A novel way to massively improve the SQ of computer audio streaming

[austinpop]

5 hours ago, str-1 said:

 

Thanks for doing that comparison. 

 

Sounds like I did right by ordering the SR4, which I have now learned is coming next week.  Choosing the SR4 ahead of the LPS-1.2 should (tonal signature aside) be a no-brainer for many in the UK given that the price difference is much smaller.  The LPS-1.2 from Vortexbox UK without mains cable is £423.80 (incl. vat and delivery), or £429.80 with mains cable, while the SR4 with silver dc cable is £450 all in.

 

Yup you Brits have a different equation working for you!

 

5 hours ago, str-1 said:

I am, however, wondering about the tonal signature of my system.  The Zenith SE is slightly warm and fuller sounding than the multiple small box system I had before.  The ISO Regen I have as one of the surviving small boxes is also slightly warm and full,  as too is the 70cm Lush USB cable I received two days ago (now with about 30 hours on it).  If the SR4 shares that tonal flavour, I might end up with a system that has too much of a good thing.  That said, I do listen to music almost exclusively through my Focal Utopia (Prion4) headphones out of what will soon be a Chord DAVE/Blu 2 combo, which may well stop things getting too warm.  I also plan to tryout the tx-USBultra very soon, which, from what I've read, should also help to redress the balance.  Only time will tell.

 

Both the DAVE and the Utopia are extremely revealing, and if you asked me, I wouldn't characterize either as warm. So I doubt you're in over-warm territory. Still, every system is different. Tonal balance is easier to correct if you find you've tilted too far in one directions, but I find resolution/micro- & macro-detail/texture/dimension to be the more elusive qualities to achieve. This is what the PSUs and clocks really deliver. And of course the Blu 2!

 

[Bruce Orr]

13 hours ago, austinpop said:

PSU Roundup: comparing the new LPS-1.2 with the LPS-1 and SR-4

 

I received my pair of new LPS-1.2s late last week, and have been burning them in ever since. They're now at about 100 hours, so I felt an initial review was in order, especially as I will be traveling for the next week or so. I'm going to revisit this again at about 200 hours, by when I hope to receive the loaner silver cable that @mozes is so graciously sending me.

 

My current chain with PSUs is:

• sCLK-EX modded switch (LPS-1) > Zenith SE > tX-USBultra (SR-4) > Codex DAC
• Energizing SMPSes are ground shunted
• Pangea AC-14SE MkII power cables to energizing supply and SR-4
• DC cables: Ghent DC01, homebrew Furutech FA 220 AWG14 with Oyaide 2.5mm plugs, Paul Hynes silver DC3FSXLR with Switchcraft 2. 1mm plug

Round 1: After 25 hours of LPS-1.2 burn-in

First the easy part. I swapped the LPS-1 powering the switch with the LPS-1.2, both with the Ghent cable, and it was no contest. The LPS-1.2 was better on every score. Bigger image, airier, deeper bass, and more dynamic. If you own and like the LPS-1, sell it immediately and buy an LPS-1.2.  It's a no-brainer.

 

Over on the tX-USBultra, I compared the LPS-1.2 (with the Furutech cable) with the SR-4 (with silver DC3FSXLR cable). This was a tougher challenge, and the LPS-1.2, while sounding excellent, seemed to have a slight tizziness and glare compared to the SR-4. I decided to let burn-in continue.

 

Round 2: After 100 hours of LPS-1.2 burn-in

For this round, the switch PSU was fixed at LPS-1.2 with Ghent cable, and I focused on the tX-USBultra PSU. Wow - that extra burn-in really helped! At this point, I would say the LPS-1.2 had pulled very close to the SR-4. If the SR-4 is a 100, the LPS-1.2 is a 95, with the LPS-1 being an 80. Let's break down this comparison in terms of the following attributes:

• Image size and air: essentially a wash. Both were equally good
• Dynamism: The SR-4 was just a smidge more dynamic
• Tonality: the SR-4 has slightly deeper bass, and the LPS-1.2, while by no means harsh, has a slightly brighter tonal signature. This may even be an advantage depending on the baseline tonal balance of the system. My own system benefits from the slightly darker, blacker, tone of the SR-4
• Noise: The SR-4 still has a slightly calmer and more relaxed presentation.

To put this in perspective, the SR-4 with the silver cable set me back close to ~$675, compared to the ~$500 for the LPS-1.2 plus Furutech cable.

 

Summary

In summary, the LPS-1.2 handily outperforms its predecessor, the LPS-1, and comes very, very close to the SQ of the much more expensive Paul Hynes SR-4. Highly recommended!

On 10/16/2017 at 4:02 PM, tapatrick said:

See here from the Grounding SMPS thread John mentioned with photos. I tried it on the input to my Aqvox switch and helped noticeably. Thanks once again John for another  simple and effective adjustment!

 

 

On 10/16/2017 at 4:02 PM, tapatrick said:

See here from the Grounding SMPS thread John mentioned with photos. I tried it on the input to my Aqvox switch and helped noticeably. Thanks once again John for another  simple and effective adjustment!

 

 

On 10/16/2017 at 4:02 PM, tapatrick said:

See here from the Grounding SMPS thread John mentioned with photos. I tried it on the input to my Aqvox switch and helped noticeably. Thanks once again John for another  simple and effective adjustment!

 

 

 

Thanks Rajiv -- I am not quite up to 100 hours yet but have already experienced same directional improvements in SQ with my two LSP-1.2s. Like you, one is driving a modded switch, and the other the tx-USBultra > Zeniith SE.  I replaced an sps-500 and LSP-1.  Great new product from Alex and the team at UpTone!  Moved the sps-500 to power my router

[Johnseye]

2 minutes ago, Bruce Orr said:

 

 

 

 

Thanks Rajiv -- I am not quite up to 100 hours yet but have already experienced same directional improvements in SQ with my two LSP-1.2s. Like you, one is driving a modded switch, and the other the tx-USBultra > Zeniith SE.  I replaced an sps-500 and LSP-1.  Great new product from Alex and the team at UpTone!  Moved the sps-500 to power my router

 

Bruce, have you found that you like the LPS-1.2 better than the sPS-500?  If so, could you explain how they compare?

 

Thanks

[Bruce Orr]

To be honest, I did not do separate A/B tests with each of the new 1.2s.  I am generally less a scientist that most on this thread. I changed out both and started listening. Generally, I was working on a general perception (unproven by me) thinking that the sps was better suited for driving the router, than the switch which was closer to the Zen/DAC.

[Confused]

16 hours ago, austinpop said:

 

Round 1: After 25 hours of LPS-1.2 burn-in

 

 

 

First the easy part. I swapped the LPS-1 powering the switch with the LPS-1.2, both with the Ghent cable, and it was no contest. The LPS-1.2 was better on every score. Bigger image, airier, deeper bass, and more dynamic. If you own and like the LPS-1, sell it immediately and buy an LPS-1.2.  It's a no-brainer.

 

Many thanks for yet another excellent write up Ragiv.  The comment above is crystal clear, but I would like to try to 'calibrate' your sound quality observations in my own mind.  To explain, we are talking here about sound quality changes resulting from comparing two similar power supplies to a network switch.  So I would expect any audible differences to be small, somewhere on the margins of perceptibility.  To be clear, I am not questioning your observations here, I am just trying to 'calibrate' your comments.  I know myself that describing observed sound quality differences in writing is a very tricky thing.

[Johnseye]

2 hours ago, Confused said:

Many thanks for yet another excellent write up Ragiv.  The comment above is crystal clear, but I would like to try to 'calibrate' your sound quality observations in my own mind.  To explain, we are talking here about sound quality changes resulting from comparing two similar power supplies to a network switch.  So I would expect any audible differences to be small, somewhere on the margins of perceptibility.  To be clear, I am not questioning your observations here, I am just trying to 'calibrate' your comments.  I know myself that describing observed sound quality differences in writing is a very tricky thing.

 

That is one of the most important points.  I've longed for a chart or breakdown that demonstrates the size of the impact to SQ, however as has been pointed out, this is not easy and can be subjective.  It can also differ by the two components being compared.  The difference in SQ between the two components can be significant, but when comparing one of those components to a completely different one the difference could be minute.  In this case we're referring to two LPSUs powering a switch.  There could be variability with the switch used; make, model, modded or not, etc.

 

I have heard a power supply make a significant impact by all standards.  Not as different when comparing speakers, but not on the margins of perception by any means.

[Confused]

@austinpop my apologies, i’ve Just noticed that I misspelt your name in my earlier post.  Shame on me!

[Confused]

26 minutes ago, Johnseye said:

 

That is one of the most important points.  I've longed for a chart or breakdown that demonstrates the size of the impact to SQ, however as has been pointed out, this is not easy and can be subjective.  It can also differ by the two components being compared.  The difference in SQ between the two components can be significant, but when comparing one of those components to a completely different one the difference could be minute.  In this case we're referring to two LPSUs powering a switch.  There could be variability with the switch used; make, model, modded or not, etc.

 

I have heard a power supply make a significant impact by all standards.  Not as different when comparing speakers, but not on the margins of perception by any means.

I was reading a speaker cable review a while ago that claimed the cable would make a 20% improvement in a $20k system.  Slightly bonkers.

[Johnseye]

Just now, Confused said:

I was reading a speaker cable review a while ago that claimed the cable would make a 20% improvement in a $20k system.  Slightly bonkers.

 

It probably does make an improvement, but not 20%.  What is a 20% improvement anyway?  Were they selling the cable for $4k

 

[Confused]

15 minutes ago, Johnseye said:

 

It probably does make an improvement, but not 20%.  What is a 20% improvement anyway?  Were they selling the cable for $4k

 

A 20% improvement is twice the sound quality increase that a 10% improvement would give you.  There’s a scale here somewhere!

 

You are near enough on price, by the way.

 

[lmitche]

3 hours ago, Johnseye said:

Not as different when comparing speakers

Try splitting your music server power across two power supplies and then we should talk about this once again.

 

Here is what one friend who tried it said:

 

"Complete blackness, vocals sound different than I've ever heard them before, there's a nuance in the voices that I've never heard. Incredible spatial separation both side to side and in depth and it seems that each instrument is perfectly marked out in space. Tonal balance is completely natural, to use a cliché very analog sounding."

 

I can't say it better myself.

[Johnseye]

41 minutes ago, lmitche said:

Try splitting your music server power across two power supplies and then we should talk about this once again.

 

Here is what one friend who tried it said:

 

"Complete blackness, vocals sound different than I've ever heard them before, there's a nuance in the voices that I've never heard. Incredible spatial separation both side to side and in depth and it seems that each instrument is perfectly marked out in space. Tonal balance is completely natural, to use a cliché very analog sounding."

 

I can't say it better myself.

 

It is split across two power supplies, the SR7 and LPS-1.  It's actually 3 if we're counting rails.  Two rails from the SR7 and one from the LPS-1.  I know I could split things up even further.  I think we both agree that the impact a power supply can have is significant, but I think speakers are the biggest factor in sound quality and signature.

 

The difference in sound that you hear from sitting in front of a pair of Personas, Sopras, Sabrinas, Legacy Focus, Vandersteen, Martin Logan, B&W, Maggies, the list goes on, is more significant than comparing any other change you could make in audio.  These are all sound signature changes.  All the speakers I mentioned are excellent at a certain price point, but they all sound noticeably different.  It is so pronounced, there is no question when side by side listening between these speakers.  I only named a few because I was auditioning those.  There are so many more, each with their own personalities.

 

I believe Rajiv said you were coming to Axpona.  You're welcome to join him at my place on Sunday for some listening and equipment comparisons.  Whether you're able to make it or not, it would be great to meet you at the event.

[lmitche]

14 minutes ago, Johnseye said:

 

It is split across two power supplies, the SR7 and LPS-1.  It's actually 3 if we're counting rails.  Two rails from the SR7 and one from the LPS-1.  I know I could split things up even further.  I think we both agree that the impact a power supply can have is significant, but I think speakers are the biggest factor in sound quality and signature.

 

The difference in sound that you hear from sitting in front of a pair of Personas, Sopras, Sabrinas, Legacy Focus, Vandersteen, Martin Logan, B&W, Maggies, the list goes on, is more significant than comparing any other change you could make in audio.  These are all sound signature changes.  All the speakers I mentioned are excellent at a certain price point, but they all sound noticeably different.  It is so pronounced, there is no question when side by side listening between these speakers.  I only named a few because I was auditioning those.  There are so many more, each with their own personalities.

 

I believe Rajiv said you were coming to Axpona.  You're welcome to join him at my place on Sunday for some listening and equipment comparisons.  Whether you're able to make it or not, it would be great to meet you at the event.

I should have been more precise, I meant splitting the motherboard power. Have you done that?

 

The quote above is from a friend in Chicago. A second friend, also in Chicago, is waiting for his power supply to arrive from China to split his motherboard power.

 

I'll introduce you to them at Axpona. As it happens, they both maintain duplicates of my music server, so perhaps you can hear what I have done here, locally in Chicago.

 

For me, I am a source first guy and always recommend that people put money there first. However, after many years I bought a pair of used Wilson speakers, they have been refoamed once, and don't feel a need to audition speakers any more.  Setup properly, the Wilsons are just terrific, with a huge soundstage both laterally and vertically, terrific extension, and excellent imaging. At 175 lbs each, as long as I don't need to move them, I really enjoy them.

 

Many thanks for the invitation. I look forward to meeting you in person.

[KingRex]

5 hours ago, Confused said:

I was reading a speaker cable review a while ago that claimed the cable would make a 20% improvement in a $20k system.  Slightly bonkers.

I just splurged for the Inakustik LS2404 Ref Air speaker cable.  It had a dramatic impact on my speakers.  It was no small margin.  When I went to do the A/B, I was bracing to have a hard listen. I assumed it would be tough to tell the differences as I had a pretty decent set of Graditeck cable.  After burning my new Inakustik for 40 hours I put my old cables back in and queue up some James Taylor, Then Cat Stevens, Then Chopin.  After playing a few tracks I put the Inakustik back in.  It was all of 2 seconds and I realized this was a dramatically better set of cables.  They took my speakers forward in performance no small degree.  The extension of highs was breathtaking.  Just so much more realism in the guitar strings.  They really sang like a live guitar.  The decay on piano notes via Chopin Nocturnes was far more real.  It really sounded more like a piano.  The bass was fast and taut, full and clean.  Could not ask for more.  The volume was also notably higher.  I had to lower by 2 clicks which is 3 db on my preamp.   Can I ascribe a number like 20% gain.  I don't know, but I was truly amazed how a good cable made that much difference in the performance of my speakers.  My stereo is a $35K system.

[mozes]

2 hours ago, lmitche said:

should have been more precise, I meant splitting the motherboard power. Have you done that?

Hi Larry

Can you pls elaborate a bit on how you can do that. In my case I have a Jetway NF596 motherboard that I feed 12V via a DC2.5mm plug.

[seeteeyou]

We need a motherboard with both 4-pin/8-pin CPU connectors and 24-pin ATX connectors in order to split the motherboard power. Most likely we won't find both sets of connectors on motherboards with NUC / 3.5" SBC / Nano-ITX / Pico-ITX form factors. Usually Both ATX and Micro-ATX should be good to go due to the size, though Mini-ITX would be more like a hit or miss depending on the implementation.

 

Supermicro X10SBA-L should be somewhat "unique" in a way that even its 4-pin CPU connector could provide power to the entire motherboard, though we could still feed both 4-pin CPU connector and 24-pin ATX connector simultaneously like what Innuos did.

[lmitche]

25 minutes ago, mozes said:

Hi Larry

Can you pls elaborate a bit on how you can do that. In my case I have a Jetway NF596 motherboard that I feed 12V via a DC2.5mm plug.

Hi Mozes,

 

Your motherboard has two 12 volt DC inputs but they cannot be used simultaneously, so it can't be done with the Jetway.

 

The Supermicro motherboard used in the Innuos machine can take two DC inputs simultaneously. In my experience, it is the only embedded cpu motherboard that can do so, but I've not researched other options.

 

My Asus motherboard supports an i7-6700k which is usually positioned at the opposite end of the computing spectrum. Typically these motherboards have a big 24 pin atx connector supplying multiple voltages, and a 8 pin molex connector supplying 12 volts direct to the cpu. I am powering these two inputs from two different power supplies, each sized according to current demand.

[seeteeyou]

BTW, the "Holy Grail" would be providing separate power to feed specific voltages to each and every input separately

• -12V
• +12V
• +5Vsb
• +5V
• +3.3V
• 1.35V for DDR3L RAM
• 1V (or so) for the processor

Obviously both 1.35V and 1V (or so) would require modifications while I don't even know what kinda processor / linear PSU combo would actually work. Let's say we're disabling 1 outta 2 cores of i3-7350K (4.2GHz with 60W TDP) while aggressively underclocking that with a single digit clock multiplier. How many watts do we actually need then? We're talking about only 1V (or so) here so the current draw would be 5A for 5 watts and then 10A for 10 watts etc. Maybe we could try something like AMB σ11 but they said that's only practical for 5V

 

https://www.amb.org/forum/sigma11-at-below-5v-t2267.html

[rickca]

37 minutes ago, seeteeyou said:

BTW, the "Holy Grail" would be providing separate power to feed specific voltages to each and every input separately

Is there no such thing as good sounding switching regulators that could be used on a motherboard?

I think we need better motherboards designed for audio because this is getting out of hand.

[austinpop]

10 hours ago, Confused said:

Many thanks for yet another excellent write up Ragiv.  The comment above is crystal clear, but I would like to try to 'calibrate' your sound quality observations in my own mind.  To explain, we are talking here about sound quality changes resulting from comparing two similar power supplies to a network switch.  So I would expect any audible differences to be small, somewhere on the margins of perceptibility.  To be clear, I am not questioning your observations here, I am just trying to 'calibrate' your comments.  I know myself that describing observed sound quality differences in writing is a very tricky thing.

 

Not sure what you're asking here. You'll recall that many people who've bought the trifecta have been astounded by the improvement due to the reclocked switch. Even now, with my Zenith SE chain, the switch makes a profound improvement on streamed music - i.e. Tidal, or music stored on my NAS.

 

The effect of the switch is by no means marginal. It is startling. I know that's not what you're asking. But once you have a component that makes a significant sonic improvement, it's not hard to grasp that changes to the PSU powering the component can be quite startling too.

 

So bottom line, moving from LPS-1 to LPS-1.2 to power the switch, the SQ improvement was substantial, not marginal.

[Johnseye]

1 hour ago, lmitche said:

Hi Mozes,

 

Your motherboard has two 12 volt DC inputs but they cannot be used simultaneously, so it can't be done with the Jetway.

 

The Supermicro motherboard used in the Innuos machine can take two DC inputs simultaneously. In my experience, it is the only embedded cpu motherboard that can do so, but I've not researched other options.

 

My Asus motherboard supports an i7-6700k which is usually positioned at the opposite end of the computing spectrum. Typically these motherboards have a big 24 pin atx connector supplying multiple voltages, and a 8 pin molex connector supplying 12 volts direct to the cpu. I am powering these two inputs from two different power supplies, each sized according to current demand.

 

That I am not doing. My separate supplies are powering the board, drive and pcie card. I'll have to look into two supplies for a board itself. Next time. 

[seeteeyou]

28 minutes ago, rickca said:

Is there no such thing as good sounding switching regulators that could be used on a motherboard?

 

Well, we do need some switching regulators for specific locations but usually we could go for linear regulators (especially LT3045) if the current draw weren't too high

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=241&tab=comments#comment-773624

On 1/31/2018 at 6:50 PM, romaz said:

Of interest, he found that in some locations on the board, switching regulators were better than linear regulators because they had faster response (I presume he's talking about the regulator for the CPU itself).

 

Anyways, replacing existing regulators with LT3045 would require quite a bit of work unless we wanna pay DFI $1,500 to build one of those custom-made motherboards. Other than that, our best bet might be designing our own carrier boards for the usual suspects

• COM Express Type 7
• Intel Compute Card
• SolidRun SolidPC Q4

[rickca]

Antipodes is working with Asus on customizing a motherboard for their products.  I wonder if they are just tuning or really trying something innovative.

[seeteeyou]

There's something else from Supermicro that might be worth a look. X10DAi should cost $300 or so now

 

https://www.supermicro.com/products/motherboard/xeon/c600/x10dai.cfm

 

Engineering sample version of Xeon E5-2628L v4 could be found on eBay / TaoBao for roughly $150 a piece

 

https://ark.intel.com/products/91775/Intel-Xeon-Processor-E5-2628L-v4-30M-Cache-1_90-GHz
https://www.ebay.com/itm/06-Intel-Xeon-E5-2628L-v4-ES-QHV8-1-5GHz-12C-LGA2011-3-Compatible-X99-i7-6950X/152816793229

https://item.taobao.com/item.htm?id=564370836490

 

According to the block diagram of X10DAi, CPU2 on the right is totally NOT bored by anything else while CPU1 on the left is connected to whole bunch of stuff that could be "ignored" if you will

 

https://www.supermicro.com/manuals/motherboard/C612/MNL-1563.pdf#page=19

 

Not sure about assigning a specific RAM module for a particular process / program under Windows, though we do have that option for a specific processor core as shown below

 

Help! My Application only runs on a Single Processor system!

https://blogs.technet.microsoft.com/askperf/2009/02/03/help-my-application-only-runs-on-a-single-processor-system/

 

Tip: Assign a Program or Process to a Specific CPU running on Windows 7

https://technet.microsoft.com/en-us/library/ee851672.aspx

 

How to Set Affinity for an Application on CPU in Windows 10

http://www.howto-connect.com/set-affinity-for-an-application-on-cpu-in-windows-10/

 

 

In other words, why don't we leave any "useless" processes / programs for CPU1 while CPU2 could be dedicated for "useful" audio-related stuff? Similarly, we could install up to 3 pieces of tX-USBexp (1 for USB DAC / 1 for storage / 1 for Ethernet) that are directly connected to CPU2 while the rest of the motherboard might not really matter anymore. Alternatively, HQPlayer could use one of the processors for upsampling while everything else could use another processor.

 

Everything that's connected to the PCH could be disabled in BIOS so that Windows wouldn't even "see" them. If we don't even need to access all 12 cores of Xeon E5-2826L v4, only enable as many cores as we actually want.

 

Most likely we could find something similar from other manufacturers but not all of them would provide the block diagram like Supermicro did.

 

1 hour ago, rickca said:

Antipodes is working with Asus on customizing a motherboard for their products.  I wonder if they are just tuning or really trying something innovative.

 

That's definitely something to look into since their DX Gen 3 seemed to be superior to (i.e. with a much heftier price tag) the ultraRendu + LPS-1 combo

 

https://www.computeraudiophile.com/forums/topic/38528-antipodes-dx-gen3-vs-innuos-zenith-mkii-se/?tab=comments#comment-786885

[Cornan]

17 hours ago, austinpop said:

 

Hi Micael,

 

i experimented with JSGT on the feeder Meanwell supplies a while back, and found it to have a small benefit - even powering other devices than a switch. As for the LPS-1.2, they have ground shunting built in. 

 

Bottom line - in my system, ground shunting has a very minor effect on SQ.

 

I’ll be interested to see how your experiments go!

 

The JSGT can both give some but also take some away in my setup as well.

 

I am looking forward to give the LPS-1.2 a try. I have also got another interesting power supply in the pipeline that will be a challenge for both my series Stammheim 3A/LT3045 and the LPS-1.2. It is a full blown 0-15v 5A/LT3045 power supply (totally 10 pcs LT3045 boards) with PI-filter (digital filter) at the output. 

I do not even have a complete setup to listen at the moment. I expect it to be set & done by this weekend though. I had to move my setup to another location in my appartment to make room for my brand new Sony A1 55" OLED TV. After hours of fine-tuning it and giving it proper clean power (DC blocker>Peaktech2240>AVR>Sony A1) I was really impressed with this TV for both the brilliant picture and great sound.

 

Quote

Pi filters and T filters

Pi filters are basically one inductor surrounded by two capacitors and arranged like the Greek letter Pi. The input capacitor is selected to offer low reactance and repel the majority of the nuisance frequencies or bands to block. Its inverse, the T filter uses two shunt inductors and a coupling capacitor. These single-stage filters can act as low pass, high pass, band pass, and band stop.

Pi filters present very-low impedances at high frequencies at both ends due to the capacitive shunting. T filters conversely have very-high impedances at high frequencies because of the inductive coupling (Figure 2).

Figure 2: Low-pass Pi filters (left) can tap into an RF-transmission path allowing just the lower frequencies to pass. Conversely, high-pass T filters (right) block the lower bands and allow the upper frequencies to pass. These parts can be used to both optimize each band’s performance and allow full-duplex operations.

An RF transceiver may use a T filter to block a shared or competing band while using a Pi filter to clean up and pass the desired frequencies. In either case proper selection of component values can match impedances on both sides to maximize power transfer between active stages, switches, and antennas.