ready and ordered first 🙂

Improving a Mean Well IRM PSU…

If you’ve been here before, you might have noticed that I have been using Mean Wells IRM-xx series PSU modules quite a bit. They are small, cheap, easy to use and available from many of the parts sources I normally buy from. Being cheap switching supplies they have quite a bit of ripple and noise which on the face of it is a problem. However, in practice many supporting applications aren’t too fussy about the quality of power and for those that are, alternative PSU arrangements can usually be found.

But what if the noise from the IRMs could be removed? One application where that would be useful (and where several people other than me have tried it already), is for Kevin Gilmores various discrete amplifiers (Dynalo, CFP etc.). They are normally powered by 16-24V DC and require a few hundred milliamps per board, so it’s pretty ideal for an 24V IRM with some additional filtering and regulation. Here I’ve made a single channel PSU intended for a 15/20W IRM-module, so output currents are in the region of 0.8-1.5A or so.

Most integrated voltage regulators have low rejection of HF-noise, but as the IRMs have a high switching frequency a passive filter seems an ideal way of damping the noise before it gets to the regulator. I can’t find any spec to state how much capacitance the IRMs will tolerate so I’ve stayed on the conservative side, but even so a small passive CRC/CLC filter is very effective at 100kHz so it should be fine.

Instead of the “usual” LT108x LDO voltage regulator I’ve gone for an LM2941 instead. This has an even lower drop-out voltage which will help if the starting point is a 15V module. The downside is a max. output voltage of 20V – 22V would have been better (and while we are at it, can we get 9V and 18 versions of the IRM20 please Mean Well :D). Actually, I’m going to try to do an LT10xx-based version as well, but for now the LM2941 works fine. For this test example I’ve set the output voltage to 19.5V and I get 19.4V even under load so that is perfect. Next step: Build two more boards for my first “real” use-case for these 🙂

A real gem…

I’ve already had my hands on a couple of designs by Richard J. Murdey, mainly here and here. However, Richard has quite a few other interesting designs on his site and one of them I’ve had my eyes on for a while, namely the “Sapphire” which is a line/headphone amplifier. Unfortunately my first order of Sapphire boards never showed up and while I was working on the Sapphire I also had a look at the Emerald MM/MC RIAA. Those boards did show up and so while I wait for the reordered Sapphire boards to (hopefully) show up soon, let’s look at the Emerald instead 🙂

The Emerald is very simple RIAA-design with two opamps per channel and selectable gain for both MM and MC cartridges. The RIAA correction is a bit special as described in Richards write-up and the design also includes an on-board discrete voltage regulator.

Richard has graciously shared the Eagle files for his designs so that was what I started from, but what was originally meant to be minor touch-ups ended up taking me a bit further away from the original than I expected. Although the schematic is still (more or less) the same, I’ve made quite a few changes to the board so that it now looks more like one of mine and uses the parts footprints that I normally use – the latter is absolutely intentional, the former a bit less so. I’ve also done quite a bit of rerouting – which I think of as optimizing the layout, but others might disagree – and managed to shrink the overall board dimensions a little in the process as well.

I haven’t really sacrificed anything from the original, except maybe by downsizing the footprint for the output cap a little bit, but that’s hopefully an acceptable compromise for most people. Project files (probably) coming shortly 😀

In search of synergy…

Slightly off-topic post, but I have written a few times about how I think that system-matching is much more important than any “absolute” sound quality, at least as far as enjoying the music is concerned. Well, today was another reminder that I still think this is the case 🙂

A couple of months ago I got new speakers, trading my old (and much-loved) standmount Sonus Fabers for some floor standing Scansonics that offered a bit more low-end slam. I was quite happy with the trade from the beginning and I have absolutely no regrets, but after a time the inevitable restlessness sets in and you start thinking about change (at least I do…). I’ve been running the Scansonics with a simple 125ASX amp on my Harman/Kardon preamp, but just to try it I dug out another ICEpower-amp from my collection, this time based on the older 200ASC-modules.

Although I would definitely still class the 125ASX as the better amp overall, the Scansonics (which are just a little bit bright) immediately benefitted from the more “closed-in” presentation style of the 200ASC, so as usual after initially listening to half a track I started to go through my normal playlist of tracks I know well and just enjoyed listening to some music that I would normally say I know back-to-front already.

To be fair I am honestly not surprised at this, because I saw the same change when I switched from the even older Elac speakers that much preferred the warmer sound of a 50ASX amp whereas the Sonus Fabers really came to life with the more lively presentation of the 125ASX. However, I still think that it’s nice to be reminded once again what really matters when putting a well-rounded system together and of course experimentation is always fun (although it can sometimes be very expensive as well…)

Building an(other) F5…

Although I recently built a new type of F5 amplifier, I haven’t completely abandoned the original F5 design 🙂 Hiding in one of my many boxes were a pair of half-finished F5 boards and some matching matching fan heatsinks that only needed the last bits of assembly and calibration. That honestly didn’t take long to do once the right parts showed up and I then managed to confirm the boards were indeed working.

The boards were originally bought from ebay and are more or less the same as my original F5 build – nothing special there. I have some matching PSU boards as well, only missing the last few parts which are now in the queue for my next order and that’s going to be a standard C-R-C type thing as well.

The mechanical design is from the same time as my JLH mono blocks, so the idea is also more or less the same. This heat sink profile is too large to fit in most enclosures though, so cracking what to do took some time but I think I have it figured out now. It’s also going to be monoblocks, but much larger ones than the JLHs. From my first tests during calibration of the boards I think a slow-speed fan should be enough to keep the heat under control, so hopefully they will be living-room friendly when they are done 🙂