Group Actions
Engagement
  • Author Author: michaelkellett
  • Date Created: Date Created
  • Views 1690 views
  • Likes 9 likes
  • Comments 7 comments
Related
Recommended

HP 3325A Synthesizer/Function Generator

michaelkellett

I recently bought this instrument on Ebay - it's quite an old design and if you fancy one the 3325B might be a better bet since it has a much more modern processor and a few extra functions (but will cost more).

It was first released in 1979, at $3000 (roughly $13k at today's prices). Like most HP instruments of that era it has full HPIB remote control.

I'm not going to describe it in detail because that is done so much better than I could in Giovanni Beccattini's beautiful (and free) book "The Great Hewlett Packard" , download (2GB) at www.k100.biz

My 3325A was advertised (and marked on the back) as fitted with option 2 (the high voltage output amplifier) but had been modified at some time and fitted with the High Stability Oscillator instead - so I'll be building a high voltage output amplifier of my own some time.

The fancy oscillator works OK - I'm not sure exactly when my 3325A was built or last calibrated - but with the output set to 10MHz the frequency is within 12ppb (ref a Racal counter using a GPS disciplined clock (Leo Bodinar miniGPS))

I haven't found anything that doesn't work yet but at £200 I expect most people would prefer something smaller and with the greater flexibility of a DDS generator. But you won't get a service manual with a modern instrument !

image

(apologies for the croc to BNC bodge, I need a long lead to reach the counter !)

image

MK

  • Yes, HP used to be the go to brand for these instruments.

    They always had the best quality.

    Great find.

  • in reply to shabaz

    Found a couple of these, they are ancient and obsolete (they were for a project that never worked out).

    image

    It has a 1 PPS clock output. I was wondering if they could be used for quick identification of not-so-good oscillators, by simply counting the oscillator (e.g. using a FPGA counter or a microcontroller) and gating on (say) 1000 seconds of pulses from the GPS module, provided that no 1-second pulses were missed over 1000 seconds (easy to check with a normal RTC). Obviously this isn't very long-term, and the oscillator could be a lot worse than reported, but a bad reading from this test would indicate an immediate reject.

    I don't know much about the timing output with GPS modules. This bit didn't make sense to me in the datasheet, since the red underlined parts contradict each other:

    image

    I guess at some point one may as well just get the ready-built GPSDO, since it's a reasonable price for what it is. 

  • in reply to michaelkellett

    Found it:  Time is "cheap" - 2nd user low cost precision oscillators  

    By coincidence I was looking at the same GPSDO that you used, at sdr-kits.net the other day!
    I have not purchased anything from them before, and was tempted, but was shopping for a SMA VNA calibration kit at the time (this one: https://www.sdr-kits.net/Female-12-GHz-Kit ). I think that arrives tomorrow, so I'll post some photos of that once I've tried it out!

    The 10 MHz oscillators on AliExpress are extremely low cost so I've ordered 20 of them (for £31 total) without PCB, and will try to measure them at some point once I have a method for that!

  • in reply to shabaz

    There is a lot to be learned from HP designs. Of course most of the parts are obsolete now but the concepts are as relevant as ever..

    I have several little Chinese ovened 10MHz oscillators from Aliexpress which would make very good references for an instrument.

    https://www.aliexpress.com/w/wholesale-10-MHz-ocxo.html?spm=a2g0o.productlist.search.0

    They are still selling them. AFAIK they are all pulls from old gear - very well aged and likely to be stable unless you are unlucky and get a duff one.

    If you buy the ones without a support board they are likely to be delivered still soldered to a a piece of ciruit board which has been sawed apart. Easy to remove from the board by clamping the oscillator can in a vice and using hot air on the scarp of board. The first one I did I stuck a thermocouple to the oscillator and it didn't get above 85C.

    I think I did  a blog about measuring these oscillators a while ago.

    MK

  • Wow, nice instrument! (and the frequency counter!). Really neat that it has such extremely high granularity. And also supports log sweep.
    I just took a peek at the service manual. I might steal a bit of the reference input circuit.

    image

    I want to build a freq synthesizer (discussed not in the blog, but in the comments here:  Using a MAX2870 Frequency Synthesizer / Signal Generator  ) and I don't have a 10 MHz external reference yet. The circuit I was going to use:

    image