AVO CT160 Calibration

[chiilyuk]

Thought I would check over this, new to me, CT160 and calibrate it. I spent a fair while reading CT160 related threads on this forum, some very experienced and knowledgeable people here.

I just wanted to relay my experiences, pass on some knowledge and foolishness.

When I first turned the CT160 on, set the selector to ~, the needle would only just reach up to the LHS of the black sector, even with the mains set to 240V, -5. I tried different times of the day, but the needle settled just to the left of the black sector. Reading around pointed to R3 and R4, but more on those in a mo.

Reading more CT160 threads pointed at a meter replacement was costly to find an original, or would take some jiggy-pokery to replace with a different sensitively meter. Not wanting to pop the meter I decided to add some meter protection as detailed here https://www.vintage-radio.net/forum/...6&postcount=58

On opening the bottom case I noticed that R9, in parallel with the meter (10k) was in parallel with a 100K resistor. I measured both by removing them.

[img]https://***********************/65535/53206168599_bdcfb7cdd4_c.jpg[/img]PXL_20230922_133153456 by Garf Arf, on Flickr

That did seem rather strange.

I replaced the 10K with a metal film 100K, 1%, adding two diode pairs and a 8uF film cap.

Switching the CT160 back on, on ~, the needle was too far to the left on the black sector. Moving the fine mains voltage setting to 0 brought the needle to the middle of the sector...kinda strange.

I decided to re-calibrate the 160's; I had bought and checked an AvO 8 mark V to do this, having read about how to measure the correct voltages over RV2 and how DVMs can mis-read the voltages.

The first test is to measure the voltage over RV2 with the link open. The test is not clear whether this test should be with the meter set on AC or DC,

(a) Set mains ‘On/Off’ switch to the ‘Off’ position.
(b) Open LK (Fig. 3) and set the panel controls as follows:–
CIRCUIT SELECTOR to TEST, ELECTRODE SELECTOR to A1 and NEG. GRID
VOLTS to 40.
(c) Connect the Testmeter across RV2 (Fig. 3). Switch on and adjust RV3
(Fig. 3) until a voltage reading of 20.8V is obtained. Remove the Testmeter.

Whilst the next section says...

(a) With LK still open, connect the Testmeter set to a suitable dc range, across
R5.
(b) Switch on and check that when the dial is advanced to its 10, 5 and 2mA/V
positions, readings of 52.5mV, 105mV and 250mV ± 3% are obtained.

Should the first test be AC or DC?

I conducted the first test with a Fluke 79 II and measured 20.8VDC. I remember reading the this should be an AC reading. So I swapped to the AVO meter, selected AC, marked the RV3 pot's position and adjusted it to get 20.8VAC.

And so I proceeded, all the voltages lined up. The problem I was having was whilst the voltages were right, I couldn't get the needle to get into the black sector with the link closed and the selector set to ~. No matter what I did to the mains selection on the front panel, 47VAC across RV2 and the needle in the sector weren't happening to together, and often I wasn't getting near the 47VDC mark.

The next step was to lift one side of R3 R4 and check their combined value; 1.31M, so they are okay. I had to be doing something wrong.

Much scratching of head and a sleep.

Then I remembered that the Fluke had read 20.8VDC when I first tested with the link open. I retraced my steps, re-opened the link, moved the selector to test, and twiddled RV3 until the Fluke measured 20.8VDC. I purposely didn't watch the position of RV3 as I adjusted it so I wasn't influenced by its final position wrt the original mark I had made at the start. Unsurprisingly RV3 was right on the mark....

I closed the link, the voltage lifted to 47VDC and the meter needle rose to LHS of the black mains sector. I fiddled with the fine mains adjustment, but I couldn't get the needle in to the black sector.

Round the loop I went again, 20.8VDC....47VDC...needle to the LHS.

Then I remembered the 100K on in parallel with the meter resistor, R9. Fished it out of the parts bin, tacked it across the meter and the needle lifted into the mains sector. Hooray!!

The two morals to this are first, check voltages before twiddling knobs, you might just find that the device is still in calibration and two, if someone with greater knowledge that oneself has fitted a mod, the chances are the mod is there for a reason.

Do 41 Comd REME fit into "greater knowledge"?

[img]https://***********************/65535/53204909107_a406f05103_c.jpg[/img]PXL_20230922_100831024~2 by Garf Arf, on Flickr

Anyway, I checked around the CT160 for other voltages and they all seemed to be close enough to those in the service manual.

I'm really happy with the CT160. Its measurements are agreeing with my friends CT160.

Apologies for a rambling post....

[2000 type]

41 Command Workshop REME was at Strensall, N. Yorkshire. It was closed in 1993.

[Chrispy57]

Hi - we all learn from experience, eventually, but can also benefit from the experience of others, so thanks for the tips and the story. BTW, did you actually mean this - "I replaced the 10K with a metal film 100K" or is it a typo?

Cheers
Chris

[chiilyuk]

Ah, indeed, should be like for like, 10k for R9.

[Wanderingfree]

Hi Chilly UK
I have the same problem, except mine seems to be worse, so to be sure of what you said. Did you put a 10k resistor in parallel with a 100k resistor?
Mike was sort of ok before I did the calibration with a Fluke 179 when I went to adjust the reading as it was was 21.9v DC so I wound it down as per manual to 20.8V and that’s when my meter problems began, it is well off the mark.

[John_BS]

The 100k should not be fitted unless there is a known problem with the actual meter sensitivity. Its presence alone may well cause the issue you've observed.

With the link in place, there are two interleaved half-wave rectified waveforms present across RV2, of different amplitudes. The mean DC of the two combined should be 47V if the input mains, rectifier diodes and mains taps are correct. From the service manual:

SECTION 4 – CHECKING THE SET ~ INDICATION
(a) Standardise the Testmeter at 47V dc. Close the link LK and set the panel controls as follows:–
CIRCUIT SELECTOR to ‘SET ~’, and ELECTRODE SELECTOR to A 1 .
(b) Connect the Testmeter across RV2 and switch on.
(c) A reading of 47V should now be obtained, whilst the meter on the the instrument panel should indicate within the ‘~’ zone.
If voltage reading is correct, but panel meter indication is outside ‘~’ zone, check Resistors R3 and R4.
(d) Switch off and remove the Testmeter.

The meter alone is 30uA, (3.23k): in // with 10k = 39.7uA , (2.44k)

So, with R3+R4+2.44k and 39.7uA*0.9 fsd we arrive at just over 47V.

John

[Alf Fisher]

You might find the AP for the CT160 gives useful info.

This may be downloaded from
https://www.blunham.com/Radar/Signal...ment/CT160.pdf

Good luck.
Alf

[David Simpson]

Like Alf, from our long off days working on RAF Test Eq. MU's. Me at 103MU & Alf at our parent MU at Henlow, I remember AP2537F.
The sinusoidal nature of CT160 Valve Tester's & their VCM sibling's working voltages often confuse folk when it comes to AVO Manuals and several AP's floating around(including an American Forces one). These Manuals & AP's often give the impression that test voltages & calibration voltages are of an RMS or DC nature. No - they are not. Best bet - get busy with a 'scope & study them. Use a safe x10 probe for Va & Vs, as full P-P Va values appear on both meter terminals under "Test".
"Set ac" for example is not in any way connected to the meter circuitry - its just a mains transformer's adjustment of the primary winding. On CT160's - this is to ensure that the(sometimes overstressed) single transformer's 55V winding is giving out just that - - 55V. If that's wrong, then the working -ve Vg peak pulse's accuracy goes to pot. That's just a 10mS pulse. AVO threw in the 66V pulse just to enable a test technician to get some sort of readable DC value when reading -veVg with his AVO7 or AVO8 MM.
Using a decent scope these days - not a heffing 13A back then - one can easily determine all the pulse's values using correct v/d on a decent metric graticule. Go on - give it a shot, chillyuk.

Regards, David