Mr.French's Dynamic Mutual Conductance Tester.

[TowerRadio]

Split from this thread:-

https://www.vintage-radio.net/forum/...d.php?t=154175

Has anyone built and used Mr.French's Dynamic Mutual Conductance tester from an old issue of of Radio Constructor? I have a HSVT but would like to know if the the former was of any use. Les

[Bazz4CQJ]

Can you post a schematic or the date of the Radio Constructor edition?

B

[Electronpusher0]

If you have a Mullard High Speed Valve Tester I started a thread
https://www.vintage-radio.net/forum/...d.php?t=149561
which describes my modifications to make it measure gm (I expanded the project to also display HT volts and Current and grid volts).
It it is still work in progress as far as casing goes but it does work.

Peter

[TowerRadio]

First published in June 1956,the article is shown complete in the National Valve Museums "Technical Articles, re-published articles,General articles" in that order.I obtained a 4KHZ solid state oscillator and was all set to make a prototype when I was struck down by CLL and things went downhill from there.It was at this time I travelled to Northampton to buy a Mullard HSVT and the repairs to that went on the back burner also.The HSVT I got going last year and it does more or less what I want but I still have the inclination to try the RC circuit. The HSVT certainly detects bad valves but how good are the "good" ones? Les

[Bazz4CQJ]

The circuit and a very brief description are attached.

I don't think I have the expertise to look at that diagram and make any comment

B

[Radio Wrangler]

Yes, that's a viable bridge circuit. Tune the pot to dip the audio and read Gm from a scale on the pot.

Gm is the amouunt of anode current variation from a 1v grid voltage variation. The anode loar R turns this into voltage variation on the anode in antiphase to the grid. The anode swing amplitude shifts the position of the null point on the pot.

David

[Bazz4CQJ]

Wot! No 50uA moving coil meters ?

[Radio Wrangler]

The posted circuit will allow a user to null a bridge and read mutual conductance.

It doesn't include leakage and gas tests, heater continuity etc.

It includes one stage of amplification in the detector path and bridge nulling could be done with headphones.

Avo wanted a machine where the operator pressed a button and read Gm off a scale. There's more labour involved in balancing a bridge before reading a scale. The leakage tests either need a sensitive meter or else a DC amplifier whose gain would have to be very accurately controlled. Do-able in the valve days, but a much more awkward and expensive process than it is today. So Avo went for 30uA meters, right at the limits of fragility.

Avo's Gm circuit is also a bridge, but uses a meter to measure the amount of unbalance caused by a change in grid voltage.

The posted circuit uses an AC bridge and reads Gm from the position of the potentiometer after balance has been manually achieved.

David

[TowerRadio]

Well David's comments have made my mind up to try this circuit.I will not be modifying my mark 4 HVST but I do have a old homebrew emission tester from which I can obtain neccessary voltages. I guess my first step is to make up a panel and layout a readable scale to match the pot I bought a good while ago.Les

[Bazz4CQJ]

Yes, it falls short of an Avo, Mullard,Taylor or Hickok, but would presumably be a much better option than a simple emission tester? The easy 'calibration' for Gm is a big plus.

Curious that the Mr French's design seems to have been largely forgotten, or hasn't re-appeared under another guise.

For perspective, the same edition of Radio Constructor carried a major article for a converter for Band III TV!

B

[Skywave]

Originally Posted by Radio Wrangler

Yes, that's a viable bridge circuit.

Would you be kind enough to re-draw the basic concept as a bridge cct. for me, please? I tried, but cannot manage it.

Thank you.
Al.

[ms660]

The signal and bridge idea goes back a fair way, here's one from General Radio c1929:

https://www.ietlabs.com/pdf/GR_Exper...y-Aug_1929.pdf

Lawrence.

[ColinTheAmpMan1]

I can see that there is a bridge circuit there with the two fixed resistors and the potentiometer. My puzzle is - how do you calibrate the potentiometer in gM units? It sounds like a bit of a faff; maybe I'm missing something?

Colin.

[Radio Wrangler]

OK, You'll need the two circuits pdf'd in post number 5 in front of you. I couldn't draw anything clearer than fig 1 to show the basics. This is a stripped-back circuit without bias and dc blocks being shown.

Let the input have an appropriate DC bias with 1v AC superimposed on it. THe 1V AC also gets fed to the bottom end of the pot, R10.

The valve sees 1V of grid-cathode AC, and so it responds by varying its anode current by Gm milliamps. Gm milliamps flowing in the 1k anode resistor makes an AC signal at the anode of Gm volts.... BUT IN THE OPPOSITE PHASE TO THE GRID VOLTS.

Let's assume the pot, R9, R10 are very large values compared to 1k, so they don't load the anode. If the valve as a Gm of 1mA/V we get 1V AC at the bottom of the pot, and the anode applies 1V in antiphase. These cancel at precisely the midpoint of the pot. put a mark there for 1mA/V

if the valve has a Gm of 2, then we get 2V at the anode There are 3V across the pot. 1v at the bottom, 2V antiphase at the top, and they cancel out one third of the way up. Mark it Gm=2

If the Gm=8mA/V then we get 8v at the anode, 8V at the top of the pot and still only 1V at the bottom so 9V across it, and our null point is now 1/9 of the way up.

So you can measure the pot with an ohmmeter and calibrate it. Check the anode resistor's OK and Bob's your aunty Fanny. Simples! and elegant.

The second circuit introduces some niceties.

Firstly the audio signal source gets a cathode follower to buffer it so it doesn't see any low load impedance. Its cathode resistor is tapped down to pick off 1/40th of the AC voltage to feed the pot. (more about why later) The valve under test gets a grid blocker and an external supply setting its grid voltage to the point on its curve where you want to measure Gm. Another external supply powers the anode to the voltage chosen for the test. There is only a scrotty little 25 Ohm anode load resistor, so voltage drop doesn't move the anode much from the chosen test voltage. If the input pot is twiddled to set an AC amplitude any will do, but let's just say 1V... we get 1/40th of a volt from the cathode tap into the bottom of our bridge balancing pot. The valve anode gets Gm milliamps of signal, and across 25 Ohms, that gives us Gm/40 volts of signal on the anode, and at the top end of our bridge balancing pot. Hey that's 1/40th again! at both ends of the balance pot. So the balance position and Gm calibration isn't affected by the 1/40th business, but the output level away from balance is weedier. That's OK, the pentode amp running at high gain makes up for that.

The 25 ohm resistors make for low impedances driving the balance pot, and so it's easy to have the balance pot a thousand times higher resistance than the pair of them. I'm not going to lose sleep over maybe 0.1% error.

Not all bridges have four resistors. This one uses two resistors (the pot) to find the balance point between two different and opposing voltages.

Smart.

With a name like Mr French's Dynamic Mutual Conductance Tester, you'll have to get dressed-up suitably to do the hunter-gatherer bit for components. A leather waistcoat, big-wing jodhpurs, a free choice of long, brown boots or puttees, a pocket watch, a pair of welding goggles and a free choice of pith helmet or short topper. The blunderbuss is absolutely essential, but must have added pipes and clock gears.

Have fun

David

[Skywave]

Originally Posted by Radio Wrangler

I couldn't draw anything clearer than fig 1 to show the basics. This is a stripped-back circuit without bias and dc blocks being shown.

O.K.: here's my incomplete (and probably incorrect) analysis.

Sketch a basic bridge cct. with 4 impedances. Using points of the compass as reference, let A be north, B be east, C be south and D be west. Then we have 4 impedances, Zab, Zbc, Zcd and Zda, where the lower case letters are the same as the uppercase letters (i.e. A = a, etc.)

Studying the 'conceptual cct.', to me, I see the following;
The detector is connected to A and C; C is at 0v.
The anode is at B.
The grid is at D.
Zab = R9 and a capacitor.
Zbc = the anode load resistor.
Zda = R10 and a capacitor
Zcd = ? Cg-a? The bridge 'source' is between C and D, D being the grid. But for a bridge cct., the source sits between B and D; detector between A and C.

Hence my remark "incomplete (and probably incorrect) analysis".
Hence, my earlier request of you.

Al.

[Radio Wrangler]

The bridge in this circuit does not have four impedances.... three at a pinch

It is balancing the ratio of two resistors against the ratio of two voltages (the grid swing and the anode swing)

One of those voltages is the product of the anode load resistor, the Gm of the device and the grid swing.

We know the pot ratio from its scale, we know the anode resistor, and rather delightfully, the grid swing cancels out. So we can calculate and fit a gm scale to the pot and it remains accurate even if the drive swinging the grid changes.

David

[Skywave]

Originally Posted by Radio Wrangler

The bridge in this circuit does not have four impedances.... three at a pinch

I see - I think - so thank you. A 'bridge' that does not have four 'arms', eh? Hmm: that is quite contradictory to what I have always understood to be an essential feature of a 'bridge' in electronics circuitry.

Al.

[dave cox]

There is a bridge at R7, R8 & R9 (with R9 viewed as 2 resistors)

dc

[Radio Wrangler]

Think of a see-saw being used to compare two weights. You are interested in the ratio of the weights of the weights, but aren't concerned in their absolute values.

You put one weight on each end of your seesaw, then you move the fulcrum along the seesaw until it balances. You now use a ruler to measure the distance from the fulcrum to each of the ends. The ratio of the distances (done the appropriate way up) is equal to the ratio of the weights of the weights.

In this case the weights are the grid and anode AC voltages, the see-saw is the pot and the position of the fulcrum is the wiper.

Take it all to Mars, where the weights will each weigh less, and the balance point stays the same.

In Mr French's circuit, reduce the AC drive, and the balance point stays the same.

David (Analogies R Us!)

[Bazz4CQJ]

Well, Les says he started work on the front panel yesterday, so hopefully, he will be reporting back on the complete tester by ... Monday? No pressure .

B