New Posts New Posts RSS Feed - 1970s Design Indulgence
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Club

 

Open to all owners plus those contemplating the purchase of a Graham Slee audio product wishing to use our loaner program: join here
Subscribe to our newsletter here (Rules on posting can be found here)


1970s Design Indulgence

 Post Reply Post Reply Page  <1 133134135136>
Author
Sylvain View Drop Down
Senior Member
Senior Member


Joined: 18 Jan 2010
Status: Offline
Points: 177
Post Options Post Options   Thanks (0) Thanks(0)   Quote Sylvain Quote  Post ReplyReply Direct Link To This Post Posted: 14 Mar 2021 at 5:53pm
..............So, to obtain a continuously rated 40WPC, the transformer must be at least 255VA, and perhaps 225VA (a "preferred value") might also hold up. 300VA will do 2 x 50W into 8 ohms, but is often the size found in 2 x 100W into 8 ohms amplifiers, and usually only 225VA..........

''.......circuit of a channel in its latest form...'''''

...was able to identify that teh diagram relates to one channel after confidence

Ok so far 
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 15 Mar 2021 at 9:27am
One of the most challenging printed circuit design tasks is the separation of functions such that destructive mutual inductance cannot occur.

There has been a tendency to ridicule printed circuit designers for laying out circuits as they appear in circuit diagrams. If a circuit diagram is drawn indicating the separate functions, then it isn't a bad practice. On the other hand, blindly following an arty circuit diagram can lead to erroneous results.

Low current circuits set inside high current board areas will suffer induced currents. However, keeping the voltage amplifier stage away from the current amplifier stage is difficult, seeing that the voltage amplifier's load is the bias source of the current amplifier.

However, we can draw the line at the voltage amplifier's base junction, such that the collector load can be inside the current area, which means the small signals are kept outside the higher current loop.

We can define four areas as being:
Input stage and VAS up to the collector
The current amplifier
The HT supply
The output components

Even so, the inter-dependency means some items must bridge a loop to "escape" that loop.

The academics way, using CAD, is to draw the schematic, and input the physical dimensions of each component, so that a PCB rats nest is generated - all the components with their interconnections crisscrossed in a tight bunch. They then pull and push each component into their ideal locations and then give the auto-router a run, in the hope of it sorting the layout for them. Inevitably, that leads to the placement of sensitive components inside current loops.

Add to the above, the thinking about ground planes and the auto-routing gets easier for the computations. The pesky ground always gets in the way. The optimum routing doesn't need to be considered - stick it on another layer.

Likewise, it can go to three layers using a power plane or four layers if the power is dual-rail.

This works perfectly at low frequencies, but experienced amplifier designers know that stability is dependent on high-frequency integrity. There has to be a gain-bandwidth product extending into MHz frequency units for there to be a frequency response suitable for high fidelity.

High-frequency ground returns flow under their outgoing traces but then have to negotiate higher current high-frequency returns. This is worse than mutual inductance in that signal strengths cross and intermingle.

Some designers then think placing cuts in ground planes help to separate sections, but when an HF return signal must go around the cut - which way? If this were fast glue logic, the timing gets blurred, so what about the amplifier's HF stability?

Edward M. Cherry, a member of the Audio Engineering Society, pointed out mutual inductance problems in power amplifiers in 1981, proposing layout methods to limit it. I believe that pre-dated the modern work on EMC, but EMC practitioners also call-on separation of functions. Unfortunately, some in EMC consider solid ground planes to be a cure-all.
Not simple enough for Google-Bot to understand...
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 29 Mar 2021 at 6:47am
Forty years ago, National Semiconductor pushed its proprietary audio chips, but they never really caught on in high fidelity applications.

Manufacturers insisted on individuality by having their designers configure audio circuits using discrete transistors and operational amplifiers--which aren't pre-designed preamps!

Skip to now, and one thing is for sure--the real designers must be dead or comfortably retired!

More amps now feature power-amp chips than before. Fifty years ago, Amstrad was sneered at for doing it, but now "respected" manufacturers don't bat an eyelid.

Admittedly, it still takes some skill to design an amp around a chip, but as most come with app-notes on how to do it, it isn't the same skill older designers had to master!

So, as the numbers using pre-designed building blocks increase, the transistor supply meant for power amps dries.

We'd better hurry up with our amplifier design here, or we'll be searching for NOS!

(NOS = new condition but old stock)
Not simple enough for Google-Bot to understand...
Back to Top
BAK View Drop Down
Senior Member
Senior Member
Avatar

Joined: 14 Mar 2010
Location: Kentucky, USA
Status: Offline
Points: 1309
Post Options Post Options   Thanks (1) Thanks(1)   Quote BAK Quote  Post ReplyReply Direct Link To This Post Posted: 29 Mar 2021 at 2:56pm
Originally posted by Graham Slee Graham Slee wrote:

So, as the numbers using pre-designed building blocks increase, the transistor supply meant for power amps dries.

We'd better hurry up with our amplifier design here, or we'll be searching for NOS!

(NOS = new condition but old stock)

 Yes, the transistors used for hifi are getting harder to find. The manufacturers are not making or reproducing them.
 Many repair technicians and DIYers are finding it hard to source the parts needed to refurbish the great-sounding amplifiers and radios of yesterday. 
 Sometimes one can find a suitable substitute, but not always. Many times the substitutes require circuit changes to maintain stability.
 That is when NOS (new old stock) becomes the only option. And as electronic parts deteriorate as they age, they need to be tested before use to verify the parts' usefulness.

 Valves (tubes) have already reached the NOS-only sourcing stage of availability. The really good valves aren't available as new anymore... only cheap copies from "over there" are.

 The same can now be said for the really good transistors. This issue is continually brought up on DIY forums. "Where can I find xxxx parts?"

 Thank you Graham for designing circuits around "seemingly" popular transistors that are abundantly available today.

 It is sad that some of the best transistors for audio outputs, the NJL3281D (NPN) and NJL1302D (PNP), are no longer readily available. These included the temperature-sensing diode inside the same transistor package... allowing the best temperature control of output idle current.

 ASICs (application specific integrated circuits) Are now becoming the manufacturers' preferred designs in all circuits... and Class D output amplifiers, for their higher efficiency, are taking over from linear amplifiers. (Class D is a switched mode amplifier.)

 I'm going to stay with discrete transistor linear amplifiers as long as I can. These do not "chop up" the audio signal. 
 This is the same reason that I prefer vinyl to reproduce my music.



Edited by BAK - 29 Mar 2021 at 2:58pm
Bruce
AT-14SA, Pickering XV-15/625, Hana EL, Technics SL-1600MK2, Reflex M or C, Lautus, Technics SH-8066, Eminence Beta 8A in custom cabs; Using Majestic DAC & Proprius
Enjoy Life Your Way!
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 02 Apr 2021 at 4:46pm
Start of PCB build. Boards laid out in their relationship when in the case.

part one of board population
Not simple enough for Google-Bot to understand...
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 05 Apr 2021 at 10:19pm
The last picture was a bit boring, perhaps?

Here are a couple of shots of the rectifier/reservoir board, which will fit between the mains transformer and the case rear.

psu board

psu track side


Not simple enough for Google-Bot to understand...
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 07 Apr 2021 at 9:03pm
Originally, the trimmer R19 was in the R19A position and value 500R.

The 10% end to end resistance tolerance of the trimmer might mean some values are 450R up to 550R, but 500R is required to maintain the designed emitter current on T2.

Being in series with R20 - 510R - the accuracy improves to 5%, but I'd prefer the emitter current to be more consistent between each amplifier built.

Making R19 a decade larger in value - 5k - in parallel with R19A - 560R - improves accuracy, while still using a 10% tolerance trimmer.

At the lower extreme 4k5 in parallel with 560R results in 498 ohms.

At the upper extreme 5k5 in parallel with 560R results in 508 ohms.

We end up with a 503R resistance of 1% tolerance, which is more like 0.5% with R20 in series. Using 1% resistors we can be confident that the emitter current will be more consistent between builds.

The total resistance making the DC bias current to T1 increases from 230k to 232.5k, which is a fraction of the resistors 1% tolerance, and the trimmer is there to adjust the DC "midpoint" anyway.

A more accurate DC offset trimmer



Not simple enough for Google-Bot to understand...
Back to Top
 Post Reply Post Reply Page  <1 133134135136>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.

This page was generated in 0.078 seconds.