[time-nuts] strange behavior of 53230A or is the light on, but nobody in?

mike cook michael.cook at sfr.fr
Wed Feb 19 03:33:32 EST 2014


Le 19 févr. 2014 à 01:05, Tom Knox a écrit :

> Thanks Tom and Bob, I have been thinking of contacting Agilent for some time. I think they are a great company with some good products, but there are a few real blind spots in some current products. I also have seen in the past a genuine interest in listening. I would be willing to approach them if I could enlist your help in addressing potential changes to improve the product. 
> Thanks;
> Thomas Knox
> 
   
   If they are steering the VCXXO,OCXO from the Ext. Ref. , then they are in effect calibrating it. Why not remember the applied EFC when they get phase lock?  That can be applied when the internal timebase is selected. 
It couldn't be that they might lose the chance to sell a signal generator ;-), as calibration needs a square wave input, and the Ext. Ref In is ignored.

> 
>> From: lists at rtty.us
>> Date: Tue, 18 Feb 2014 18:00:17 -0500
>> To: time-nuts at febo.com
>> Subject: Re: [time-nuts] strange behavior of 53230A or is the light on,	but nobody in?
>> 
>> Hi
>> 
>> Well at least this got me digging a little. 
>> 
>> If you grab a copy of the 53230A spec sheet and look under the external reference input, it’s pretty well described. It will accept 1, 5,10 MHz as an external reference. It will lock over a 1 ppm range with the XO option and 0.1 ppm with the OCXO option. Based on that I’d guess they are still using the same basic PLL approach as on the older counters (5335 era). 
>> 
>> The “Microsoft Windows inside” sticker on the back of the counter was a bit of a surprise ….

  No sticker on mine. 

>> 
>> Bob
>> 
>> On Feb 18, 2014, at 11:51 AM, Tom Van Baak (lab) <tvb at leapsecond.com> wrote:
>> 
>>> TomK,
>>> 
>>> If anyone has technical contacts deep within Agilent, let's see if this issue can be resolved. I would have bought a 53230A when it came out a few years ago but my eval units showed this clock noise problem. That plus the poor quality of the ref out made me think the designers were cutting corners, or had little experience in metrology, or maybe they thought this was "ok" for a bench instrument.
>>> 
>>> Otherwise it's a really nice counter; the first one from Agilent than can actually do ADEV properly (since it is a time stamping counter).
>>> 
>>> I should dig out my old data and send it to you. Maybe as group we can help them fix the problem.
>>> 
>>> /tvb
>>> 
>>>> On Feb 18, 2014, at 12:10 AM, Tom Knox <actast at hotmail.com> wrote:
>>>> 
>>>> I have asked Agilent 
>>>> if stock versions of the 53230A and 53132A switched the internal oscillator out of circuit with an Ext Ref signal 
>>>> applied. I thought 
>>>> Agilent's engineer was intentionally vague but said the oscillators were
>>>> indeed switched out of circuit on the counter with Ext Ref signal applied. These questions were related to several 53132A's I have seen configured with a small board back near the Ext Ref input (OPT H01 I think) that appeared to Switch the internal reference out of circuit. Agilent would not share information on the option. My question to Agilent is why sell an option and be unwilling to say what it does or how your stock unit functions?
>>>> Thomas Knox
>>>> 
>>>> 
>>>> 
>>>>> From: tvb at leapsecond.com
>>>>> Date: Mon, 17 Feb 2014 09:38:28 -1000
>>>>> To: time-nuts at febo.com
>>>>> Subject: Re: [time-nuts] strange behavior of 53230A or is the light on,    but nobody in?
>>>>> 
>>>>> Bob,
>>>>> 
>>>>> I'm wondering if you (or any else) has measured the PLL performance of the 53230-series?
>>>>> 
>>>>> I agree it will "clean up the crud" but this assumes the ext ref is dirtier than the internal osc.
>>>>> 
>>>>> What I found instead was that if you use a good external ref the PLL actually makes it worse. This was very disappointing. The XO version of the counter performed worse than the OCXO version even with a maser as the ext reference. Did your reading of the schematic show a way to directly use the ext ref, bypassing the noisy PLL?
>>>>> 
>>>>> The other thing I found was that the ref out signal was a very polluted copy of the ref in.
>>>>> 
>>>>> /tvb (i5s)
>>>>> 
>>>>>> On Feb 17, 2014, at 7:04 AM, Bob Camp <lists at rtty.us> wrote:
>>>>>> 
>>>>>> Hi
>>>>>> 
>>>>>> If you dig into the schematics (when they supplied them … ):
>>>>>> 
>>>>>> The external reference goes into a phase detector. It’s one of those digital ones that can lock up to many inputs. You could feed 3.33333333 MHz in as a standard input as well as 0.5, 1, 2.5, 5, and 10 MHz. The internal oscillator (or an internal oscillator) is phase locked to the external input through a fairly narrow analog loop. The idea is to clean up the crud on the standard line. 
>>>>>> 
>>>>>> With no external reference, the PLL drops out and you go back to what ever the local reference is. 
>>>>>> 
>>>>>> Yes there’s a little more to it than that and no the circuit is not exactly the same on every counter HP ever made. 
>>>>>> 
>>>>>> Bob
>>>>>> 
>>>>>>> On Feb 17, 2014, at 7:55 AM, wb6bnq <wb6bnq at cox.net> wrote:
>>>>>>> 
>>>>>>> Hi Mike,
>>>>>>> 
>>>>>>> The most likely answer is when you select external time base for an input, it disables the connection for the internal oscillator.  The external input signal is probably also routed straight to the reference output jack.
>>>>>>> 
>>>>>>> However, it would be good to read the manual, as they usually cover how those connections work.  Otherwise, perhaps someone that owns one could provide further insight.
>>>>>>> 
>>>>>>> Bill....WB6BNQ
>>>>>>> 
>>>>>>> mike cook wrote:
>>>>>>> 
>>>>>>>> Something that must be simple to explain, but that I can't get my head round.
>>>>>>>> 
>>>>>>>> I got a new 53230A.
>>>>>>>> When first using it, I measured my T-Bolt 10MHz using the internal 10MHz timebase and it came up short of 10MHz, 9.999 998 5xx. I wasn't worried about it as the counter only has a TCXO internal oscillator. So I fired up my PRS10 and after leaving that on for some time, connected it to  Ext Ref. , changed to the ext time base and measured again. This time 10.000.000.00x. Then I switched the two references, measuring the PRS10 against the T-Bolt. Again I got 10MHz down to the 11th digit.
>>>>>>>> All that looked good so I have been using it with either the PRS10 locked to GPS, or the T-Bolt as the external time base.
>>>>>>>> 
>>>>>>>> After leaving it on (but not inactive) for a month, I did an Autocal. No problem.
>>>>>>>> I was wondering if that would have changed the internal time base frequency, but no, using that still gave similar figures to the above.
>>>>>>>> 
>>>>>>>> So at that point I decided to measure the Internal TB against my reference. So I connected the Int. Ref. Out to channel 1, connected my PRS10 ref to Ext. Ref In, selected the EXT time base and found that the count was 10MHz dead on?????  I don't get that at all.
>>>>>>>> 
>>>>>>>> in summary:
>>>>>>>> DUT against internal TB counts < 10MHz.    To me that means that the internal timebase is a bit fast. Is that assumption correct?
>>>>>>>> DUT against Ext.Ref counts 10MHz
>>>>>>>> Internal TB against Ext.Ref counts 10MHz.       If my assumption above is correct, the count should be greater than 10MHz, no?
>>>>>>>> 
>>>>>>>> Can anyone shed any light on that?
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> _______________________________________________
>>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>>>>> and follow the instructions there.
>>>>>>> 
>>>>>>> _______________________________________________
>>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>>>> and follow the instructions there.
>>>>>> 
>>>>>> _______________________________________________
>>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>>> and follow the instructions there.
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at febo.com
>>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>>> and follow the instructions there.
>>>> 
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>>> and follow the instructions there.
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>> 
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> 		 	   		  
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.



More information about the time-nuts mailing list