A foray into the world of glowy stuff.

chris661 · Post by **chris661** » Wed Jul 11, 2012 3:58 pm

Okay, so the following is educated guesswork on my part...

I can work out the turns ratio using DC resistances.

If I want the amplifier to see 3kohm, I (think I) multiply the primary:secondary turns ratio by the speaker impedance.

This would explain the lack of power out of this, too: 1/8w isn't much. In fact, I think my iPod will drive more than that out its headphone socket.

Mike H · Post by **Mike H** » Wed Jul 11, 2012 4:42 pm

chris661 wrote:I can work out the turns ratio using DC resistances.

No you can't.

All you'll get is what is the resistances of the lengths of wire used, which entirely depends on gauge. (Thickness, cross section)

The way to do it is apply, say, 5VAC (i.e. 5V heater transformer, or could use 6V etc.) across the primary in series with a multimeter capable of measuring AC microamps / milliamps.

With no load on the secondary, the current will be as per the incremental inductance. Divide your AC Volts (known) by the microamps / whatever value to get the reactance. (AC impedance)

You know it's 50 Hz so if you do the calculation or use a freq. reactance chart you'll get the equivalent Henries.

Put a dummy load resistor on the secondary, do it again, produces a larger AC current this time.

The AC Volts (known) divided by that gives you the primary impedance for whatever value load resistor on the secondary.

Seemples!

ed · Post by ed » Wed Jul 11, 2012 5:18 pm

rubbery expranation Mike!

chris661 · Post by **chris661** » Wed Jul 11, 2012 10:16 pm

Mike H wrote:
chris661 wrote:I can work out the turns ratio using DC resistances.
No you can't.

All you'll get is what is the resistances of the lengths of wire used, which entirely depends on gauge. (Thickness, cross section)

Ah yes, I'd forgotten they might use different thicknesses of wire. You can tell its the holidays, eh?

Tomorrow's a new day - perhaps I'll be more focused then.

Cheers for the explanation Mike. I'll get the multimeter out tomorrow and get it figured out.

Mike H · Post by **Mike H** » Thu Jul 12, 2012 2:58 pm

Yes, primary tends to be thin, secondary tends to be thick. Because of course, while the Voltage is step-down, there's a current step-up.

chris661 · Post by **chris661** » Fri Jul 13, 2012 10:49 am

Some good reading here:

http://www.radioremembered.org/outimp.htm

Might be worth putting that over in the "Stacks" section methinks.

Mike H · Post by **Mike H** » Fri Jul 13, 2012 6:13 pm

There ya go, that's another way of doing it, find the turns ratio by measuring the secondary Voltage.

chris661 · Post by **chris661** » Mon Jul 16, 2012 9:30 pm

Hmmmm...

Found the turns ratio to be 36:1, by applying 13mV to the secondary, and getting 0.47V on the primary.

Square of the ratio gives 1296:1 impedance ratio. Feeding an 8ohm speaker, the primary impedance will be ~10kohm

I can (for a test) get 5kohm impedance by using a 4ohm speaker. I suppose 2ohm is possible with some clever wiring, which would give 2k5ohm impedance.

How critical is the output transformer primary impedance?

chris661 · Post by **chris661** » Tue Jul 17, 2012 12:00 pm

Okay, so I went for a 2k5ohm transformer impedance (ie, 2ohm speaker), as I figured that a speaker of 2ohm nominal can vary considerably (its actually 4x 8ohm PA speakers wired in parallel - one is a dual-coned McKenzie so there's a decent full-range sound), giving pretty much a 3k load.

Of course, removing the resistors on the output transformer has meant more current through the output valve. The HT has now dropped to ~190V, which is far too low to get the full power out of this design.

I suspect a larger, higher voltage (and current) mains transformer is needed as a next step. To eBay!

For now though, it'll swing about 1V RMS, giving an output power of half a watt into 2ohm. Given the speakers I'm using (each with an efficiency of >95dB@1w), I can get some fairly loud sounds out of this

Chris

Mike H · Post by **Mike H** » Tue Jul 17, 2012 3:44 pm

chris661 wrote:Found the turns ratio to be 36:1, by applying 13mV to the secondary, and getting 0.47V on the primary.

Square of the ratio gives 1296:1 impedance ratio. Feeding an 8ohm speaker, the primary impedance will be ~10kohm

Not too surprised, since for the original purpose it were only a titchy o/p valve.

How critical is the output transformer primary impedance?

Depends what the o/p valve you're using wants, in general, more impedance means less distortion for the same valve type.

I reckon 5k is it for your purposes, 2.5k starts getting into KT88 territory, kind of thing and the gain will be worse.

In this original radio I wonder if the speaker is/was 3 or 4 Ohms? Which would make perfect sense. Have you got this speaker? Just to check....

Paul Barker · Post by **Paul Barker** » Wed Jul 18, 2012 5:25 am

If it is the one i gave him it is off a tape player.

Paul Barker · Post by **Paul Barker** » Wed Jul 18, 2012 5:25 am

If it is the one i gave him it is off a tape player.

andrew Ivimey · Post by **andrew Ivimey** » Wed Jul 18, 2012 8:35 am

ah, stereo then!

chris661 · Post by **chris661** » Wed Jul 18, 2012 9:21 am

The output transformer is still the one off the old radio I set off with. The next thing to try is the other output transformer.
The old speaker (a Celestion, I might add) measures 3ohm DC resistance, so estimate in the 4ohm nominal range.

Looks like I might have found the right thing for the HT:

http://www.ebay.co.uk/itm/Mains-transfo ... 19d332f7bf

280v RMS at 900mA should give me 400v DC with plenty of current.
Looks like I'd have to keep the current transformer too for rectifier heating, though.

Aha, wait a second.

Page 3 of http://www.r-type.org/pdfs/5r4gy.pdf indicates to me that some voltage will be lost across the rectifier plates as current is drawn. About 40v lost would take the HT down to ~360v. That'd be better than it is at the moment, but still not quite enough.

I could go to silicon diodes instead of a valve rectifier (this would get the voltage back up to the correct number), but I understand a saggy HT is part of what gave the old valve guitar amps their sound.

I think the best bet is to get that mains transformer, and try valve vs SS rectifiers.
Of course, a SS rectifier would need a HT switch or some sort of delay circuit to ensure everything's warmed up to prevent damage to valves.

Chris

Mike H · Post by **Mike H** » Wed Jul 18, 2012 11:12 am

chris661 wrote:The output transformer is still the one off the old radio I set off with. The next thing to try is the other output transformer.
The old speaker (a Celestion, I might add) measures 3ohm DC resistance, so estimate in the 4ohm nominal range.

There ya go, so primary is 5k then.

Not sure how "saggy" a SE amp can get though?

'S easy for p-p where you can drive it into class B with 'insufficient' supply current behind it

audio-talk

A foray into the world of glowy stuff.

#136

#137

#138

#139

#140

#141

#142

#143

#144

#145

#146

#147

#148

#149

#150