TGM10 - based on NAIM by Julian Vereker

Status
Not open for further replies.
Jump to Latest
#442
I don't think any commersial designer will learn anything here except tastes and prujudices of a buyer. As I said Crimson seemed to be a product that everyone liked. Kendle Perry who was at Burr Brown last I heard taught me electronics from the Crimson circuit. It was exactly where to start and impressed Kendle. John Deans my friend from SSL also had worked with Crimson and said it blew away anything he knew from the Pro Audio World. To be able to satify both the technical and subjective side of the industry is to truely win. The technical part is not the problem. It is knowing how to spend the money and to know you have the right to believe in your product. Sugden amps really were the unsung hero's.
 
#452
mjona said:
Possibly diamond transistors if you look at the price🙂
Click to expand...


You may remember years ago I joined the NAP 140 kit thread when one builder had too many volts from his transformer. I suggested a regulated supply to all but the NPN output transistors. I used the very circuit that in 1972 college told me was for historical interest only. An amplified zener. I argued if the transformer large enough this might be better than the NAP250. I was exspecting at least one of you to try this and get very excited. Not a word. The irony is Julian thought his thoughts this way. In cooking this mistake is not knowing how to cook. You have alll the exact right things. You might even have a cook book. What you never had was time with a chef. The largest improvement to a Naim you ever could make is to better Julians power supplies. When Holden and Fisher went bust Julian was very worried others would get their hands on his idea. As far as I know this was making DCR as low as one dares go on the transformer primary. How " you " might do this is the largest transformer your money might buy ( > 500 VA ). You might have to fit a 10R 16A thermistor and 30A contact relay to remove the thermistor after >1 second to allow power up ( circa 1M 1 uF if using NE555 ). That's a bit of a guess as to where you would go with that. Julian was always very willing to talk about things. He told me he always told everyone everything as he knew they would never believe it. He even gave factory tours to competitors. He then would borrow an example of their product a year later and find they never had listened to a word. Whilst you might think Julain had a very big head I would say he was genuinely trying the raise standards. Like the Naim sound/look or not the engineering was excellent within it's concept. That was not one penny spent where it wouldn't help the sound. He liked 3 other components not made at Naim. Linn LP12, Quad ESL 57, and Sugden amplifers. The latter for a man like Julian was unexpected admiration.

Never make the power supply capacitors too large without hearing the Naim version first. You might find the amplifier sounds slower. I reduced the size of caps in an NAP 250 as that's all my friend could afford. We used BBH 10 000 uF 100V which fit like a glove ( on specail from Farnell at £10 at the time, high ripple current type with screw terminals ). The amplifer sounded more open with a slightly faster bass quality. On the whole it was better. I suspect the high frequency performance of this capacitor was better as many high voltage types are. They have to be to prevent flash over. If you can't get high grade caps often high voltage ones will not be far off.

Andrew. Do you have any idea of a good Vbe bias circuit when about 2V ? Even then I am not sure how best to track an all NPN output. All I know is the Naim one isn't the best. I don't hate the Sinclair/A60 idea of a fixed resistor and the current source/sink sensor. Arcam felt it worked best 1 mm from the heat sink. If not thermal modulation could happen as it does in chip amps.
 
#453
If you mean me, then I read about all the complicated multi-transistor style of Bias Voltage generators.
I have used a couple that came with the PCBs I have assembled.
But I have never designed anything more than a single transistor version for my own generators.
D.Self and R.Cordell have much to say on this.

The only thing I will demand is to ensure the bias is not undercompensated for temperature changes.
 
#454
That's exactly what I found and thank you Andrew and it was you I hoped would answer. I have built the ones shown by Douglas Self and can not say they were as he described, they did drift down with time I seem to remember which is a positive. I like MOSFET's for this ( e.g. BUZ900/905 ) as they are not fussy and can be set to taste.

I have real doubts I should ask you this. Without saying exactly if you don't want to , what was your favourite Civil Engineering project. I am sure people here would like to know a little. An engineer is an engineer and Civil engineering is not easy. One of my bosses was a Civil engineer although he graduated as a mechanical engineer. In the Army they adapted his skills to build the bridges when pushing into Germany. Finally he became an electro-mechanical engineer at Smiths speedometers. He taught me mechanical engineering. When I read the turntable section at DIY Audio I am so pleased he taught me. It was mostly what not to do. Like be careful with PTFE as it is not always a good engineering solution.
 
#455
Nigel,

V interesting!

Hugh Dean also discovered there is a proper value for the rail caps. But back in the AKSA days I read builders finding a proper size for the transformers not necessarily the biggest though.

The NAP 140 bias comp is a kludge for sure. I'm not sure why. But in the project of this thread I'll be making what I believe is a better way. In simulations it works well and in practice it will have adjustable temp comp so I can tailor it as needed through a simple resistance change.

But a better design for the Vbe multiplier comes from Haggerman - read about it on my TGM5 thread.

Yes, I am also cautious with PTFE as it is stretchy!!
 
Last edited:
#458
nigel pearson said:
I don't hate the Sinclair/A60 idea of a fixed resistor and the current source/sink sensor. Arcam felt it worked best 1 mm from the heat sink. If not thermal modulation could happen as it does in chip amps.
Click to expand...

The following page to that I posted earlier illustrates this approach. However the output stage is a triple. I have seen the same used in conjunction with complementary feedback pair output stages in another Linsley-Hood circuit.

He used a conventional generator for his quasi-complementary design. There was a mild tweak which seems to have inspired other designers in this regard.

I rate the Leach Low TIM amplifier as having a good bias generator scheme. I could never induce mine to get very much above room temperature.
 

Attachments

Status
Not open for further replies.
Top Bottom