FYI JV talks about how he got started and touches on design decisions.
Malcolm Stewart interviews Julian Vereker - Audio Counsel
Malcolm Stewart interviews Julian Vereker - Audio Counsel
The use of a sensitive clip indicator which also stretches the pulse to100msec or so, will easily show if the amp is unable to drive the load for any reason. If you need to differentiate between voltage clipping and protection clipping it is not difficult to use an LED as part of the VAS buffer current limiter. This needs to have the correct threshold so that voltage clipping is unable to light it. If this LED lights alongside the clipping LED you have a protection circuit problem.
There is no reason why single slope VI protection needs to impact on the sound as long as the VI line is in the right place.
There is no reason why single slope VI protection needs to impact on the sound as long as the VI line is in the right place.
Voltage and Current sensing to keep the output transistors within their safe area operating range - SOAR being a parameter given in power transistor data sheets.
The standard protection circuit was devised by A.R.Bailey in the mid 60's and widely adopted.
The system works but the SOAR of a transistor plot has a dog leg look about it. The problem is dynamic with more than one risk base to cover. Covering all in a blanket approach can be unnecessarily restrictive - particularly on current.
To avoid nuisance triggering Naim added tantalum capacitors to slow the response - making the trigger less hairy and allowing higher current delivery for a brief period.
Member
Joined 2009
Paid Member
Member
Joined 2009
Paid Member
Member
Joined 2009
Paid Member
Images of NAP 140's show a tinned wire between the capacitor terminals the form the earth with various earth returns soldered to it.
Andrew Russell - pseudonym "Bonsai" has a self explanatory piece on how to sort earthing which I prefer see http://hifisonix.com/wordpress/wp-c.../How-to-wire-up-a-Power-Amplifier_Updated.pdf
I substituted .models for the parts you used for your "My Naim" circuit - there is a little rounding in the square wave image as in the case of my other simulation.
The square wave can be improved by changing some capacitor values (and one resistor value) as I mentioned before - if one dares, meaning if you still had your 140 clone with the "dark" tone you dislike.
Unlike most amplifiers there are no output stage local decoupling capacitors in the NAP140 (C10 and C12 in the diagram in post #1).
I have been down that route with an amplifier I built. C10 and C12 were pigtail electrolytic capacitors that were going out of stock for over the counter sales.
Unfortunately I dropped a screw driver inside which caused one of these to blow up - I removed them all to re-audition and left things that way as the change softened the presentation and made it more relaxing.
From a quick look, Bonsai's grounding diagram is pretty much the same as Self's. Including the T for less distortion. Nothing new under the sun... In the old Naim brochures is written that the amplifier earth is in the preamp. The power amp is basically floating. That's how I did it. Just to make sure that we talk about the same thing, I call earth the point where the amp "ground" (or the "return") goes to the wall earth pin.
As I mentioned before, changing Cdom or making input or otput (the Zobel) to allow for higher frequencies didn't impress me. I don't miss the band between 20kHz and 50kHz. It's something else maybe in the low treble or higher midrange that wasn't right. Most likely poor distortion distribution, unusual harmonic structure or who knows what. Otherwise, when measured across a load resistor, the amplifier was perfectly linear.
Anyway, this boards are gone and I don't want to contaminate this thread any more. It's all written many times.
I'm much more curious to hear about listening and comparative tests of Bigun's new project.
Last edited:
Hello Gareth
Initial observations only. There is no output inductor or allowance for speaker cable inductance and R47 and R28 are missing a zero in value.
Michael
Last edited:
Member
Joined 2009
Paid Member
Yes, I've found that the output inductor reduces phase change at h.f. but the values of R47 and R28 (part numbers from simulation not from pcb) don't provide much benefit, more of fine tweaking. What I don't like is the bod plot does not show a nice monotonically decreasing OLG near the limit of unity gain bandwidth.
Power Supply
---------------
I've looked at simulated PSRR (attached), giving the main rail caps parasitic series resistance of 100mR and parasitic series inductance of 1uH. I've compared a) the original NAP 110 / 140 type circuit where there are just two main rail caps (I guessed 4,700uF but could be an underestimate) after the rectifier and no isolation of the amplifier front-end with RC, with b) my version with CRC after the rectifier (both caps 4,700uF) and 220R + 330uF to isolate the amplifier front end. The latter has the most impact on improving PSRR over the main audible range and I'll use a high quality cap here - probably electrolytic rather than tantalum.
This will surely impact the sound, my clone may not sound as 'relaxed' in the bass region as the original if I populate the improved front end isolation but hopefully it will retain a lot of the overall character in the mid-range where I expect this amplifier to shine the most.
Attachments
Gareth,
I have had to look for my references from Australia's Silicon Chip Magazine. I have some links to their 2011 article about amplifier stability that may be of interest to you.
A Look At Amplifier Stability & Compensation - July 2011 - Silicon Chip Online
The first two pages are available online.
I posted two more in separate posts on another thread www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1739.html#post4192526 and
www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1752.html (post 17515)
Michael
I have had to look for my references from Australia's Silicon Chip Magazine. I have some links to their 2011 article about amplifier stability that may be of interest to you.
A Look At Amplifier Stability & Compensation - July 2011 - Silicon Chip Online
The first two pages are available online.
I posted two more in separate posts on another thread www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1739.html#post4192526 and
www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-1752.html (post 17515)
Michael
Member
Joined 2009
Paid Member
There's no way this type amplifier has OLG of 120dB, no way at all, I'm not sure where your expectations are coming from. You need a very different topology for those kind of gains. I think 50dB to 60dB is already pretty good (similar to AKSA amplifier by the way). You can push it up to 80dB with an EF buffer on the VAS. Of course, the Sziklai lower-output has local feedback, probably around 40dB but we don't 'see' that in the main loop.
Hey, thanks for the links, I'll check them out.... whoa, one of these threads has 19,000 posts already, where the heck does one begin to find nuggets in that one ?
Last edited:
Member
Joined 2009
Paid Member
This thread moves fast. Somehow I forgot the Bode plot you posted and in "Senior Moment" looked at the Supply simulation by mistake. I can laugh at myself.
Last edited:
Funny add. Those extruded aluminium (aluminum to you) shells are pretty sturdy.
Hey, when are your PCBs coming?
Member
Joined 2009
Paid Member
The pcb house is shut-down for Chinese New Year (it's a week long celebration). They come back to work at end of this week I believe. Then they'll start processing the boards for me. They they'll ship them to me. I reckon it'll be around 3 weeks. During that time, if I'm not too lazy, I should prepare the chassis (lots of drilling and filing will be required), get the parts in-house, perhaps install the power transformer and connectors etc. Actually, as I think about it, I'll be hard pressed to get it all ready before the pcb's show up due to day job and kids.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.