Golborne Vintage Radio

Mu is pretty much a function of tube geometry and fairly constant, but the other two vary. Gm can vary by a factor of three, for perfectly reasonable operating points.

(17-07-2015, 08:53 PM)Alan Douglas Wrote: [ -> ]Mu is pretty much a function of tube geometry and fairly constant, but the other two vary.  Gm can vary by a factor of three, for perfectly reasonable operating points.

Indeed. This leaves Ra inversely proportional to Gm, which, as you've noticed, can vary a lot, meaning Ra is also highly variable...

New tube-tester owners always want to know "the" Gm of a tube, and why it varies from one machine to another, and how to calibrate their pet tester to read the "right" value. Fun to watch from a safe distance.

I'm trying to think my way around a few problems and hope you may have some thoughts for a simple and clever solution to each. The problems relate to the LT supply for the curve plotter, which if you recall I'm proposing to derive from the AC mains supply.

To make operation slick I thought about employing fixed and trimmer resistors to produce 1.5, 2, 4, 6 and 6.3v heater supplies, selected via a rotary switch. The problem however is the range of filament current, which for 2v valves alone can vary from 0.07A to 0.2A. Other than reverting to a pot I don't know how to get around this. Any thoughts?

The next problem is filament protection from surge. Feeding the LT from the mains is all well and good, but is there a risk that a small power surge may blow the filament of a 1.5v valve? If so, what's the easiest solution?

I'm also concerned about valve performance. Throughout the duration of the tests I need a constant filament voltage. No problem using a DC cell, but I'm proposing to use a mains supply. How do I guarantee an even supply?

I really don't want to use batteries to supply the LT circuit.

Thanks.

Nick.

Well one way would be to find an old transformer, of enough VA, without pitch or any encapsulation and pull the laminations out of it (carefully noting how they go and keeping the sides in order, they are coated to prevent eddy currents). Hopefully you will be lucky and the heater windings will be on the outside of the former (that is usually the case).

Say there are a couple of heater windings one 6.3 and the other 5V. Unwind them both noting the number of turns for one of them. This will give you the turns / volt of that transformer (normally around 5). Then rewind with suitable new wire (enough for the current) bringing out taps for 1.5, 2, 4 etc. Just bring out a loop and carry on winding.

You would need to devise a method of terminating them or take them direct to a switch or better IMHO sockets and have a fly plug for selection.

For varying mains voltage, supply the primary from a Varaic with an accurate meter on the heater winding in use.

No need to worry about mains surges and spikes that's how they are in radios and they are pretty robust devices (why they still work 80 years on).

It might occur to you that you could use a 6.3V transformer and just drop that voltage with the Variac, its possible if the winding has enough current capability but more dangerous than selecting a voltage on the secondary side.

Gary

There is another way that I have used but for lower currents than you are into.  It means getting a large Toridal transformer of suitable VA..  It want to be large so that it has a big hole in the middle

Then get some flexible wire, the silicon covered sold at swaps would be a good bet.  Wind say 8 turns through the hole and measure what volts you get, again you get the turns / volt.  Then you need a long length of wire and wind it through the hole bringing out taps as before.  To make it easier to handle the wire you could make a long thin spool.

Terminate, supply and measure as before.

Of course you could go DC and use a series regulator:  there are plenty of designs around for Battery Eliminators.  With those its possible by just using different voltage setting resistors to give you the steps you need.

Supply and measure as before.

Gary

I don't bother too much about Gm when testing, nice if its close to AVO data but emission is the important thing (well after no shorts and gas).

Gary

Mmm, those are elegant solutions Gary, but perhaps a bit beyond where I am at. It has however given me an idea.

If I bought a 100VA voltage control transformer with AC outs of 6.3V 24V 36V and 220V could I then feed the 6.3v tap through a continuously variable stabilized output power supply like the one here (http://www.electronics-lab.com/project/0...0-002-3-a/), but adapt it to provide a maximum 6.3v / 3A output?

I could then use the transformer's 36v tap to feed the GB and the 220V to feed the HT.

The transformer is this one http://www.ebay.co.uk/itm/191633460257?_...1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
There's only a couple left so I'd have to bag one quickly.

Just a thought....

Nick

How about an LM338...minimum component count.

Lawrence.

Yep, that would do too. Is the approach a good one though Lawrence i.e. ref the transformer?
Thanks
Nick