A look at the Harbeth P3 crossover

[gx502]

I've watched some of the videos on this channel. I don't think the work that is done by this guy is totally without merit. He shows response curves, pre/post upgrades, and gives us his impressions of what he feels are issues in the original speaker that his kits address.

(But in general I do agree with the overall comments in this thread, that the part swapping approach is risky, you may end up with something feral and unbalanced. A safer approach is just demo and buy a better off the shelf speaker!)

Of course, just swapping in boutique parts willy nilly, is not going to match a well considered development and voicing process. But I do find it very interesting when he shows us the innards of speakers at various price points.

And to be fair, i think he concludes, he cannot do anything with respect to part swaps on the P3 anyway (due to lack of space). So you are left to conclude if you like the speaker in its original form then just enjoy it as is.

IIRC from one of his previous videos, didn't this guy do some design work on Usher loudspeakers ?


Lolz EDIT: perhaps that's why Harbeth think all competently designed amplifiers sound the same, its due to the lack of transparency in their crossover circuitry --- j/k.

 

[John Phillips]

In LspCAD (crossover design software) there's no option to add parasitic elements to resistors in the model. I'm sure I read somewhere that the small amount of inductance in sandcast resistors wasn't worth worrying about (could be wrong though). You can add ESR to capacitors in the pro version of LspCAD, but it's a fixed value.

I would be curious to know if there is any crossover design software out there that allows you to go into that much detail.

Looking at the LspCAD web site it looks like you can enter your own component parasitic elements as separate elements in the crossover network. However, as you write, the question is whether it is worthwhile and whether replacing components declared as "cheesy" is worth it per se.

I get the feeling that perhaps with resistors and capacitors it's not as big a deal as it is portrayed in the video. I still suspect that with inductors their series resistance is much more important (e.g. compared to the resistance of a bass driver you might get noticeable signal loss). But then I wonder about if/when/how core saturation in inductors matters. Many thanks for your comments. This is an area of design new to me and I'm probably being over-curious.

 

[Fatmarley]

Looking at the LspCAD web site it looks like you can enter your own component parasitic elements as separate elements in the crossover network. However, as you write, the question is whether it is worthwhile and whether replacing components declared as "cheesy" is worth it per se.

I get the feeling that perhaps with resistors and capacitors it's not as big a deal as it is portrayed in the video. I still suspect that with inductors their series resistance is much more important (e.g. compared to the resistance of a bass driver you might get noticeable signal loss). But then I wonder about if/when/how core saturation in inductors matters. Many thanks for your comments. This is an area of design new to me and I'm probably being over-curious.

Didn't even notice that, probably because I've never heard of anyone adding parasitic elements to a model, and I haven't seen it mentioned in any tutorials, but now you've got me curious... Inductor series resistance and ESR of capacitors definitely make a difference, but not many people even add ESR (takes some guesswork with a fixed number though).

Looking at the photo in post #60 It looks like Alan Shaw thought it was worthwhile replacing cheesy components (capacitors), but we don't know if the difference in sound is due to that, or if he changed some component values. It's quite possibly a combination of both.

 

[irb]

Looking at the LspCAD web site it looks like you can enter your own component parasitic elements as separate elements in the crossover network. However, as you write, the question is whether it is worthwhile and whether replacing components declared as "cheesy" is worth it per se.

I get the feeling that perhaps with resistors and capacitors it's not as big a deal as it is portrayed in the video. I still suspect that with inductors their series resistance is much more important (e.g. compared to the resistance of a bass driver you might get noticeable signal loss). But then I wonder about if/when/how core saturation in inductors matters. Many thanks for your comments. This is an area of design new to me and I'm probably being over-curious.

Yes, series resistance of inductors is important, and any half-decent design software allows you to enter it.
If you fancy modelling simple crossovers, I can recommend Basta, an excellent free download which allows you to play around with all the variables - driver, box/baffle, active and passive crossovers. It's highly educational. You can find it here: https://www.tolvan.com/index.php?page=/basta/basta.php

 

[barryb]

Didn't even notice that, probably because I've never heard of anyone adding parasitic elements to a model, and I haven't seen it mentioned in any tutorials, but now you've got me curious... Inductor series resistance and ESR of capacitors definitely make a difference, but not many people even add ESR (takes some guesswork with a fixed number though).

Looking at the photo in post #60 It looks like Alan Shaw thought it was worthwhile replacing cheesy components (capacitors), but we don't know if the difference in sound is due to that, or if he changed some component values. It's quite possibly a combination of both.

Yes, there was some revisions to the circuit too, I believe. I've yet to read an account of those very familiar with both models who preferred the fancy anniversaries.

 

[Tony L]

Fancy film caps are not a ‘drop in’ replacement IME. Fine if you are designing from scratch, but just replacing electrolytics in vintage speakers with the same value polys really changes the voicing IME. It really killed my JR149s, lost all the warmth and naturalness. Initially “cleaner, more detailed”, but after some acclimatisation just hopelessly wrong compared to fresh electrolytics.

 

[Fatmarley]

Fancy film caps are not a ‘drop in’ replacement IME. Fine if you are designing from scratch, but just replacing electrolytics in vintage speakers with the same value polys really changes the voicing IME. It really killed my JR149s, lost all the warmth and naturalness. Initially “cleaner, more detailed”, but after some acclimatisation just hopelessly wrong compared to fresh electrolytics.

There can be enough difference in ESR to literally change the frequency response, and that can include how the crossover sums. Some argue that it's not enough to be audible, but when voicing, a fraction of a db over a wide enough bandwidth can make a significant subjective difference IMO.

 

[tuga]

Fancy film caps are not a ‘drop in’ replacement IME. Fine if you are designing from scratch, but just replacing electrolytics in vintage speakers with the same value polys really changes the voicing IME. It really killed my JR149s, lost all the warmth and naturalness. Initially “cleaner, more detailed”, but after some acclimatisation just hopelessly wrong compared to fresh electrolytics.

The problem with reconditioning a very old speaker is that the sound will change, more or less depending on the level of ageeing of the electronic components and the replacements, and the owner might not like it.
Even if one has bought a 30 or 40 year old speaker when it came out we will have become accustomed to the gradual change in sound due to ageeing components, so we may prefer the sound of the "broken" speaker to that of the original.

I agree that it might be wiser to have the crossover and speaker measured to make sure that the replacement parts (when different from the original ones) are not changing the original target performance (too much).

 

[Tony L]

The problem with reconditioning a very old speaker is that the sound will change, more or less depending on the level of ageeing of the electronic components and the replacements, and the owner might not like it.
Even if one has bought a 30 or 40 year old speaker when it came out we will have become accustomed to the gradual change in sound due to ageeing components, so we may prefer the sound of the "broken" speaker to that of the original.

True, to a degree, though worse is worse, and any experienced listener will pick that up given time. It is why it makes sense IMO to rebuild like for like using components as close to the original spec as possible. As stated above ESR etc can be very different and that will impact the way the crossover behaves. It is extraordinarily arrogant to assume one even knows as much, let alone more than the people who actually designed the speakers that stand today as landmark products. It actually often makes me laugh at the idiots that think they do despite zero proven record of their own. My goal is only to get things to perform as they should, no more. If I want a different sound I’ll just go and buy something else.

PS FWIW my JR149s are as close to a brand new pair as it is possible to find in 2021. They are as close to original spec as a brand new pair of Falcon LS3/5As. Basically every part has been replaced with a current direct equivalent.

 

[Sue Pertwee-Tyr]

PS FWIW my JR149s are as close to a brand new pair as it is possible to find in 2021. They are as close to original spec as a brand new pair of Falcon LS3/5As. Basically every part has been replaced with a current direct equivalent.

Trigger's loudspeakers...

One thing that isn't often discussed is the effect that crossovers have on phase relationships in the signal. I love single-driver loudspeakers for their directness and 'believability'. Any shortcomings as to bandwidth are effectively minor colorations which I'm often happy to trade for that coherence thing. I don't use them for the main system, but for me I've always found speakers that try to minimise phase distortions sound more engaging than those that attempt a flat response at the expense of phase.

 

[Tony L]

One thing that isn't often discussed is the effect that crossovers have on phase relationships in the signal.

I agree. Time alignment is a critical parameter IMO and I suspect is a key reason I end up drawn to mini-monitors, point source etc, and my tendency to actively dislike large speakers with multiple drivers spread across a tall baffle. I’ve yet to find a single driver speaker I can live with, but hearing a good pair of Lowthers left a lasting impression (just so coherent and dynamic). I love ESL63s too which remain the benchmark when it comes to phase/time alignment.

 

[narabdela]

I’ve yet to find a single driver speaker I can live with, but hearing a good pair of Lowthers left a lasting impression (just so coherent and dynamic).

Lowthers tend to get a bad press on H-Fi forums, but I agree with you. I wonder how many naysayers have actually listened to them. They have their faults, but all speakers are a compromise.

 

[Tony L]

Lowthers tend to get a bad press on H-Fi forums, but I agree with you. I wonder how many naysayers have actually listened to them. They have their faults, but all speakers are a compromise.

It was jazz drumming that sold me. They just captured every nuance of brushing etc, just so alive and free. The exact opposite of so many modern high-mass multi-driver MDF boxes with reactive crossovers that just suck the life out of everything no matter how much solid state power they demand.

 

[tuga]

It is extraordinarily arrogant to assume one even knows as much, let alone more than the people who actually designed the speakers that stand today as landmark products.

I see it slightly differently. In my view all commercial equipment has potential for improvement (some more, some less). To believe that "I" can do it might be extraordinarily arrogant/mistaken but there are those who do know how to improve performance. To use a car analogy, look at Mercedes/AMG, Porsche/Ruf, BMW/Alpina.
On top of that, there have been advances in the science (electronics, acoustics, pyschoacoustics), in component performance and in the measuring equipment all of which contribute to increasing the effectiveness of modifications and, consequently, of performance.

 

[ssimon]

The cynic in me says that all manufacturers want to minimize their costs, so if cheap components are good enough to sell the product then why use better components.
When I saw that my JBL K2 speaker crossovers used all chinese components, I upgraded the capacitors to good effect.
Interestingly JBL made a MKII version, the only difference being that they removed the bypass caps. They saved money and I'm sure said it was a better speaker.

 

[James]

I agree. Time alignment is a critical parameter IMO and I suspect is a key reason I end up drawn to mini-monitors, point source etc, and my tendency to actively dislike large speakers with multiple drivers spread across a tall baffle. I’ve yet to find a single driver speaker I can live with, but hearing a good pair of Lowthers left a lasting impression (just so coherent and dynamic). I love ESL63s too which remain the benchmark when it comes to phase/time alignment.

Time and phase are related, but they are not the same thing. You could have perfectly aligned drivers so that sound emanating from them arrives at your ears at exactly the same time, but still sound wrong because of divergent relative phase. Similarly, drivers could have a different flight path, but perfect summation that sounds seamless at your ears. If that isn't the case, then sitting non-equidistant from loudspeaker pairs would seriously screw up your listening enjoyment.

 

[Tony L]

Yes, I fully understand that. It is arguably the major failing of multi-driver speakers. They all fail to some degree.

 

[James]

Yes, I fully understand that. It is arguably the major failing of multi-driver speakers. They all fail to some degree.

I agree. That's the nature of something that cannot be perfect.