Amplification Best Approach for Efficient Active Speakers

What people are working on at the moment
User avatar
rowuk
Old Hand
Posts: 453
Joined: Sun Jun 01, 2014 2:50 pm
Location: Germany

#16 harmonic distortion?

Post by rowuk »

If we cross the tweeter over at 10k, the first harmonic is 20k and no issue.....

Horns are at best decade devices. Their useful range is max 3 octaves. Cross them over too low, and you get unhealthy honk.

Direct radiator woofers have a tough job if we expect them to provide infrasonics. It normally makes sense to have a separate channel for under 40Hz.

Best amps for those efficient speakers does depend on what one calls "good". Large scale symphonic music is much different than chamber music, is different than jazz, rock or pop. As no driver is really Neutral, amps must be selected for complementary behaviour. Using high order DSP requires a different approach than first order passive.

Probably the biggest difference is the learning curve for our ears.

Alex Kitic who has quite a few interesting posts here is really the only designer with a big family of amps with similar sonic signature. His RH amplifiers would be a great starting place for multiple amplification.
Whenever I feel blue, I start breathing again.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8869
Joined: Mon May 21, 2007 9:42 pm

#17

Post by Paul Barker »

Yes Ed I was listening for the piano harmonics on the behmas at owston but piano was absent. Someone either played an electric piano or the beymhas did not deliver the harmonics in that room position with that Amp and source. I was quite happy with their rendition of strings. There is plenty of capability for life like instrumentation from them. But they didn't reach the necessary heights at owston with sufficient amplitude to match the behavior of the acoustic instrument. Plasma tweeter might be the match to take over and provide the piano with its vast spread of harmonic tones.

None of this means that I could bring speakers to owston which would do the job. Also doesn't mean the Behmas don't do something great within their best frequency range at Owston and at a wider frequency range in a smaller room.

They aren't a lot different to the first impressions of Andrew I's AMT's at Witton. "wow that is a cool sound at that limited range? how are you ever going to match that with your other drivers?"
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#18

Post by steve s »

2/3rd of the a pianos 1st harmonic are under the radar of the beyma's, so you cant really blame the Beyma for that, what it does highlight is that the 15"driver is not really dynamic enough just below that.
I had no issues with the 12 inch drivers in the 400-1kh range...i've been putting some thought to that and already have some plans
After adding some super tweeters after Owston there was subtle differences at the top end too.
I can see me being back with 4 ways again soon,

As 2 ways, the beyma's are still bloody good base, but at a level that somebody always picks up on the discrepancies, and sections of the frequency range that are a little quiet,
and it turns out to be program dependant... so its really easy easy to miss..
I will get there...
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
User avatar
Nick
Site Admin
Posts: 15707
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#19 Re: harmonic distortion?

Post by Nick »

rowuk wrote:If we cross the tweeter over at 10k, the first harmonic is 20k and no issue.....
That is assuming a brick wall filter. And if its a passive filter after the amplifier then you will get loads of harmonics you probably would rather avoid.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#20

Post by vinylnvalves »

I know its going off topic, but marrying up drivers with the TPL isn't easy.
I had tried running my Altec 414'z with the TPL's as a 2 way ( 1100 -1600hz) it was not bad, but lacked the authority and realism that putting the JBL 2482's in between adds. ( even with the issues of driver spacing to accomodate a 700mm horn)
I tried some audax PR170MO's in place of the JBL's to get a more compact arrangement. OB with a short wave guide they were almost as good as the JBL's - and are a solution i could live with.

The altec is supposed to be as good as it gets for a 12" going above 1200Hz.
I was paranoid about the top end of the Beyma as it decays from 10k. I tried adding some JBL 2405's, but found that i prefered it without. I also contoured the reponse adding a shelf from 10k in the DSP to bring the response back upto flat. Both myself and "golden ears" Dave could hear notice any difference.

Back on topic i noticed Bottlehead are going a little 4 watt D Class amp - http://bottlehead.com/quicksand-battery ... ier-specs/
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#21

Post by vinylnvalves »

Having now taken in what everyone has said on this forum (and others) There are a couple of options.
1) Build/have built a valve multichannel amp to cover Mf and Hf. Space constraints dictate only 2 pairs of monoblocks. (one of those driving the bass channels) Cost if built for me, time and knowledge if i build it
2) Build a multichannel SS amp using established designs kits JLH 10 watt, Le Monstre. Kits off ebay - add battery/SMPS/ linear power supply
3) Purchase an existing commercial product that fits requirements.

I know that the sound quality will almost be be in the same order as above.

I am almost going for option 2 if i can convince myself that the quality of the replacement semi-conductors are sonically as good as those impossible to source originals. I cannot afford the time or cost to build a good valve amp currently ( have an ongoing health issue which is zapping all my spare energy)

A curve ball for option 3 has appeared in the last couple of days which may be a good compromise. Whilst looking at low powered SS amps. I can across "Monarchy SP 70 Pro" which are a 25w SS zero feed back Mosfet amp, which is meant to be very musical (according to reviews). I think i remember these shoe box amps from shows a while back. Coincidentally there happens to be a pair on ebay (not a bargain price though).
Having read a review regarding the issues with noise with K horns and the manufacturers solution, i contacted the manufacturer and got a positive response back to my issues and a potential solution. The amps use OP amps in the driver stage, and the suggestion is the change the feedback in this stage to reduce the gain and improve the s/n ratio. Looking at the Op amp spec sheet it goes seem that they can be used from a buffer upto a gain of 100, so tuneable.

Does modifying the gain of the input stage (feedback or changing OP amp) make sense sonically?

I know this option is potentially sonically the worst or the options. but i feel that i would get more sonically more out of finishing some elliptical 200Hz horns i am designing to reduce driver spacing. (More of a stop gap - until my health is restored.

Ps - If anyone has a pair of these amps gathering dust let me know.
User avatar
rowuk
Old Hand
Posts: 453
Joined: Sun Jun 01, 2014 2:50 pm
Location: Germany

#22 Considerations?

Post by rowuk »

Actually if you are considering multichannel (and in my opinion should be), you won't find consumer 3 channel amps to really fit your bill. You need something "substantial" for driving the woofer, the rest does not need horsepower to move lots of air, but does need the sonic qualities.

If it were a multichannel valve amp, 1x EL84 (Pentode like Alex Kitic RH84) + 2x PCL86 would do it for the rest at a very modest cost. The EL84 would need an output transformer with higher inductance, but the horn mids and tweeter get by with much smaller ones.

Rereading what you originally posted, there is a lot that does not make sense. The amps that you have should offer at least "very good" based on what you have posted. To straighten things out, you need to look at one problem at a time.

The noise on the 6BG4 amps - could be the amp, but could also be the DSP. The only way to know is to short the RCA inputs on the amp (alligator clips on the RCA inteconnects) making essentially a perfectly quiet input. There is no danger for the amp when doing this. On the midrange horn, no hum should be audible through the speaker as it would be out of the pass band of the driver and horn. Any amp induced hiss would be present. If these amps pass the midrange test (and should), then you can plug them into the woofer and see if there is any hum. If there is, then it should not be difficult or expensive to fix this.

You can test all of your other amps on the mid and woofer too.

There are plenty of issues with DSP that are far more serious than any unfounded feelings about dynamics. If we use the digital gain control, we increase noise for those channels. There is a great mod to the Behringer DSP that adds analogue volume controls to prevent the digital mess.

With a bit of common sense, and perhaps a small investment, you can probably stay with what you have and get great results.

A quick and easy method of working out crossover points (regardless of amp technology) is to invest in a MiniDSP. It is not the absolute, ultimate in sonic quality but is inexpensive and lets you model everything that others only conjecture about. 48dB/8va crossovers are easily modelled, up to 128dB/8va can be realized using biquad programming. In my case, after the MiniDSP, I ended up with passive 6dB/8va crossovers based on what I learned.

When "reviews" call an amp musical, I usually run in the opposite direction. That is normally a description that the amp has serious problems and they needed something good to say because the manufacturer took out an ad in the magazine..............


I read an excellent recommendation at The Goodsoundclub.com website that matches my listening experience: SE valve amps for high efficiency speakers with AlNiCo magnets, SS for ferrite magnets, Neodymium requires experimentation.

If a manufacturer uses a noisy op amp, then what can you expect from the rest? A more feedback bandaid does not make the amp better. Noise should NEVER be an issue in good SS design - regardless if it is discrete or opamp. SS can have other issues, but noise has been solved for decades in quality stuff. Your original post complained about noise - why even think about tolerating it here?
vinylnvalves wrote:Having now taken in what everyone has said on this forum (and others) There are a couple of options.
1) Build/have built a valve multichannel amp to cover Mf and Hf. Space constraints dictate only 2 pairs of monoblocks. (one of those driving the bass channels) Cost if built for me, time and knowledge if i build it
2) Build a multichannel SS amp using established designs kits JLH 10 watt, Le Monstre. Kits off ebay - add battery/SMPS/ linear power supply
3) Purchase an existing commercial product that fits requirements.

I know that the sound quality will almost be be in the same order as above.

I am almost going for option 2 if i can convince myself that the quality of the replacement semi-conductors are sonically as good as those impossible to source originals. I cannot afford the time or cost to build a good valve amp currently ( have an ongoing health issue which is zapping all my spare energy)

A curve ball for option 3 has appeared in the last couple of days which may be a good compromise. Whilst looking at low powered SS amps. I can across "Monarchy SP 70 Pro" which are a 25w SS zero feed back Mosfet amp, which is meant to be very musical (according to reviews). I think i remember these shoe box amps from shows a while back. Coincidentally there happens to be a pair on ebay (not a bargain price though).
Having read a review regarding the issues with noise with K horns and the manufacturers solution, i contacted the manufacturer and got a positive response back to my issues and a potential solution. The amps use OP amps in the driver stage, and the suggestion is the change the feedback in this stage to reduce the gain and improve the s/n ratio. Looking at the Op amp spec sheet it goes seem that they can be used from a buffer upto a gain of 100, so tuneable.

Does modifying the gain of the input stage (feedback or changing OP amp) make sense sonically?

I know this option is potentially sonically the worst or the options. but i feel that i would get more sonically more out of finishing some elliptical 200Hz horns i am designing to reduce driver spacing. (More of a stop gap - until my health is restored.

Ps - If anyone has a pair of these amps gathering dust let me know.
Whenever I feel blue, I start breathing again.
brig001
Old Hand
Posts: 550
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

#23

Post by brig001 »

With the efficiencies you are talking about here, could you do something like this and not have a power amp at all? http://www.audio-talk.co.uk/phpBB2/viewtopic.php?t=4891

The op-amp in the schematic is actually the last op-amp in the active crossover and the current buffering is inside the last op-amps feedback loop. The simulated distortion is basically that of the op-amp itself, and I expect that the noise is too. If you can't get to that op-amp in your DSP, there's nothing to stop you adding one.

I use it with my 98dB @ 1W speakers and it goes plenty loud for a house. Pity I couldn't make Owston really so you could have a look and listen.

Brian
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#24

Post by IslandPink »

Dear V-n-V .
If you're going to use a 'compromise' solution & maybe an off the shelf amp, just make sure it's going to work well at low levels . You need low noise and low distortion at small signal levels . Many SS amps do NOT have low distortion at <0.1W level . Any that run output devices in Class A should be OK, but do some research first.
"Once you find out ... the Circumstances ; then you can go out"
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#25

Post by vinylnvalves »

Did not get the monarchy amps as they went for almost as much as a baby Krell on another forum. Wasn't worth the risk, as my logic is if i buy something reasonably priced then i can move it on without much financial hardship.

Having been in contact with someone who has some BD Swings, and has tried lots of amps with them (more money than me) We came up with a short list,
Hagmann BYOB, Le Monstre 20w, 10Y SET amp or something using 4P1L. So a mixture of chip SS and SET amps, not familiar with the valves in the latter two. Building from scratch again, expect for the first one which is too expensive for me.

Another curve ball - noticed a Edison 12 for sale on another forum at a reasonable price. Fondish memories of one of these a while back. Would this amp be a good platform to modify to get something tweeked to my requirements (triode strapping, removing gain stage adding a choke to the psu etc) I know the transformers are not UL but were meant to be good quality.

The logic behind this through is these come up frequently so getting a second one to do my 4 channels would not be an issue. The box is big so can get other bits in like DC heater psu's etc. The only concern i have is that its PCB based, not as easy to mode.
But the million dollar question is it a sow ear, or can it be a silk purse?
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#26

Post by vinylnvalves »

Having looked around i have drawn the conclusion that there isn't a commercially available amp (that i can afford which will do what i want)

Having looked around the options suggested EL84's or DHT, i cannot see much more complexity with DHT's, and have been put off going down the exotic tube routes just due to cost. However one suggestion i had off another site, and i see its been discussed on here, but not sure if anyone built one is the 4P1L SE amp. As its weedier than a 45 would probably go for a PSE, as then it seems i am into fewer winding OPTS's.

Before i go and start this project i need to get a stable design to start from, I have seen numerous designs which all seem to be evolving, is there a schematic i can build with confidence.

The biggest expense is of course going to be the OPT's, and recommendations, bearing in mind i will be buying 4. I have looked at the usual suspects Llundhals, Tributes, but maybe would want to use something a little cheaper, around £100 per channel. Anyone got experience of the James ones, or Transcendar ones, aware of hidden costs. alsohave seem some on ebay from canada "Tomiko" ones for about the same monies. Any recommendations - or if anyone has any on a shelf gathering dust....

Almost ready to go an order a dozen valves or so. (don't want to fall foul or things getting popular and either being to expensive or unobtainable when i need replacement tubes)
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8869
Joined: Mon May 21, 2007 9:42 pm

#27

Post by Paul Barker »

Mark "Pink Label" likes the O'Netics ones. Bud Purvine the maker is a member here. He has given me lots of helpful klnowledge and shared a great deal with the whole diy community.

At an entry level amplifier the different transformers don't have a huge amount between them, there are small differences. you have to spend fortunes to get really big differences. Chasing differences elsewhere can be cheaper, you know I mean resistor choices and coupling caps..

I am sure your chosen route is plenty good but I would advise a beginner to build with more standard dht's like 2a3 6b4g 300b. Almost any make is worth having these days.

again you can chase differences but once again massive cost hike to get small difference.

NOS 6b4g a good place to start. Or if more power needed cheap 300b.

Nothing wrong with your chosen valve. Just seems odd to start with something off on a tangent with few time proved schematics.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#28

Post by IslandPink »

I wouldn't discourage anyone from building with 4P1L's given that they sound great and cost about £20 for an 8-pack ! PSE will require a bit of care in matching though . I use 4P1L in pentode mode as a driver for DHT's, and really like them . NOS 6B4G's and Sovtek 2A3's would be my other choices ( as Paul suggests ) - very good but quite a bit more expensive .
For cheap decent transformers , JamesD always recommends 'AskJanFirst' , who has a lot of off-the shelf audio gear . Just google his site. I've bought Sylvania 6B4G's from him - very helpful guy . For 2x 4P1L I suppose you'd only want a standard 3K transformer .

I've seen a lot of circuits on DIY Audio for 4P1L's, but people will insist on stuffing them full of Mosfets, which sort of defeats the object of having a DHT amp . We can give you some pointers . Andy Evans has got his head screwed on, check his circuits ; but something fairly straightforward like a 6N1P at 10ma should be able to drive a pair of 4P1L's and give a decent amount of gain given the mu of 10 for the 4P1L . If you need less gain for the upper frequency channels then some obvious driver options are 6N6P, 7A4 , 6SN7, 12SN7 etc . Using 4P1L as driver is an option but will require a very clean current-sourced heater like the ones Andrew has developed . You should have a couple of pairs of those anyway for the output stage if you want to get the best of the DHT ! That's one of the 'difficult' bits for a DHT circuit, which are otherwise quite simple . The heater can be make-or-break for the sound .

But please check with BudP what his price for a standard 'level 1' pair of 2A3 transformers would be - they are not at all expensive - I had a pair of level 1 PP transformers for 2x300B and I think they were less than the Lundahls . They completely outclassed and embarrassed the LL1620's !
"Once you find out ... the Circumstances ; then you can go out"
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#29

Post by vinylnvalves »

Taking all the advise on board and contacting various people, i think i now have a foundation to start from. Discussing it with Andy Evans here is his latest version, which has only had minor tweeks over the last year or so.

Having gone around the OPT's - the Transcendars seem to good to be true, looks like i will be able to get a couple of pairs for £360 including customs. The James ones are a little more for the 6123's, and as heavier cost more to ship. The Onectics are twice the price again. Maybe i will try and pick up (borrow) some cheaper OPT's for the initial testing. Don't mind paying for good ones after i know i like the sound and see it as a goer.

The DHT heaters seems straight forward enough, Rod Colman has a very comprehensive set of instructions.

The HT PSU i would like a bit of inspiration for.
Was wondering whether i could initially use the mains transformers from my old mullard 5/20's
0-410V recitfied through a hybrid semi/EZ80 would give me around 300V (might just be enough)

If i cannot utlise what i have got then i may go SS regulation with an HT delay. As i am sort of with Morgan Jones there isn't a good reason not to. As one of my design requirements is to minimise heat generation, its a serious consideration.
Attachments
4P1L_PSE.jpg
4P1L_PSE.jpg (10.7 KiB) Viewed 7105 times
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#30

Post by vinylnvalves »

Got a pair of Lundhal interstage transformers off ebay, for the 4P1L PSE amp. What i want to know is there a way of checking these transformers ( make sure coils aren't shorted etc) before i give the feedback. I know using a multimeter will probably magnetise the core, so is there another way? I don't have a scope or signal generator to hand though.
Post Reply