Re: A dirty way to get more frequency resolution out of a signal generator

#103557

I should point out I have not attempted this as I had no reason to, why I am sharing this is that it appears to be an unusual exploit should you desire a more fine frequency control though not without its obvious draw backs. I am anticipating there might be people in this group who wouldn't recommend this, I haven't tried it and I certainly don't encourage others other than simply being a interesting discussion.

Re: [WANTED] HP 85700A 32kB RAM card

#103556

... Relying on super caps is not 100%.? I have found bad ones.
... It's better to be safe than always assume goodness.

I never assume goodness, hence my statement to always do a backup.

A backup should be done anyway as a matter of good care and feeding, even if not replacing the battery.

As long as I had a backup in hand, I would personally not worry about an external power connection.? It's more poking at the board with a soldering iron, now with wires in the way, and more time and opportunity for something to go wrong.

If the super-cap didn't hold, it should be replaced.? HP designed it onto the board for the sole purpose of easy battery replacement.

-mark

A dirty way to get more frequency resolution out of a signal generator

#103555

To put simply, take a function generator such as a HP 3325B, set the output to 0dBm at 10 Mhz, plug that into the reference input of the signal generator. Then if you want a frequency of 100 Mhz just dial that into the signal generator. If you desire more precision with the frequency than what the signal generator is capable of, just make very slight adjustments to the function frequency on the generator supplying the the reference 10 Mhz.

I would think this could work somewhat however it might not scale linear at the various frequencies set on the signal generator, and the phase noise would suffer. And to determine the actual frequency a frequency counter would be needed.

Re: HP3325 - no output, but sync works

#103554

Hi,

If you do not hear the relays clicking, first thing to check is the
ribbon cable between the attenuator board and the A14 board. They are
notoriously bad, to the level that on one of my 3325A I replaced the
pin headers and the ribbon cables. The newer ones are ok, but the
older version which somewhat looks like a flatflex cable, they are
bad.
If that is not the case, then on some versions  there is a fuse on the
power supply board, which fuses unregulated 15V power to the relays.
it is quite easy to blow, if accidentally shorting the separate cable
that is supplying it.

Otherwise follow the service manual of the troubleshooting and follow
the signal path.

Szabolcs

Randy Newman <randy.n.at.home@...> ezt írta (id?pont: 2020.
márc. 1., V, 2:30):

Show quoted text

New to the group, but have used HP gear for decades. I have a 3325 which appears to have an output attenuators issue. I can see the sync output on my Tek (555) scope, but do not see anything from the signal output connector. I have tried switching amplitudes, but do not hear any relay clicking (I had read some other post about the relays being audible). This unit was purchased years back as surplus, and had a red “not economical to repair” tag on it. I think I paid $25 for it. All the 7-segment displays work, and I can set for various frequencies, and see the sync....just no output,
Any ideas appreciated!

Re: [WANTED] HP 85700A 32kB RAM card

#103553

Relying on super caps is not 100%.? I have found bad ones.? It's better to be safe than always assume goodness.

Greg

Re: Fuses for 5350B input 2 [lower frequency input]

#103552

Don,

If I’m not mistaken, HP part number 2110-0301 crosses to a Littelfuse part number 0275.125 under their Picofuse product brand.? This fuse was obsoleted in 2007 along with many others in this product line.? Some other Littelfuse part numbers that are compatible and still available are:

0251.125 – Picofuse II, 125V, 0.125A, very fast acting

0263.125 - Picofuse II, 250V, 0.125A, very fast acting

?

There are others but the remainder are time-lag or slow-blow types which are not recommended for inputs such as those found on test equipment.? You want rapid overload protection for the circuitry that lies behind the front panel connector.

?

If you Google the above part numbers you will find distributors that handle the fuses for prices much less than what HP/Agilent/Keysight will charge you for the very same device.

?

Greg

Re: HP3336b

#103551

On Sun, Mar 1, 2020 at 12:01 PM, Nigel wrote:

In general terms I could never understand why the American electronics and radio industry, from hobby level all the way through to military, embraced these so universally as RF connectors, although I suspect cost would have been a significant factor.

What type of connector are you referring to? Apart from the fact that "RF phono" seems an oxymoron to me, I am not aware of HP widely using low-cost connectors. Are you talking about RCA/cinch connectors?

Raymond

Re: Upcoming Stuff Day event at Sphere Research

#103550

What does it mean by Keysight qualified part?
Apparently the 34401-49321 part is listed this way and it won't let me purchase it.?
Thanxs?

Re: Upcoming Stuff Day event at Sphere Research

#103549

I used Chrome and it seems to work, now I have to wait for verification?

Re: [WANTED] HP 85700A 32kB RAM card

#103548

The additional effort to attach an external power supply to keep the memory contents is really not needed.

In the referenced Assembly Repair manual there's a section on how to replace the A16A1BT1 (page 337).? The memory board has a super-cap which keeps the memory powered for up to 8 hours while you replace the battery.

The RTC will be reset when you remove the memory board.? Don't panic.? This is because the RTC is on the main board and get its power from the battery on the memory board.

But ALWAYS do a a paper backup of the constants anyway!? User data needs to be backed up via GPIB or an SRAM card.? I've been able to backup DLLs via GPIB, but have not been successful via an SRAM card (at least not for DLLs loaded by the EMC personality card).

If you do lose the memory for whatever reason, after the restoration procedure you can use the undocumented command "SERSET nnnn;" entered via GPIB or EXECUTE TITLE to restore the serial number (if you care about it being set).

You don't need to poke at memory for anything, but since it was mentioned, here are the GPIB commands for the 8590 series:

? MRDB <addr> - returns single byte in decimal
? MRD <addr> - returns single word in decimal
? MBRD <addr>,<count>? - returns a block of memory in binary

<addr> and <count> are in decimal.? The SRAM is 524288 (0x80000) bytes long and starts at 0xf80000.? It's a Motorola processor, making storage big-endian.? I/O is memory mapped, so there's a lot to explore outside the SRAM range.

For write, I found the commands MWRB, MWR, and MBWR.? No doubt these are the analogous operations for writing, but I did not experiment with them.

For example, the model number exists at 0xffbfee, 16760814 decimal, and is stored as decimal:

? $ spq MRD 16760814
? 8595

And after that, the next byte:

? $spq MRDB 16760816
? 69

69 is decimal for "E" --> "8595E".

Have fun!

-mark

WTB signal generator / multiplier above 110 Ghz

#103547

开云体育

High all

looking for a signal generator / multiplier above 110 Ghz? any one have something on the shelf requiring a new home

Regards Paul

Re: [WANTED] HP 85700A 32kB RAM card

#103546

I just tested the ZSETADDR and ZRDWR? commands given there, on my HP8595E. They seem to do something, but not what I'd have expected. Subsidiary problem: My 8595E display centering is a little off, such that the top line (where any error messages appear) is mostly cut off, making it very hard to read the messages. I suppose I can do the HPGL screen dump to my KISS-488, but that rather slows down the process. Is there a straightforward way to adjust the centering? I recall many years ago watching my dad adjust centering on a TV by moving the magnets around the tube yoke on a TV - hoping that's not the method here!

Steve Hendrix

At 2020-03-01 05:19 AM, Askild wrote:

Show quoted text

In this thread on eevblog forum:
It is listed undocumented gpib commands for the 856xE series, one can be used to read any memory location.
I have no idea if this works with the 859xE series. These commands is not accepted by my 8562A.

Re: HP3336b

#103545

That does sound like a drifted master oscillator, or perhaps further down the chain, but so far I can't find a table of error messages for the 3336.

However, on a more general note when working on these......

Much as I love the 3336 series, for their versatility and especially as the A and B versions are so easy to modify to C, and despite the often unreliable attenuators, I am definitely not in love with HP's use of phono connectors for the many internal coax connections.

In general terms I could never understand why the American electronics and radio industry, from hobby level all the way through to military, embraced these so universally as RF connectors, although I suspect cost would have been a significant factor.

Whilst there are so called "professional" phono RF connectors, and I have used these, most that I've come across fitted to equipment seem no different to the basic audio version.

When new they generally do seem to be ok but intermittent contact on older kit has always been a problem and nowhere more so, in my experience anyway, than with the 3336 internal interconnects.

Sometimes a bit of work on the socket can fix it but I seem to remember the socket moulding on the HP PCB versions made this quite awkward at times and I've had to replace the leads using plugs with a slightly fatter inner in order to maintain reliable contact.

Probably not the issue in this instance but certainly something to be aware of!

Still a great bit of kit though:-)

Nigel GM8PZR

?

From: Tim
Date: Sat, 29 Feb 2020 05:06:33 PST

Hello Everyone. Been a lurker here for awhile. Great stuff here. I find myself reading about equipment I don't even own. I need a little direction. I have s HP3336b that when I turn it on will show a Fail Oscil for about 5 seconds and then go away. It seems to work find after that. Every once in away I will have to turn it off to get rid of the error. Can anyone point me the direction to start? Thanks

Re: Fuses for 5350B input 2 [lower frequency input]

#103544

Hi Don,

In my SM J2F1 is listed as 0.125A 125V .281X.093 (hp number 2110-0301)

Br,

Askild

Show quoted text

On Sun, Mar 1, 2020 at 6:47 AM donald collie <donaldbcollie@...> wrote:

Hello group. Can anyone advise me the rating, and where to get, replacement fuses for the 525MHz channel input protection on a 5350B? The manual that I have has this information missing.
Thanking you for your consideration!.............................................................................................................................................................................................................Don C.

Re: [WANTED] HP 85700A 32kB RAM card

#103543

Hi,

In this thread on eevblog forum:

It is listed undocumented gpib commands for the 856xE series, one can be used to read any memory location.

I have no idea if this works with the 859xE series. These commands is not accepted by my 8562A.

But as Greg writes, I would do the pen and pencil backup for the important data first.

Br,

Askild

Show quoted text

On Sun, Mar 1, 2020 at 2:10 AM Greg Muir via Groups.Io <big_sky_explorer=[email protected]> wrote:

Ron,

?

The 85700A isn’t going to help you much if you plan to replace the NVRAM battery B1 on the video/processor board.? The calibration constants have to be manually logged per the instructions in section 3 of the “HP 8590 Series Assembly-Level Repair Service Guide” titled “Backing Up And Reloading Correction Constants.”

()?? The method to retain the correction constants falls back on good ol’ paper and pencil using the forms included in that section.

?

As for battery replacement without losing the constants, I use a procedure utilizing a bench supply.? I first remove the memory subassembly from the analyzer since it contains both the memory and backup battery.? Then I tack a ~1k ohm resistor to the positive side of the battery trace and a wire to the negative side (both away from the solder pads that attach the battery) then connect the resistor and lead to the power supply set to near the battery voltage.? This will allow you to retain the battery voltage while you replace the battery.? Use extreme care so as not to short out the supply lines to the memory lest you lose the contents.

?

It is also a good idea to add a diode in series with the resistor to prevent any backflow of current should you be using a power supply with a capacitive filtered output and forgot to turn the supply on before connecting it to the + and – sides of the memory board.? A discharged capacitor in the power supply could cause a brief drop in the RAM supply line depending upon where you connect your leads to the board.\

?

In all, using care you can easily replace the battery.? I do not recommend trying to replace the battery with the RAM board under power in the spectrum analyzer.? First, it is quite difficult and second, if you use any type of soldering iron grounded to the third wire outlet ground, you can short out the analyzer power supply voltage and not only lose your correction constants but also could do additional damage elsewhere.

?

Greg

Re: 8510C: 8340B vs 83631B

#103542

Hi Vladan,

Not sure it's possible to calibrate the 8360 with the front panel (really don't know...), normally you do this with the software running on an HP98x6 or a R382 computer and HP power meter plus the 8360 connected to the GPIB bus all together.
It takes if I remember a long time and sometime you need to do it twice.

73's
--
Pierre-Fran?ois, F5BQP
JN18AN

Fuses for 5350B input 2 [lower frequency input]

#103541

Hello group. Can anyone advise me the rating, and where to get, replacement fuses for the 525MHz channel input protection on a 5350B? The manual that I have has this information missing.
Thanking you for your consideration!.............................................................................................................................................................................................................Don C.

Re: HP3325 - no output, but sync works

#103540

Hi Randy,

Welcome to the group! I assume that your unit is a 3325A, but both the A and B models are quite similar. Place your scope probe (high impedance) on the A16/A26 AMP OUT pin. If you have signal here, then the issue is in the attenuator board or output cables.

If there is no signal at AMP OUT, then the fuse F3 on the A16/A26 amplifier board could be open. Since you are not hearing clicking of the relays, the problem could be a few other things:

1) Bad or missing ribbon cable between the attenuator board and the A26 amplifier board.
2) Missing relay power
3) Open relay coils (not likely that all coils are open)
3) Faulty logic from the A14 board that enables relay switching.

Hope that helps! Let me know what you find and we can dig a little deeper..

Regards,

Joe

Re: Upcoming Stuff Day event at Sphere Research

#103539

开云体育

Did you try more than one browser?

Show quoted text

On Feb 29, 2020, at 16:15, Dana Smith via Groups.Io <raidermc01@...> wrote:

When I try to submit nothing happens.
I work for a custom magnetics company and tried that as well.
I can't get past the registration page to order.

HP3325 - no output, but sync works

#103538

New to the group, but have used HP gear for decades. I have a 3325 which appears to have an output attenuators issue. I can see the sync output on my Tek (555) scope, but do not see anything from the signal output connector. I have tried switching amplitudes, but do not hear any relay clicking (I had read some other post about the relays being audible). This unit was purchased years back as surplus, and had a red “not economical to repair” tag on it. I think I paid $25 for it. All the 7-segment displays work, and I can set for various frequencies, and see the sync....just no output,
Any ideas appreciated!