Bold move! Thank you for your support.
I definitely recommend the aluminum covers. They're much easier to drill than the steel ones.
That's what I did in my 4U x 400 chassis layout. The only drawback is that sometimes there's a bulge on the heat sink where the two heat sinks meet which pushes the two heat sinks further apart than what's in the CAD drawings. You may have to file off that bulge. It's not a show stopper, but it is a bit annoying.
Tom
Thanks for answering my e-mail the other day. I asked about Modulus-686 design considerations. I've since read all the relevant documentation on the website (except for the Modulus-686 design documentation that gets referenced, I can't find the page for that).
The graph from the quote has got me wondering: does the supply voltage influence the THD+N performance as measured in the graph? The lower supply voltage option that gets mentioned on the product page would be very interesting to me. I want to build a custom case and don't need the full output power, so having less heat to deal with helps my design. If I'd lose performance because of it though, I'll opt for a full power build.
I deliver the design doc in the form of a download link sent by email when your order ships. That's why you don't see it on my website. I consider the documentation to be part of the product.
The THD+N is entirely noise-dominated, so it does not change with supply voltage. I love the "Safe-n-Sane" build. 3U x 330 x 300 mm chassis from ModuShop and a pair of Mean Well RPS-400-27-C. I used to have a handful of those amps, but I've sold them. These days I'd control them with an Intelligent Soft Start.
Tom
The THD+N is entirely noise-dominated, so it does not change with supply voltage. I love the "Safe-n-Sane" build. 3U x 330 x 300 mm chassis from ModuShop and a pair of Mean Well RPS-400-27-C. I used to have a handful of those amps, but I've sold them. These days I'd control them with an Intelligent Soft Start.
Tom
I completed my Modulus 686 build at the weekend. Well, it is not 100% complete as I need to fit the Molex connectors to the Power 686 boards, when the crimps arrive.
I want to thank Tom for his support, and superb design/build documents, and for answering my stupid questions.
I will post some pictures up later in the day, but I am very happy with my first DIY build. It is not perfect, as will be revealed in the pictures, and if I were to do this again, it woukd be very slightly different.
The differences would include getting the mounting holes for the ISS V2 correct, and increasing the speaker binding post distance between centres from 19mm to 30mm, as ut gives you a better choice of posts.
I would also like to thank people who have already posted on this thread for the build ideas and inspiration.
I want to thank Tom for his support, and superb design/build documents, and for answering my stupid questions.
I will post some pictures up later in the day, but I am very happy with my first DIY build. It is not perfect, as will be revealed in the pictures, and if I were to do this again, it woukd be very slightly different.
The differences would include getting the mounting holes for the ISS V2 correct, and increasing the speaker binding post distance between centres from 19mm to 30mm, as ut gives you a better choice of posts.
I would also like to thank people who have already posted on this thread for the build ideas and inspiration.
Here are a few of the photos from the build.
As you can see, the ISS V2 is much smaller than the ISS V1. Unfortunately, I missed this point when I did the CAD files for the case, so my ISS V2 was never going to fit.
So, I made a mounting plate out of an old PC case of the right colour to avoid drilling new holes in the case.
After wiring all the power cabling up, I powered the amp up, to check the voltages on the Power 686 boards to make sure everything was correct before connecting the amplifiers to the power supply.
hen I connected up all the cabling, and checked it all worked. Then tidied the internal cables up a bit, but I forgot to take a picture of that stage, so you will have to make do with the picture of the untidy case instead.
What can I say about this amp? My previous amps were a set of Modulus -86 mono blocks I bought on here. The 686 does what the Modulus -86 does, but it just does it better. It is certainly better suited to my Triangle Esprit Antal EZ speakers. It is certainly one of the better amps I have heard. I am looking forward to the end of this pandemic, so I can go and see a friend who owns a Benchmark ABH2 and see if there is a detectable difference between the two amplifiers.
As you can see, the ISS V2 is much smaller than the ISS V1. Unfortunately, I missed this point when I did the CAD files for the case, so my ISS V2 was never going to fit.
So, I made a mounting plate out of an old PC case of the right colour to avoid drilling new holes in the case.
After wiring all the power cabling up, I powered the amp up, to check the voltages on the Power 686 boards to make sure everything was correct before connecting the amplifiers to the power supply.
hen I connected up all the cabling, and checked it all worked. Then tidied the internal cables up a bit, but I forgot to take a picture of that stage, so you will have to make do with the picture of the untidy case instead.
What can I say about this amp? My previous amps were a set of Modulus -86 mono blocks I bought on here. The 686 does what the Modulus -86 does, but it just does it better. It is certainly better suited to my Triangle Esprit Antal EZ speakers. It is certainly one of the better amps I have heard. I am looking forward to the end of this pandemic, so I can go and see a friend who owns a Benchmark ABH2 and see if there is a detectable difference between the two amplifiers.
Looks like it came together nicely for you, even if the ISS shrunk a bit. This is the first build picture I've seen with the new ISS.
I'm glad you like it. Welcome to the Modulus-686 Club.
I'd be curious to hear how it compares against the ABH2.Tom
Way back at the beginning of June this year I ordered lots of nice stuff from Tom – 2 x Modulus-686 modules, 2 x Power-686 modules, 2 x Guardian-686 modules and an ISS module. Here I had the basic parts to build a Dual Mono 686, all I needed in addition were some parts, a couple of huge toroidal 400VA 25V transformers from Toroidy in Poland, a nice case and some sundries like switches and sockets.
Toroidy offer 3 levels of transformer – standard, audio and supreme, I plumped for the Audio grade – the Supreme were out of my budget at £115 each!
Parts all ordered from Mouser including some nice Tellurium copper binding posts (BU-P3770-0), Neutrik XLR sockets (NC3FD-LX-B) and a Schaffner medical grade power entry module (FN9262B-10-06).
Case also selected and ordered from AliExpress. I had to wait and wait and wait and wait and finally it was delivered …… to an address 400 miles from me. Replacement case ordered and again I had to wait and wait but this time it did manage to get delivered to me.
During this time I’d assembled all the boards but hadn’t tested anything. As soon as the case was here I set about drilling all the required holes in the base and backplate. Toroids mounted, Power-686’s mounted, ISS and power inlet mounted, wired up and switched on. Phew, no magic smoke and all voltages look just fine. Heatsinks then drilled for the Modulus-686 and the Guardian modules. Modules fitted along with all the remaining sockets and connectors.
Wiring checked, doubled checked and triple checked – all looks goods. Left hand channel tested first with no problems. Right hand channel tested and no problems (I’m using a current limiter all this time).
Current limiter removed and bang, magic smoke from the right hand channel. WTF??? A retest of the module now shows there is no output. There then ensued several emails back and forth to Tom to try and figure out what had gone wrong. Tom got me to check a number of items, all of which pointed to the possibility of a problem in or around U1 or U2 as it appeared there was an issue with the inputs being shorted but only on power on. I was on the verge of asking Tom if I could send the suspect module back to him to see if he could figure out what the issue was. Before I did that I eye-balled the module with a magnifier and noticed that R44 looked a bit worse for wear and it was in fact open circuit. Replacement part ordered for R44 and once it was replaced the module appeared to be working again. Slowly put everything back together again and it all still worked. To this day I have no idea how or why R44 got blown but that was the only issue on the board – music could now he heard!
There was a bit of an ‘oh no’ moment when I first switched it on – I couldn’t hear anything, not even the faintest whisper of hiss from the speakers and I did actually check I’d plugged the speakers in! It is silent, totally silent, no humm, no buzz, just lovely silence.
My intension had been to get the listening tests done in my system way back at the end of October beginning of November but unfortunately my dad took ill (he’d actually been quite unwell for the past couple of years) and sadly he passed away a couple of weeks ago so the hifi has been quite far down my list or priorities. I’ve had music playing but not to listen to seriously, just background music while I’ve been doing other stuff around the house. At least this has meant the 686 amp has been powered up and running so any burn-in effects should be well and truly over by now.
The past couple of days though I’ve had a bit of time and sit down for a serious listen. I have my Monitor Audio PL300’s and Benchmark DAC2 HGC connected directly to the amp. Music is supplied to the Benchmark from an Allo DigiOne/Raspberry Pi powered by a nice linear power supply.
First track to get played was ‘Fill The Hat’ from the Atmos album Headcleaner. Probably not a very well known track but it can certainly give your system a bit of a work out. If you went to clubs around 2000 you may well know this track, takes me back to my days when I still did the club thing now and again in Glasgow – way too old for that now! Lots of belly thumping bass to shake anything in your room that isn’t tied down but also a great mix – samples of Tangerine Dream can be heard way down in the mix and I was able to clearly pick them out as I tried to keep myself in my seat rather than closing all the curtains, digging out the glow sticks, and dancing around my listening room. Really wide soundstage and all the synth tracks can be easily picked out and followed. Good start then. Box 1 ticked, amp can do Atmos!
Let’s try something a million miles from Atmos.
Next up was ‘Arlington’ from The Wailin’ Jennys album 40 Days. They’re an all female folk/bluegrass group from Canada – thought it was an appropriate selection given that’s Neurochome’s home. The striking thing about this track is how beautifully the singer’s voices come together at the start of the second verse and the huge soundstage created by the instruments. They’re hanging in the air, wonderfully placed on the stage and the violins sound just perfectly tuned. I’ve had amps that struggle creating the huge soundstage I know this track is capable of producing but no problems here for the 686., It was so good I replayed this track 3 times in a row, it was than enjoyable.
Something a bit more mainstream now. Pink Floyd - The Wall. You can’t pick a track out from this album to play on it’s own, the only way to listen to it is to listen from start to finish in one sitting. Wonderful, just wonderful. ‘Nobody Home’ on CD2 really stood out for me though. The emotion in Roger Waters voice had me gripped by the proverbial’s and the hair’s on the back of my neck were standing to attention - the piano and tv playing in the background could be easily focused on, again the separation of the instruments was first class.
I’ve listened to a fair old amount over the past couple of days and I can say with hand on heart this is a damn fine amplifier. I’ve owned quite a number of very well respected amps over the years EAR 509, EAR 534, Luxman L550 Aii, Chapter Audio Precis 200S to name but a few and I currently use a Perreaux Eloquence 250i 40th Anniversary.
So how does the 686 stack up against the £8k Perreaux? You know what, I could happily live with either of them. The Perreaux is a different beast – it’s an integrated, it’s beautifully finished just as you would expect from an £8k product, it has a really nice DAC fitted, it’s computer controlled and has a bunch of handy features like limiting the volume limit so you can’t accidentally blow the speakers. It grips the music but lets it flow so naturally.
It neutral but not stark. It really is a superb amp.
All really nice, but there’s a but.
I didn’t build it, I just took it out of a box and plugged it in. I have no idea how it’s put together. I’d have no idea how to fix it if it went pfffttt. In 20 years time it might well become a boat anchor if Perreaux stop supporting it and the micro controller board dies and I can’t get a replacement. Having said that it is the best amp I’ve ever owned.
My 686 cost me around £1400 to build plus my time. Yes it’s not finished to the standard of the Perreaux, but it does look nice even though I say so myself, and yes it doesn’t have all those natty features. But I built it. With my bare hands. I ‘think’ I know how it works and could have a go at fixing it if it went wrong. I am proud of what I’ve managed to produce and I’ll love letting people listen to it and saying ‘I built that’!
It’s not ‘better’ than the Perreaux but it certainly isn’t out performed by any realistic margin. I’m not even sure in blind testing I’d be able to tell them apart. It really is that good.
I’d like to finish by saying I wasn’t approached by Tom or anyone else to write this. These are my thoughts and my thoughts alone. Tom was fantastic to deal with, parts arrived in no time at all and he was super helpful when I had a problem. There really isn’t anything else you could ask for from a designer and supplier. 10 out of 10 Tom, many thanks.
All that’s left for me to do now is see how Tom’s 2x8 Preamp does against my DCG3 Pre, the Benchmark with it’s inbuilt pre and the Perreaux’s. That’ll be an interesting shoot out.
Merry Christmas. Stuart
Toroidy offer 3 levels of transformer – standard, audio and supreme, I plumped for the Audio grade – the Supreme were out of my budget at £115 each!
Parts all ordered from Mouser including some nice Tellurium copper binding posts (BU-P3770-0), Neutrik XLR sockets (NC3FD-LX-B) and a Schaffner medical grade power entry module (FN9262B-10-06).
Case also selected and ordered from AliExpress. I had to wait and wait and wait and wait and finally it was delivered …… to an address 400 miles from me. Replacement case ordered and again I had to wait and wait but this time it did manage to get delivered to me.
During this time I’d assembled all the boards but hadn’t tested anything. As soon as the case was here I set about drilling all the required holes in the base and backplate. Toroids mounted, Power-686’s mounted, ISS and power inlet mounted, wired up and switched on. Phew, no magic smoke and all voltages look just fine. Heatsinks then drilled for the Modulus-686 and the Guardian modules. Modules fitted along with all the remaining sockets and connectors.
Wiring checked, doubled checked and triple checked – all looks goods. Left hand channel tested first with no problems. Right hand channel tested and no problems (I’m using a current limiter all this time).
Current limiter removed and bang, magic smoke from the right hand channel. WTF??? A retest of the module now shows there is no output. There then ensued several emails back and forth to Tom to try and figure out what had gone wrong. Tom got me to check a number of items, all of which pointed to the possibility of a problem in or around U1 or U2 as it appeared there was an issue with the inputs being shorted but only on power on. I was on the verge of asking Tom if I could send the suspect module back to him to see if he could figure out what the issue was. Before I did that I eye-balled the module with a magnifier and noticed that R44 looked a bit worse for wear and it was in fact open circuit. Replacement part ordered for R44 and once it was replaced the module appeared to be working again. Slowly put everything back together again and it all still worked. To this day I have no idea how or why R44 got blown but that was the only issue on the board – music could now he heard!
There was a bit of an ‘oh no’ moment when I first switched it on – I couldn’t hear anything, not even the faintest whisper of hiss from the speakers and I did actually check I’d plugged the speakers in! It is silent, totally silent, no humm, no buzz, just lovely silence.
My intension had been to get the listening tests done in my system way back at the end of October beginning of November but unfortunately my dad took ill (he’d actually been quite unwell for the past couple of years) and sadly he passed away a couple of weeks ago so the hifi has been quite far down my list or priorities. I’ve had music playing but not to listen to seriously, just background music while I’ve been doing other stuff around the house. At least this has meant the 686 amp has been powered up and running so any burn-in effects should be well and truly over by now.
The past couple of days though I’ve had a bit of time and sit down for a serious listen. I have my Monitor Audio PL300’s and Benchmark DAC2 HGC connected directly to the amp. Music is supplied to the Benchmark from an Allo DigiOne/Raspberry Pi powered by a nice linear power supply.
First track to get played was ‘Fill The Hat’ from the Atmos album Headcleaner. Probably not a very well known track but it can certainly give your system a bit of a work out. If you went to clubs around 2000 you may well know this track, takes me back to my days when I still did the club thing now and again in Glasgow – way too old for that now! Lots of belly thumping bass to shake anything in your room that isn’t tied down but also a great mix – samples of Tangerine Dream can be heard way down in the mix and I was able to clearly pick them out as I tried to keep myself in my seat rather than closing all the curtains, digging out the glow sticks, and dancing around my listening room. Really wide soundstage and all the synth tracks can be easily picked out and followed. Good start then. Box 1 ticked, amp can do Atmos!
Let’s try something a million miles from Atmos.
Next up was ‘Arlington’ from The Wailin’ Jennys album 40 Days. They’re an all female folk/bluegrass group from Canada – thought it was an appropriate selection given that’s Neurochome’s home. The striking thing about this track is how beautifully the singer’s voices come together at the start of the second verse and the huge soundstage created by the instruments. They’re hanging in the air, wonderfully placed on the stage and the violins sound just perfectly tuned. I’ve had amps that struggle creating the huge soundstage I know this track is capable of producing but no problems here for the 686., It was so good I replayed this track 3 times in a row, it was than enjoyable.
Something a bit more mainstream now. Pink Floyd - The Wall. You can’t pick a track out from this album to play on it’s own, the only way to listen to it is to listen from start to finish in one sitting. Wonderful, just wonderful. ‘Nobody Home’ on CD2 really stood out for me though. The emotion in Roger Waters voice had me gripped by the proverbial’s and the hair’s on the back of my neck were standing to attention - the piano and tv playing in the background could be easily focused on, again the separation of the instruments was first class.
I’ve listened to a fair old amount over the past couple of days and I can say with hand on heart this is a damn fine amplifier. I’ve owned quite a number of very well respected amps over the years EAR 509, EAR 534, Luxman L550 Aii, Chapter Audio Precis 200S to name but a few and I currently use a Perreaux Eloquence 250i 40th Anniversary.
So how does the 686 stack up against the £8k Perreaux? You know what, I could happily live with either of them. The Perreaux is a different beast – it’s an integrated, it’s beautifully finished just as you would expect from an £8k product, it has a really nice DAC fitted, it’s computer controlled and has a bunch of handy features like limiting the volume limit so you can’t accidentally blow the speakers. It grips the music but lets it flow so naturally.
It neutral but not stark. It really is a superb amp.
All really nice, but there’s a but.
I didn’t build it, I just took it out of a box and plugged it in. I have no idea how it’s put together. I’d have no idea how to fix it if it went pfffttt. In 20 years time it might well become a boat anchor if Perreaux stop supporting it and the micro controller board dies and I can’t get a replacement. Having said that it is the best amp I’ve ever owned.
My 686 cost me around £1400 to build plus my time. Yes it’s not finished to the standard of the Perreaux, but it does look nice even though I say so myself, and yes it doesn’t have all those natty features. But I built it. With my bare hands. I ‘think’ I know how it works and could have a go at fixing it if it went wrong. I am proud of what I’ve managed to produce and I’ll love letting people listen to it and saying ‘I built that’!
It’s not ‘better’ than the Perreaux but it certainly isn’t out performed by any realistic margin. I’m not even sure in blind testing I’d be able to tell them apart. It really is that good.
I’d like to finish by saying I wasn’t approached by Tom or anyone else to write this. These are my thoughts and my thoughts alone. Tom was fantastic to deal with, parts arrived in no time at all and he was super helpful when I had a problem. There really isn’t anything else you could ask for from a designer and supplier. 10 out of 10 Tom, many thanks.
All that’s left for me to do now is see how Tom’s 2x8 Preamp does against my DCG3 Pre, the Benchmark with it’s inbuilt pre and the Perreaux’s. That’ll be an interesting shoot out.
Merry Christmas. Stuart
Last edited:
Very nice builds fellas! A pleasure to read and see! I knew this amp was definitely in high company once Do built it and gave it the thumbs up (see post 1 of this thread for links)! Just look at his amp arsenal (and his build quality), he can make some very cogent comparisons.
Best,
Anand.
Best,
Anand.
Got a couple of Modulus-686's on the way for my next build.
Now I'm trying to figure out what to do for the power supply.
I'm thinking about running two Mean Well UHP-500-36's or UHP-750-36's, with an ISS to turn them on.
The reason I like the UHP instead of the RPS is because I want my amplifier chassis to be completely closed, with no vent holes. The UHP is designed to dissipate its heat through conduction to its chassis rather than through convection like the RPS. So, I figure I can bolt them to the bottom of my chassis and it will be fine.
However, it seems like the UHP-500 and UHP-750 both use a lower switching frequency than the RPS. The UHP-500 switches at 65 kHz and the 750 switches at 55 kHz, whereas the RPS switches at 100 kHz.
My question is: will this lower switching frequency be a cause for concern?
What does the PSRR of the Modulus-686 look like at 55 kHz? Or are these frequencies so high that they will be filtered by the supply rail caps on the amplifier boards?
I realize the PSRR probably drops as the frequency rises, but I also feel like 100 kHz is more likely to be "absorbed" by the capacitors compared to 55 kHz, so 55 kHz noise may be more reliant on the amplifier's PSRR to be filtered out.
I'm 100% speculating here.
Mean Well actually publishes full DVT results for all of these supplies, and it looks like the UHP-750 has about 40 millivolts of switching noise under full load. It also performs much better than the RPS-500 on the transient load test.
The only thing to worry about now is whether these UHP-500/700 supplies do the whole "pulse skipping" thing under low load scenarios, and whether the idle dissipation of each amp board is enough to keep the supplies out of that mode.
Now I'm trying to figure out what to do for the power supply.
I'm thinking about running two Mean Well UHP-500-36's or UHP-750-36's, with an ISS to turn them on.
The reason I like the UHP instead of the RPS is because I want my amplifier chassis to be completely closed, with no vent holes. The UHP is designed to dissipate its heat through conduction to its chassis rather than through convection like the RPS. So, I figure I can bolt them to the bottom of my chassis and it will be fine.
However, it seems like the UHP-500 and UHP-750 both use a lower switching frequency than the RPS. The UHP-500 switches at 65 kHz and the 750 switches at 55 kHz, whereas the RPS switches at 100 kHz.
My question is: will this lower switching frequency be a cause for concern?
What does the PSRR of the Modulus-686 look like at 55 kHz? Or are these frequencies so high that they will be filtered by the supply rail caps on the amplifier boards?
I realize the PSRR probably drops as the frequency rises, but I also feel like 100 kHz is more likely to be "absorbed" by the capacitors compared to 55 kHz, so 55 kHz noise may be more reliant on the amplifier's PSRR to be filtered out.
I'm 100% speculating here.
Mean Well actually publishes full DVT results for all of these supplies, and it looks like the UHP-750 has about 40 millivolts of switching noise under full load. It also performs much better than the RPS-500 on the transient load test.
The only thing to worry about now is whether these UHP-500/700 supplies do the whole "pulse skipping" thing under low load scenarios, and whether the idle dissipation of each amp board is enough to keep the supplies out of that mode.
I have not tried the UHP-series. Nice find, though!
The UHP-750-36 can provide up to 21 A of output current. That's not quite enough for continuous sine wave operation, but I would think that it's enough for music reproduction.
One trick that I employed in my Safe-n-Sane build with the RPS-400-27-C was to invert one channel on both input and output. Assuming that the signals in the two channels are correlated, this should balance the load between the supplies, thereby avoiding bus pumping.
The big question is whether the UHP-750 is quiet with no (or low) load. Each MOD686 draws about 350 mA. I would buy the supplies from somewhere with a return policy (like Mouser) and test the supplies upon arrival to make sure they work.
Some had issues with the 36 V version of the RPS-400. Others didn't. Maybe you'll get lucky.
I wouldn't worry about the PSRR at 50 kHz vs 100 kHz. The PSRR is probably 5 dB lower at 100 kHz than at 50 kHz, but as you say, there are many variables in that equation.
Tom
The UHP-750-36 can provide up to 21 A of output current. That's not quite enough for continuous sine wave operation, but I would think that it's enough for music reproduction.
One trick that I employed in my Safe-n-Sane build with the RPS-400-27-C was to invert one channel on both input and output. Assuming that the signals in the two channels are correlated, this should balance the load between the supplies, thereby avoiding bus pumping.
The big question is whether the UHP-750 is quiet with no (or low) load. Each MOD686 draws about 350 mA. I would buy the supplies from somewhere with a return policy (like Mouser) and test the supplies upon arrival to make sure they work.
Some had issues with the 36 V version of the RPS-400. Others didn't. Maybe you'll get lucky.
I wouldn't worry about the PSRR at 50 kHz vs 100 kHz. The PSRR is probably 5 dB lower at 100 kHz than at 50 kHz, but as you say, there are many variables in that equation.
Tom
Hi there Tom.
I am planning on attempting a build of your Modulus 686 amps to make a stereo system. Being that I haven’t touched electronics in component form since i was in College, I am naturally a little apprehensive but will forge forward.
Is it easier for me to email you directly about the ordering of components and selection of power supplies or to keep it in the open forum?
Regards
Al
I reached out to Mean Well. The guy confirmed that the UHP-500/700 (and probably others) do the pulse skipping thing at low loads, which we already knew.
However, he wasn't sure what the threshold was for it to enter into this state, so they're actually going to run some tests and get back to me.
Pretty remarkable support considering I'm inquiring about an order quantity of two. They seem like a solid company.
I'll report back here if/when he gets back to me. I hope the threshold is low enough such that I can get away with just the idle load of the amplifier boards, or maybe a few watts worth of resistor dummy load.
The UHP-750-36 might be a little shy of being able to handle the theoretical max power of two channels at 4 ohms, but like you said, if I do the phase reversal thing on one of the boards, it will realistically be fine for music...unless the music happens to contain full level sine waves on both channels at the same time that also happen to be 180 deg out of phase lol
There's also the UHP-1000, but that's getting up there in price.
If the UHP-750 turns out to be ill-suited for this, then I'll probably just run twin Antek AS-4225's and Power-686 boards. Or maybe one big 800VA one...depends on how I do the chassis layout.
What we found with the RPS-400 units is that the objectionable noise behavior is variable across the product family. The lower-output-voltage units didn't seem to have the noise issue, even when running in the low-current mode. The 36V units are noisy (electrically and audibly) at their nominal output voltage, but require only a small amount of loading to avoid the noise IF the output adjustment trim pot is set to full output voltage (about 39.5V). Full details are in one of Tom's thread's on this site but I don't remember which. As far as I know, there's no comment from Mean Well about this behavior, though I don't know if anyone has asked.
The Mean Well RPS-400-36(-C) is the whiny one. 50% of builders have had issues with audible whine both from the supply itself and from the speakers. In the 5-6 cases I've heard of, adjusting the voltage to 39-40 V has made the whine go away. That said, I cannot recommend the RPS-400-36 as I cannot guarantee that turning up the voltage will solve the issue for everybody.
The RPS-400-27(-C) that I used in the "Safe-n-sane" Modulus-686 build works great. I've gone through 10 of those supplies and none have whined.
I did ask Mean Well about this and they did respond. Basically they said to put a minimum load on the RPS-400-36. I forget the exact load they mentioned, but I think it was 10-20% of maximum load, so 1-2 A. It isn't specified in the data sheet.
Basically, I cannot recommend using the RPS-400-36(-C). Those who have them and have issues with audible whine, try turning up the output voltage. I can recommend the RPS-400-27-C.
Tom
The RPS-400-27(-C) that I used in the "Safe-n-sane" Modulus-686 build works great. I've gone through 10 of those supplies and none have whined.
I did ask Mean Well about this and they did respond. Basically they said to put a minimum load on the RPS-400-36. I forget the exact load they mentioned, but I think it was 10-20% of maximum load, so 1-2 A. It isn't specified in the data sheet.
Basically, I cannot recommend using the RPS-400-36(-C). Those who have them and have issues with audible whine, try turning up the output voltage. I can recommend the RPS-400-27-C.
Tom