TDS3054B fails timebase calibration. XTAL bad.

[Jwalling]

TDS3054B failing calibration at the cal lab, timebase. With 10MHz input, scope measuring 9.980 to 9.985MHz. On this failing one, Y500 an oscillator marked AE803LVA 75.757575MHz but measures 75.8890MHz which would explain the low readings. Can't find a datasheet. No stock anywhere. Grrr… 

Closest I can find is http://www.digikey.com/product-detail/en/fox-electronics/FXO-HC726R-75.75/FXO-HC726R-75.75-ND/2155522

Not quite the right package, and no stock either. What an odd frequency they chose... Usually they are a nice even # like 100MHz. FU tektronix!

[nctnico]

Isn't this a case of a bad XTAL loading capacitor?

[MagicSmoker]

Isn't this a case of a bad XTAL loading capacitor?

This is what I thought might be happening, too, but the frequency is being pulled about 130kHz and that's about 10x more than is typically possible just from messing with the load capacitance. Trying to pull a crystal too far off frequency either causes it to operate on an overtone, not resonate at all, or crack.

[Sceptre]

Perhaps this one?
http://www.aextal.com/display-oscillators.asp?ID=29
Anderson Electronics 7 x 5 x 1.7mm (6 pad) 10 MHz to 800 MHz XO with LDVS output
When Google returned nothing for 'AE803LVA oscillator', I shortened it to 'AE803 oscillator' and came across these.

I expect that you will have to contact them directly, as Octopart has no hits for AE803L.

--Sceptre

[CJay]

I found a couple of Chinese sites with a search for AE803LV but that's about it.

None of the big four I use list a 75.7575 MHz crystal or oscillator module.

It may be worth looking into having a module made, I've had 'special' ECL oscillator modules made in the past when a certain high performance video card had one fail, it was surprisingly cheap at around £35 for a one off IIRC.

Might it be possible to replace it with one of the tiny programmable parts from Silicon Labs?

I seem to remember they had some really cute programmable oscillators in a tiny footprint that had EEPROM to store a startup configuration, marketed as a crystal replacement and could be field programmed or supplied programmed to your choice of frequency.

I'm not sure how good the stability was though.

(Edit.)

These things, best stability is 10PPM so it may not be good enough.

http://www.digikey.com/product-search/en/crystals-and-oscillators/programmable-oscillators/852047?k=SiTime&FV=fff4000d%2Cfff8004f&stock=1

[Jwalling]

Isn't this a case of a bad XTAL loading capacitor?

Nah, this is a 4 lead SMD can OSC. Attached a picture of it and the entire ACQ board.

Interesting to note that the A/D converters have the same part # between both models. ADG361D. There are no caps from the analog input hybrids to the A/D converters either, like in Jay_Diddy_B's thread here:
https://www.eevblog.com/forum/testgear/conversion-of-500mhz-tds744a-to-1ghz-tds784a/

I compared both ACQ boards and I didn't see any strapping resistor differences either. This leads me to believe that the 3014B could be upgraded to a 5054B with programming only...

I'll place the order with Digikey and I'll see what happens. For now I salvaged one from a TDS3014B that doesn't finish booting which I suspect has corrupted firmware.

Thanks everyone!