UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (https://www.vintage-radio.net/forum/index.php)
-   Components and Circuits (https://www.vintage-radio.net/forum/forumdisplay.php?f=40)
-   -   Single-ended Transistor amps. (https://www.vintage-radio.net/forum/showthread.php?t=160609)

G6Tanuki 16th Oct 2019 8:17 pm

Single-ended Transistor amps.
 
I've grown bored of doing yet-another 6V6/EL84 5-Watt valve audio-amp. My mind is turning towards solid-state - and single-ended.

I've got a bucketful of OC35/NKT404 TO3-cased transistors - which can surely be visualised as the logical 1960s-successors for the old single-ended valve amps.

Big problem is - output-transformers! The classic 'single-ended" amps in car-radios of the time used a really rather small transformer, but they were only expected to deliver a couple of Watts.

I vaguely remember a circuit in a Practical Wireless book where the output-transformer was actually an autotransformer (around 2:1 ratio) with the somewhere-along-the-winding taken to the speaker through a DC-blocking capacitor.

Pragmatically, I'm thinking of doing the same but using a '100V line' transformer's secondary winding, which has 4- and 8-Ohm taps.

Ideas? Suggestions? I'm also pondering doing 'sliding-bias' to reduce the idling dissipation in the output transistor.

dave_n_t 16th Oct 2019 9:46 pm

Re: Single-ended Transistor amps.
 
I seem to recall an early Elektor project (possibly 'Stentor', after the classical character with a loud voice) which used the secondary of a valve output transformer as an AF choke and fed the spkr via a capacitor from the output transistor collector. Can't remember whether it was Si or Ge-based, tho'.


dave

Radio Wrangler 16th Oct 2019 10:08 pm

Re: Single-ended Transistor amps.
 
I'd not want to use those OC35/NKT403 transistors. Their greatest value is in keeping classic equipment which used them going.

For a new amplifier, use currently available parts.

If you're going to do sliding bias then that is 90% of the way to variable bias class-A

The linsley-hood class-A design was a bit crude. Here's the challenge, better it!
The active bias saves the need for magnetics as well.

David

kalee20 16th Oct 2019 10:40 pm

Re: Single-ended Transistor amps.
 
It's debatable whether a single transistor operating in Class A, with DC via another transistor operating as a constant-current sink, can truly be considered as 'single ended.' A choke or output transformer is what is needed, and would work well, but this is the OP's sticking point!

Agree with David, no need to use the limited-resource of obsolete germanium devices, when current-production silicon would be just as applicable and would work better.

joebog1 16th Oct 2019 11:47 pm

Re: Single-ended Transistor amps.
 
When I worked at the CSIRO, one of the engineers made a single ended class A amp.
It produced 20 watts of audio and my guess would be 200 watts of heat.
No output transformer though, was capacitivly coupled to the speaker.
From memory it used 2N3055's. It did sound rather good though.

Joe

Maarten 16th Oct 2019 11:54 pm

Re: Single-ended Transistor amps.
 
Philips used such a transistor single endedly in their B5W32AT table radio.

Argus25 17th Oct 2019 3:04 am

Re: Single-ended Transistor amps.
 
What a great idea !

I'm very fond of germanium transistor Class-A running from a 12V supply.

It is dead easy to design the tapped output choke , butt stacked. You just select a suitable core cross section for the power you require and have enough inductance so that the Bass response is maintained and suitable wire to fill the bobbin so that the DC resistance is not too excessive.

You can assume with a 12V supply that the peak collector voltage at full power before clipping will be around 10V, so from that and your required output power you can calculate the collector impedance and set the tap on the choke to transform that impedance of your speaker into the collector.

It pays to read up on the factors which determine thermal stability in a class A transistor stage. Of note it will always be stable if the collector voltage is half the supply voltage (or less) with a resistive collector load. However this is not the case with a transformer or choke load, so you need to be sure the heat sink is adequate, the temperature rise above ambient is not excessive and there is some emitter degeneration.

The classic circuits are in the old Philips manuals. It seems like the car radio transformers were physically small but this of course is one of the great advantages of an autotransformer.

Have a look a the underside of this radio chassis, you can see the 22:8 Ohm choke I wound if you zoom up on the image:

http://worldphaco.com/uploads/THE_EF98-OC16_RADIO.pdf

One other thing is that its very easy to drive the B-E junction of the output transistor with another transformer, which drives the B-E junction and the emitter degeneration is excluded from lowering the AC gain. And with a driver transformer, easily match the driver stage. One great thing about 12V operation you can use batteries or many existing 12V supplies.

A great transistor for the task is the AD149 or similar. AC188's make a great driver.

Luckily these old germaniums are still plentiful and the audio people haven't got as interested as a 6V6 and the prices are still good.

The original AD149 made by philips or mullard had a beautiful bright nickel plated (on copper) finish nobody makes TO-3 transistors as beautiful looking that that anymore do they ? and the TO-3 case is obsolete. Or you could use a boring T0-220 cased epoxy silicon device.....not much of a contest is it ?

I have a collection of really nice ADZ12's that I plan to make into an audio amp one day.

Argus25 17th Oct 2019 5:43 am

Re: Single-ended Transistor amps.
 
.......so obviously from what I said above, at around full power the rms collector voltage prior to clipping will be around 7V and if you just have the choke (with no tap) and an 8 Ohm speaker you will get about 6W out of it.

G6Tanuki 17th Oct 2019 3:20 pm

Re: Single-ended Transistor amps.
 
Thanks for the suggestions Argus25: your thinking is much the same as mine - though I'm trying to avoid the need for a driver transformer [even if in doing so I need to use an extra small-signal transistor or two].

Apart from the output transformer - which I'll probably have to wind myself (I've got some old wall-wart transformers that can be repurposed) my one concern is thermal stabilisation; the 'classic' brute-force way to do it - as found in car-radios etc - is to include a low-value decoupled emitter-resistor of course - but that uses-up some of the 'HT' volts.I got to pondering about reducing this resistor to a point where it's only developing 0.2V or so, then using this voltage to provide base-bias to a transistor whose collector is connected to the output-transistor's forward-bias network in such a way that if the emitter-voltage rises it turns the little transistor on and 'robs' the main transistor of some of its forward bias - so providing a simple DC feedback-loop. I've got a load of 'unmarked' OC71-style Germaniums here waiting...

ADZ12 - that brings back memories - they were used in the transmit HT-inverter circuit of the Pye Vanguard two-way radios in the 60s.

kalee20 17th Oct 2019 4:15 pm

Re: Single-ended Transistor amps.
 
LM10, which is a combined op-amp and 200mV reference, to sense and stabilise bias? A super little IC, a Bob Widlar triumph!

Though, you might take a dim view of a single-transistor in the output stage and 88 ancillary transistors to stabilise its bias...

G6Tanuki 17th Oct 2019 4:31 pm

Re: Single-ended Transistor amps.
 
I *could* use something like a LM10, but I don't have one in my junk-box. This amp-design idea is very much based around 'getting loudness on the cheap', which investing in any new bits would rather go against.

Bazz4CQJ 17th Oct 2019 4:50 pm

Re: Single-ended Transistor amps.
 
Quote:

Originally Posted by G6Tanuki (Post 1184495)
I *could* use something like a LM10, but I don't have one in my junk-box. This amp-design idea is very much based around 'getting loudness on the cheap', which investing in any new bits would rather go against.

I seem to recall, talking about bucketfuls of NKT404's etc, that you once waxed all lyrical about the wonderful "HiFi Amp" you created from the modulator section of a Pye Vanguard. Not single-ended, but I wonder if subconsciously, that's at the back of your mind ;) ?

B

G6Tanuki 17th Oct 2019 5:24 pm

Re: Single-ended Transistor amps.
 
Yes, my "Vanguard Amp" was never in any way hi-fi - but I don't have 'golden ears' either. I can't hear "quality" but I *can* hear *Loud*!

[The AM Vanguards came equipped with a transformer that included a secondary winding designed to feed a loudhailer - if you look at photos of 1960s police cars you often see the loudhailer speaker mounted on the front bumper]

Argus25 17th Oct 2019 10:07 pm

Re: Single-ended Transistor amps.
 
2 Attachment(s)
Quote:

Originally Posted by G6Tanuki (Post 1184484)
ADZ12 - that brings back memories - they were used in the transmit HT-inverter circuit of the Pye Vanguard two-way radios in the 60s.

Some things from electronics history are just so beautiful and so well made it is very hard to ignore them (ADZ12 image attached) even if a modern epoxy cased device completely outclasses them at every turn on the specs. The have the mojo that turns my head. It is hard to beat these transistors though in a Royer converter, more modern parts need the HF response tamed down.

(On the subject of thermal stability in a class A stage its worth noting that amplifier circuits where the collector voltage is Vcc/2 or less are thermally safe because increases in current lower the dissipation. Also stages run at low current with low values of stability factor, dIc/dIco, or S<10 are rarely going to get thermal runaway. But power amplifiers usually have a low emitter resistance for power efficiency and this results in a high stability factor S. Also due to the leakage properties of germanium devices the base bias source needs a low resistance. The Ico for a germanium transistor is much larger than for a silicon a factor of 1000. Also Ico doubles every 10 degC and Vbe drops -2.5mV/degC.)

Argus25 18th Oct 2019 12:56 am

Re: Single-ended Transistor amps.
 
Quote:

Originally Posted by Bazz4CQJ (Post 1184497)
I seem to recall, talking about bucketfuls of NKT404's etc, that you once waxed all lyrical about the wonderful "HiFi Amp" you created from the modulator section of a Pye Vanguard.

The Newmarket 404 was a great 1960's germanium power transistor workhorse featured in a number of RSGB projects. They make good audio amps and good DC/DC converters. I have a large box of them and still use them, from time to time.

The major competitor with Mullard's ADZ12 on the opposite side of the Atlantic, was RCA's 2N174, with not exact but similar specs. These are still really easy to get. They were made by RCA, Motorola and Delco, there are even mil spec versions as they got used in aerospace projects. A more lowly Vce rated type, the 2N441, was ubiquitous in hybrid card radios.

When the 2N3055 was released in the early 1960's, it was game over for the power germaniums.

FERNSEH 18th Oct 2019 1:06 am

Re: Single-ended Transistor amps.
 
1 Attachment(s)
Example of a class A single transistor output stage.

Pye stereo radiograms models 1206 and 1208.

Output transistor is AD140, replacement is AD149.

DFWB.

Argus25 18th Oct 2019 1:48 am

Re: Single-ended Transistor amps.
 
....Oh one thing I forgot to mention with these germanium choke output amplifiers:

If you accidentally short the transistor's collector to ground, it fully energizes the core of the choke, then when the short is released, the very high voltage high energy transient sends the transistor to la-la land.

In most of my amplifiers with this configuration and are +ve ground I now fit a bidirectional transil across the choke, that is normally just 4 or 5v volts out of conduction at full output. Often the transistor body is on insulators and its easier to short it to ground than you might guess.

One good solution, is to ground the collector (of these PNP power devices) and use neg ground. This way you don't need an insulator and the thermal coupling is better.

It is easy if you have a transformer driving the base circuit as your drive is isolated, and then you put the output choke in "what looks like" (but isn't) the emitter circuit, it is really still in series with the power supply and the collector circuit.

This way there is no risk of shorting the transistor body to the instrument case as it is connected there anyway.

I used this trick with the ASZ17's for my ZC1 radio, so they could be bolted directly to the aluminium cases, see page 1 at the top where there is a photo of 5 units that I manufactured:

http://worldphaco.com/uploads/ZC1_MK...R_SUPPLIES.pdf

Argus25 18th Oct 2019 2:24 am

Re: Single-ended Transistor amps.
 
Quote:

Originally Posted by joebog1 (Post 1184385)
When I worked at the CSIRO, one of the engineers made a single ended class A amp.
It produced 20 watts of audio and my guess would be 200 watts of heat.
No output transformer though, was capacitivly coupled to the speaker.
From memory it used 2N3055's. It did sound rather good though.

That ratio sounds extreme. I think it turns out that the heat dissipation for a class A power amp is very roughly the same as what it produces as a maximum power audio output (possibly excluding bias resistors and heaters in the valve case),or maybe double, that of course is when a transformer or choke is used with low copper losses.

Those massive losses must have been in the load resistor. The great advantage of the choke or output transformer is the stored magnetic energy which also allows the anode or collector to exceed the power supply voltage, which it cannot do with the resistor. But as noted that can threaten the output device especially if the speaker is not connected too and there is little damping.

Output transformers /chokes are much easier to design for transistors than valves, very little issues with self winding capacitances, resonances etc, they have much thicker wire and far fewer turns making them easy to wind too.Also in the choke/autotransformer case they can be very compact and efficient.

joebog1 18th Oct 2019 3:32 am

Re: Single-ended Transistor amps.
 
It was very long ago Argus, 40 plus years. I can remember BIG heatsinks with perhaps five or six transistors mounted thereon. After switchon it took about 20 mins to "stabilise"
By that time my "reckoning" would have the heatsink at JUST bearable temperature to the hand. The heatsink was one of the early Mullard types with very large size ( about 12" X 12" ) but the fins were only maybe 3/4" deep, even with them being both sides of the extrusion. I must say that the temperature didnt change after that "warmup" time.

Joe

terrybull 18th Oct 2019 6:34 am

Re: Single-ended Transistor amps.
 
The John Lindsey-Hood original version 1969 class A uses 2n3055’s and sounds very good indeed. There is a version (for electrostatic speakers) with paralleled output transistors which produced 20 watts and a later version with + - power supply to remove the capacitor from the output.
I have built several of the original and the performance belies the component count.


All times are GMT +1. The time now is 12:50 pm.

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