Millett LR + TP Opus + AMB α24 reboot

What people are working on at the moment
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#121 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

PCB done...

Image
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#122 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

Now to wire it all together. The step-up transformers look like they are on stilts because they are plugged into component sockets, rather than soldered directly to the circuit board. I'm intending to buy Dom Harper's cartridge next year, and it has less than half the output voltage of the VdH, so will need a 1:16 step-up rather than a 1:8. So I also need to be able to exchange the LL1931's for LL1941's as well.

Image
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#123 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

Well, it's complete and currently up and running in the system, and sounding quite good, although it does seem to be a little rolled off on top with some recordings.

Of most concern though, is the fact there is a significant amount of background hiss, accompanied by a high-pitched whistle. At low-ish to medium listening levels it *just* disappears below the level of audibility from the listening position, but can be easily heard close to the speaker. This is obviously going to be adding significant intermodulation distortion to the top end, possibly this is why it sounds rolled off?

I can only think that this is coming from the SilentSwitcher - I have tried turning off the DacMagic and music server to see if that made any difference, sadly no effect. I remember when I powered it up on the bench to test it, I could hear a very soft whistle coming out of the board itself.

This is rather disappointing and somewhat at odds with the good things I had read about this product, notably a few people had used it for their Millett LR phono builds and had sung its praises, which is why I selected it for the job. As you can see from the picture, it's not right up next to the other boards, but despite this I think electromagnetic interference is finding its way into the inductors or step-up transformers.

I'm going to see if I can locate it a little further away in the chassis to see if that makes a difference.

EDIT - on reflection maybe I'm being a little hard on the SilentSwitcher - those step up transformers are dealing with microvolt signals, and thinking about it, the other builds I saw using the Silentswitcher didn't have the step-ups in the chassis. I bet if I shifted it to the other end of the chassis, that would help alleviate things, so tomorrow I'm going to get it back on the bench and play with the orientation etc.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Nick
Site Admin
Posts: 15707
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#124 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Nick »

Also may be worth putting the supply in a shielding can?
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Nick
Site Admin
Posts: 15707
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#125 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Nick »

To me the idea of using a USB plug to power a phono stage seems like a recipe for disaster. I would expect it to allow the ingress of all sorts of noise. Also the step ups on a PCB will remove the option of using shielded cable or twisted pairs to connect it all together.

TBH, with the ferrite cores in the phono I would think about putting the entire board into a shielded can with maybe some mu metal.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#126 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

Well it's getting its power from the battery packs in the preamp via an XLR4, so the incoming voltage to the switcher should be as clean as it's possible to get. The switcher itself is supposed to have only a few microvolts of noise on its output, but clearly it must be throwing out some amount of HF electromagnetic noise of its own, why else would my phonostage have a heterodyne whistle? I'd rather have hum than this.

To be honest, I'm at the stage where I'm starting to wonder if this whole expedition into switched supplies hasn't just been pure folly. Maybe I've made some fundamental error in my implementation, but if someone shows me a tiny PCB that is supposed to have a fraction of the noise output of a typical conventional linear supply, along with the related savings in space, transformer cost, having to muck about with safety grounding and all the rest of it, then yes I'm going to give it a go.

Thing is - at this point I'm just not convinced that it sounds very good, at least when it's powering the stuff I have, and again maybe it's down to the way I've implemented it because a lot of other people have had nothing but success - the guy behind the Paradise phono sang its praises. I've been telling myself that it's my imagination, but I just cannot help feeling that the linestage took on an edgy, grey, slightly flat quality since I changed the power supply over. Before, it was a faultlessly transparent conduit for my sources to whatever power amp I was using at the time.

Maybe this is a lesson that there are things other than low noise that affect the influence a power supply has on the sound of a particular circuit. The sigma 22 power supply that I used before was noisier by a factor of 50, and yet I'm seriously considering not only reinstating that, but also getting another for the phono stage (which was my original intent). Either that or powering the circuits from batteries directly rather than going through a switcher.

Time for some pondering.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Nick
Site Admin
Posts: 15707
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#127 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Nick »

why else would my phonostage have a heterodyne whistle
You have op amps and inductors, I can think of any number of reasons. I would try measuring and see where the tone comes from.

Without knowing where its from you may be chasing your tail.
so the incoming voltage to the switcher should be as clean as it's possible to get
And it probably is, but without checking its a guess.
because a lot of other people have had nothing but success
There may be an important internet based lesson here.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#128 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

Thanks Nick. I suspect you're right about the internet based lesson.

Despite the noise we were listening to a few records with it last night and Meredith commented that she really likes the sound of it, so it might have legs. To me, on the Ana Moura LP it sounded like all the top end had vanished on her voice (which is quite close miked), but the effect didn't seem apparent with other LP's, which seemed rather odd.

Other than that it seems to make sense of the music in a nice way, but given I'm not 100% happy with the linestage at the moment, it's tricky making a call on it.

I might take a look later today but Meredith and I had our third Covid jabs yesterday, and unlike the others, this one hit us hard - we were both knocked out with a fever and aches overnight and this morning. I'm inclined to just take a day's break from the thing.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#129 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

I figured there was at least no harm in trying it out with a decent linear power supply, given that I had a spare one knocking about which was configured to the correct voltage.

So today was my first Cardboard Box Chassis moment. I grabbed the AMB Sigma 22 PCB that had previously powered the linestage, along with its transformer, and enclosed both in a tastefully pink box that Meredith had ordered something in, and which happened to be about the right size.

Image

I've just set it up in the system and applied power (after having reset the DC offset again, it does seem to drift a bit). Turning up the volume to max, there were absolutely no horrible whines or whistles, just a gentle hiss and slight hum, the hum being no louder than it was before despite there being a mains transformer in closer proximity. I've left it powered on to warm up a bit until I get a chance to listen to some LP's, which I think is going to involve prying Meredith away from the Robson Green thing she's been watching all afternoon. So watch this space to see if it actually sounds any good.

However, that does seem to clearly indicate where the noise has been originating - my suspicions were first aroused when I could actually hear a similar noise coming from the switcher board itself when power was applied. This in turn has thrown the whole battery operation project into doubt - the whole point of the exercise was to supply the phono stage with the quietest power possible, isolated from the mains. Clearly the SilentSwitcher has not turned out to quite be the panacea that I thought it would be.

I really wanted this to succeed, and it may well be that this still a good product for use in something like a DAC where the voltages are hundreds of times higher, but a phono stage? I'm not sure that I can recommend it for that application, certainly not if there are inductors or other iron nearby. The board is supposed to have been designed to minimise EM interference but it's clearly still significant enough to cause issues of its own in the wrong environment.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#130 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Ray P »

I'll declare up front that I've had a good experience with a silent switcher and perhaps I'm therefore a tad defensive but it appears to me that the conclusion that the internet learning thing was too easily concluded - has any consideration gone into the possibility that the Slent Switcher in question might be faulty or that the implementation was problematic? Why not reach out to Jan Didden, he's very approachable and knows his stuff?

https://www.diyaudio.com/community/thre ... er.297147/
Sorry, I couldn't resist!
User avatar
Ray P
No idea why I do this anymore
Posts: 6294
Joined: Thu Nov 22, 2007 5:18 pm
Location: Somerset

#131 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Ray P »

BTW, did you try the Silent Switcher in the pink box too?
Sorry, I couldn't resist!
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#132 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

(All valid points Ray, I'll pick them up later on).

But for now ... drum roll... it's review time! :D

So, what about the LR phono - the all-important question, how does it stand up against the Whest PS30RDT that itself is widely regarded as a reference? Pete himself doesn't get drawn into the debate about whether LR stages are superior (see page 2 of his ETF13 presentation) whereas the Whest company crow from the rooftops that they make the best phonostages in the world (and charge accordingly).

Well, for starters, the LR is not as quiet - the Whest has been engineered to have an almost vanishingly low level of background noise, so to a certain extent I've been rather spoiled in that regard. However, whilst I don't currently have one to compare it against, from memory I don't think it's any worse (in fact I suspect it's better) than a lot of valve designs, and it's not at a level that is in any way intrusive. But if the output from your cartridge is on the low side or really black... BLACK! BLACK! You lock me in the cellar and feed me pins! ... ahem, sorry ... black backgrounds are your thing, then maybe look elsewhere.

I did read one online criticism regarding the choice of op-amp - there are alternatives to the AD797 that have lower current-noise specs, maybe that could improve the noise performance? But I suspect that Pete chose the AD797 for a reason and I'm not going to second-guess him.

The second problem stems (I think) from the lack of capacitors in the signal path - it has a lot of DC-coupled gain, so there are trim pots on the first set of opamps to null out the offset. I have read stories where some samples of AD797 cannot even be nulled to zero, but the first four I plugged in (out of the six that I bought) were fine. It's important to have the inputs shorted out (or a cartridge / MC transformer connected) when setting the null point.

Initially there was evidence to suggest that my initial null setting drifted off, as there were pops from the speakers when I clicked the volume up and down. I re-adjusted the offset before plugging it back into the system last night, and the pops were much more subdued, so I think things have now stabilised a bit. The AMB linestage doesn't mind DC offset on the source, so performance is not affected.

When we were listening a few nights ago with the switched supply, despite the noise injection Meredith commented that the sound seemed smoother and more 'real' than the Whest. All I could focus on was the 7kHz carrier frequency, but it did initially sound quite musical although lacking some air and sparkle compared to what I was used to.

Last night we listened with the stage powered by my AMB σ22 regulated power supply, which formerly saw duty on my linestage circuit (and likely will do so again). Tech highlights here, it's a seriously good piece of kit.

Initial impressions with music I was familiar with - it certainly didn't sound the same as the Whest. Again, I felt at first that some air and sparkle was missing, it sounded darker, although conversely there seemed to be more detail coming through in the midrange. Bass was fantastic - really tuneful and tight. However, it became clear the longer we listened, that my perceived sparkle shortage was in fact a reduction in artificial-sounding sibilance. The crunch came when we put on records that I wasn't as familiar with, because I didn't have any pre-existing expectations of how they should sound - and they all just sounded, how can I put this - right. The music just seems to hang together better, it makes the other stage sound disjointed by comparison. My impressions seem to tie in with others who've heard LR or LCR stages.

My word is it musical. I ended up listening to LP's until about 3 or 4 in the morning, which is always a good sign that something is going in the right direction.

So this is a very pleasing result - I've heard the Whest before up against a number of other very capable stages, and it held the bar pretty high. The only stage I've heard that put any real distance ahead was Nick's reference stage, and it's near-certain that the Millett stage still wouldn't touch what that can do.

What helps - really good step-up iron, the best you can afford. I'm thoroughly sold on Lundahl after using one of their line input transformers on the 300B. Also the power supply - power supplies do matter, and it worked so brilliantly with the AMB board, that I've now decided that that's how I'm going to power it - the supply will have to be in a separate chassis of course.

This is a really good route into the world of inductor-based phono stages without the usual cost and grief. Some components may present a challenge though - it uses a model of tapped inductor that's made by Cinemag and the experience of ordering direct from them outside America is reported to be... variable (in the end I got a friend of Meredith's to buy them for me and ship them over). They can be bought from Don Audio, but the price ends up being about double what Cinemag charge.

The other problematical component right now is the AD797ANZ op-amp - for most of this year they've been unobtainium from all of the proper sources, there is fresh stock now but the price seems to have rocketed even higher (not quite sure why they are so expensive). I eventually got mine from a UK-based eBay seller who I questioned thoroughly as to their provenance, eventually I concluded he and the product were genuine. Do NOT buy from China - they WILL be fake.

The latest word is that Pete has recently updated the design to use Lundahl inductors, but I don't know if it's using different op-amps - the details are not in the wild yet.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
Ant
Shed dweller
Posts: 2332
Joined: Mon Jul 31, 2017 6:45 pm
Location: Yorkshire

#133 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Ant »

Sounds like it does the 'thing' that the cole does which makes the thing sound right despite its shortcomings
Shame that the silentswitcher bit has its issues and that another box is required for the psu, but its all good fun
Also starring Rex Hamilton as Abraham Lincoln

www.bte-designs.weebly.com
User avatar
Thermionic Idler
Old Hand
Posts: 1018
Joined: Sun Apr 22, 2012 8:22 pm
Location: Southsea

#134 Re: Millett LR + TP Opus + AMB α24 reboot

Post by Thermionic Idler »

Ray P wrote: Thu Dec 30, 2021 7:27 pm I'll declare up front that I've had a good experience with a silent switcher and perhaps I'm therefore a tad defensive but it appears to me that the conclusion that the internet learning thing was too easily concluded - has any consideration gone into the possibility that the Slent Switcher in question might be faulty or that the implementation was problematic?
Well, there is certainly that possibility, although all the output voltages were bang on and rock-steady, so if any part was failing, I'd anticipate some effect being seen on the output... possibly? I have no idea when it comes to SMPS engineering, and like most mere mortals I don't possess measuring equipment sensitive enough to verify whether the output noise levels are within spec - I have to just take it on trust. I would assume these undergo some kind of testing before they leave the factory - wouldn't they?

I do have three of them so my plan is, at some point, to do a bench test on all three to see if they exhibit the same behaviour. They will certainly be kept for use in future projects, for example I'm still very likely to try it out for the Opus DAC, although that can has been kicked even further down the road now (see below).

With regard to the implementation, I seem to recall reading a post by Jan that the silent switcher pcb shouldn't be located too far away from what it supplies power to, and I'd seen pictures of other builds of Pete's phono stage where the pcb was a lot closer to the signal board than where I had mine positioned. However, I don't think any of these had moving coil step-up iron in them, so it may well be that step-ups don't like it - we all know the signal levels going through that point are less than one thousanth of the voltages that you find flowing around a typical DAC.

I realise I could have tried positioning it further away in the chassis, but I was still uncomfortable with the continuing presence of something that could cause that kind of noise issue, and was just curious to see what it would be like with the old linear supply out of the linestage - it was a pretty easy thing to knock up in an afternoon. As it happens, it sounded absolutely spectacular, head and shoulders above the Whest.

How much difference that power supply made to the performance, I've no idea. But the interesting thing is that it was originally designed to power AMB's small β22 power amplifier, so it's quite a bit over-capacity for a phono stage that only consumes about 50mA. However, it did make me think of the principle behind the NVA Phono 1 and 2, which (correct me if I'm wrong) are pretty much the same RIAA circuit, just that the Phono 2 has a much bigger power supply in a separate chassis which is hugely overspec (something like 240va across the two channels?). I know he's no longer around to give me a good slap for accidentally misquoting him (and I know he wouldn't like the step-ups), but I seem to remember Richard saying that every time he made the power supply bigger, the better it sounded, and there didn't seem to be an end point (didn't he make a BBPS - big bastard power supply - version?). They both use op-amps, but nobody seems to have tried the same trick with Pete's phono stage - they either use batteries, the SilentSwitcher or Pete's own supply which looks a bit 'meh' to me. So I'm curious to see what will happen.

Sorry, this has turned into another ramble, but I think where I'm getting to here is that I was so impressed by how it sounded with that supply that that's the route I want to follow now - my plan is to follow Richard's example and build up another sigma22 power supply ("borrowing" some parts of the case I had made for the DAC project), with higher quality electrolytics and a bigger power transformer on the front of it.

Along with this, I'll be restoring the previous power scheme to the AMB linestage - the sound definitely didn't improve after the low voltage conversion, in fact I'm quite certain it degraded slightly - over time, I've just got the feeling that things became a bit more grey and flat sounding (although still good enough to assess the change in phono stage), and there's sometimes a slight shriekiness in the upper mids that I don't think was there before. And I'm fine with that, I don't think it's indicative of a fault or poor performance on the part of the SilentSwitcher, just that in this particular application, the linear supplies simply sounded better to me, even though they are theoretically noisier (same could be said of belt drive versus idler drive turntables).

Added to this of course are the reliability issues I've had with the control software since I made the change - I've taken on board Nick's point about the reset button track on the PCB acting as an aerial, but the fact is I had four years of faultless operation when it had its own dedicated linear supply. Its latest party piece is flipping the volume down to minimum and staying there if you happen to touch the control pot on the front panel. Lovely if you want to drop the volume momentarily when cueing up an LP. Thankfully the mute button function still works.
Last edited by Thermionic Idler on Fri Dec 31, 2021 5:21 pm, edited 1 time in total.
Deck: Garrard 301 - Audio Origami PU7 - NW Analogue DH3S
Phono: Pete Millett LR, Lundahl SUT, AMB σ22 PSU
Linestage: Bruno Putzeys Balanced Pre + Hans Polak relay mod, LCDuino control
Power amps: 300B PP 'Symmetric Reichert'
Speakers: FF225WK Big Mets
User avatar
jack
Thermionic Monk Status
Posts: 5493
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#135 Re: Millett LR + TP Opus + AMB α24 reboot

Post by jack »

Just thoughts:

Put a ferrite on the psu outputs , try mu-metal screening? If you've a bit of blank fr4 make a box out of that for the switcher and use feed-through caps etc.

Check your earth paths and if there is any noise on the earth - a noisy earth can inject all sorts of crap. Actually some of the problems sound like a flakey earth.

I.e. take a few ideas from the RF cookbook.

Make sure all unused inputs are shorted. Scope the power rails - note scoping tiny voltages in an earth-return-sensitive environment is an art in itself.

Follow the signal through the circuit with the scope. See where the noise is being injected.

Is anything oscillating? Such oscillations are generally at several MHz so a good analog scope really helps.

Does it vary with the volume setting? Is it only one type of noise?

The PSSR values for modern opamps are generally very good, do I'd be surprised if they were the issue.

Change only one thing at a time.
Last edited by jack on Fri Dec 31, 2021 6:32 pm, edited 5 times in total.
Vivitur ingenio, caetera mortis erunt
Post Reply