Hulp gevraagd voor AVO CT-160 (Techniek Overig)

door Leo Snoeren , Dongen, 05.08.2023, 20:53 (266 dagen geleden) @ Patrick2
Bewerkt door Leo Snoeren, 05.08.2023, 21:04

Dag Patrick,

Bij deze het verhaal (in het Engels):

A practical method of replacing the Set Vg control, RV2, on the Avo CT160
Mike Phelan 3/2005

Having acquired one of these instruments with RV2 missing, sourcing its replacement was the first thing that was needed in the restoration. As I have not met the CT160 before, I naively looked at the circuit, and breather a sigh of relief that this pot had a standard value – 10k.
My euphoria was not to last, as those more knowledgeable (almost everyone!) informed me that RV2 was certainly not a standard component – the track was a one off!
Looking at the set Vg scale on the control, that which I had missed became obvious. The scale was divided into three 90 degree sections, each linearly calibrated from 0-5v, 5-15v, and 15-40v respectively. Ho hum.
A quick calculation showed that the equal thirds of the track would need 1.25k, 2.5k and 5.25k values.
We needed a plan.

First thought was to fit a 10k linear pot, and a new scale. This was discounted because the bottom third of the scale would be 0-13.3v, and therefore very cramped.

Next thought was to ‘bend’ the curve by a resistor from the wiper to one end. I knocked up a computer model of this, but the best result still gave a 7.1% error – unacceptable.

I then considered a rotary switch and chain of resistors. The valve data book states Vg in smallest increments of 0.1v, so we would need 50 divisions in the first section – extrapolated to 150 poles in 270 degrees.
I could have fairly easily used my clock wheel cutting set up to cut radial grooves in a piece of laminate and fabricated a rotor; the real problem was connecting 149 resistors to this lot. Plan C scrapped!

A bit of lateral thinking; if I could get a higher value wirewound pot, would it be possible to ‘tap’ the track and connect resistors in parallel with each section to get the correct original values? Nichrome resistance wire would have been difficult to solder to neatly, but how about conductive silver paint?
As it would have to be a wirewound, and the highest value generally available is 25k, I acquired a Reliance 25k pot with 280 degrees rotation from a fellow enthusiast.
Next thing was to mount the CT160 scale on it and carefully advance from the minimum end until the reading started to increase, which it did not do immediately – there was a ‘dead’ spot of a couple of degrees. At this point the scale was set to zero, and advanced to the 5, 15 and 40 volt settings, and the position of the wiper marked on the track with a fine pen.
These points corresponded to resistances of 7.62k, 15.73k and 24k. To get the calibration right we would need three fixed resistors and three presets across each section. 3k9, 12k and 1m for the fixed, 5k, 10k and 50k for the presets.
I dismantled the pot by carefully removing the circlip on the spindle followed by the spindle.
The track ends are connected by being wedged between the case and two retainers. The track ends have holes into which it was possible to insert a small screwdriver to ease it out.
The track was covered in insulating paint except for the last 2mm or so on the edge where the wiper contacts.

The next thing was to remove the paint where I had marked the track earlier, by scraping carefully with a Stanley knife, so the connections could be made.
A couple of strands from some hook-up wire were taped on to the track, following the direction of the winding, where the paint had been cleaned off, and under a high power magnifying glass the connection was made with conductive paint, taking care not to leave too wide a ‘footprint’.
As this arrangement had no mechanical strength, I followed up with a blob of PVA adhesive over the connection. When this dried, the tape was removed.

Now we needed a way for the connections to reach the outside world – achieved by cutting a couple of slots in the pot body.
It was now possible to reassemble the pot with the wiper connection ring cleaned and the spindle greased, after shortening it. Great care was needed not to damage the connections made earlier; they were taped to the outside of the body, and reinforced further with a tie-wrap around the body.

Only remaining job was to put the three presets and fixed resistors on a piece of stripboard, and make a bracket to attach this to the bar that the pot is mounted on. All connections were completed now.

Calibration
Even though the supply is rectified AC in the instrument, it is easier to calibrate it with a supply of 40 volts DC, so the readings correspond to the scale calibrations, so this is how we did it:
Turn pot to zero, and attach scale at zero, but only tighten grubscrew lightly. Measuring resistance from wiper to lower end, rotate pot until the reading only just starts to increase, then turn it back until reading drops to zero (or practically, a few ohms). Note scale reading, probably less than 0.5v, and rotate scale anticlockwise against stop so the reading starts to increase as soon as you move off zero on the scale. Will need a few tries; when you have succeeded, tighten grubscrew, then put a meter between the wiper and lower end and:

Set scale at Adjust
40v Supply
20v 50k preset
14v 10k preset
4.5v 5k preset

Repeat all these adjustments several times until no further adjustment is needed to correct readings. You can then check the calibration throughout the range – this is what I got:

Scale Actual % Error
1 0.96 -0.04
2 1.95 -0.03
3 3.09 0.03
4 4.09 0.02
5 5.05 0.01
7 7.04 0.01
9 9.04 0.00
11 10.8 -0.02
13 13.35 0.03
15 14.58 -0.03
20 20.07 0.00
25 24.8 -0.01
30 29.9 0.00
35 34.8 -0.01
40 40 0.00

After everything has assembled, the calibration was all checked as per the manual, and worked fine.

Deze oplossing heb ik al meerdere malen toegepast en werkt uitstekend. Voorwaarde is dat je een bruikbare potmeter hebt, liefst Engels i.v.m. de asmaat (6,3mm). Indien gewenst: ik kan je bet ook sturen als Word bestand, stuur me in dat geval een email.

De potmeters zijn nog steeds te vinden op beurzen en zo.....

Succes!

Edit:

Ik heb er ooit een printje voor gemaakt, dit is het schema:

En dit is de layout:

Maar op een stukje experimenteerbord gaat het ook. Daarnaast had ik toevallig 10 slagen potmeters liggen, maar met gewonen instelpotmeters werkt het ook.

--
Met vriendelijke groet,

Leo Snoeren

https://www.rrd.link/ (onder constructie)

leosnoeren(apenslingertje)freedom.nl