As an example of the madness that is passive speaker design. I designed a mini monitor, having spent alot of time listening working on the drive units, cabinet construction, baffle, panel damping, internal suffing, port length and dimensions and the crossover, the listening team and I were happy that I had a result that worked.
So I had a pilot production run done, it was only a few units but they were made with all of the production processes by the OEM manufacturer. I set it up to listen to it and whilst it had a very similar characteristic to the prototype some of the “realism” or “life” of the sound had gone, it is very hard to put into words exactly what it was but all the listeners agreed something had been lost. I spent a long time checking the design and all of the measure parameters matched the prototype, the enclosure behaved in the same way (I checked it with a surface accelerometer) the bass port gave the same output and the response in frequency, polar and time decay all looked equivalent.
Being extreemly frustrated I even tried calling a member of the listening team back and actively tried to convince him I had fixed it (I had made no change) to make sure it wasn’t listening bias. He was not confused for even a second, aftera few seconds listening he turned it off and said sorry that hasn’t done it.
So I decided I had to dissasemble the speaker and the entire crossover and check every components values and tolerance and compair the characteristics to the prototype. I did this and could not find any significant differences in the components they were all well within tolerance and very close to nominal values. I put it back together rather disillusioned and was doing a listening testing to ensure I had put it together right, one of the listening team came in and imediately said “Thats much better, not back to where you were but getting there”. I listened more actively and agreed some of the “life” was back.
So I took the other speakers crossover apart checked all the values and put it back together then reassembled the speaker. I called a member of the listening team back in and we both agreed the problem was solved. Now all I had to do was work out what I had changed it taking it apart and putting it back together.
After a very long investigation it came down to the hotmelt glue that had been used to ensure the bass inductor didn’t come lose in shipping, I had used a tie wrap but the manufacturers were not happy that it was sufficient and had added hot melt glue. They didnt say anything because “why would it make any difference” and its hard to argue with that rational. In the end we compromised on two tie wraps, the OEM was sure I was mad but we were the customer and if we insisted it had to be two tie wraps then thats what they would do. I never did track down what it changed, I did lots of time decay waterfall plots and thought it might be changing the decay sligtly, but if it was it was so close to the noise that I could not find a definative argument. I also dont know if it was euphonic, i.e an error being added that made it sound nicer or removing an error. Inductors are sligtly microphonic so it could have been the PCB vibrating in the energy in the box casing an error to be added or it could have been that with the inductor looser it was vibrating and adding a sort of reverb that sounded nice or it could have reduced the energy transfer by not being coupled to the PCB. Or it could have been nothing to do with that and something else was going on!
This was not a massively expensive speaker about £400 a pair, I would never have belived such a small change could have such a dramatic effect if I had not spent about 6 weeks trying to track it down.
So I had a pilot production run done, it was only a few units but they were made with all of the production processes by the OEM manufacturer. I set it up to listen to it and whilst it had a very similar characteristic to the prototype some of the “realism” or “life” of the sound had gone, it is very hard to put into words exactly what it was but all the listeners agreed something had been lost. I spent a long time checking the design and all of the measure parameters matched the prototype, the enclosure behaved in the same way (I checked it with a surface accelerometer) the bass port gave the same output and the response in frequency, polar and time decay all looked equivalent.
Being extreemly frustrated I even tried calling a member of the listening team back and actively tried to convince him I had fixed it (I had made no change) to make sure it wasn’t listening bias. He was not confused for even a second, aftera few seconds listening he turned it off and said sorry that hasn’t done it.
So I decided I had to dissasemble the speaker and the entire crossover and check every components values and tolerance and compair the characteristics to the prototype. I did this and could not find any significant differences in the components they were all well within tolerance and very close to nominal values. I put it back together rather disillusioned and was doing a listening testing to ensure I had put it together right, one of the listening team came in and imediately said “Thats much better, not back to where you were but getting there”. I listened more actively and agreed some of the “life” was back.
So I took the other speakers crossover apart checked all the values and put it back together then reassembled the speaker. I called a member of the listening team back in and we both agreed the problem was solved. Now all I had to do was work out what I had changed it taking it apart and putting it back together.
After a very long investigation it came down to the hotmelt glue that had been used to ensure the bass inductor didn’t come lose in shipping, I had used a tie wrap but the manufacturers were not happy that it was sufficient and had added hot melt glue. They didnt say anything because “why would it make any difference” and its hard to argue with that rational. In the end we compromised on two tie wraps, the OEM was sure I was mad but we were the customer and if we insisted it had to be two tie wraps then thats what they would do. I never did track down what it changed, I did lots of time decay waterfall plots and thought it might be changing the decay sligtly, but if it was it was so close to the noise that I could not find a definative argument. I also dont know if it was euphonic, i.e an error being added that made it sound nicer or removing an error. Inductors are sligtly microphonic so it could have been the PCB vibrating in the energy in the box casing an error to be added or it could have been that with the inductor looser it was vibrating and adding a sort of reverb that sounded nice or it could have reduced the energy transfer by not being coupled to the PCB. Or it could have been nothing to do with that and something else was going on!
This was not a massively expensive speaker about £400 a pair, I would never have belived such a small change could have such a dramatic effect if I had not spent about 6 weeks trying to track it down.
It depends what people understand by high-end. You can edit your first post at any time if you want to make it clearer
"I believe we already started dig deeper than thread starter wants to, before his first steps in to audio."
I agree and disagree with this statement. I agree you are digging more than the meaning of my post because I didn't mean to bring a discussion about big $$$$ speakers against what people can humbly build, even when buying the most expensive not always means happiness. I disagree about your opinion these are my first steps on audio. If you mean, I have to own a pair of Wilson Audio to consider myself and audio lover, you should think broader. Audio is your, and what everybody else can reach too. I listen to music. I don't seat to see the lights of tube gear on or to judge the way my equipment sound. I try to enjoy not to be sick about the faults of the sound I listen to.
I agree and disagree with this statement. I agree you are digging more than the meaning of my post because I didn't mean to bring a discussion about big $$$$ speakers against what people can humbly build, even when buying the most expensive not always means happiness. I disagree about your opinion these are my first steps on audio. If you mean, I have to own a pair of Wilson Audio to consider myself and audio lover, you should think broader. Audio is your, and what everybody else can reach too. I listen to music. I don't seat to see the lights of tube gear on or to judge the way my equipment sound. I try to enjoy not to be sick about the faults of the sound I listen to.
Hi dvjorge, sorry if our high end speakers description are diferent.
I also I didnt mean that price is key factor to sound quality, but I most best performance human being things made(be it speaker, car or whatever you name it ) is (unfortunately) expensive. You cant have it cheaper, its just not possible.This is not my fault.
On the other hand there is a lot of things which is both expensive AND poor performance. But we are not discussing those.
So I correct myself- yes you can built DIY speaker which outperform some comercial "high -end" speaker brands. I wish you a lot of fun with your project! All the best.
I also I didnt mean that price is key factor to sound quality, but I most best performance human being things made(be it speaker, car or whatever you name it ) is (unfortunately) expensive. You cant have it cheaper, its just not possible.This is not my fault.
On the other hand there is a lot of things which is both expensive AND poor performance. But we are not discussing those.
So I correct myself- yes you can built DIY speaker which outperform some comercial "high -end" speaker brands. I wish you a lot of fun with your project! All the best.
dvjorge - one other aspect that hasn't been brought up, but may be obvious to you....
A common theme of DIYers tends to be the desire (need?) to continually tweak and upgrade. While you can often get a decent price for used commercial products, you will recoup very little of your investment if you decide to upgrade and attempt to sell your DIY speakers.
A common theme of DIYers tends to be the desire (need?) to continually tweak and upgrade. While you can often get a decent price for used commercial products, you will recoup very little of your investment if you decide to upgrade and attempt to sell your DIY speakers.
Hi dvjorge, sorry if our high end speakers description are diferent.
I also I didnt mean that price is key factor to sound quality, but I most best performance human being things made(be it speaker, car or whatever you name it ) is (unfortunately) expensive. You cant have it cheaper, its just not possible.This is not my fault.
On the other hand there is a lot of things which is both expensive AND poor performance. But we are not discussing those.
So I correct myself- yes you can built DIY speaker which outperform some comercial "high -end" speaker brands. I wish you a lot of fun with your project! All the best.
I don't have anything against you or your points.! I probable didn't write the tittle correctly when I opened this post. I know what you mean and also know you are correct when we talk about the ultra hi end lines. But, again, it was not my intention to open a debate about what is best. I wanted to get an answer about "improving" the sound you get from budget bookshelf even when those are made by brands that also make hi-end models. Let say B&W, Kef, etc. I have owned many entry level speakers to mention here, and honestly, I can spend some money buying a good pair ($5000 or $6000 and even more if it was my priority) I love audio but have other priorities. In spite of it, I didn't mean to be better than those big names, I see there are good drivers in the market, people that know what they are doing such as this generous gentleman who has done the technical part for me using the drivers I chose, so, this keeps my wish of experiment and to adventure alive. I see entry level bookshelf from many brands have nothing special that a DIY project can not overcome. I have opened many of them to see inside and the poor quality is visible . I saw the bass driver of one B&W 801 one of my friend owns, and you wish to take it to the garbage, and yes, a B&W 801. I know this isn't the case of Wilson, Focal (top line), Avalon etc. Focal also has entry level speakers that aren't high quality. My reaction to the way this thread turned around is because it will be an endless debate arguing who is right or wrong about the best quality speakers and some people may feel hurt if they can not afford a pair of Wilson but still wish to listen to a decent sound. I am very grateful of having a place like this where I could write my intentions and where some members do all they can to support and help without look for a profit. Very grateful of this forum and what this British gentleman has done for me. If I used a broad term mentioning hi-end speakers, I apologize for it. It was not my intention to make people think about the top limits.
You are absolutely right.! However, I never lose the drivers and probably won't build totally finished enclosures until I be sure they sound how I want. There should be a way of taking the best sound of quality drivers even when you may fail several times. In my case, I have hope they sound good, at least expecting it since someone with knowledge and experience has offered me an starting point. If they need some adjust, it will be done. If all fail, I will end selling the drivers in Ebay but happy of doing what I love, even losing money. It won't be the first time I lose.
Of course you don't mind loosing some money when you pay around 100$ for all of the drivers! I think you'll not be so happy about knowing that you tried to build your own speakers , spent 500-1000$ in drivers , another 100-150$ in crossover parts and then you realize that you got to a dead end. You could sell used drivers on eBay maybe for half of what you originally paid. Maybe.
Last edited:
Yes, I also disagree about the opinion of the other forum member who called a $5000 or $6000 speaker a mediocre one. This is the vision of someone who owns (or wish) a much more expensive one and disqualify anything lower than what he agrees to pay or dream. This is a public forum and nobody should call garbage what others own or what most people can pay, in any case. It doesn't hurt me but as a respect to all, if that were my case, I was more humble and sensible.
Last edited:
Listen to Globalplayer.! Also, you always can fix them if you aren't happy. When you buy some drivers and plan a project, you have to be aware of what you are buying and getting.! Don't expect to buy a KIA engine to assemble a Ferrari.
Not really dvjorge, I don't want to discourage you, just my experience for your info: not all drivers match and matching drivers and crossovers is another story unless you build the crossovers yourself or use a digital crossover. Then the drivers and the enclosures must match as well.
So if you buy all, you may end up with a bunch of different drivers and crossovers. For me, they are not the worst things but the cabinets, which take up a lot of space in the storage room.
So if you buy all, you may end up with a bunch of different drivers and crossovers. For me, they are not the worst things but the cabinets, which take up a lot of space in the storage room.
Last edited:
por
Did you read back in this thread ? If not, I suggest you to do it. You are going to find surprising things.
Last edited:
Andrew, I remember Arcam had a reputable satellite speaker/subwoofer combo. The Solo subwoofer was part of this package. Are you the designer of this sub ?? It has a high reputation as very powerful and good sounding subwoofer. Thanks
Thanks for sharing your anecdotes, made me smile.
Touching on a few:
- I was always concerned with using a shunt rc in parallel with a tweeter, with no series tweeter pad: impedance can drop quite low at HFs and with class D amps or any ultrasonic content (RF etc), that could require the amp to sink an awful lot of current.
- That's funny about the inductor, your theory may be right. I had similar issues with passives in a particular build. All caps measured perfectly on a capacitance bridge, but then I took each and measured complex impedance to 30 kHz using MLSSA and found a bum cap whose impedance only went wonky at higher frequencies!
- A long time I wondered if it would be best to try and mimic the radiation pattern of instruments. But always felt there were two big issues: which to pick? and more importantly, doesn't matter due to the unknown reference issue, the "circle of confusion" as Floyd Toole has branded it. Bob was clever to try and get them out the door fast using an obvious colouration!
- Couldn't agree more on the importance of having a good reference when voicing. I had spent months casually voicing a design and thought I had nailed it. A quick trip to the local Hi Fi shop and a listen to some Revel Salons told me there was a bit more work to do on the crossover
I always wondered how good the DIY designs sold by Meniscus and PE etc sound. They get designed so fast. It takes me up to 2 years to finish voicing my own personal stuff, but that's at a casual pace, maybe only trying 5 to 10 iterations listening to each a few months.
The beauty of DIY though is you get to tune it for your own room. This usually means trading off on and off axis response for the right tonal balance to work with the room signature, assuming everything else was taken care of. And why voicing takes so long! Commercial couldn't do this as the result probably wouldn't sound great in a wide variety of rooms.
Last edited:
Yes I designed both the Muso satellite speaker, the active crossover and acoustics for the LOGO sub. The Muso was the speaker with the odd inductor sound issue, possibly because it is such a small cabinet there is a lot of energy inside it. The LOGO uses an active Linkwitz transform to extend the bass response and adjust the Q of the bass roll off. The roll off is set to a Bessel alignment (Maximally flat in the time domain - i.e. minimum group delay) as in listening tests I found this provided the easiest in room integration and most realistic bass timing.
One of our listening panel was (probably still is) a bass guitar player and he was very well "tuned" to the sound of deep bass and helped with some of the listening tests on both the Muso and Logo. These were the last commercial designs I did, I did a bit more research on an all in one design but this wasn't taken forward.
I am glad you enjoyed my various musings on the HiFi industry.
In answer to the R-C tweeter shunt I did consider the load on a class D amp. The 4.7R with the tweeter inductance forms a minimum load at 40KHz of 3.3R. In fact the load will be higher as the 1uF is still a significant impedance at this point, this just gives a worst case look. There should be very little energy in this region, a switching amplifier should have a capacitive reconstruction filter anyway and the switching frequency is typically much higher (Eg TK2050 is 650KHz it has a 15uH output inductor and a 220nF capacitive load) where the inductance of the tweeter will have reduced the load even further. A competent class D amplifier should have no problem with a HF 3.3R load. However if it does, in a DIY design you have the option to just remove it.
One of our listening panel was (probably still is) a bass guitar player and he was very well "tuned" to the sound of deep bass and helped with some of the listening tests on both the Muso and Logo. These were the last commercial designs I did, I did a bit more research on an all in one design but this wasn't taken forward.
I am glad you enjoyed my various musings on the HiFi industry.
In answer to the R-C tweeter shunt I did consider the load on a class D amp. The 4.7R with the tweeter inductance forms a minimum load at 40KHz of 3.3R. In fact the load will be higher as the 1uF is still a significant impedance at this point, this just gives a worst case look. There should be very little energy in this region, a switching amplifier should have a capacitive reconstruction filter anyway and the switching frequency is typically much higher (Eg TK2050 is 650KHz it has a 15uH output inductor and a 220nF capacitive load) where the inductance of the tweeter will have reduced the load even further. A competent class D amplifier should have no problem with a HF 3.3R load. However if it does, in a DIY design you have the option to just remove it.
Wow.! It was given 5 starts by the What Hi-Fi Magazine:
Arcam Muso/Logo review | What Hi-Fi?
ARCAM MUSO: Tough Speakers With Cuddly Corners - The Audiophile Man
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.