Wavetek 1002 calibration

[Rax]

I'm working on a DOA Wavetek 1002 that has at this point most functions working. Some of the things I am seeing that are not quite "square," though.

• Linearity seems pretty bad during sweeps, at least by what I see on my scopes screens. This needs a grains of salt, as my scopes (500MHz HP 54522A and about 250MHz 7704A) don't reach as high as the Wavetek - or maybe I should say they all reach as high and I'm not sure who's linearity I am seeing (or not seeing). But there's fluctuations of level during the sweep up and down and up again (should be within 0.25dB). Reason for thinking it needs recalibration
• I can't see any signs of life from the markers (there's a 1MHz and a 10MHz). Tried spectrum, scope, nothing at those two frequencies I can discern. I didn't try to measure the modules themselves, just the RF out. Maybe these work in a different way than I imagine - quite possible.

Now, calibration went mostly as expected (I am at 5.4.7 per Service Manual).

One thing I am not happy with, and would love thoughts on, is at 5.4.6 SWEEP RATE ADJUSTMENTS (module M1H). On p.5-9, second column, with VAR/MANUAL fully CCW the waveform looks like this (see HP plot). The service manual shows a pretty vanilla triangular waveform, but mine isn't at all pretty. Is seems there some pretty severe limitation of the top level?

As shown, I was able to adjust to the required 120ms, but I am wondering if some of this stuff I'm seeing isn't responsible for the overall non linearity, and so maybe I need to go beyond adjustments, and some parts need replacing.

[Rax]

For a bit more clarification, I don't see the tilting of that peak on the entire range of the variable sweep time control. May be that turning logarithmic on the lower part of the range, as this is manual sweeping after all, is deemed to be OK?

Anyway, another a bit irregular an issue I am observing is at the next step (5.4.6 step #3). I enclose another scope screenshot, and need to add adjusting the wait and the clamp voltage are a bit antagonistic. I shot for the best compromise, and it was about -15.75 (clamp) vs. about 0.5ms (wait) last night. I am seeing what's on the attachment now. Probably OK.

[Rax]

Going pretty well this far - except some persistent intermittence on the module sockets - but I hit a snag. 5.4.8 is a level adjustment that requires a power meter (SM recommends HP 432A), which I don't have.

I wonder if anyone would have some thoughts on using the HP 54522A possibly for this measurement (attach a load at output?...). For my unit (1002), the measurement is supposed to be performed at 300MHz (easily doable for the scope). Certainly, for one thing, I'd probably stick with the 10x probing, and maybe use a Tektronix 50ohm passthrough which is rated at 2W as Waketek load.

Can't think of anything else I have here that may fit the bill in some way.

[Rax]

I went ahead and used the setup I described. Results seem to confirm I am well within the ballpark - I get less than 0.58dB off flat response across the entire range of +13dBm to -7dBm (per SM).

[Rax]

I also think I'm done with the part of the calibration I feel I need to do (through 5.4.8 ). From that point on (= series of XY measurements) I'm unsure a 250MHz analog scope can perform linearly enough. But flatness has been restored through the prior steps, and it's now well within specs, I feel.

As I completed this last night late, and proceeded to pour a well deserved single malt, I suddenly lost the sweep function. I don't recall if I was buttoning or whatnot, but it'd repeatedly initiate and complete the sweep sequence, without any actual sweeping (solid sine at what CW would have it). Some modules seemed very warm to the touch and the thoughts on intermittence came back to mind, but then I dove back into the SM and some late-night thinking on it - no further work, I decided - the possibility that S105 may have stopped switching started getting a solid profile, so I encircled the likely culprit on my schematic to explore in the morning.

Today I rigged some jumpers to verify my theory, and there it was - sweep was back in all its glory. Linear too, per calibration the prior day. So until I get a replacement pot (10k linear w/ DPDT or DPST) - how this switch could fail both poles at the same time eludes me - I decided to temporarily remove the "Ext. ALC in" jack and place a switch on there for toggling between CW/sweep.

Some pictures of this saga enclosed. One knowing the innards on this will see a green, a red, a black, and a white wire with some zip ties.