Opamp suggestion for DC-servo in an Alexander amp

Status
Not open for further replies.
Jump to Latest
peranders

peranders

Member
Joined 2002
sjostromaudio.com
Paid Member
#1
#6
Hi Peranders

The danish HI-FI magazine Highfidelity once had an project on this amp, the people who builded it all complained about lack of deep bass but were much impressed at it`s hights. I wewre close to build it but didn´t. I am very interested in your oppinion on the sound on this one. Mark Alexanders article is very interesing indeed.

Regards
 
#7
I helped a friend in the 80's and the output stage used MOSFET's but I don't remember which type, could have been IRF some type which ELFA had.

I never got a chance to really test this amp but it measured and behaved good.

When somebody is used to certian sound and them tries something new it's hard to say what the problem really is.

Thin base and good high range may also be an indicator of that the normal amp has bad treble performance.

Deep bass (technically) isn't a problem because the amp is totally DC connected. Bass down to 0 Hz is possible without a problem.

So at the moment I have a neutral attitude towards the Alexander amp. Current feedback is interesting and I will dive down more into the subject. I have found lot's of info on the net.

BTW: I have also got the real original article from the AES.
 
#8
opa-277? very low offset, Iq < 1 mA , Ib quite small.

elfa. 73-387-59

I was also considering building this amp (according to the HF articles) until I heard about how many who had problems with the output stage oscillating and sourcing the output transistors.
could also have been due to the fqact that the output stage was connected to the main board with wires e t c...

Anyway, I have also seen a 200 W version of this amp with IRF 640/9640 in the output somewhere on the net, some years ago. ¨
I am almost 100 % certain I have the schematic printed (IRL ) on paper...

/rickard
 
#10
Wow! Someone actually building the Alexander amp.
I'm waiting with baited breath to hear your reactions.

Patent: You need no permission from ADI to build this
amp as long as it's not "commercial". I believe that they
expressly give permission for any person to build one
for their own hobby use.

Servo amp: Low quiescent current is important as you
want the main op-amp to modulate as much of the current
in the current mirrors as possible. That's why the OP97 was
chosen.

Main op-amp: The SSM2131 was SSM's "improved" version
of the NE5534, designed for high end pro audio equipment.
If I were to build one of these, I'd consider an OPA627 or
AD8610, etc.

PCB layout: Alexander's paper has a very small and tidy
PCB layout; one board for the low-level stuff and the other
for the output devices. Also, good hints for physical layout
and routing of power wires -- same stuff that Doug Self
recommends.

BTW, "SSM" stands for "Solid State Music", a small audio-
oriented semiconductor company that PMI ("Precision
Monolithics, Incorporated", for you youngsters) purchased
shortly before ADI purchased them.

Does anyone know the wereabouts of Mark Alexander?
 
#11
Per:

>The Alexander amp is a high speed design so I think the physical design is really important. No hard wiring!<

If you combine hardwiring (component leads only, no separate wires) with low-impedance ground and power planes, and design a tight layout, you can make some pretty high-bandwidth circuits. I've designed and built all-discrete 600MHz amplifiers like this.

If anything, I find it easier to go faster using 3-dimensional circuit structures rather than keeping everything on the pcb (2D). A calculated mixture of smd, leaded board and 3D also works well, and is how I normally design.

hth, jonathan carr
 
#12
Alexander amp

I have also build this amplifier a year ago with six pairs of IRFP240
and IRFP9240.
The driver transistors are ultra fast BF758 and BF762 they have bean used in tv as uhf amplifiers, i think they are discontinued.
I have used OPA604AP from Buur Brown as gain stage instead
of SSM2131.
I did not use DC servo beacose offset was always beatwin +/-
30 to 50mV.You have to place trim pot beatwin pins 1,5 for offset triming.
i dont know why peaple say that amp oscilates, its very stable
i also didnt use zobel network its not nesesery this is high bandwith amp,and you dont want to put low pas filter in the output.
you also must take care of grounding of the driver board,i did
had a problem with input left open it is very sensitive beacose
input capacitor(750pF) is a filter which make this amp goes to
1MHz.
The amp is unbelivably fast(transient reasponse) if you put some
CD with acustic gitar or piano or some pasage with some good solo drums it realy scare you.
The amp reali have total control of the speakers, if you never heard amp with high bandwith and high slew rate(250V/us)then you peaple dont know what you are mising.
i wil send some foto's in a few days (driver and power print).
I am also suprised how few peaple have build this amp.
 
#14
Emsad said:
I have also build this amplifier a year ago with six pairs of IRFP240
and IRFP9240.
The driver transistors are ultra fast BF758 and BF762 they have bean used in tv as uhf amplifiers, i think they are discontinued.
I have used OPA604AP from Buur Brown as gain stage instead
of SSM2131.
I did not use DC servo beacose offset was always beatwin +/-
30 to 50mV.You have to place trim pot beatwin pins 1,5 for offset triming.
i dont know why peaple say that amp oscilates, its very stable
i also didnt use zobel network its not nesesery this is high bandwith amp,and you dont want to put low pas filter in the output.
you also must take care of grounding of the driver board,i did
had a problem with input left open it is very sensitive beacose
input capacitor(750pF) is a filter which make this amp goes to
1MHz.
The amp is unbelivably fast(transient reasponse) if you put some
CD with acustic gitar or piano or some pasage with some good solo drums it realy scare you.
The amp reali have total control of the speakers, if you never heard amp with high bandwith and high slew rate(250V/us)then you peaple dont know what you are mising.
i wil send some foto's in a few days (driver and power print).
I am also suprised how few peaple have build this amp.
Click to expand...

Old tread but still interesting.

Ik heb deze gebouwd met ring emitters en een super emittervolger, opa604 en opa07 servo, heb de rc tijd vergoot ivm aantasting bas. Ik gebruikte geen current mirror icm bjt ipv mosfets.

Ik heb nog steeds geen betere gehoort ook geenb buizen, impuls muziek als gitaar, er is geen betere.

I have build these with ring emitters and a super emittor follower opa 604 and opa 07 servo, have increase the rc time becausem it did affect the bas. I did not use the current mirrors because of the bjt and there current driving, not voltages.

I do not now is this also protected by patent because it is quite changed for a part.
 

Attachments

  • ScreenHunter_2357 May. 04 13.18.jpg
    ScreenHunter_2357 May. 04 13.18.jpg
    492 KB · Views: 192
Last edited:
Status
Not open for further replies.
Top Bottom