View Full Version : Preamp to Poweramp impedance matching
Hadn't really thought much about the above, up till now any preamp I've had has one set of XLR outputs to the power amp.
Changed my Preamp recently to a Holo Audio Serene KTE, which has TWO XLR outputs.
Both run the same output voltage, the difference being one is a low but 'normal' 40 Ohms impedance, the other though is an extremely low 3 Ohms, and allows the amp to be run as a headphone amplifier as well as a normal preamp.
Cool.
My original intention was to use the low impedance XLR output with a converter cable to let me use the Serene with my Headphones?
I'm reconsidering that plan though!
Last night I watched a random YouTube video from Paul McGowan where someone asked him what to look out for when matching up a Pre and Power amp from different manufacturers?
He spoke of the "10x Rule" of impedance matching, I hadn't heard of this before and got me curious.
He advised the Power amp's INPUT impedance be at least Ten times the OUTPUT impedance of the Pre Amp.
i.e. Preamp_ImpOUT (a) < Poweramp_ImpIN (10 x a)
BUT... he said Ten times was an arbitrary ammount - Only a good baseline - and that "the bigger the difference the better" as more of your preamps voltage (and effectively the musical signal put out) would get through to the power amp and onto the speakers.
My Preamp...
Has two XLR outputs, was using the 40 Ohm one as the lower impedance 3 Ohm one was reserved for headphones.
My Meridian 557 power amps input impedance is 32K Ohms.
A big difference in the impedance delta then between the Serene's two different XLR outputs?
So hang about says I...
McGowan suggested the bigger the difference between Preamp output impedance and Power Amp Input impedance the better...
Why dont I try feeding the big Meridian with the lower 3 Ohm XLR output, as the difference then would be substantially bigger.
:eek: Colour me shocked.
Effortless and unforced.
Real transparency, realism and impact now, noticably more than before running it from the 40 Ohm output.
The Serene Pre is known as an "open window" to what your system is doing and Im really hearing that now.
Went back and forth a few times to verify it, yup, its a very obvious improvment.
So my headphone plans might change now for 2026:
Considering using the low inpedance XLR outputs of my Preamp for normal listening through the speakers, and add a dedicated balanced headphone amp to my system?
ALTHOUGH:
I could just swap the cables out when listening to headphones?
Easy access to the rear panel, takes 2 mins to swap them out and I would get access to that amazing sound from the Serene KTE preamp.
I would guess I'd need to spend big to get a headphone amp to noticably beat it, and headphones would only be used for occasional night listening so not a priority purchase.
One less box to faff around with and money saved too.
Just wondering if anyone else has played around with impedance matching between Pre and Power, and a recommendation as well to give it some thought if shopping for a new Pre or new Power amp, have a think about the matching impedances it can really make an appreciable difference.
I'm surprised you didn't know this. Like you say it can make a massive difference.
Most of the time though anything not boutique, antique or valve is compatible.
Ten times the input to the output is really the bare minimum. Also it's not a constant, input and output impedance both vary with frequency so matching on paper is not always watertight.
The other one is voltage output and input sensitivity. You don't want to clip the input - remember how some people thought CD was harsh in the early days?
I'm surprised a simple increase in input/output impedance ratio should make such a difference. We're talking about a change of ratio of 800x to one of approximately 11,000x. There must be more to it, as a ratio of 800 ought to be quite sufficient. There must be some additional circuitry involved with the 3 Ohm headphone output.
In my system the preamp has a measured output impedance of 0.015 Ohm, and the power amplifiers have an input impedance of 14kOhm (in parallel with 50H), both via balanced-line XLR connectors.
Barry could you give me an estimate of a preamp output impedance that would suit my Croft series 7 power amp which has an input impedance of 470kohm?
Lawrence001
02-01-2026, 18:14
I'm with Barry on this, if the 3 ohm output was so much better in every case, why not just make both pairs of outputs 3 ohm?
I'm with Barry on this, if the 3 ohm output was so much better in every case, why not just make both pairs of outputs 3 ohm?
I think one of the outputs was different so it would be suitable to drive headphones.
I surprise myself a lot by how much I dont know. Daily. :lol:
Different takes on the Serene's "best" XLR output online too, and the fact I was locked 'n loaded for the 3 Ohm output being for headphones only meant I never gave it a thought.
Sounded terrific as it was.
Obviously with a Pre/Power combo it can be case dependant, so no hard and fast answer online given different Amp impedances with different people systems.
Also,
theres always the old "ok its compatible" Vs what actually works better in practice thing?
If its about lowering the Voltage drop between Pre and Power with the lowest Impedance possible (Getting as much of the 'Juice' from your Preamp to the Poweramp without any loss) then lower impedance as better makes sense?
Also,
the Holo gear is a bit niche, so definately no absolute truth on the Internet about using one in practice.
Why I always say try for yourself.
Also,
never had a pre with more than one XLR output before so never gave it a thought or had reason to know. :)
Took me a lot of bad CD players to finally get one that didnt sound harsh!
My old Rotel RCD-965 was the CD first player I had that didnt have the Digital Glare I didnt know was there till it was gone.
Knew what to look out for after that.
I'm surprised you didn't know this. Like you say it can make a massive difference.
Most of the time though anything not boutique, antique or valve is compatible.
Ten times the input to the output is really the bare minimum. Also it's not a constant, input and output impedance both vary with frequency so matching on paper is not always watertight.
The other one is voltage output and input sensitivity. You don't want to clip the input - remember how some people thought CD was harsh in the early days?
I'm with Barry on this, if the 3 ohm output was so much better in every case, why not just make both pairs of outputs 3 ohm?
Fair comment, would have preferred that m'self!
I've no idea of the design philosophy behind the amp, but It's designer seems to have thought of most stuff that matters in advance, not many compromises that I can see?
800x vs 11,000x impedance difference though does seem a substantial jump?
... and maybe that accounts for the difference?
Who knows, its better and any free upgrade is a big win. https://theartofsound.net/images/smilies/smile.gif
I've no need to understand the technical reasons, I just trust my ears.
Obviously lower impedance means a difference in design between the two XLR circuits?
Maybe the BOM cost between the two was significant enough to limit to one output and market it as a feature, dunno?
In equipment design/manufacturing theres always hitting a final pricepoint per Unit and choosing design priorities to meet it.
Like any Design I guess,
you can always throw more design budget at it, but that can drive the cost up.
Barry could you give me an estimate of a preamp output impedance that would suit my Croft series 7 power amp which has an input impedance of 470kohm?
As Martin has pointed out, the 10:1 rule is a bare minimum. If we take a figure of 1000:1, then your preamp could have an output impedance of 470 Ohm. I think that any preamp would easily have an impedance of less than this, including valve designs.
The preamp output impedance forms a simple voltage divider with the input impedance of the power amp. In other words, the higher the ratio of the two impedances is, the less power is dissipated or wasted in the preamp. In Gary's case the change in ratio in moving from one output to the other is 13.3x, so the improvement is 11.25dB, which is quite noticeable.
Thanks Barry
Yeah, was genuinely taken aback with the improvement.
Just wanted other folks to be aware if they were considering a new Pre or Power amp in there own systems, that it was a thing worth considering.
I've given this a bit more thought, and I think I have mislead you. With an output impedance of 40 Ohm, the transmission loss into 32K Ohm is -29.03dB. That is the power dissipated in the preamp is -29dB compared with the power dissipated in the power amp, and is small (about one thousandth).
With an output impedance of 3 Ohm, the transmission loss is -40.28dB, which is considerably smaller (again about one ten thousandth). So although the difference is -11.25dB, I doubt you would be able to hear this difference; it certainly wouldn't be obvious.
There must be another reason for the audible difference between the two output sockets.
Whatever it is,
having been shocked by it its REALLY obvious :)
I've ran the numbers myself, and as you say on paper 11.25dB IS a significant difference?
Measurements aren't everything, thats why (for me) relying on your own ears in your own System has the final say?
Any improvement in Transmission Loss though is a good thing, and -11.25dB is no small chunk of change.
Always say though as with most of this stuff in Hifi: "It Depends"
I spoke earlier to BOM costs, and Circuit design differences leading to impedance differences in different XLR outputs so I dont think your wrong in saying there more at work than than just the Impedance Specs?
Maybe the 3 Ohm XLR output was just flat out designed to be the Amps best sounding one, dunno?
Anyhoo,
Happy days.
What it is, I have no clue :)
Im running on Faith that Jeff Zu of Holo Audio is smarter than me and He's worked it all out in advance.
As I said, main point of the post was just to say if shopping for a new Pre or Power amp, have a thought about the Impedance between the two amps.
The preamp output impedance forms a simple voltage divider with the input impedance of the power amp. In other words, the higher the ratio of the two impedances is, the less power is dissipated or wasted in the preamp. In Gary's case the change in ratio in moving from one output to the other is 13.3x, so the improvement is 11.25dB, which is quite noticeable.
Yup,
Interestingly the power amp input impedance being 10x more than the Pre’s output, gives 10% voltage drop, so 90% of the preamps musical goodness is getting through.
So any increased difference between Pre and Power Impedance reduces the Transmission loss.
Which is never bad.
Lawrence001
03-01-2026, 02:01
I wonder if there's a slight difference in the gain, do you have to adjust the volume to compensate or do they sound the same? A gain of 0.5db can sound inaudible in volume terms but give the impression of a "fuller" sound quality with more dynamics and drive.
Yup,
Interestingly the power amp input impedance being 10x more than the Pre’s output, gives 10% voltage drop, so 90% of the preamps musical goodness is getting through.
So any increased difference between Pre and Power Impedance reduces the Transmission loss.
Which is never bad.
With a 40 Ohm preamp output impedance, your power amp receives 99.9% of the signal. With a 3 Ohm output impedance, the power amp receives 99.99% of the signal. Or if you like a reduction in the transmission loss from 0.01dB to 0.001dB. I doubt if anyone could hear the difference.
As Lawrence asks, is the volume control set to exactly the same level in each case?
Firebottle
03-01-2026, 09:13
Very good result here Gaz, I like it.
It could be as simple as the 40 ohm output including a series resistor to mitigate any upset if encountering a short circuit condition.
I would guess there is different circuitry for the 3 ohm output.
I would guess there is different circuitry for the 3 ohm output.
That is what I think is happening here.
… :doh:
I give up
With a 40 Ohm preamp output impedance, your power amp receives 99.9% of the signal. With a 3 Ohm output impedance, the power amp receives 99.99% of the signal. Or if you like a reduction in the transmission loss from 0.01dB to 0.001dB. I doubt if anyone could hear the difference.
As Lawrence asks, is the volume control set to exactly the same level in each case?
… :doh:
I give up
As do I! There is no doubting you hear what you hear, but based on the figures alone I can't explain why.
So to prevent us going round and around in circles, I suggest you stick with the 3Ohm output, sit back and enjoy. :)
On further inspection,
Turns out one of the reasons for the difference between the XLR outputs is that the Heaphone friendly (better sounding one) doesn't have any series Load Resistors in circuit.
Genuinely curious as to why this XRL output sounds better, even through speakers...
Would enjoy any circuit design Boffo's to pitch in?
Alan?
Firebottle
06-01-2026, 08:32
OK Gaz, I don't think I can offer an explanation other than 'everything matters'.
I know you understand this and it does seem weird that one resistor changes things.
Did you look inside? Do you know what value resistor is in the 'standard' XLR output?
Wouldn’t dare to look inside Alan, would invalidate my warranty!
Internal pic here:
https://eliseaudio.com/cdn/shop/files/Serene-17-2048x1366.jpg?v=1695027510&width=1946
“Everything matters” always works for me. :thumbsup:
Because of the discussion here I swapped the XLRs back to the other output as a sanity check, yup, the difference is there.
It is what it is, hey… I’ll take it.
Wouldn’t dare to look inside Alan, would invalidate my warranty!
Internal pic here:
https://eliseaudio.com/cdn/shop/files/Serene-17-2048x1366.jpg?v=1695027510&width=1946
“Everything matters” always works for me. :thumbsup:
Because of the discussion here I swapped the XLRs back to the other output as a sanity check, yup, the difference is there.
It is what it is, hey… I’ll take it.
It's perfectly understandable that you don't want to invalidate the warranty. Sadly it is not possible to enlarge the photo sufficiently to read the value of the series resistors used in the two sets of balanced XLR outputs.
Again no-one is challenging you over the audible difference between the 40 Ohm and the 3 Ohm outputs. It's still hard to explain why; and as I have tried to show regarding the transmission loss, the difference in values cannot account for the audible difference.
Regardless of a lack of explanation, do what anyone else would do, that is choose the output that sounds the best to you.
Been doing that from way back since I discovered the difference between the two outputs Barry. :)
I just wanted to flag up that it’s worth thinking about impedance matching between Pre and Power, using my own experience as an example.
I believe Glenn Croft factored this into his amp designs?
I disagree by the way that impedance value differences don’t make a difference here… my Signature seems apt?… but hey, let’s not fall out. :)
.
I believe Glenn Croft factored this into his amp designs?
)
of course he did. Really it's only a potential issue when matching components from different manufacturers, or same manufacturer but years apart. Or with some boutique products where the designer does not seem to think it matters and concentrates on his hobby horse ideas instead.
It can be complex and I agree that a back of the envelope calculation won't necessarily mirror the real world outcome.
of course he did. Really it's only a potential issue when matching components from different manufacturers, or same manufacturer but years apart. Or with some boutique products where the designer does not seem to think it matters and concentrates on his hobby horse ideas instead.
It can be complex and I agree that a back of the envelope calculation won't necessarily mirror the real world outcome.
Yup, 100%
Theres a quote I enjoy that touches on that?
"What is a complete set of measurements that would uniquely allow me to identify good sound? That I do not know.
What I do know is that scientific pursuit is about accurately describing what we observe, not aligning observation with model-driven expectations."
Lawrence001
06-01-2026, 21:42
On further inspection,
Turns out one of the reasons for the difference between the XLR outputs is that the Heaphone friendly (better sounding one) doesn't have any series Load Resistors in circuit.
Genuinely curious as to why this XRL output sounds better, even through speakers...
Would enjoy any circuit design Boffo's to pitch in?
Alan?Can't explain exactly why but when I tried attenuating plugs from 2 well regarded manufacturers and volume matched with a digital volume control they sounded terrible compared to when they were out of circuit. Mainly a flattening if depth and dynamics. So it's not a surprise to me that a resistor in series can make the sound perceptibly worse.
Both sets of XLR output sockets of the Holo Serene preamplifier are fitted with series resistors. I assume the XLR1 set have 3 Ohm series resistors and the XLR2 set have 40 Ohm series resistors.
The XLR1 output is designed to drive headphone loads of 10 - 150 Ohm, but more importantly is designed to provide a high current drive to achieve sufficiently high listening levels.
Reading around on the web about the Holo Serene preamplifier, it appears that some owners prefer the use of XLR2 as a feed to their power amplifiers. "Curiouser and curiouser" :scratch:
Different systems different ears I guess?
Could you link to the site Barry your quoting that gives that technical breakdown, be interested to read it.
Thanks
Different systems different ears I guess?
Could you link to the site Barry your quoting that gives that technical breakdown, be interested to read it.
Thanks
I typed in a lot of things such as:
"Why does the headphone output of the Holo Serene preamplifier sound better than the line output when feeding a power amplifier?" and amongst other things I found this
https://www.magnahifi.com/shop/holo-serene-level-1-18#attribute_values=43,63,64,72
as well as several on-line reviews of the preamp. Some of the reviewers said they used the headphone output to power active speakers, but most did not discuss the merits or otherwise of either output.
https://www.head-fi.org/showcase/holo-audio-serene-preamplifier.27365/reviews
https://www.headfonia.com/holoaudio-serene-kte-review/
https://www.youtube.com/watch?v=8TZBa51ANSc
https://www.youtube.com/watch?v=mdxCksgbScM
https://www.youtube.com/watch?v=lD2Fj1tTu8Y
There was also the What Hi Fi Forum website where someone asked if one could use a headphone output to power an amplifier, but there was confusion between a true headphone amplifier and the headphone socket on an integrated amplifier. https://forums.whathifi.com/threads/why-is-using-a-headphone-output-into-a-hi-fi-amp-frowned-upon.107156/
The Holo Audio Owners Club on Facebook https://www.facebook.com/groups/666904327315858/ I thought might be a good place to look, but you have to wade through scores of posts concerning the Red and other Holo DACs, without any real discussion about the choice of outputs.
However on the HeadFi forum https://www.head-fi.org/showcase/holo-audio-serene-preamplifier.27365/reviews , the two outputs are discussed in a Member review, with findings that conflict with yours.
This forum: Audiophile Style https://audiophilestyle.com/forums/topic/62142-holo-audio-serene-preamplifier/page/6/#comments has some comments concerning the two outputs of the Serene preamplifier (page 7), which agree that XLR1 should be used with active speakers, and by implication all power amplifiers; so agreeing with you.
The general consensus is that in switching between the two outputs, there is a noticeable change in volume for any setting of the volume control, which suggests the headphone output via XLR1 has gone through some additional circuitry.
Excellent!
Better than Chat GPT
Will dig into these links, and hopefully learn something more, always a bonus!
Thanks Barry :)
Elise Audio (bought all my Holo Audio gear from them) specifically speak to XLR Output 1 being the preferred one, unless you have a funky Power Amp.
https://eliseaudio.com/products/holo-audio-serene?srsltid=AfmBOopf9LvJXmNYoAPD2HMjj9AaXLoMRTV 5oUlC5rAHtcP6Mwdv4_9k
Hadnt noticed that till now.
Only mentioned the XLR output differences as I was knocked back when I directly compared them with my own amp and system.
So do they have different output voltages as well as different output impedance? Does the one thing automatically follow the other due to Ohm's Law?
Ali Tait
09-01-2026, 09:42
Yes, Power (Watts) = Amps x Volts.
Elise Audio (bought all my Holo Audio gear from them) specifically speak to XLR Output 1 being the preferred one, unless you have a funky Power Amp)
https://eliseaudio.com/products/holo-audio-serene?srsltid=AfmBOopf9LvJXmNYoAPD2HMjj9AaXLoMRTV 5oUlC5rAHtcP6Mwdv4_9k
Hadnt noticed that till now.
Only mentioned the XLR output differences as I was knocked back when I directly compared them with my own amp and system.
One of the exceptions I referred to was from a user who much preferred to drive his McIntosh power amp from XLR 2 (40 Ohm output impedance).
But as you say "Different ears, different gear". Jeff Zhu has been contacted for his advice, but to date no reply.
Yeah, agree.
Really need to hear stuff for yourself.
One of the exceptions I referred to was from a user who much preferred to drive his McIntosh power amp from XLR 2 (40 Ohm output impedance).
But as you say "Different ears, different gear". Jeff Zhu has been contacted for his advice, but to date no reply.
Also,
McIntosh amps might qualify for those edge case “high impedance” amps where XLR2 might be better? Mac amp’s Autoformers present a different load to most amps.
The MC462 has an input impedance of 44K Ohms, which is what I think the guy in that Forum post was using?
(I’m familiar with it… nearly bought one before I decided the big Meridian was more my thing)
Different strokes for different Folks, like a lot of stuff in life? :)
It was a McIntosh MC462 power amplifier, and has an input impedance of 22K (as quoted by the poster). With the 40 Ohm output impedance source, the impedance ratio is 550:1, implying a transmission loss of 0.008dB.
With the 3 Ohm output impedance XLR, the transmission loss is 0.0006dB. Either of these two figures is utterly inaudible.
It was a McIntosh MC462 power amplifier, and has an input impedance of 22K (as quoted by the poster). With the 40 Ohm output impedance source, the impedance ratio is 550:1, implying a transmission loss of 0.008dB.
With the 3 Ohm output impedance XLR, the transmission loss is 0.0006dB. Either of these two figures is utterly inaudible.
As I said, very familiar with this amp. Nearly bought one so know it well.
22K was the unbalanced input impedance, Balanced is double that say, say 44K
And not to be unkind, but your just quoting Specs you read online, only one of us have heard and compared the equipment in question.
And its not you Barry.
Happy to leave it here Sir, no disrespect intended.
Noticed this a few posts back, but took the view its just banter?
And of course it is. :)
As I said, very familiar with this amp. Nearly bought one so know it well.
22K was the unbalanced input impedance, Balanced is double that say, say 44K
And not to be unkind, but your just quoting Specs you read online, only one of us have heard and compared the equipment in question.
And its not you Barry.
Happy to leave it here Sir, no disrespect intended.
Noticed this a few posts back, but took the view its just banter?
And of course it is. :)
22K is the input impedance for both the balanced and unbalanced inputs, as quoted by McIntosh. I looked at a pdf of the user's manual for the MC462. (page18), just to make sure.
https://www.mcintoshlabs.com/-/media/Files/mcintoshlabs/DocumentMaster/us/mc462om01.ashx
No disrespect taken.
Ok.
Thanks for the link
*deleted by me*
Firebottle
10-01-2026, 09:03
Cool it Gaz, I didn't detect any dissing in the thread. :)
From another engineers perspective I see Barry just posting electrical/matching/transmission figures.
Knowing a little of Barry's background I would say he is a very talented engineer. If posting such information could explain differences in sound that would be pretty fantastic.
However as we all know specifications tell us nothing about how things actually sound and the sheer variety of ancillary equipment just amplifies the possibility of differences in sound. :scratch:
As you were and back to the sound, we can all agree on that. :cool:
Yeah. :)
Cool it Gaz, I didn't detect any dissing in the thread. :)
From another engineers perspective I see Barry just posting electrical/matching/transmission figures.
Knowing a little of Barry's background I would say he is a very talented engineer. If posting such information could explain differences in sound that would be pretty fantastic.
However as we all know specifications tell us nothing about how things actually sound and the sheer variety of ancillary equipment just amplifies the possibility of differences in sound. :scratch:
As you were and back to the sound, we can all agree on that. :cool:
Firebottle
10-01-2026, 20:16
Thought this might fit here. How you know you are an engineer:
http://www.youtube.com/watch?v=QatYbWusODk
:)
Thanks Alan.
Heard, understood and acknowledged.
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.