3325A SYNTH/FunGen connectors, ampl CAL ?

[jonpaul]

Bonjour a tous,


Needed to check AC V CAL of my Keysight 34465A DVM, my best choice was  1982 HP 3325A synth. as I have no Fluke 5100 etc CAL.

Unearth the HP unused for about 15-20 years...turns on ...BLANK DISPLAY!   

By the time I found the bound loose-leaf orig serv man, and powered on off a few time...IT WORKED!

Opened up, super clean, gold plated boards made in USA 1982 clean as new!

After self check, warmup and CAL, seems 100%!

 So, checked sine  results at 1 mV ...1V RMS and 1 kHz..300 kHz.

All my ACV meters, HP 400 FL, HP 3400A Fluke 8920 and 8921 TRMS. also with  Fluke 8842A DVM
 consistent results with error ~ 1.0033 for a 1.000 VRMS 3325 setting.

1/  The amplitude adjust procedure in the  service manual seems to be  complex. Anyone ever CAL the amplitude on 3325A?

2/ These seem prone to flaky Molex or RCA connectors used for interboard connections. Exercise them, contact cleaner, connector plus etc?


Appreciate any experience or any tips.

Many thanks


Jon

[trobbins]

Yes it may be worthwhile 'exercising' the various connections as a start.

It doesn't sound like there are gross problems (I had a few related to particular control functions not working due to faulty LM310's).  I did have an rms output that was not accurate.  My notes say that "Multiplier U11 adjusted for R60 Y offset, and R68 Offset out, using cal procedure in 5-13", and then levels were within spec.  But I agree, the manual can be complex to navigate if a fault may be related to various sections and you have to track through to locate the various sections of relevance.  And I still have one noticeable non-spec issue that the sine distortion is not within spec, and that fault requires tracking through various sections and I got as far as confirming it originates before the Mixer stage but then gave up.

[jonpaul]

Trobbons many thanks, I can only marvel at the fine electronic, system and mechanical design of this instrument!


I agree re connector exercising.

Looking at the block and amplitude control I see why the CAL for AMP is a bit unintuitive, but 5-13  procedure is not difficult.

Besides the ampl   1.003V for 1.000 RMS indicated  other specs   FREQ, THD   seem fine.

The 120V fan is too loud, I am adding a series AC 1 MF cap to reduce the speed.

I plan to place a Leo Bodnar GPS OSC at 1 or 10 MHz as an ext ref.

Open to all advise and comments!

Bon soiree,


Jon

[SilverSolder]

Perhaps try measuring the 50 ohm load you are feeding - if the 50 ohm load is not exactly 50 ohms, the RMS voltage won't be precise to several decimal places either?

[jonpaul]

SilverSolder: KUDOS! great diagnosis!

Of  20 feedthru BNC terms  most nearly  51, a few 50.5 or 49.5 and some   55 Ohms! (ALL marked 50 Ohm)

Finally I discovered an EGG TF50 perhaps from 1960s, which is indeed 50.00 Ohms.

The ACV error is now corresponding less! The other meters confirm the 34465A and the 3325A HP seems much closer to CAL!

If the HP 3325A IS exactly 50.0 Ohm Zo, then a lot is explained. When I check the ACV out with NO term   (at 1 kHz) I got ~  2.000 V.

VOILA! This is remarkable and a lesson learned by this old guy.

I shall measure all the terms and keep the 50.00 EGG in a special standards box!

Using  EXT REF from  Leo Bodnar GPS at 1 MHz 

Next ..to  DIY OPT 001:  A 15V PSU  and  OCXO  like HP10544A.

 Many thanks again to SilverSolder!

Bon weekend,

Jon

[jonpaul]

Rebonjour, the HP 3325A was checked at first with the off cal 50 Ohm terms at the end of long coax.

Removing the term to get 2X V is not accurate due to the lumped cable C and Zo. At 1 kHz it will affect  open ckt readings.

New setup:

replace coax with short test leads, run test at 100 Hz not 1 kHz, check Zo of the 3325A with an R box.

Result: Better CAL check of 3325A well within spec with no CAL required.

Finally the Zo of the 3325A is 50.5 Ohm, a standard 1% R value.

Bon weekend,

Jon

[SilverSolder]

I really enjoy using the 3325A.   To me, the user experience is like opening the garage door on a sunny morning, and backing out a 1960's muscle car and taking it for a drive in the countryside.   Of course, you could take your much smaller and more efficient Prius for a trip instead, and that is also a nice experience, but it is not the same experience! 

That said, I don't think the 3325A is a paragon of amplitude stability - the amplitude can drift up and down slightly over time (fives to tens of minutes).   Most of the time that doesn't matter (i.e. if you are making relative measurements downstream) but it is something to be aware of if you are working at the limits of what it can do...

[mawyatt]

Those were our "goto" function generators before AWGs came along and got all the "kinks' worked out, recall the 1st AWGs had numerous issues which took some time to get corrected.

The method the HP3325 used to create triangle waveforms was absolutely brilliant, tribute to the typical HP equipment of those past days, and now apparantly lost forever

Best,

[trobbins]

I replaced the fan on my unit as it was noisy (ie. old bearings).  I'd caution against dropping the fan cooling rate as certain power supply parts get quite toasty - especially the diodes on the power supply pcb, and the devices heatsunk to the side bar.  Perhaps check that all is ok with the incumbent cooling, such a checking the heatsinking has appropriate thermal paste to minimise thermal resistance.  The temperature rise is nothing those parts can't operate at in that rear power supply region, but for long term use it could be a concern for reliability and e-cap life.