Supersonic Conquest radio's PR80 and PR80C

[Mike Watterson]

In my first year of studying Electronics I had a slide rule. In 2nd year I added a basic calculator, so still needed slide rule. Filter and transformer designing using tables, including log tables, and for a while in the BBC Communications Dept.

Mostly I've designed transformers for RF and some for SMPSUs. Only repaired / rewound audio and mains using original cores etc. Mains, audio, RF and SMPSU are four different areas of transformers, though there is some overlap between mains and audio, especially if the audio is push pull. There is some slight overlap between SMPSU and RF, but RF also can be divided to about four areas: LF/MF, HF (sometimes no cores), VHF/UHF (cores that increase, or decrease inductance and air) and UHF to microwave using striplines and printed coils.

The computer controlled mill for PCBs was nice

I was involved for about 4 years designing programming languages, systems and controllers for industrial and communication applications.

[Podbuilder]

Hi Mike, What are your thoughts on the coil windings in this audio output circuit to suit 8 ohms or 35ohms? I've spoken to a few guys with original circuit boards that state the radio had a 35ohm speaker yet they're using 2 x 16ohm speakers in parallel with 1mfd cap between them. I assumed there must be 8 ohm winding on the seconday coil but apparnetly not. The original paperwork suggest 140:35ohm impendance which suggests a 2:1 winding ratio.

[Mike Watterson]

"yet they're using 2 x 16ohm speakers in parallel with 1mfd cap between them"

That doesn't make sense. Which speaker is connected direct?
2 x 16 Ohms in series is 32, but I don't know what the 1uF is for.

A 1uF is high pass, as full LF is 100 uF to 1000uF depending on speaker impedance.

The turns ratio depends on supply voltage and power and on speaker impedance of an output stage. The transistors or valves are unimportant as long as they can do the current required for the required power at the supply voltage.

Roughly for class B push pull the rating of the transistors each is a minimum of 1/4 of the output power. For Class A it's 1/2.
Most push-pull are somewhere between B and A. Depends on the bias current. Silicon transistors allow easier setting of bias accurately. There is now no reason to use Germanium except to repair an old amp or radio without changing the bias. Germanium needs a thermistor in the bias circuit.

The power limit. apart from rating of the transistors or valves, is set by core size at the lowest frequency used. More power needs a bigger core, so does 20 Hz (HiFi) rather than 300 Hz (communication or phone) lower frequency.

The speaker driver size and cabinet volume also set a lower limit on frequency.

The primary inductance sets a lower limit on frequency.

"140:35ohm impendance which suggests a 2:1 winding ratio."

But is that for full primary or half primary (half is used on each half cycle)? The max 1% distortion power (or maybe 10%) and supply voltage allows calculation of the effective primary impedance. Since it's push-pull, you double the supply voltage.

[Podbuilder]

15ohm woofer and 15 ohm tweeter in parallel not series for what I can see.  Yeah.. very confusing.  Remember this was thrown together when John Deacon found the circuit board in a rubbish bin.  He put it in the cab, but did he know the impedance of the audio transformer... probably not. Im attempting to make a replica and just trying to make sense of it.

[Mark Hennessy]

The 1uF is just a basic high pass filter (crossover) for the tweeter. Very common in basic speakers like that.

[Mike Watterson]

Yes, clear from the photo the 1uF is for a tweeter (HF speaker). So not two 15 ohm speakers in parallel. The theory of having no coil in series with the main unit is to save cost and that the main speaker has poor output hat HF.

[Mike Watterson]

The effect of the wrong impedance speaker if lower, is to reduce bass response. A lower value may also allow more power at the expense of more distortion beyond a certain amount and risk of damage to transistors or transformer. Obviously 4 Ohms is much worse than 8 Ohms, than 16 Ohms on a system designed for 32 Ohms.

The speaker ohms rating is also only a nominal value. It varies with frequency. A large 15 Ohm main driver in a larger box with a tweeter instead of a small nominally 32 Ohm in a smaller case (maybe plastic) will often sound better and make little difference to the apparent volume. It would likely be acoustically more power at same electrical power, and have a much increased output below 200 Hz.

[Mike Watterson]

Those bipolar electrolytics dry out and then there is no output on the tweeter. The main unit isn't very large, so the value of a tweeter is marginal, especially for an electric guitar. I've never seen old valve amp-speaker combos for guitar with a tweeter, though no doubt someone had them.

I was in shop this last week with real hifi and speakers, which is rare (contents, not me visiting a shop). They also sold keyboards and guitars and a wide variety of guitar amps with built in speakers from practice, to medium and full size. I don't think any had tweeters. If you were re-purposing a generic bookshelf or similar speaker unit you'd not bother unbolting the tweeter (and having to fill the hole?) though simply cutting one wire is plausible, but not worth it. They also had amps with DSP to give any sound, add on boxes with same and loads of knobs and an entire case of effects pedals, inc one like the dumb one I have that has two on/off buttons and two jack cables and the other one I have that's broad with only "on while held down" (or off, not sure if it is a changeover). Some were valves and one claimed to be the lightest 50W valve amp ever with a "lift me" label!

I was surprised they also had a ribbon mic for sale. One reason you scratch, not blow to test, because you never know!

[Podbuilder]

Here's where Im up to now...  a little more progress.  I wound the audio output transformer as an autotransformer with the secondary coil as a feedback loop.  9 ohms DCR across the coil as per the original specs with taps on both sides of the centre tap measuring 5 ohms DCR between the speaker taps.  I put some pins in the PCB so I could try it out without fixing it in place.  This would give me the oportunity to reverse wires if required which turned out to be a good decision.  I had the inputs for the driver coil reversed and the feedback loop on the audio transformer reversed both resulting in squeal.  I'll reconfigure the pins if possible or just wind another and place the coil taps in the correct position.  I've also started working on the cab. 

[Podbuilder]

I thought I'd try the 2 x 16ohm speakers with 1 mfd cap as per the original deacy.  Sounds much better.