Author Topic: HP8644A Calibrating.... (Read 3347 times)

tkamiya · « **on:** November 09, 2018, 09:26:19 pm »

I have a new-to-me HP signal generator, 8644A. I've addressed obvious problems and it seems to be working fine. That is, except for "calibrating...." after every power cycling. I know this is usually due to NiCad battery on A16 board going bad, so I replaced it.

Then I ran Cal: 171 to let it do self-calibration. It ended with result=0.

It's been 4 hours now but it still does "calibrating...." after every power cycle. Would someone please confirm, there is no step to SAVE this self-calibration or anything else? I've done this before but it was over a year ago and I don't remember.

I thought replacing the battery and power cycling twice was the process.
(first power on will result in "battery content lost" error)

tkamiya · « **Reply #1 on:** November 10, 2018, 03:00:02 am »

Hum..... 6 hours later, still the same thing.

Just to make sure, replaced the battery with another one. STILL calibrates each and every power cycling.

tkamiya · « **Reply #2 on:** November 10, 2018, 04:31:00 am »

Just to record my progress:

It still does calibration every single time the power is turned on.
note: it is NOT every time the unit is plugged in, but every time power switch is pushed

BIOS is Rev 2.18.0
Self diagnostic shows result code = 15,035,113 - this means I/O assembly J1, test 12B
message says: I/O Board failure
then: Calibration error

Huh?

The unit seems to work pretty well despite this. Power supply mode WAS replaced with a donor board due to fan circuit failure. I wonder if this is some kind of compatibility issue? Or did I screw this up?

tkamiya · « **Reply #3 on:** November 10, 2018, 04:55:54 am »

Chart says to check +5v

It reads 5.177v (reading by calibrated 6 1/2)
Standard is 5.0 to 5.3
Looks OK to me.

tkamiya · « **Reply #4 on:** November 10, 2018, 05:33:36 am »

Tested functionality with a calibrated SpecAn. Level and frequency is spot on.... No errors except for "Calibrating...." at beginning.

I am beginning to think there may be an issue in Static Ram backup circuit, and controller board failure is bogus.

Does anyone have CLIP for 8644A??

tkamiya · « **Reply #5 on:** November 10, 2018, 05:45:21 am »

Captains log: too late to be doing this kind of thing....

Found a notation in manual that DCU needs +21V for memory (battery backup) WHAT? (page 2-5)
+21V for a stupid CMOS memory backup???

Does anyone have a CLIP?

tkamiya · « **Reply #6 on:** November 10, 2018, 05:57:51 pm »

Spent few hours looking everywhere. Seems HP8644 literatures are harder to come by than other units. I wonder why??

Also, noticed something interesting.

Last night, I changed the CMOS battery and played around. Set sweep function, then turned it off. This morning, when I turned it back on, it did the usual "Calibrating...." but went right into the exact state I left last night. If battery backup wasn't working, this would not have happened.

I wonder if I have a bug in firmware? 2.18.0? I have another unit, 2.19.0 and it works fine.

Or, is this somehow my stupidity? How do I save self-cal data? I recall doing something last time but I cannot recall.

Can anyone assist me?

tkamiya · « **Reply #7 on:** November 11, 2018, 03:40:10 am »

FIXED!

I removed the interface board which is right above power supply section. On PCB trace on back of the board, there was a big gouge. One of it was clearly broken under microscopic observations. Fixed that and others that were looking suspecious. Now diagnostic reports no errors and it remembers calibration.

Yay!

By the way, there IS a CLIP at Artec manuals and it is filed under 8643. It is either identical or very close.

KE5FX · « **Reply #8 on:** November 11, 2018, 05:38:41 am »

Good work, that's a nice generator.

tkamiya · « **Reply #9 on:** November 11, 2018, 04:37:31 pm »

Thank you. This is a 4th one! (one was sold)

I wish frequency went a bit higher though.

analogRF · « **Reply #10 on:** November 12, 2018, 03:03:37 pm »

was it 8664a or 8644a? the title says 8664 but it seems you were talking about 8644?

tkamiya · « **Reply #11 on:** November 12, 2018, 04:48:08 pm »

I get them mixed up often. It's 8644A

tkamiya · « **Reply #12 on:** November 12, 2018, 05:04:49 pm »

I must be somewhat dyslexic. I've been doing a lot lately. Product name being just a sequence of numbers really doesn't help.

Anyway, it was quite a surprise "Calibrating...." wasn't solved with a new battery. Yet, it remembered the last setting. This particular unit which I bought cheap was the most troublesome one. First, fan didn't work. It took the fan driving circuit with it. Second, battery was dead. Then replacing the battery revealed a second problem.

I love this signal generator. It's my 4th one. Every now and then, one shows up on eBay cheap.

trukresom · « **Reply #13 on:** November 13, 2018, 09:12:03 am »

Hello tkamiya,

May I ask You a question ?
I am owner of a HP8644B. Look at the screenshot what happens at the RF output
when I manually adjust the frequency with the knob. Note that this behavior
is independent of the setting of special function #105 (used to enable/disable
the muting function). It is also independent of the frequency increment.
Do you observe the same effect on your machine ?
None of my other signal generators (HP8662, R&S SMIQ03) shows a similar effect.
This is very annoying when fine adjusting a frequency.

tkamiya · « **Reply #14 on:** November 13, 2018, 03:53:51 pm »

I just tried it with all 3 of mine.

They act exactly the same as yours. My guess is, while PLL is not locked, the output is stopped. I've never noticed this..... That means, of course, I need another set of 3 signal generators without this behavior.... muahahaha.... Kind of makes sense though. When PLL is not locked, output can be anywhere and swinging. I would not want that coming out.

Seriously, all of mine do the same and it has never bothered me.

I wonder if there is a setting in special menu somewhere? "Output while unlock" or something? (I'm guessing here) I'll take a look later.

tkamiya · « **Reply #15 on:** November 13, 2018, 04:34:53 pm »

I tested for the same thing in "SWEEP" mode. Wow.... To my surprise, there are this gap. Oddly though, for a really big span, it does only 8 times or so. So it appears there is some "catching up" time required for sweeping and PLL locking.

I was talking with a technician while back. The difference between HP models sold as sweeper and generator is when PLL locking takes place. Sweeper does it in more controlled manner where as regular generator tries to do it every time. I have not been able to verify this since I don't have a PLL sweeper.

trukresom · « **Reply #16 on:** November 13, 2018, 09:57:04 pm »

Hi tkamyia,

thank you to having given a look on this effect for me.
Somewhere I hoped that my generator was faulty. I this case there had been
eventually a possibility to repair this. But now, beside using this gear as
a fixed frequency low phase noise reference, this machine is almost useless
for me. One thing which is not clear to me is the fact that the special
function 105 shows no difference.

tkamiya · « **Reply #17 on:** November 13, 2018, 09:59:50 pm »

I was a bit surprised to see all of mine do the same. Yet, it didn't surprise me.

Good thing is, these things hold value very well. I'm sure, if you wanted to sell, you'll get your money back. Or, if you want to throw it away, you should do it near Orlando, Florida USA, and let me know where it is.

tkamiya · « **Reply #18 on:** November 13, 2018, 10:07:57 pm »

I wonder, why is it useless for you though? The frequency changes as requested. It's just that while it's going through steps, it blanks out the output to NOT put uncertainty in the signal. Since PLL, by nature, is step-by-step method, if they didn't blank it out, it could be anything out there.

Many years ago, I have designed a primitive (by today's standard) PLL signal generator. Literally, when I change the frequency, the input to charge pump does up, down, up, down,until it settles down. It will go outside of before, and after, frequency. I think some people call it "ringing".

I always thought option 105 is to enable/disable the OUTPUT button. I've never tried though.

trukresom · « **Reply #19 on:** November 14, 2018, 09:47:31 am »

I agree with you that eventually it makes sense to mute the output when doing large frequency hops.
But when I compare the outputs of a 8662 vs the 8644 with an oscilloscope one can adjust the frequency
of the 8662 in increments of 0.1 Hz without amplitude variations and without phase jumps. The phase
remains always coherent while finetuning. There is absolutely no need to switch the output off.

In the fastest mode (mode 1) the output of the 8644 switches off during 10 ms. If this is the time needed
to settle the frequency its not famous. For the 8662 it is specified : 0.4 ms

And yes, I don't like the 8644: as sweeper -> no; internal audio source -> much too complicated;
weight -> horrible ! 35 kg (with options 002/011/010)

But the 8644 is not the only instrument which serves only as a base load on my racks.

trukresom · « **Reply #20 on:** November 14, 2018, 10:50:35 am »

Hi tkamiya,

Good news ! I found the problem with Special Function 105:
I always used the keystroke sequence
[SPECIAL] [1]

[5] [OFF] to try to switch off this damned muting function.

This was wrong ! When I enter:
[SPECIAL] [1]

[5] [ON] [OFF] [FREQ] it works !

This setting gives a smooth behavior when tuning the frequency like on
all other signel generators I have.

This setting also disables the muting function when sweeping. But there I see
sometimes undefined signal transients on some frequencies especially during
the retrace from stop to start frequency.

tkamiya · « **Reply #21 on:** November 14, 2018, 03:43:42 pm »

Good to know! Thanks!


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: HP8644A Calibrating.... (Read 3347 times)

Share me