I mean it was fully complementary and balanced circuitry for me after that for anything I did for myself. For my job, I designed for a company called ILP, making modules. There, cost, cost and cost were the main drivers.
Is there an ILP " like " circuit we could see? I could imagine Ian ( ? ) wouldn't like that. The ILP was encapsulated.
HY2000 ILP ELECTRONICS, Audio Power Amplifier, 1 Channel, 40 W, 7 Pins | Farnell element14
HY2000 ILP ELECTRONICS, Audio Power Amplifier, 1 Channel, 40 W, 7 Pins | Farnell element14
No, I can't show you a circuit and it would not be the same as the more modern module that was referred to. Ian would definitely not like it, especially as I decided to set up a power supply company to sell to one of his customers.
However as an example, the 100W module used fully complementary darlingtons (BDX34/35?). The output stage bias was fixed by design at 0mA (no adjustments allowed). The was a single dominant pole for stability, but this was from the very output to the base of the VAS as it would allow the VAS collector to slew quickly in the dead zone. We could then meet <0.1% distortion at full output. (1% at 1W!!!). There were three stages of gain and the front end was a LTP.
PCB, heatsink, potting compound, all components, label and packing box cost less than £1 (1980 prices).
Quite frankly I'm surprised the failure rate wasn't astronomical. I would never take one home, let alone pay for them as I valued my loudspeakers.
Member
Joined 2009
Paid Member
Last edited:
Member
Joined 2009
Paid Member
That is the sort of comment many would either accept in good faith if you were a Naim devotee. A good many would find that unconvincing.
Anyway, Silicon has very few free electrons.
To improve conductivity an impurity or dopant has to be infused - commonly Phosphorus which has more free electrons to form N material - Aluminium to create a deficiency of electrons to form P material. Boron can also be used.
N or P regions can be made highly conductive depending on the doping levels used.
Phosphorus and Aluminium differ in atomic structure. Each has 2 atoms in first energy orbit and 8 in the second.
In the lowest energy 3rd order there are 5 and 3 atoms respectively. Consequently the doping levels for N material and P material cannot be the same for the same conductivity level.
Also electrons in N material have greater mobility than deficiencies or holes in P material.
Member
Joined 2009
Paid Member
I see no value in worrying about perfect complimentarity as the main result if the transistors are well behaved is more 2nd harmonic distortion. Having a bit of 2nd order is something that we have said was a design goal here, ot at least we suspect that JV was not adverse to this. The bigger concern are the higher order harmonics. They should quickly diminish in significance.
Last edited:
Well now. I have read several pages of his forum from the link you gave. I am afraid that I do not think I agree with any of the suggestions he makes there. Each idea is not bad in and of itself in a different context, and could be found in an electronics textbook. That's not the problem. The problem is that he doesn't appear to understand the Naim design and so misjudges the effect of those changes. He presents with total confidence and some poor sods are swallowing it whole and are keen to maim their own Naims.
"Naim maiming" 😀
I was alluding to reasons for choosing a quasi complementary output stage - there is a bit more on this in an article by John Linsley-Hood see attached.
On the LTP one has to consider what orders of harmonics an output stage is capable of generating and out of phase signals generated by a dynamic load such as a loudspeaker.
You have an output coil and a zobel network to mitigate against the latter. The original Naim circuit had a 0R22 resistor instead of the coil.
An LTP is good at reducing even order distortions however the issue of high frequency odd order nfb signals generated by the output stage is another matter.
From what I have read JV sought to use output transistors which had the best switching speed - analogous to removing this to a higher frequency domain - enabling a safer cut off frequency ceiling for the amplifier to work below
Adding the 22k collector resistor to the inverting input transistor gives rise to a large Miller capacitance between collector and base which makes it necessary for high frequency signals of whatever order (small anyway) to be of higher magnitude to get make it through to the error signal correction process.
The LTP is balanced since the Miller capacitance between collector and base requires a charging current - which reduces the current flow in the of the inverting amplifier collector resistor. Miller capacitance depends on the voltage seen at the collector during operation and is thus a dynamic factor
To me the intent of the design is to preserve the features of voltage feedback at frequencies where it matters - and - by means of some current feedback at the inverting input transistor - to subvert unwanted artifacts at higher frequencies from upsetting the nfb process.
This will affect even order harmonics if these are present at the same high frequencies where odd order artifacts are present.
There is better matching between NPN devices at higher frequencies if Linsley-Hood's assessment is correct.
Last edited:
Yes, showing that was the main reason for drawing attention to the thread. There have been many disputes here, over subjective v objective sound quality standards and showing how skilled manipulation of distortion can subjectively improve the sound quality of very ordinary sounding domestic power amplifiers. Of course, professional engineering folk detest and mock the practice but for personal use, audio products can be quite different (e.g. the revival of tube and hybrid amps) to those in pro. applications.
I am not involved in any commercial arrangement.
I resent the way Members use the Forum for commercial gain.
Are you not an electrical engineer Andrew ? Or time served in some way?
I am so happy our designer from ILP is here. ILP transformers set the benchmark for our industry. The power amps for all of their faults were exceptional value and so easy to use. All the utility of a Quad 405 for pennies and very good specs. I belw up a few until talking to the boss of ILP and him giving me my first indepth understanding of the thermal problems. What I didn't realise was he was retelling what he had learnt from " gholl". His wife gave me to understand Ian spent his life designing things and perhaps designed for the sake of it. It seems that was not exactly true. There was no DIY Audio then to pass experiance on so we learnt where we could.
If the wife of our designer was working at ILP then maybe I spoke with her.
I am so happy our designer from ILP is here. ILP transformers set the benchmark for our industry. The power amps for all of their faults were exceptional value and so easy to use. All the utility of a Quad 405 for pennies and very good specs. I belw up a few until talking to the boss of ILP and him giving me my first indepth understanding of the thermal problems. What I didn't realise was he was retelling what he had learnt from " gholl". His wife gave me to understand Ian spent his life designing things and perhaps designed for the sake of it. It seems that was not exactly true. There was no DIY Audio then to pass experiance on so we learnt where we could.
If the wife of our designer was working at ILP then maybe I spoke with her.
If a special quasi device had been made no one would have known if like a darlington it wasn't PNP. Precaustions for stablity taken. A T03 pakage would have been large enough.
Here is an ILP like device.
100V Power Op Amps, High Voltage Op Amps, High Current Op Amps
Here is an ILP like device.
100V Power Op Amps, High Voltage Op Amps, High Current Op Amps
My profile tells everyone that looks that I am a retired Civil Engineer.
Yes, time served but to be a Civil Engineer.
But the reason you are asking is: can a practising Electrical Engineer be a Member of the Forum and NOT use the Forum for commercial gain. Yes, they can. And many do, they give their time and expertise for free.
Last edited:
- Status
- Not open for further replies.