any reccomended mods for the elliot P3A?

[demogorgon]

Hi!

i'm in the process of (trying to) making a pcb layout of the P3a by Rod Elliot, and wonder if there are any mods that will lift the performance of the amp up a notch or two, or anything i need to know before making the amp?

I guess i should build the original amp first, and then mod it, but since i'm making my own pcb, i want to be able to incorperate the mods to it, and not etch a new pcb, de and re-solder all teh parts..

all help is appreciated.
marius

[hienrich]

its a bootsraped vas, try constant current sourcing the vas.


but that original vas works dependably need not be change.

or may be, try a higher voltage rail.



hienrich

[AndrewT]

Hi,
leave room for components that have a different footprint, extend traces to allow caps that have different pin spacing. Allow space for a full 4 component output zobel, although most can be fitted to the output terminals. Diodes across the NFB cap. 1extra NFB resistor with //cap location. Feedback link from VAS collector to -ve input with room for series cap & resistor. link // to LTP load resistor to run series cap & resistor from Vrail to VAS base.
LTP and VAS emitter degeneration resistors. All electrolytics // film cap. I agree ccs for VAS and add VAS protection transistor.
All these are optional and can be left blank or short link.
There are probably many others that may turn up at debugging time
regards Andrew T.

[demogorgon]

Originally posted by hienrich


its a bootsraped vas, try constant current sourcing the vas.


but that original vas works dependably need not be change.

or may be, try a higher voltage rail.



hienrich

bootstrapping.. have no idea of what that is.. going to have to get some literature on that..
been poking around at the school library and asking the clerk to buy in some books on amplyfier construction and principples. I'm looking forward to digging in on knowledge..
isn't bootstrapping putting a small cap from colektor to basis?
"google, here i come"

,

Originally posted by AndrewT
Hi,
leave room for components that have a different footprint, extend traces to allow caps that have different pin spacing. Allow space for a full 4 component output zobel, although most can be fitted to the output terminals. Diodes across the NFB cap. 1extra NFB resistor with //cap location. Feedback link from VAS collector to -ve input with room for series cap & resistor. link // to LTP load resistor to run series cap & resistor from Vrail to VAS base.
LTP and VAS emitter degeneration resistors. All electrolytics // film cap. I agree ccs for VAS and add VAS protection transistor.
All these are optional and can be left blank or short link.
There are probably many others that may turn up at debugging time
regards Andrew T.

thanks Andrew, all your tips will be kept close at hand and googled on extensivly.


regards
marius

[Greg Erskine]

Originally posted by demogorgon
bootstrapping.. have no idea of what that is.. going to have to get some literature on that..
been poking around at the school library and asking the clerk to buy in some books on amplyfier construction and principples. I'm looking forward to digging in on knowledge..
isn't bootstrapping putting a small cap from colektor to basis?
"google, here i come"

While your library is getting you the books, have a look at the Leach Amp documentation. It explains the workings of a amp from beginning to end. It is nicely structured and well written.

http://users.ece.gatech.edu/~mleach/lowtim/

The Leach and P3A are at the opposite ends of the complexity scale and I would try to keep things simple for your first go.

On the P3A, R9/R10 and C5 forms VAS bootstrap. The amps I like best seem to have a VAS bootstraps, it could be a coincidence though. So don't throw them away without trying them first.

BTW: I've never heard a Leach so I can't comment on good it is.

Also, Elliott has 2 t's but I've noticed that nobody ever gets it right.

[demogorgon]

Originally posted by grege


While your library is getting you the books, have a look at the Leach Amp documentation. It explains the workings of a amp from beginning to end. It is nicely structured and well written.

http://users.ece.gatech.edu/~mleach/lowtim/

The Leach and P3A are at the opposite ends of the complexity scale and I would try to keep things simple for your first go.

On the P3A, R9/R10 and C5 forms VAS bootstrap. The amps I like best seem to have a VAS bootstraps, it could be a coincidence though. So don't throw them away without trying them first.

BTW: I've never heard a Leach so I can't comment on good it is.

Also, Elliott has 2 t's but I've noticed that nobody ever gets it right.

Grege!

thanks for your reply!
i didn't quite understand the reason for bootstrapping, what does it do? oh well, dont bother, i'm shure there is tons of literature allready written on the subject just waiting for me.

I'l be looking into the leach amp for shure, maby it's even a worthy amp building? but the p3a first.

and for misspelling mr. elliotts name, both you and he just have to exuse me, two t's at the end of a name is a gramatical phenomenon not commonly observed in my part of the world

-Marius

[PB2]

Originally posted by demogorgon

i didn't quite understand the reason for bootstrapping, what does it do? oh well, dont bother, i'm shure there is tons of literature allready written on the subject just waiting for me.

I'l be looking into the leach amp for shure, maby it's even a worthy amp building? but the p3a first.

The P3A is very much like the Citation 12 that we've been discussing in the other thread but the P3A has been made complementary CFP as I suggested. It's interesting that in simulating a quasi complementary design like the Citation one can observe the differences between emitter follower (EF) and CFP. CFP has many advantages which I've also seen in testing on the bench.
I think you might find Self's page more applicable since he offers a detailed analysis of this common topology and offers enhancements. I'd use Self's diff amp and darlington VAS. I would add diff amp emitter degeneration, but that lowers forward gain, an emitter follwer before the VAS helps, but a current mirror is needed to make it effective and raise the forward gain back up. The bootstraped VAS and the CFP output are fine in the P3A. I'd double up on output devices if you plan to use low impedance speakers:
http://www.dself.dsl.pipex.com/ampins/dipa/dipa.htm
The changes I suggest would require fine tuning and stability analysis.

[sam9]

If you like bass, you can reduce low freq distortion by increasing the value of the input cap and the NFB cap. This doesn't alter their basic function which is DC blocking. I first came across the practice of useing really big caps in these positions when I took out a sub-woof plate amp. So checked the effect in a SPICE simiulation and later in real life. The issue is not just bass extenstion. Small values (4.7uF at the input and 220uF at the NFB tend to dominate THD by raising the noise floor all the way yp into mid-bass. I think I can hear a difference but others say no.

Self, for instance used 1000uF on the NFB cap. This brings up some compromises about foot print vs. working voltage. Self compromises by using a bunch of protection diodes.

At 1000uF I would think a small value (0.1uF) ceramic or metal foil cap in parrallel ight be a good idea as a precaution to keep the NFB behaving at very high frequencies. I've never heard any mention of this issue elsewhere, but it certainly can do no harm.

[demogorgon]

Originally posted by PB2


The P3A is very much like the Citation 12 that we've been discussing in the other thread but the P3A has been made complementary CFP as I suggested. It's interesting that in simulating a quasi complementary design like the Citation one can observe the differences between emitter follower (EF) and CFP. CFP has many advantages which I've also seen in testing on the bench.
I think you might find Self's page more applicable since he offers a detailed analysis of this common topology and offers enhancements. I'd use Self's diff amp and darlington VAS. I would add diff amp emitter degeneration, but that lowers forward gain, an emitter follwer before the VAS helps, but a current mirror is needed to make it effective and raise the forward gain back up. The bootstraped VAS and the CFP output are fine in the P3A. I'd double up on output devices if you plan to use low impedance speakers:
http://www.dself.dsl.pipex.com/ampins/dipa/dipa.htm
The changes I suggest would require fine tuning and stability analysis.

Originally posted by sam9
If you like bass, you can reduce low freq distortion by increasing the value of the input cap and the NFB cap. This doesn't alter their basic function which is DC blocking. I first came across the practice of useing really big caps in these positions when I took out a sub-woof plate amp. So checked the effect in a SPICE simiulation and later in real life. The issue is not just bass extenstion. Small values (4.7uF at the input and 220uF at the NFB tend to dominate THD by raising the noise floor all the way yp into mid-bass. I think I can hear a difference but others say no.

Self, for instance used 1000uF on the NFB cap. This brings up some compromises about foot print vs. working voltage. Self compromises by using a bunch of protection diodes.

At 1000uF I would think a small value (0.1uF) ceramic or metal foil cap in parrallel ight be a good idea as a precaution to keep the NFB behaving at very high frequencies. I've never heard any mention of this issue elsewhere, but it certainly can do no harm.

you both are talking above my head now, i'm still not on a level that allows me to fully understand what you are saying, but getting there. the thread is being cept for future refferance though offcource.

until later
Marius