How To rise sensitivity of aleph -x?

[elviukai]

hi everyone, I having a problem . my dac and cdp do not drive aleph x to its full potential.it is (22V rails, 12A Bias per channel) they are able to privide about 1.5-3V( at least manufacturer said so) and I would say that this voltage drives aleph to 40W or so

I want to rise aleph x input sensitivity but not decrease input impedanse. how to this? somebody advice to increase 100Kohm value I increased to 130kohm and get 11V DC at speakers output.
I wrilly need to increase, because I dont use preamp.. aleph 5 drives with no problem with 3v line output.
please help and thanks in advance

 

[gl]

Hi elviukai,

You need to increase both R16 and R30 from 100K to a larger value. Did you change both resistors or just one? The 100K value gives 20dB of gain for a balanced input. Increasing these resitors to 220K increases the gain to about 26dB. Increasing the resistors to 130K probably won't produce too noticeable a change.

Cheers,
Graeme

 

[elviukai]

hi gl, thanks for help. of course I increase both 100kohm resitors (in both sides,) or You meaning some others resisors? I was surprised to see dc showing at outputs.. to what value I should go 150kohm, 220kohm?. I know that lowering too much will be negative afects??? I am about to redesign my dac to pass labs dac1 which will have 150ohm output impedanse, so then maybe I can reduce 10kohm input resistors thus lowering input impedanse an rising sensitivity. or I am wrong?

 

[gl]

Hi elviukai,

Changing only one feedback resistor might have been the cause of your 11V offset. Have you tried re-adjusting the input diff pair CCS pot V2. I wouldn't change the input resistors because that would change other things that shouldn't be changed. You are using 22V rails so I would put in 220K resistors if it was my amp. In fact that's what I have in mine. Search for AX100.

Cheers,
Graeme

 

[elviukai]

hi I read all your thread(I wonder why I did not saw it before) and find VERY useful things.

but I notice few diferences with my amps. seems no one has this problem. dc ofset (between plus and minus of spekers outputs) depends and wht source I conect. for example amp with conecded one cdp has stabile 30mV and when I am conecting my dac I get 8V! dac has coupling caps.

also my aleph xa monoblocks are big and they run at lot of bias. 2x 2kva toroids 22V rails. 24 devices per channel with 0.5ohm source resistors. 12 amps do yours new schematic is ok For highly biased aleph x? I will experiment with this maklohan resitor and output to ground. because amp will not blow. others things I will better leave for electronic proofs, because just too many fets was blowed..
also question - for source resitors I use precision vishay dale wirewound resistors 10W. 1ohm . because I have bunch of them and I paraleled them. so total 0.5 ohm 20W. is this bad? maybe I can remoove one resitor and left only 1ohm per device but run not 0.5V but 1V throught this resistor? how to do this?

I would love to do this because -
they are very big
each of them costed 8EUR(10usd) 48 devices and 96 resitors is to litle expensive

 

[gl]

Hi elviukai,

I will try to answer your questions.

First, if you have only 30mv of relative offset with the CDP then then 11V problem is coming from the DAC. Perhaps there is a grounding problem.

You can use ideas from my amplifier if you wish. It really is very little different from the GR original or the Hifizen schematic. I don't like Kari's schematic. My advice (and you won't like it) is to read through the original Aleph-X thread and the WIKI and use that information as the basis of your design. The Aleph-X is a very advanced DIY project. You can see by the many Aleph-X threads that in general people with more experience have better luck with it than people with with less experience.

The maximum voltage you can use across the source resistors is .7 volts because that is the Vbe of the control transistor. So your idea of using 1 ohm resistors at the same bias level won't work.

I read your other threads and I don't understand why you are using so much bias. Your 22V rails will result in a voltage limited output somewhere between 90W - 100W. With 12A of bias this means that you can drive a load of less than 2 ohms to this output level. I will assume that you have a good reason for this.

I hope this helps.

Cheers,
Graeme

 

[elviukai]

well reason is that my speakers has 2.5-3 ohms impedanse at 20-40hz. exactly where it has problems- lean low octaves. evrything from 60 to 13khz can ripp the wals off with big orchestras. but not a 20-40hz. I tried standart 22v 5A bias aleph x and was surprised in how low qty bass fell in. but it was tought and faster than my aleph 5. thats I value very much. so I did experiment. with the same 8 fets(transformer was single 1.6kva) per channel I wind down 5round of secondaries and get about 16v at load.and rise bias up to 8-10A channel . I just paraleled 0.39ohm with extra 0.39ohm. results was very interesting- at LOW levels amp sounds very good, and at all regions- bass become more weighty and puncy, sounstage become even more airy and focuses. and sound was very delicate and dark. but playing at louder levels - 4ow or so it souned like its clipping.and discortions rise VERY much. no need for measuring devices. actualy I listened only 10-20 minutes(when I fix problem- at first my rectifiers do not like 20A countiniuos current and after 30-60 seconds wires reached soldering temeperature and soldered out, I tide them) after 20 minutes heatsinks reached unaceptable temperature. then turned amp of and lowered bias to 6A and discortion dissaper. I hope it was because fets was dissipating too much bias. abecause generaly I liked what I heard, so investigate litle time and money and bought 2 x 2kva 18v trafos build 2 bigs chasis an directly mount 24 devices per channel . today I started only one channel. I having to much wires.. there goes about 20 wires in one monoblock from pcb to fets. all devices fets, source gate resitors, output, output to ground are not in pcb) I need redesign this.

what makes me sad that I saw a 4.7uf capacitor at the input in yours schematic. I understand that it blocks dc voltage and improove things but try to explain this to sick audiphileS (like me) who dream nightmare about capacitors .lot of caps in series with signal.. capacitor like drug. it must be there where is not posible to survive without it. otherwise using too much drug can increase ilness posibility. . I already have capacitors in signal in speakers, in CDP( DAC actualy soon will be capacitor free with output transformer) and luckily I have no preamp but adding one more(no matter how expensive and how nice cap logo) in power amp can be dangerous for my listening expierences.

speaking about aleph amps and especially x I wounder how they deliver such a punchy and speed beeing SE and not PP. all others SE comercials,diy amps I have heard no matter tube or SS lacks of dynamic, mostly due to low power and other things. and Tube and SS PP designs are very good in home theater to reproduce canon shot. but not buterfly fly. pass amps can reproduce 70% of canon shot and 90% of buterfly.

sorry for such a speaking but I am newbie at solid statte elctronic and even do not know basic princips.


I just calculated 12A and 22V rails to target best this 2.5-3ohms region..

 

[gl]

Hi elviukai,

Thank you for the information.

You are making huge changes to your amplifiers and spending large amounts of time and money. It is also clear that there are things you are still learning. You say that you are a newbie. I don't want to discourage you but you would probably be better off to simply copy one of the existing successful designs and use that even though it doesn't have the current output you need - at least until you learn enough to understand more of the intricacies of the Aleph-X circuit.

In all of your posts you do not mention adjusting V1, V2, and V3 or of optimizing the Aleph CCS current gain. For instance, if you left R12 and R34 at the values specified in Karis schematic while you made all the changes you describe, then that is probably where a large amount of your distortion is coming from. I am suspicious that Kari's gain resistors aren't right even for his amplifier as is.

My advice would be to cut down the bias to 6 - 7 amps and spend some time optimizing the CCS current gain and other operating parameters.

Worry about the capacitors after everything else is working optimally. People make too big a fuss about this. Pass Labs uses polypropylene film caps on the input of the XA series amplifiers and those units get rave reviews from all over the world. This is good enough for me - at least for now until I run out of other things to tweek.

Cheers,
Graeme

 

[elviukai]

hi again and thanks for helping. I dont know what is v1 v2 v3 in yours shematic, but lets say V1 and v2 is for each side. and v3 for dc ofset output. after warm up I targeting to adjust the same current across source resistors in both sides(a 0.5V or so) with theese and then with v3 I adjust output dc ofset to lowes value simulating or conecting real speker and with conected input)

I read 30pages of karis thread and find that he used Kristijans schematic(like I do) but I didint find values and where stands
r12 r34??
someone told me changing bias do not need to change anything but only source ressitors. seems it is not true. looking at schematic I see that mainly onlymac milan resitors are changing,and source resistors.
if you tell me (basic things) or what voltages where I must get I will try to redesign.
I contruct exacxt Kristians alepx x with good parts and 22.5v rails and 5 A bias. and I dont like sound very muc. after adding bias things change..
now would be silly to leave things out. too big toroids and too big chasiss. any info regarding what parts need to change when ading more mosfets/uping bias are very welcome.

I dont think I will be qualified on solid state amps. I can build(to friend and for myself I build about 25 variuos alephs) but I can not unerstand basic operation princips too litle time. all free time I spending on speakers designing. greets

 

[elviukai]

hi again if i am not mistaken on comercial ax 160 I see 10x 1kohm resistors in paralel making nice 30W 100hm resitors from output to ground. and also 10x 1ohms forming 0.125ohm 30w resistor at output.
on comercial xa 100 10x resitors output to ground and 6x at output.at river boards I see 3 capacitors more than in DIY aleph -x. 2 electrolytic and one near diferential pair and single irf9610 and input film cap.

 

[gl]

Hi elviukai,

I will try to answer all your questions one ata time.

First I need to apologize. I did not mean Kari's I meant Kristian's (sp?) schematic in my previous comments. My apologies to Kari as well - I am not familiar with his work. There is no feedback from builders (including Kristian) that the values on Kristians schematic produce an optimal amplifier, For instance with a CCS that runs at 50%. On the other hand many people have demonstrated that they have built Grey's ar Chad's (Hifizen's) exact schematic and have verified that things work and sound right.

All the parts references I use like R16 and R34 are from the original Grey Rollins or Hifizen schematics. That is what people on this forum use as a common refence for the Aleph-X.

In general I start by assuming that any change to the Aleph CCS will change the current gain and that I need to measure the gain and adjust it. Note that changing the bias current through the output mosfets will change their transconductance and therefore the current gain. This all discussed in detail in the big Aleph-X threads.

The bottom line here is that Aleph-X is simple and straightforward to build if you build a proven version. As soon as you start to change things whether it's the number or value of the components or change the bias level then things quickly become very complicated especially for someone without specialized electronic knowledge. That's why I gave you the advice I did. I am trying to help you get the most out of the time and money you have invested.

Many people have modified Grey's original Aleph-X for various reasons of their own and created a new Aleph-X implementation. I believe that there are two keys to doing this successfully. First, a person needs to read and understand every single word written on the subject. Second, a certain almost artistic balance must be maintained regarding the electrical operating parameters or properties of the circuit. Stay within the allowable limits and you have a state-of-the-art power amplifier. Go outside the limits to any degree and you don't.

Cheers,
Graeme

 

[elviukai]

hi GL , really thanks for helping.. all I need is to build two mono amps to mach speakers.I can use basic one but its just doesnt mach to woofers. woofers are capable o lot of more speed and deep. with standart version 74kg speakers with 4x dynaudio woofers each 29mm max sounds like litle monitors. I tried them biamp with no success. also tryed triamp(one aleph 5 for hights and mids, and dedicated aleph 5 amps to every woofer) triamp helps a lot. but this is nonsense to spent money on three amps. also my dac did not like feed 3 amps. there no way to go back. I already spent about 1700eur on parts and have to build this. if I will not have succes I rebuild this monoblocks to simple standart 2 stereo aleph x amps, or built aleph 1.2 or similar. frankly speaking I dont need any power higher than 80W/8ohms speakers has real 90db sensitivity, but I need enought power in low impedanses. and dont want to loose midrange and hight purity. how to shoot two rabits at once?
seeing shematic there is no big diferences in aleph x schematic varying ,only few resistors. . actulay it reminds me very much alephs but with symetric balanced schematic.
all I need is few encouraging words, and maybe help a bit if where to moove if I will went wrong. I can hear diferences in sound, but my ears can not subsititut osciloscope. I can reduce bias to 10Amps but can this really help.my amps has 24 output devices and that is really diferent from 4 greys or 8 kristijans fets. yours is more or less more close. also find one in pass diy galery close to yours. but seems no one tried higher bias on aleph x. everybody talking that increasing the bias sound is better. in ol aleph -the same. when I asked nelson about driving 2 ohms, his answer was short- "want to drive 2 ohms- double output devices"

lets say for starting value - using 24 fets , 22V rails what ressitor values I need to change. except source resitors of course. maybe when diferential pair drive more outpur devices it need diferent values resistors. comparing old alephs I notice that.
so any value to start??
BTW what does input caps in signal if in cdp or dac stands output cap also, why two times use for the same reason?

sorry for so much questions.