UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 25th Aug 2017, 1:05 am   #1
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Golden Ear Laboratory Amplifier Build

So I have progressed slightly with my new "toy".
I have posted some stuff already on this build, but now I have all the bits n pieces needed.
I ordered two chassis this morn - pics when delivered.
I have redrawn the circuits using somewhat more modern devices and components, although the design itself remains essentially unchanged.

I have attached jpg's of the circuits, so that all can comment on the idea in general, AND point out errors that might have sneaked past me ( easy to do at my age )
You might notice that my KiCad wont load the switch.lib library

Those that have looked at the origional design from Radio Electronics
August and September 1956 might also find errors from back then.

I will do a write up of how I go, along with pics and explanatory notes in case someone out there is as crazy as me

Joe
Attached Thumbnails
Click image for larger version

Name:	Golden Ear-1.jpg
Views:	899
Size:	57.1 KB
ID:	148357   Click image for larger version

Name:	Golden Ear-2.jpg
Views:	532
Size:	41.0 KB
ID:	148358  
joebog1 is offline  
Old 25th Aug 2017, 10:29 am   #2
paulsherwin
Moderator
 
paulsherwin's Avatar
 
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,785
Default Re: Golden Ear Laboratory Amplifier Build

Joe, if you want to stop the forum software resizing your images, put them in a PDF or ZIP container and upload that. You will need to observe the file size limits for these types.
paulsherwin is online now  
Old 25th Aug 2017, 5:30 pm   #3
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,951
Default Re: Golden Ear Laboratory Amplifier Build

That's an interesting design: I commend the use of the 12BH7 as the driver, it has a bit more power-dissipation capability than the usual ECC81/2/3 types, indeed some years back (well, some decades...) I built a little tape-deck output amp using the 2 halves of a 12BH7 in push-pull and it sounded really good.
G6Tanuki is offline  
Old 25th Aug 2017, 11:57 pm   #4
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Thanks Mr Sherwin

I did know that jpg's would be compressed, but didnt think that far.

Anyway, there are far better engineers present than I so all suggestions and criticisms will be appreciated. The output transformers are my own design and build, being 9 section wound on a 1 1/2" X 3" stack of 5% Si GO iron. I figure it should be very low distortion even at max power ( around 50 watts).
Primary inductance is well over 120 henries ( my own inductance meter so I dont know exactly) primary to half secondary is a few milli henries.
Power transformers are big toroids overwound 360 watt cores but delivering a maximum of 218 VA, so should be low noise as well. Power supply will be built on a separate chassis, but bias supply will be constructed inside the amplifier chassis. For umbilicals I will be using standard circular connectors.

For those interested they were purchased out of China for less than $10 for a mating pair, but I had to modify them to make male and female mounting by removing the circlip and placing the insulation and pins/sockets to prevent a nasty 420 volt surprise if a finger was accidentally poked into a socket.
It is a simple exercise BUT the insulation comes in two halves and holds the pins/sockets and is VERY easily disengaged and you end up with pins/sockets everywhere .

They are beautifully made!! supposedly rated @ 500 volts and 5 amps per pin. I would think the voltage rating would be AC though, but I "reckon" they will handle 420 volts DC.

Joe
Attached Thumbnails
Click image for larger version

Name:	connectors.jpg
Views:	435
Size:	52.9 KB
ID:	148429  
Attached Files
File Type: pdf Goldenear.pdf (155.2 KB, 221 views)
joebog1 is offline  
Old 27th Aug 2017, 6:51 pm   #5
John Caswell
Hexode
 
Join Date: Jul 2009
Location: Wokingham, Berkshire, UK.
Posts: 437
Default Re: Golden Ear Laboratory Amplifier Build

Hi Joe,
In your power supply you can't just stick a gas regulator straight across the supply, you will need some form of dropper resistor to limit the current through it.
Build the bias supply without the regulator and place a resistor of some value across the supply to draw approx 10mA. Then measure what the rail voltage is then calculate the value of the resistor in series with the regulator to provide the required voltage drop to the regulator.
So if the unregulated supply is say 150V and the regulated supply required is a 105V the voltage drop is 45V at 10mA ie 45/0.01 = 4500R say 4700R = 4K7.
You may have to adjust the R if the bias line is drawing a lot of current, but looking at the cct I don't think so.

John
John Caswell is offline  
Old 27th Aug 2017, 7:14 pm   #6
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Golden Ear Laboratory Amplifier Build

In the third long-tailed pair section there appears to be positive feedback, from each anode to the opposite side's grid via capacitors. If very small they might be ued to cancel input capacitance but it looks unusual.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 28th Aug 2017, 1:21 am   #7
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Thanks John!
Yes I understand that. I have small EI style traffos for bias that will be fed by the 12 volt heater supply. The traffos will be used backwards i.e. I will feed 12 volts into an 18 volt primary (originally the secondary) and adjust as necessary.

Wrangler you are also correct, the lil caps are to compensate for input capacitance of the 12AX7/ECC83.
Its actually one of the ideas that intrigued me, and started me thinking of constructing a stereo version. The design is old, I agree, but Joseph Marshall seems to have analyzed every possible source of error ( distortion) or even DC unbalance in the design.

Sooo, I designed some output transformers, wound them up, and had special very low flux density toroids wound to my design and purchased all the other "bits". Chinese made ceramic valve bases, ( of which I have used many hundreds and not yet had even a single failure or flashover) as 1% 2 watt metal film resistors, 3 watt ten turn pots for splitter balance ( V1)pre driver bias ( V3) 5 watt driver bias pot (V4) and as I couldnt find anything smaller, 25 watt balance pot ( V5-V8), the last two being single turn.
Most valves are either English or Australian NOS ( because I have them ).
Two other components I havent mentioned are the coupling capacitors between V3 and V4. I have a collection to try, ranging from Russian military aluminium cased waxies, to Audio phool French copper foil plastic "super capacitors". Perhaps I can end once and for all the truth/myth of capacitor "sonic" signatures.
I spent 30 years designing circuit boards and have discovered one thing!! circuit boards and valves is a dead set no-no, UNLESS of course you like repairing dry joints .
I am still working on layouts, but 90% of components will be valve pin, pot lug to valve pin, and of course I will have a minimum of tagstrips/ single tags as no matter how "clever" one is, there is always a "long wire" and no!! thats not antenna.

And thanks to both for feedback.

Regards
Joe
joebog1 is offline  
Old 28th Aug 2017, 6:36 am   #8
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,636
Default Re: Golden Ear Laboratory Amplifier Build

Couldn't figure out the circuit, but presume it has a paralled differential/balanced input. I got lost after that. Could Joe/someone walk us throw the circuit and tell us why you need all those valves up front? And what is an 4X5B/254? A search threw up a water cooled beam tetrode as a possible??

Swish connectors Joe.

For gas regulators I thought you calculate R/value for 5mA for the VR valve and what ever the circuits need. I have an STC book with specs etc on common VR valves if needed.

Nice to see you making something Joe. What is the leafage inductance of your OPT's BTW?

Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 28th Aug 2017, 9:27 am   #9
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Golden Ear Laboratory Amplifier Build

I think it translate as four valves, each 5B/254M

Which was an 807 built into a smaller envelope and base by STC. They aren't as common as the 807 itself, and as they are closely identical electrically, I'd build it with the 807.

The first stage is a pair of cathode followers (ECC82 =12AU7() and there is a twiddler to vary current between the two halves But the connection to ground for ther cathode current is missing. There seems to be a link not shown. The second stage is a pair of ECC83 =12AX7 with grids and cathodes connected across the outputs of the first stage. There are added cathode resistors, and the first two stages rely on the ECC82 running more current than the ECC83.

This second stage has capacitance-cancelling capacitors from cathode to anode on each side.

The third stage is another pair of ECC83 triodes as a voltage amplifier running from higher HT. This time there is a long-tail shared cathode resistor, and a small network to give HF lift.

All the first three stages are DC coupled.

Two OA2 voltage regs are used to make lower HT for the first two stages. THey are fed 410v through a 10K 10W resistor

The signal now gets AC coupled to a push pull pair of 12BH7 cathode followers running from 410v HT and using a -105v supply for their cathodes. These then DC couple into a push pull stage with mini 807s in parallel pairs on each side. The use of generous cathode-follower drive suggests that the output stage is going to get driven into significant grid current, and with four 807s on-board, this is a powerful amp.

Full marks for putting in parasitic stopper resistors, but disqualified for the lack of any current balancing arrangement for the output bottles. Lots of negative marks for running the entire DC current of four powerful valves all through the slider of a single pot.

This amplifier is a valve - roller's delight. Any owner will need a good valve tester and a big stock of 5B/254Ms just to keep the output stage in balance.

I love the input capacitance compensation on cathode input nodes with values got from grid capacitance on data sheets, with no allowance for cathode to heater capacitance.

I know where I'd put this circuit:

“But the plans were on display…”
“On display? I eventually had to go down to the cellar to find them.”
“That’s the display department.”
“With a flashlight.”
“Ah, well, the lights had probably gone.”
“So had the stairs.”
“But look, you found the notice, didn’t you?”
“Yes,” said Arthur, “yes I did. It was on display in the bottom of a locked filing cabinet stuck in a disused lavatory with a sign on the door saying ‘Beware of the Leopard.”

I'm sorry, this circuit contains clear traces that fashion not engineering is dominant. Worth looking at out of curiosity, but not worth building. It does however, fulfil the goals for the campaign for honesty in circuit names

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Old 29th Aug 2017, 12:20 am   #10
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Thanks to David for the explanation, heres a photograph of a three pound note in appreciation.

I understand your criticisms on most points, but I beg to differ on a couple of points.

The output stage is only JUST outside class A, so the big drive capability of the 12BH7 is there to keep everything "just loafing along". After all with my Tannoy monitor gold speakers I will never get to the 50 watts or so of power available.

The output balance DOES in fact run through the single pot wiper which is a carbon brush of 4 mm X 4mm size. It is in contact with five turns of resistance wire, and a failure is not likely. ( I do understand that there is an outside chance of a failure though.)
I think I have enough bottles to get a fairly close balance.

I do have 807's as well, still packed in wax by the aussie army, sometime in the 30's or 40's.

Yes its a "tube rollers delight" and I have a couple of power supplies to select the output quads. If I cannot find enough close examples, I have access to a few thousand CV428's.

The compensation caps on the input pair are better than no compensation at all, and I am very interested to see if it really makes any difference.
( if ANY ).

I have a few valves laying around and LOTS of time, so this little project will keep me off the streets.

Thanks again for your insights, and explanation of the circuit.

Joe
Attached Thumbnails
Click image for larger version

Name:	CV428.jpg
Views:	284
Size:	72.9 KB
ID:	148579   Click image for larger version

Name:	pots.jpg
Views:	278
Size:	57.5 KB
ID:	148580   Click image for larger version

Name:	807.jpg
Views:	280
Size:	110.7 KB
ID:	148581  
joebog1 is offline  
Old 30th Aug 2017, 3:27 am   #11
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Now that I have "switches.lib" loaded and working,
A revised circuit diagram of the Golden Ear amp with a switch to show how the input stage is switched.

Joe
Attached Files
File Type: pdf Revised_circuit.pdf (92.6 KB, 211 views)
File Type: pdf Radio-Electronics-1956-08.pdf (523.9 KB, 292 views)
File Type: pdf Radio-Electronics-1956-08_1.pdf (474.6 KB, 178 views)
File Type: pdf Radio-Electronics-1956-08_2.pdf (468.6 KB, 179 views)
joebog1 is offline  
Old 13th Oct 2017, 12:37 am   #12
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

Some progress again!! I picked up the chassis at 06:30 AM this morning.
1.6mm zincalum steel. Its mild steel that has been electroplated with a mixture of zinc and aluminium. It cant be soldered !!! ,but of course a dremel to remove the plating and a hot iron, and the steel will solder very well. I only need one point in each chassis, and that is the "star" point.

The power supply pic is quite self explanatory:
A single toroid for each channel, feeding some glass passivated fast recovery diodes in a bridge configuration. These feed the 100 uF F&T main filter caps, one each side of the chokes inside the rectangular and potted boxes.
(recovered from a Royal Flying Doctor base station transmitter, 8H @ 300 mA )

Pic three is the amplifier layout ( just to assure myself it will all fit )
It looks a bit busy, but remember that the pots will be under chassis, the valve bases will of course be mounted from under the chassis, and the transformers have nice covers which will be held in with button head Allen key bolts.

Extra decoupling of the power supply is taken care of with an additional pair of F&T 100 uF 500 volt caps. There will be a small screen to prevent the heat from the 5B/254 output bottles drying them out. Underneath the chassis is an additional HT choke of 15 henries @ 70 mA.

Now I have a couple of days with a rule, a pencil and an eraser marking out exact dimensions of cutouts and valve positions. I always drill 12 holes around my valve bases for two reasons!! One is to reposition a valve to reduce lead length or facilitate a neater layout, the second of course is to allow cooling air to flow past the bottle, especially in the case of 5B/254, as they tend to run stinking hot anyway ( what happens when you shove an 807 into a tiny bulb).

I will post more pics as I progress. If anybody needs better explanations of why or how, I will attempt to blunder through a mo-bedda explanation!!.
Alas I am a very poor writer.

Joe
Attached Thumbnails
Click image for larger version

Name:	Chassis.jpg
Views:	266
Size:	40.8 KB
ID:	150666   Click image for larger version

Name:	PSU.jpg
Views:	283
Size:	62.5 KB
ID:	150667   Click image for larger version

Name:	amp.jpg
Views:	350
Size:	90.6 KB
ID:	150668  
joebog1 is offline  
Old 13th Oct 2017, 5:17 pm   #13
Stevie342000
Hexode
 
Join Date: Jan 2006
Location: Greater Manchester, UK.
Posts: 373
Default Re: Golden Ear Laboratory Amplifier Build

Personally the amplifier chassis looks too crowded and I would use two of those chassis and spread it all out a lot.

It's a big amplifier and there is going to be a lot heat there and I may be wrong but it does look too crowded above and below in my head.

I am familiar with the amplifier and the original article, I am following this as this is one of the power amplifiers I would like to build.

Interested to see how it turns out and what you think it sounds like. It's going to be a bit of a beast.

Keep us posted on how it goes.
Stevie342000 is offline  
Old 13th Oct 2017, 6:25 pm   #14
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Golden Ear Laboratory Amplifier Build

Applying feedback over five stages seems to be asking for trouble.
G8HQP Dave is offline  
Old 13th Oct 2017, 7:24 pm   #15
GrimJosef
Dekatron
 
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,310
Default Re: Golden Ear Laboratory Amplifier Build

I suppose quite a few stages are direct-coupled which will keep the overall phase-shift low. But it's hard to predict much about stability. So much depends on phase-shifts in the output transformer. Harold Leak managed four stages back in the 1940's but abandoned that design for the more stable three stage TL/12.

Cheers,

GJ
__________________
http://www.ampregen.com
GrimJosef is offline  
Old 13th Oct 2017, 8:15 pm   #16
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Golden Ear Laboratory Amplifier Build

Direct coupling keeps LF phase shift low, but will not affect HF phase shift. Having two signal paths through the amplifier also raises the possibility of non-minimum phase response i.e. phase response worse than that which could be predicted from the frequency response.

To me the design just looks strange, but I can't be bothered to critique it in detail.
G8HQP Dave is offline  
Old 13th Oct 2017, 10:31 pm   #17
PJL
Dekatron
 
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
Default Re: Golden Ear Laboratory Amplifier Build

One stage is a cathode follower. Both earlier stages have miller effect compensation. It is basically a differential amplifier but what does the 'inverter' achieve? Why the common 22K in the anode supplies? Why the 13K between the cathodes of the driver stage? Are these required for stability?
PJL is offline  
Old 13th Oct 2017, 10:41 pm   #18
Argus25
No Longer a Member
 
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
Default Re: Golden Ear Laboratory Amplifier Build

I have a theory on how this circuit came about.

The entire front end of it, which is an HT stabilized DC coupled differential amplifier, with the small cross coupling capacitors is a classic oscilloscope design. Tektronix used the technique of cross coupling with small capacitors in differential amplifiers.(the purpose being to cancel the miller effect)

It looks as though somebody decided they liked or were enamored with the linearity and wide bandwidth of an oscilloscope style amplifier, wanted to apply the concepts to audio and they knew at some point they would have to have a phase splitter anyway to drive their output stage, so figured why not go differential right to the input.

With this sort of design it would be better if it was used as a differential input and that function (with useful ground loop and common mode noise rejection) wasn't wasted by connecting one input (one of the 12AU7 grids) to ground.

So whatever signal source is going to be driving this amp, ideally it would be a differential signal feed coupled to both grids of the 12AU7, and the negative feedback should ideally also be a differential feed (which can be acquired from two terminals of the output transformer) split equally to the differential input.

Last edited by Argus25; 13th Oct 2017 at 10:49 pm.
Argus25 is offline  
Old 13th Oct 2017, 11:27 pm   #19
joebog1
Banned
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
Default Re: Golden Ear Laboratory Amplifier Build

copied from another current thread:
Some would ask why I am building somebody else s amp design. I have designed and built many amps ( I cant remember how many) both vacuum AND sand, and even ventured into modern mosfets for a time.
WHY ?
I used to own a small audio manufacturing business. I am NOT a brilliant engineer by any means, BUT I am extremely proud of my output transformers. Also my toroid design.
I just happen to have an abundance of brand new, unused big fat output transformers, AND as it happens, a fairly large ready made supply of power toroids. Add to this a couple of hundred weight of passive components.
I have retired from working mostly, except for a few repairs here and there for beer money.
WHY will I build the Golden ears amplifier? I still love building stuff. I still build model aeroplanes ( yes balsa and tissue style) BUT I dont fly them.
It keeps me off the street, and I derive huge pleasure from drilling and filing and soldering, then desoldering after discovering a better layout. At the end of the day, my kids will have a stonkin' amplifier to listen to, ( or do rock festivals with ).
I have a seven watt amp at present that I have displayed here, and I almost don't use it at all.
I still love my construction projects though!!!.

Stevie342000, Yes it does look busy as I noted myself I redid the top pic without the pots, the pot nuts will be about all one will see when looking at the amp.
The chassis will be 50% flat black so it will also tone down the brightness.
I guess you live in one of those country estates where real estate in the lounge room is easy to find, hence your recommendation of two chassis.
I don't have that. Plus I have two "sexy lookin" power meters recovered from some Japanese amp I stripped that will be fitted into the front end of the chassis. There will not be any controls at all on the power amp, just the meters that will be backlit to show the amp is on. I have half hard brass plates for the front of the amp and the power supply for that bit of extra "bling".

There are lots of comments about feedback and stability, so I attach the original articles for those who have doubts about the design.
As there are only two capacitors in the main amp circuit, the phase shift will be quite small, and as I already stated, I am very proud of my transformer designs. Leakage is VERY small!!!. My inductance meter is home made and designed so I cannot give exact measurements, as I had nothing available to use as a reference. BUT when measuring an Acrosound TO330 transformer, and comparing it to mine, my traffos exhibit less than half the leakage between pri and sec, and half pri to sec of the Acrosound versions. Other specs for both transformers are similar with regards to primary inductance although my versions are around one kilogram heavier, remembering mine are not in a metal can nor do they have tagboards for termination, meaning I have a larger stack. This doesn't affect leakage inductance however.

AS already stated, I drill 12 mounting holes for each valve, 3mm for the noval valves and 4mm for the loctals, two being used for mounting, ten are left vacant for cooling air to circulate around components under the chassis, and exiting around the valves and carrying away a large amount of the heat. My amplifiers always used punched steel sheet mesh and at least 25mm high feet to allow the air to enter the chassis in the first place. I have built many biggish guitar amps, and have found that this lil trick makes the amp run about 50% cooler than one without cooling.

GrimJosef, I think today's iron [sic] is far superior to that available to Harold Leak,
But I will show performance when it's complete.

So far under the chassis I have managed to mount just about all components from pin to pin, in true point to point fashion, and so far managing to keep signal leads very short. I haven't been so lucky with bias and balance pot wiring however, as the larger pots require LOTS of space and therefore a longer lead length. AS this is all DC and in my normal manner, will keep it as far away from any signal path, I am hoping it wont add to any instability. REMEMBERING, that so far its only on paper

Joe
Attached Thumbnails
Click image for larger version

Name:	without_pots.jpg
Views:	223
Size:	77.2 KB
ID:	150723   Click image for larger version

Name:	meters.jpg
Views:	196
Size:	54.4 KB
ID:	150724  
Attached Files
File Type: pdf Radio-Electronics-1956-08.pdf (523.9 KB, 128 views)
File Type: pdf Radio-Electronics-1956-09.pdf (500.1 KB, 129 views)
joebog1 is offline  
Old 13th Oct 2017, 11:37 pm   #20
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Golden Ear Laboratory Amplifier Build

Valves are transconductance beasts, giving a current output which has to drive a resistive load to be converted into voltage in order to be able to operate the following stage. The stray capacitance on this node, together with the resistor make a pole in the response... a roll off in the amplitude response and a roll to 90 degrees more phase shift.

If you reduce the resistance, the pole shifts up in frequency heading out of the audio band, but the gain drops. You need more stages to put the gain back. So the choice of number of stages trades off between fewer, lower frequency poles against more but higher frequency poles.

Pushing all these poles up in frequency moves them closer to where the unwanted, uncontrolled 'trash' poles congregate.

Usual design for stable amplifiers involves having one or two poles which do the business (with an added zero if there are two poles in play, then all the others are swept up in frequency far enough that the loop gain has fallen below 0dB before they start to have any significant effect.

Trying to get good control of the performance of an amplifier by feedback, across the full audio range means you want plenty of open loop gain to push the 0dB point up in frequency, you want to push up the frequency of the controllable poles, but the uncontrollable ones don't shift, and your intended dominant pole is not quite so dominant and the safety margins of gain and phase which you wanted to assure stability and refined behaviour have gone.

This is what Harold leak ran into when trying to push his amplifiers further. He'd already got more feedback working safely across the audio range than his competitors through artful transformer design pushing its stray poles a bit more out of the way.

High transconductance devices can help by running lower anode impedance and still getting the gain. Running stages in parallel like in this design can boost gain by 6dB without pushing down pole frequencies. But you can soon run out of tricks.

A long time ago I did an amplifier where the gain stages looked like the sort of thing you'd have found in a wideband oscilloscope. The output transformer had to go and it achieved wide bandwidth with lots of feedback giving very low distortion. High gm valves and paralleling of entire stages in the voltage amplifier stages gave the gain and bandwidth. Massive paralleling of TV line output devices in the output stage replaced the output transformer. This was a big departure from classical valve architecture, but no-one today would stomach it with hundreds of watts of heater power as well as dual power supplies supplying positive and negative HT. I only ever built one mono one, just to see what I could do, and it came apart once I'd compared it to a couple of other designs.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is online now  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:00 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.