Elliott Sound Products More Reader Feedback 

Some of the responses to the article on bi-amping (edited to protect the writers' privacy, etc). Many of these also qualify as FAQs, since many readers have questions, some of which I can (or will) answer, and others I leave open-ended.

Not all e-mails warrant a response in this page, but all correspondence is answered. Volumes have increased to the point where it is not possible to reprint everything, but if it is likely to be of interest to a fairly wide audience, I will include it eventually.

Please note that reference to any brand of product shall not be regarded as an endorsement or criticism of the product. Where included, such references are from a reader, or mentioned in my reply, and are included only for informational purposes.

If an e-mail from you has already been included and you prefer not to be published, please let me know immediately, and your correspondence will be removed.


HomeMain Index
ContactContact ESP


Readers' Feedback and Responses

C Comment, R Response, Q Question, A Answer


C Thanks for your comments. Bi-amping is something that not very many people understand, no less the average bride or groom looking for a DJ for their wedding reception. You are exactly correct when you said it doesn't tell the whole story. If we went into the detail that your web site has, people would get completely lost. Most don't understand two speakers versus four, let alone bi-amping.

Bottom line, bi-amping creates a better sound, you know it, I know it and we are one of only a few DJ services in the Twin Cities that does it this way.



C I just read your article on bi-amping, and I'm quite impressed. I don't think you left anything out! I am writing to congratulate you an a particularly easy (and entertaining) to read article. I am a devotee of the bi-amping principle myself, and I am currently building my third set of loudspeakers. This time, it's a no-compromise design. It will, naturally, be bi-amped, though I'm beginning to think that it'll have to be tri-amped.

Q What a fascinating web site! Rarely do I come across folks on the web who know how to articulate and present an argument as well as you have done.

This is not to say that I agree with everything you have said, I am still digesting some of your comments about damping factors and remain a little sceptical about the advantages of bi-wiring but your web site was immediately bookmarked under the "good" list!

In your career have you done any lecturing or other form of teaching?

My own 5.1 HT system at home runs 2nd order crossovers somewhere in the region of 2200Hz and the intelligibility is woeful (not helped by poor choice of operating bands for the drivers). I am going to bite the bullet and do a Bi-Amped system. actually, Tri-Amped, since I already run an active 4th order LP crossover into the sub channel.

I am very interested in hearing more about your idea's on minimum phase active crossovers - any further light you can shed on this? I am an Electronics Engineer by Profession by definitely *not* by experience, so my expertise is a little limited. I work in the Telecommunications industry in management now...


A I don't recall suggesting that you agree, but thanks for your comments.

I am (at the time of writing - this has since changed) a "technical training engineer" (telecoms speak for teacher), but I was previously invloved in teaching real electronics and audio.

Regarding bi-wiring - I know I bagged it, but it is still better than running the whole shebang down the one set of cable. It does rely on the amp having a good damping factor, but if you look at the specs closely, you will see that damping factor dies as the frequency increases. This is because of internal frequency limiting in the power amp (required to prevent the amp's phase shift exceeding 360 degrees while it still has gain - Murphy's Law - amplifers will oscillate, and oscillators will amplify).

There are some phase coherent designs, but many rely on subtraction - it works, but gives a lopsided crossover response (e.g. 12db/6db) - these were originally described in various magazines over the years, but they can a lumpy response across the xover frequency.


C Just stumbled across your website. Very interesting, especially since I am currently designing a system that conforms to (most of) your beliefs expressed on the page!

I already built an active filter (simple 2th order Sallen-Key, Butterworth) that works but I'm always looking for improvement! The next version might be implemented on a DSP, if that doesn't bite my budget too much.

Keep up the good work.


Q I have been thinking about implementing active cross-overs in my system and came across your web-site. It makes interesting reading and, as I intended starting with only an active low-pass for bass frequencies, is exactly in line with my intentions. I see that you have designed and built your own active crossover; is your circuit publicly available? My speakers (KEF R105) have a completely separate low-pass filter in the low frequency driver enclosure, and I would like to eliminate this completely but am a bit reluctant to because of the nature of the filter. I believe that the filter is shaping the loudspeaker response as well as limiting its supply of higher-frequency signals. Have you taken this into account? It's an interesting article you've written and I look forward to your updates.


A As regards active crossovers - go for it. Project 09 is the recommended design. Regarding your speakers, all filters shape the response of subsequent stages, it's a matter of minimising the "damage", and this is where active crossovers can make a huge difference. For starters, you get rid of the large bass frequency transitions from the amp, so intermodulation distortion is reduced - this is one of the main contibutors to a "muddy" sound. Problem is that it gets a bit expensive, but that's hi-fi.

C Thank you for your very thorough and interesting WEB article "Benefits of Bi-Amping". After reading "Benefits of Bi-Amping", I have implemented the following Bi-Amplification HIFI system to drive a pair of JM Megane loudspeakers.
  • Integrated Amp: NAD 317 (with Vampire P#YF RCA Adapter on the PRE-OUT connectors to allow multiple power amplifier connection).
  • Second Power Amp: NAD 214 ( duplicate of the power amplifier in a NAD 317).
  • The JM Megane has two drivers, a 1.5 in. inverted dome tweeter and a 7 in, woofer with dual voice coils configured as a 3-way system with crossover frequencies at 3.5 kHz (18 dB/octave) and 1 kHz (6 dB/octave).
This HIFI system began with the NAD 317 integrated amplifier - It was very acceptable beginning ... a good sounding system. I then implemented Bi-Wiring.

Bi-Wiring produced a noticeable improvement in sound stage and detail. I have learned that Bi-Amplifcation should allow for an even more dramatic improvment in sound because this technique is much more efficient in reducing Intermodulation Distortion (undesirable modulation of the high frequency signal by the low frequency signal) than simple Bi-Wiring.

So far, I have observed that the Bi-Amplifcation has dramatically improved Bass (2nd Amp??) and additional improvement in sound stage over the Bi-Wire system. Bi-Amplifcation also seems to have changed the character of the sound presentation ...

I think I now know what is meant by sound having a liquid like quality.


R I'm not at all surprised that you experienced a fairly dramatic change in sound quality, this is exactly what I was waffling on about in the article. As for further improvements, I can only suggest that you keep experimenting - there are so many variables in any amplifier/speaker system that you can probably keep yourself occupied for years!

One thing to be careful about when bi-amping a two-way:- Make absolutely sure that no DC is present at the amplifier speaker terminals (this can arise when the amp is powered on or off, and is usually only fairly brief). Tweeters are not forgiving of even small amounts of DC, and to be safe you should use a capacitor in series (I can't suggest an accurate value, since I don't know the crossover frequency, but I would think that you would need about 20uF or so as an absolute minimum. This will create its own crossover at about 995Hz (for an 8 ohm speaker) and would be (just) ok if you are crossing over at 5kHz or above. If the value is too low, it will create its own phase problems, and if too high, will allow low frequency "surges" to pass through to the tweeter. Only problem is that suitable high value polyester type caps are expensive !!!



C Thank you for writing an excellent article concerning biamping and putting it on to the web for all to see/understand/enjoy. You have confirmed my belief in biamping as an alternative to standard passive xovering. One of the problems I've seen (as a speaker builder) is the complex passive xovers with all the Zobel junk in them. I've seen speaker designs that the xover cost MORE than either one of the drivers (sometimes more than both the drivers). Seems kinda crazy to me. All that loss with the inductors, etc you get with xovers greater than 6 db/ octave.

I'm currently designing a biamped (and eventually a tri amped) system. I'm trying to fully understand the whole section concerning xover freq. I understand the critical range of human hearing part as an area to try to avoid an xover. I'm planning to use one of Vifa's nice 6 1/2" woofers (P17WJ-00-08) xovered to a Morel MDT30 tweeter. The tweeter is good down to just below 2Khz. The Vifa looks good (flat response) to almost 5Khz. Your recommendations make it sound like I should xover at the higher freq, like 4Khz or even 5khz. The pblm I see with most woofers is the not so good response curves for off axis. The Vifa's published specs show the 30 degree response curve start to break away about 3Khz from the on axis curve. By 5Khz, they are approaching 10 db difference. Therefore, before reading your article, I was planning a 3Khz active xover to reduce the off axis response pblms. But after reading it, I'm now thinking closer to 4Khz. Whatta think?

Hey, I'm on a budget, and can't justify spending the big bucks for an "audiophile" system. I have a friend here in town (XXXX, Texas) that has more money in this system than in his house. (Magnapans, All Mark Levisons, including a pair of the monster Levison amps... I think 33s.. I don't remember off hand). I love listening to his system, but little (currently) cheap system is not that much worse than his. But has given me the upgrade fever. I just what to do it at a reasonable price. I guess it is like wine, once you spend a little money on a bottle on wine, you have good wine, the expensive stuff is only marginally better. Almost an exponential curve.


R Glad you liked the article. In my opinion, there is really no better way to go, since as you said, all the nonsense you have to go through with passive xovers with all their matching networks etc just does not seem worth the trouble.

I had a similar question from another reader, and suggested that he would need at least 20uF caps. 20uF will create its own crossover at about 995Hz, and will introduce phase shifts above and below (as always). Only problem is that suitable polyester caps are expensive, but I do believe that using electros is a bad idea in high level signal paths.

I agree with your comments on the Vifa (or any other driver with a cone diameter greater than about 100mm (4")) having to provide decent dispersion at higher frequencies. This has been a major hurdle for me in my quest for the Holy Grail of speakers, and so far I have not solved it either. A vertical array of two 100mm drivers has some appeal, but then you start having impedance problems (4 ohms is not an amplifier friendly impedance. Decisions, decisions.

If you can get to 4kHz xover frequency, this is preferable, but at the expense of HF dispersion and lobing.

With your power supply, consider adding a few extra caps (10,000uF 50V) to ensure minimum hum, and this also provides real current capability in the short term. There are many amps about now which use massive amounts of capacitance for just this purpose, and all the reviews and articles I have read indicate that there is a definite improvement in the sound.



Q I am about to purchase some music equipment (I'm not sure if it qualifies as hi-fi) I bought the loudspeakers first. Contrary to my normal buying habits I didn't do a lot of market research. I listened to a range of speakers in one of the few stores (2 actually, at least that I know of) equipped with hi-fi speakers and ended up with buying a pair of Chario Hiper 1000 towers with the following characteristics:
    Low frequency load: vented NRS/2pi
    Configuration: 3 way vertical array (with the subwoofer mounted in the bottom of the tower)
    Drivers: 130 mm doped paper subwoofer
    130 mm doped paper midrange
    27 mm textile dome tweeter
    Sensitivity: 88 dB SPL 2.83 V/m/W
    Frequency @ -3dB: 55 Hz
    Acoustic crossover freq.: 135-1350 Hz
    Acoustic alignment: LKR4
    Rated impedance: 4 Ohm
    Suggested amps: rated for 50-120 W/8 Ohm
    Cabinet: 20 mm thick MDF (no idea what that means, but I guess it is some kind of pressed wood fiber)

They surely sounded the best of all the speakers available in that price category ( about 900 EURO), but the choice was very limited (one or two competing models). Any technical assistance from the sales crew was absent, although they remained friendly and were at least willing to switch back and forth between the different speaker systems while I was testing them.

From the instruction manual that came with the speakers I learned that they were suited for bi-wiring or bi-amping. I tried to learn some more about these possibilities via the net and came across the article of Rod Elliott. The possibilities of bi-amping sound attractive. According to the analysis (or what I understand from it) the basic idea is to connect an electronic crossover and 2 amplifiers.

My questions to you (and/or anybody else who is willing to help me out on this) are:

1) given that (for the time being) I don't want to spent more than ± 1000 EURO on the equipment (speakers not included) is there any gain to expect from bi-amping my system, say using two commercial low- end amps (e.g. SONY TA-FB920R) or even using my old Philips amplifier (1984) (I know that to some (most) of you this probably sounds horrible).

    N.B.: The other components of the system will be: Sony MiniDisk MDS-JB920, Sony CD player CDP-XB920 (product codes refer to those from the Sony Europe catalogue). Please also feel free to comment on the choice of Sony. Would Kenwood be a better (less worse) option?
2) if the answer to the above question is a yes, I will then need an electronic crossover. Can you get these in specialized shops? do you have to built these yourself? What does it cost?

3) Should I eventually stick with bi-wiring, do I need some sort of signal splitting device (cf. fig 5 at http://sound.whsites.net/bi-amp.htm).

4) can you give me a rough idea what I would pay for (average quality) speaker wires.

5) are there any introductory books written on the issues in Hi-Fidellity and that pay much attention to the basic concepts of the underlying electronics.


A Thanks for your response to my web page. First (and foremost, I guess) has to be a comment about Sony. ( ... this section has been removed - it is not very complimentary and is based on personal experience ...) Otherwise, as I indicated in my article, I am unwilling to make recommendations (or otherwise) about any specific brand or configuration. However - the speakers you mention sound a little suspicious: Sensitivity is 88 dB SPL 2.83V / 1m

Since the speakers are rated at 4 ohms, 2.83V equates to 2W, so they have a real sensitivity of 83 dB /W/m (i.e. 85 dB SPL at 1 Watt at 1 metre). This is a fairly low sensitivity, even by current standards, so you will need as many watts as they will take to get a reasonable SPL in your listening room. Also with a -3 dB of 55 Hz, you WILL need a subwoofer to fill in the missing bottom octave. As for bi-wiring versus bi-amping. Bi-wiring will provide some benefits to overall imagery and clarity - although the benefits are fairly subtle. Bi-amping will provide a potentially vast improvement in imagery, perceived sound level, and overall clarity.

I will not (apart from my first-hand comments about Sony) pass judgement about your choice of amps (see my disclaimer in the web page). What I will suggest is that you be prepared to experiment - most of the advances made in audio have been from just that. If you stay with bi-wiring, your speakers are already set up for this (based on your description). To bi-amp, you will need an electronic crossover. There are many manufacturers of these, but you should try to get hold of one which provides phase coherency. Have a further read of my article to see why.

As for whether it is hi-fi or not - who cares! If you like it, and it makes you feel good (along with the music you like), then all requirements are satisfied. I know this is an heretical statement, but it really is up to you. If you enjoy what you hear, that's all that really matters - think about it.

MDF stands for "Medium Density Fibreboard", which is about 700 orders of magnitude better than "chipboard" - the basic ratty stuff that people make into speaker cabinets, dog houses, junk boxes, etc.


C Thanks for your quick and elaborate reply.

First on how I feel about my system. Well ... after having read that the only components I currently own rate as "suspicious" I think I'll ask myself this question again in the morning.

I got suspicious too, though, after I having found no significant information on this brand of speakers on the net. It's an Italian brand which exists for about 20 years. The company made it's first appearance on the international market in 1986. Their most prestigious line is the Academy series. And I must admit that the more I look at them (I haven't listened to them though) the more I come to appreciate their design. Their walnut veneered finishing goes well with the rest of the furniture. But all this is of minor importance when finally I will be disappointed by the way they sound. But what also matters to me is that the sensitivity of my speakers is what one could expect of speakers within the price category mentioned (I paid them about 900 US$). Anyway, I can't turn them back, so I better learn to love them.

On your calculation. I admit honestly that I forgot all about the basics of electricity (and that for someone who even got a course on MOSFET amplifiers. But that has now been 10 years ago). Their most prestigious speaker has a sensitivity of 91 dB SPL (all the rest being the same). So much for their claim "Acknowledged as the world's best speaker by specialized journalists all over the world". The fact that I "WILL" need a subwoofer sound very depressing since they already have one.

As for bi-amping my system. I will buy a new amplifier anyway, so eventually I just might and try to bi-amp the system using my old Philips amplifier. The only thing is the price of the electronic crossover. It should be worth while buying one in the face of using two fairly low-end amplifiers for bi-amplification. So my basic question to you is, whether the mere splitting of the signal to the different drivers already gives considerable improvement, regardless of the quality of the amplifiers used.


R To give you an idea, to obtain an average SPL (at 1 metre) of 97dB (which is fairly loud, but not ridiculous), you will need an average amplifier power of about 16 Watts.

Since this is the average, you have to assume an absolute minimum of 10dB headroom (20dB is better). 10dB above 16 Watts is 160 Watts, so if you have an amp rated at about 100W into 8 ohms, this will give the 160 Watts into 4 ohms that you need - provided the amplifier is able to drive 4 ohms! Have a look at the table below.

This is in agreement with the manufacturer's claim that amps rated up to 120 W into 8 ohms are suitable (such an amp should be able to give 200 Watts into 4 ohms, or thereabouts). Also note that the above average 97dB SPL is at one metre. As you move further away, you lose another 6dB each time the distance doubles until you are into the reverberant field. This is entirely dependent on your listening room, but you could hazard a guess that for an "average" room (whatever that is), you would enter the reverberant field at somewhere between 2 and 4 metres.

As for the quality of the amps - bi-amping will not make bad amps sound good. The overall result may be better than conventional operation (internal passive crossover), but if possible, the better of the amps should be used for the mid-high section. The better this amplifier is, the nicer the combination will sound. Some distortion (harmonic and frequency) in the bottom end will not sound as bad as in the mids and highs, because the low frequency distortion is "masked" by the clean output of the top end, and the woofer will not be able to reproduce the really nasty harmonics very well anyway due to its restricted frequency range (sounds good in theory, at least).

As for the manufacturer's claims - don't dismiss them until you have had a listen (I don't know these speakers, so I cannot comment - not that I would anyway).

Based on the sensitivity of your speakers, the following shows roughly what you can expect:
 

dB SPL
(at 1 metre)
Single Amp
(Watts - each channel)
Bi-Amped
(per amplifier)
8510.25
8820.5
9141
9482
97164
100328
1036416
10612832
10925664
112512128

This assumes that the crossover is at the "equal power distribution" frequency of about 280 to 300Hz. You will need to experiment to see if your internal sub-woofer can get that high without sounding horrible if you want to get the maximum benefit of the increased SPL due to frequency splitting with an electronic crossover.

This also assumes that you want (need?) high SPL. If you don't, then you will probably not have an issue with power, so can have more flexibility with crossover frequencies.



Q Your article is excellent- very easy to read yet informative! I graduated from the University of Idaho ten years ago - majored in Electrical Engineering. I am a power engineer and am a little nervous about tweaking with my system. I work with kilo-volts not milli-volts. I have Infinity 8A Kappa speakers, 1 set of Adcom GFA-565 mono amps and an Adcom GFP-565 pre-amp.

I currently have the set bi-wired. I noticed a definite improvement after bi-wiring the system. Saw a bigger sound improvement after replacing my 10-year old Denon CD player with a good transport and D/A converter. I am now kicking around the idea of bi-amplifying my set. If I understand correctly, I would have to remove the passive low-pass crossover in series with the woofer circuit and remove the passive high-pass circuit in series with my mid-base coupler, mid-range and tweeters. Then I would install an electronic crossover to separate the signals to the high and low frequency amplifiers. This is correct?

The speaker manufacturer's literature claims that the woofers passive crossover is tuned to extend the bass response by approx. 1/2 octave to 33Hz. Wouldn't I lose the lower frequencies by removing the passive crossover and using an electronic crossover in its place? I know you don't get something for nothing. What is the downside of having the woofer circuit tuned to extend the low end?

Does it make any sense to install an electronic crossover to split the signals sent to each amp and then keep the passive crossover circuits in the speakers? I expect this would probably cause to much of a roll-off near the crossover frequencies.

I appreciate any response you might have. My wife is particularly nervous about me pulling the speakers apart and snipping inductor coils and capacitors out of the circuits. "Don't you think they knew what they were doing when they designed those speakers???!!!"

I know you don't give advice on equipment brands. If this were your system, however, would you consider upgrading to a bi-amplified system or upgrading one of the other components first?


A I'm glad you enjoyed my article. You are absolutely right about not getting something for nothing! The bass extension filter almost certainly reduces the effeciency of the woofer at higher frequencies, so it can "give it back" when it is needed in the last 1/2 octave. If you were to biamp the system, this part of the crossover circuit would need to be retained, or a matching bass boost could be done electronically (which is better, but requires more circuitry and effort, and also means that you must know the exact tuning frequency used, and the amount of loss introduced in the higher frequencies).

All the above is possible, but will take some research and maybe a bit of lateral thinking! A signal generator and oscilloscope or good multimeter would not go astray if you can lay your hands on them - this way you can plot the response of the existing crossover unit - remember to terminate the crossover outputs with a load resistor equal to the loudspeaker impedance, or the results will be meaningless.

If you leave the speakers connected, speaker and cabinet resonances will make meaningful measurements more difficult. If you want to do it this way, make sure that the box sealing is not compromised, and that you drive the system with a power amp to ensure a constant-voltage source. A signal level of 1 or 2 volts is sufficient, otherwise your ears will hate you!

You are also right about having to disable the existing crossover from bass to mid+high. If this is left in place the effect will not be a happy one! Remember that the mid to high crossover must be retained unless you are contemplating tri-amping. Too much too soon I would think.

As for your wife's consternation, she is right, but with a caveat ....

The speakers - like all speakers - were manufactured to suit the majority market in their price range. Few people will jump at the chance to purchase an expensive pair of speakers knowing that they will also have to buy an electronic crossover and another power amplifier to make it work. As a result, the final design is a compromise between simple economics and harsh reality.

Biamping WILL create a dramatic improvement, but it must be done with great care and attention to details, or a disaster will surely be the result. I would not recommend attacking the existing low to mid+high crossover with anything other than a screwdriver (for its removal intact) - this way you can always revert to the original if it doesn't work out, or while you perfect the next stage of your project. (After all, who wants to be without their sounds?)

If this were my system I would biamp first, then experiment with the result until I was satisfied that it could be improved no more. Only then would I look at the other components (which sound as if you have a reasonable investment in them). I hope this helps.


Q I enjoyed your article very much. I am interested in using a Bryston 3B for the bottom end and a single ended triode amp for the mids and upper. can this setup be made to work?

A I'm glad you enjoyed the article. I don't know the Bryston 3B, but I'm sure that the combo will work as long as the crossover point is selected to get equal power distribution. You might find that this will require a crossover in the middle of the "intelligence band", but if well done this can still work - thousands of speaker manufacturers do it all the time, so why not you too.


Q I have acquired a QUAD ELS-57 speakers, that I like very much. Unfortunately, a while ago my old amplifier stopped functioning, and I was advised to buy a new one, instead of paying for a repair. I have searched on Deja-vu News, and found that people suggest certain amplifiers as more appropriate for the QUADs then others. (Bedini 25/25, Spectral DMA-XX were some examples.)

Given the fact, that the input impedance of the QUADs is highly capacitive, and ranging from 2 to 34 ohms, I believe that some amplifiers are more suited than others. What I also noticed was that some, apparently unrelated, people recommended a "fast," "class A," "low power" amplifiers. Do those modifiers have any substance or are they only buzzwords?

Given the price of these amplifiers, is there a DIY amplifier I could attempt to build? Thank you very much.


A Congratulations on your ownership of the QUAD speakers. They are one of the all time classics, but they are hard to drive. Most amplifiers do not like the load presented by electrostatics or anything else which is highly reactive, so I'm afraid that you really do need something a little different in the amplifier department.

As for the "recommendations", they are all meaningful in one way or another, but are completely unrelated to your needs - "fast" is not a requirement, nor is Class-A or low power. If I remember correctly, the original QUAD amp was about 50 or 60W (???), but the biggest problem is finding an amp that will drive the capacitive load without a mental breakdown.

I would tend to suggest an amp which is capable of driving a two ohm load (these are not as common as amps which can handle 4 ohms) since such an amp will probably be also capable of driving highly reactive loads such as the ESLs.

The suggestions you mentioned may well be good, as no doubt they have been tried and tested with the ELS series speakers. Having said that, I cannot make a specific recommendation - although P101 has been tested with good results.

Valve (vacuum tube) amps are usually much happier about driving really difficult loads without undue distress, but I have heard of tests conducted on high power PA amps (Australian Monitor), when they were shown to be completely stable driving a 5uF load. Not a recommendation, but it shows that "solid-state" amps are quite capable of driving strange load impedances.

You did not say what sort of amp you were using - and did it die while driving the ESLs? If so, this shows that the load can indeed "blow up" otherwise good amps. If it is a valve amp - get it fixed. Anyone who tells you not to bother has never heard one. My project pages also have many amps, but other than P101 I have never tested them with electrostatics.



Q Hi, enjoyed your web site. i have built jlh 10 watter using 2n3055 &tip35 but as yet have done no subjective comparison. may also try mjl21194 (bit costly in australia). what do you think

A Did you build the amp from the plans on my site? (I guess you probably did, or you would not be responding to me ...). I know the design is old, but if you find that you like it I'd really like to hear more (I re-published the design, but have not built one yet). A bit of good feedback is always nice, and if you don't mind I will include your comments after a comparison in the Readers' Page.

Using the MJL21194 is probably overkill, but these are one of the most linear transistors on the market - as such I would expect the performance to be about as good as it can be. But yes - they are *expensive*.

Glad you enjoy my site - visit often, there are new things appearing all the time.


Q HI, JUST RECENTLY I PURCHASED A SET OF B&W SPEAKER (DM-603) AND CONNECTED THEM TO MY ONKYO M-508 POWER AMP (2 CHANNEL). I'M A NEWBIE IN HI-FI HOWEVER I'VE LEARNED QUITE A BIT BY READING MAGAZINES AND ARTICLES AND ESPECIALLY YOUR WEBSITE REGARDING BI-AMPING. I'M STILL CONFUSED THOUGH. IS IT POSSIBLE TO BI-AMP MY DM-603 USING ONLY ONE POWER AMP? CAN I USE SPEAKER "A" FOR HI/MIDS, AND SPEAKER "B" FOR LOWS INSTEAD OF BUYING ANOTHER POWER AMP? YOUR ASSISTANCE IS GREATLY APPRECIATED. I BELIEVE IN YOU.

A No, you cannot bi-amp with only one amp. You can bi-wire, but the results are nowhere near as good. I would strongly suggest that you live with your system for a while yet, because bi-amping is NOT trivial - it is possible to damage speakers (or amps) if you are not completely confident, and are experimenting.

Experimenting is good, but please do it with a cheap speaker system and an amplifier you can afford to lose.



Q Was wondering if you had anymore information about designing Capacitance Multiplier power supplies. I am planning on building a pair of Leach amps and am looking at various power supplies. Supply is +/- 57V with average current of around 5 Amps, peak as high as I can reasonably get (The amp itself has current limiting). Any help,information, or recommended reading you could provide would be greatly appreciated.

A I assume that you read the article in my project pages, and this should have enough info for you to do your own design. Most of it is pretty straightforward, and requires not much more than ohm's law and a bit of calculation of capacitive reactance.

Have a thorough read of the article, and e-mail me with any specific questions. I don't know of any other sites offering more info than I have.



Q Its been quite awhile since I've done any analogue design, so I'm a but rusty. I assume for increased current the main thing would be to ensure the driver transistor and diode can handle the increased load. Any suggestions for suitable transistors??

The only thing I'm having trouble wrapping my head around is modifying this for the higher voltage. By the looks of it the 200R and caps remain the same and its just a matter of adjusting the 12k resistor. Do I want a 10V drop between the voltage rail and the driver base, or a 10V drop between the base and ground?? Of course I could be completely wrong :o)

A A re-think on your problem: Since you are running a fairly high voltage and current, you will need some fairly "beefy" devices. However for the Leach amp, I think that anything other than a well filtered supply is not needed - 10,000uF caps are not cheap, but you will need close to that anyway before the filter.

The Leach amp (like most modern designs) will have quite good power supply rejection, and I would only consider using a capacitance multiplier where hum becomes audible.

I have run some tests on my own power amp, and even with quite high ripple on the DC supply, this does not make itself evident until the amp clips. Below this, my distortion meter shows zero 100Hz supply ripple at any power level. I doubt that the Leach will be any different. Save your self some money (and energy) and try it out with a simple supply first. If you are not happy, then look at using additional filtering - I bet you will never bother.

If you want to do it anyway, I would suggest MJ15003/4 for the power transistors (250W) and MJE340/350 for the drivers. The rest of the circuit should work as shown, but you might need to use higher wattage resistors. The 10V drop is between the input and output of the supply, and the idea is to ensure that this voltage is higher than any momentary droops (including ripple) as the amp is driven. This is adjusted with the 12k resistor as you guessed. The diodes do not need to carry the current - they are there to stop reverse voltage if the main filter cap manages to discharge too quickly. They don't need to be changed.


C Vulnerability of Tweeter to DC Transient:

There used to an organisation in the UK called the Active Loudspeaker Systems Organisation (ALSO) which had set a standard for active crossovers, active speakers and amplifiers manufactured for bi-amping. The active speakers are without any passive crossover at all to protect the tweeters from DC transcient. Basically, when complying with ALSO standard, there should be no danger having the tweeters blown out.

I am still using the following antiquated British system which are ALSO compliance:
Active Crossover and power amplifier for one channel:
A&R Cambridge SA60X
Control and power amplier for another channel:
A&R Cambridge A60P
Active speakers:
ARC 050 two way speakers with detachable passive crossovers

A&R Cambridge is now Arcam of UK, and ARC is now defunct. I am not sure if ALSO is still active.


R I did not know of the ALSO organisation, but my suggestion to use protection was based on the fact that most users will not even be aware that a DC "transient" is possible. Many amps produce small transients at power on and off, and these CAN damage tweeters. I prefer to err on the side of safety.


Q I'm designing a subwoofer module, including an electronic crossover. Could you tell me how to take a speaker level input and have that signal go through an electronic crossover? Can I use a voltage divider or is there another way to attenuate the high level signal?

A Generally a simple voltage divider is all you need. This can be set up so it will sum the outputs of left and right amps, and give you mono line level for the sub amp.

If you have (say) 100W / 8 Ohm amps, the maximum voltage is 28V RMS, so you need a divider of about 28:1 - since a lot of bass info is mono anyway, you may need more (you did not say if your sub amp or crossover has a level contol - you will need one). Design of voltage dividers is covered in an article on the ESP site.



Q Your projects page is just what I've been looking for! Do you have PCB's for projects #08 and #09 ? I logged on to Microsofts web page, trying to locate the VB40032.DLL file so that I can use your Linkwitz-Riley Component Calculator Program. However I couldn't find any information on the file. Please, tell me how I can find it. With your help, " My Rig " can sound much better. Thanks.

A Many project PCBs are available (although not at the time of writing).

The VB40032.DLL is available from Microsoft


Q Congratulations on a very good page. undoubtably one of the better presented and more informative on the web. <end gratuitous sycophancy>

I was considering building your project 2 - the preamp and just had a quick question. I wish to eliminate the balance control from the design. As I understnd it, this pot seem to add some negative 'feedforward' which reduces the gain for a given channel. If I simply remove this component and sever the link from the balance pot(s) between the two channels will the circuit stillfunction correctly?

Thanks in advance, and keep up the great work!


A If you don't want the balance pot, just leave it out of the circuit. You will have a little more gain (about 3 dB), but need do nothing else. By the way, I don't really recommend leaving it out, as it can be useful. If you are sure - no problem. There is no feedforward or other nasty stuff involved - it is simply a passive control.

Q Finally just what I've been looking for! A simple, but pratical amp for my turntable and well under $2,000! Tubes are nice but I already have a tube amp with mmphono inputs! I seem to recall seeing something about circuit boards and or parts mentioned on one of your pages. I am interested in a parts pcb package, minus the heavy stuff. Thank you for the excellant info! well I see that you are away for a few weeks, looking forward to hearing from you, I think I'll put my poor bloodshot eyes to bed!

A At the moment, I am not in a position to provide kits in any way shape or form (sorry). I am looking into this, and will do so if there is enough interest (at the moment, there is not, unfortunately).

Which amp were you looking at? I am currently examining the possibilities of a new amp (based on the 60W design) which should be capable of 100 to 150W into 8 Ohms. It's only on the drawing board at present, but should be quite nice when completed. (This is the P3A amp, which has been available for some time now.)



Q Hi, i read your article on bi-wiring/amping and am very interested in it. i was thinking of getting the paradigm moniter 7's or 9's, which can be bi-wired. i have a yamaha r-v1105 reciever and an old denon avr-1200 reciever. the yamaha r-v1105 has preouts, which i can connect to the denon avr-1200. now, is this and some speaker wire all that i need to biamp????

so now am i right in assuming that the volume dial on one of the recievers would control treble, and the other the bass???

i am really unsure if i know what i am talking about so i would appreciate any information. thankyou very much.


A Be very careful here. The concepts of bi-amping and bi-wiring are very different. The bi-amp approach requires that you do not use the crossovers in the speaker cabinets at all - you need an electronic crossover. It is possible to bi-wire using separate amps, and there might be some advantages over "simple" bi-wiring, but you will not get the full benefit without the electronic crossover.

The whole idea is to make one amp and speaker combination responsible *only* for the bass section of the frequency range, and the other only handles the mid+high component. I suggest that you re-read the article very carefully - there are lots of traps for the unwary! If you do it this way, then what you suggest is correct. This is not really bi-amping though, since all frequencies are still handled by each amp. You may get marginally less amplifier intermodulation distortion, but will not gain the real advantage of true bi-amping.



C First of all, great site! While I really haven't got much experience with bi-wiring, I do love to bi-amp. I build my own speakers and basically am put off by the cost and technical complexity of low frequency passive crossovers.

All the info you've provided is wonderful. I'm thinking of building that phono pre-amp (if I ever aquire the requisite technical expertise to do so).

Here's a couple of jokes for your humour section (maybe?):
A conversation between two audio research technicians (the field of study, not the company):
Tech 1: Gee, this new horn speaker is efficient. At what db level does it become dangerous?
Tech 2: What?

Okay, that was bad. But the following is a play on words (or a saying) and technically accurate.
Q: What did the chemist say when his assistant fell into a vat of acid?
A: Apparently, the problem is, you are part of the solution.

Well, thanks for a great site. I thoroughly enjoyed the visit.


R Thanks for the vote of confidence (and the jokes).

It's not only the complexity of passive crossovers, you are just losing so much other good stuff - power, clarity, phase (etc, etc ....)



Q I've been combing the web, looking for an answer, but just can't find anything. Your page appears to be the best on just about everything else, so I guess I'll check if you know.

Basically, I want to construct a Hafler Dynaquad style surround matrix processor. The catch is, I don't want to do it at speaker level. I'd like to do it at line level. I have enough amplification available, and I figured it wouldn't be a difficult item to construct.

The problem is, I can't find any info to guide me. There was, within the past five or so years, a product called the Phasearound, which operated on this principle. According to a review I found, it used an audio transformer to derive the center and surround channels, but the article didn't elaborate beyond that. If I knew more, that would've been enough. I don't need a center channel, as I intend to use it in an audio only system. Any help you might be able to offer would be appreciated.


A I think I know exactly what you are looking for. The circuit is easy to build (it was also published in an Australian electronics magazine very recently), and I will look at making this a new project within the next month or so (if you can wait that long).

The magazine article also describes a digital delay - this apparently enhances the "surround" effect quite a bit, by delaying the rear channels. I will not be including this in my project (or I might, I haven't decided yet). (See projects index for details.)


Q I was searching through altavista and found your site, which is definitely the best among the many I've found. I'm not a analog engineer, so there are many things I don't understand, but I find your site very informative. I just have one question. It might have been answered in your site already, but after reading it over and over, I couldn't come up with a definite answer.

Would bi-amping without an active crossover be possible? (source - 2 amps - passive crossover built into the speaker which is bi-wireable) and would it help?

It'd be like a "temporary" path between bi-wiring and true bi-amping. I'm not sure if the HI and LO input of the speaker are truly independent or if there are some interactions between them.

BTW, I have Magnepan 1.6s, which are quite obscenely (what a word!) inefficient at 83dB, 1W, 1m and I find my Bryston 3B-ST struggling (or so I feel anyway). Thanks for your time.


A You are right - what you are looking for is not in there (at the time - the information has been there for some time now).

The approach you are asking about has been suggested by a few other readers as well. It is not really bi-amping in the true sense of the term, because the amps still have to reproduce the full frequency range - its just that there is no actual power being produced at the out of band frequencies. There is probably some merit to doing this, but you will *not* get the power advantage, which is important with really inefficient speakers.



Q Thank you very much for your informative article on Amplifiers. I am an Engineering student in Washington. Over the next few weeks I will have a great deal of time to experiment and would like to built an Amp. I was wondering if you have a complete schematic of your design as it seemed the design was complete but in pieces that were spread across the document.

I am also interested in whether your design could be applied multiple times off of one power supply as I would like to build a 6 channel Amp for Dolby Digital and THX. If you have the information available what is the approximate cost of all of the components for your Amp design. Again thank you for taking the time to write your guide.


A Thanks for the nice words. Which article were you interested in? There is one that describes the general design goals, but that is fragmented (as you said). If you want complete schematics, you need to look in my projects pages, since this is where all the complete constructional articles live.

I suspect that this is not where you were looking, since these are complete in all respects.

Any number of amplifiers can be run from the one supply (with some interactions of course), but as you do need to ensure that the transformer, rectifier and filter caps are sufficient for the total loading. One trick used by a lot of manufacturers is to use a *really* big tranny, and then use separate rectifiers and filter caps for each amp.

To give you an idea, using 5 x 60W amps needs a transformer of about 500VA for continuous operation at full power (I would not suggest anything less than this). The filter caps for each amp should be at least 4,700uF for +ve and -ve (i.e. 2 of them) for each amp (10,000uF is better, but expensive). Rectifiers should be at least 10A continuous rating - most manufacturers use less that the above, but you get better reliability when you use bigger rectifiers and caps. I hope you have some money, 'cause this will not be cheap! (It will be lots of fun, though.)


HomeMain Index
ContactContact ESP