HP 8663A question

[Mrt12]

Hi guys,
I recently scored a HP 8663A signal generator, and it basically works fine, but I have two small questions, also asked on the HP/Agilent mailing list, but so far nobody answered. So I also ask here, probably someone has an idea!

a) according to the service manual, the signal generator should display a 2 digit status code in the far right of the frequency display while the STATUS button is pushed. My generator shows something else: it has a 1, followed by a space, and then 00. 00 seems to be fine, but the manual doesn't tell anything about an additional leading status code, so I wonder whether this is actually code 100 (which doesn't exist) or whether it is code 1 AND code 00 (which wouldn't make any sense either). Probably someone can check with his 8663A?

b) I noticed that when I manually change the frequency from < 100MHz to >100MHz, there is a relay chattering inside the generator. No problem with that; output frequency and power seem to be OK. But when I do a sweep in this range, the relay does not chatter. Why? malfunction? later I realised that it does no longer chatter around 100MHz, but between <200MHz and >200MHz. I cannot reproduce exactly under which circumstances the relay becomes active at 100MHz or at 200MHz, but it is kind of strange. However, the output frequency and amplitude are fine. I am just wondering whether this looks like my signal generator is having some problem or whether this is just fine.

[KE5FX]

That's correct, the unit should show "1 00" if all is OK (code 00).  I don't know what the '1' means but it presumably isn't important, given that the manual doesn't talk about it.

It is also normal for the attenuator not to click in sweep mode.  That's done to save wear and tear during long-term continuous sweeping.  The amplitude accuracy specs are derated slightly in sweep mode as a result.  See footnote 6 in the 5953-8376 brochure.

You can enable amplitude correction in sweep mode with special function 87 as described in the ops manual, but according to the manual it will use the AM modulator rather than the mechanical attenuator to compensate for flatness error.  If it's still being actuated as the sweep crosses certain frequency boundaries, I'm not sure what's up.  But it is probably not an issue.

[Mrt12]

Hi John,

many thanks, I was really wondering what "1 00" means, as you said, there is absolutely nothing about this in the manual!

Also thanks for the hint with the amplitude correction. The attenuator I am talking about is not actuated in sweep mode (as it should be, according to you), so this is normal. The only thing I found strange was that the attenuator sometimes gets actuated when switching from 99MHz to 100MHz, and sometimes it gets actuated when switching from 199MHz to 200MHz - this is quite odd, but apparently, it has something to do with the amplitude correction and since the generator is working fine it is probably really not an issue.

Soon I will also pull the EPROMs and make backup copies. I noticed that there is a further EPROM which is soldered in and apparently contains some correction constants for the attenuators - do you know whether it is worh to de-solder this EPROM and make a copy as well, or could the contents be rebuild somehow in case the EPROM will fail some day?

[KE5FX]

Hi John,

many thanks, I was really wondering what "1 00" means, as you said, there is absolutely nothing about this in the manual!

Also thanks for the hint with the amplitude correction. The attenuator I am talking about is not actuated in sweep mode (as it should be, according to you), so this is normal. The only thing I found strange was that the attenuator sometimes gets actuated when switching from 99MHz to 100MHz, and sometimes it gets actuated when switching from 199MHz to 200MHz - this is quite odd, but apparently, it has something to do with the amplitude correction and since the generator is working fine it is probably really not an issue.

Soon I will also pull the EPROMs and make backup copies. I noticed that there is a further EPROM which is soldered in and apparently contains some correction constants for the attenuators - do you know whether it is worh to de-solder this EPROM and make a copy as well, or could the contents be rebuild somehow in case the EPROM will fail some day?

I posted copies of 08663-87007 and 08663-60104 (the attenuator ROM) on ko4bb.com a few years ago, so you may not need to bother unless yours has a different part # suffix.  If you do a hex dump on 08663-60104, it looks like a ramp of relatively small signed 8-bit values from -3 right after the initial header to +7 near offset 0x2D0.  I would guess those are tenths of a dB.

IMO you would never notice the effect of installing the 'wrong' copy of 08663-60104 as a replacement part in a different generator.  You would have to check it with a calibrated power meter while paying careful attention to the power meter's own frequency response.  So if you unsoldered yours and damaged it, it wouldn't be the end of the world.  It would be interesting to compare the binary data.

The most important thing to do is get the old NiCd batteries out of there, if you haven't already.  The battery pack can be replaced by a 3-cell cordless phone NiCd battery of similar physical size.