HP3577A 5Hz-200Mhz VNA Teardown

[Cubdriver]

The key caps should just pull straight off the switches.

-Pat

[rx8pilot]

For power supply/PDN stuff you probably won't be working above 10MHz, anyway. When you do, come back and tell us about it! 

Exactly - 10Mhz will be a typical cutoff for those efforts.

I have some Samtec SMA's that I can use to solder a splitter together rather quickly. I feel like this learning curve will visually reveal characteristics of components and circuits that are otherwise rather mysterious and hard to understand. Somehow.....this is fun even though it is technically for work. As it is said.....'love what you do and never work another day in your life'

[precaud]

Exactly - 10Mhz will be a typical cutoff for those efforts.

I have some Samtec SMA's that I can use to solder a splitter together rather quickly. I feel like this learning curve will visually reveal characteristics of components and circuits that are otherwise rather mysterious and hard to understand. Somehow.....this is fun even though it is technically for work. As it is said.....'love what you do and never work another day in your life'

Totally agree on all points!   

[bson]

May I ask which active probes you use exactly?

I have a ZS1000, AP020, a differential AP034, and an AP022.  They see use in roughly that order...  With the powered BNC adapter, mainly the AP020 or AP022.  (The AP022 isn't recommended for use with the ADPPS, due to lack of offset zeroing, but it seems to zero just fine for me... But I wouldn't get it for that purpose.)

[sixtimesseven]

Does anyone know which style of panel N-Connector HP used for the 35677-63504 / 35677-63502 "Directional Bridge Assemblies" (75/50Ohm) and if somthing similar is still available?

[sixtimesseven]

Today my replacement fan arrived. I could not get the Delta AFB1224HE anywhere so I went for the Sanyo 9WP1224H1011 which was in stock at farnell uk, specs:

99CFM / 0.283in
2600RPM
39dB

datasheet: https://www.mouser.ch/datasheet/2/471/San_Ace_120WP38_E-1286779.pdf

The fan is quiete. Hard to say if it is 10x quieter but it is an improvment. At the 27.5V from the HP3577 the temperatures from the termal camera seem to be about identical to the Muffin fan, at least if I measure at the heatsink of the power supply transistor which is just visible trough the left perforated side (about 51degC).

I do some further testing trying to reduce the fan voltage & noise. Also I will install some vibration dampeners for the fan since I hear some vibrations at the back panel.

[bson]

I'll experiment as well.  Ordered a relatively inexpensive fan and dampeners:

https://www.digikey.com/product-detail/en/G1238L24B1-FSR-EM/1570-1785-ND/8120198/?itemSeq=268069836
https://www.digikey.com/product-detail/en/LZ550/381-2709-ND/5043883/?itemSeq=268069888

[sixtimesseven]

I'll experiment as well.  Ordered a relatively inexpensive fan and dampeners:

https://www.digikey.com/product-detail/en/G1238L24B1-FSR-EM/1570-1785-ND/8120198/?itemSeq=268069836
https://www.digikey.com/product-detail/en/LZ550/381-2709-ND/5043883/?itemSeq=268069888

Good choice! I looked at that fan as well. However, my Digik. list was almost empty and tax&handling is very expensive for me so the Sanyo from farnell was about as expensive all included 
I look forward to your findings!

Also, how will you attach the filter at the back with those rubber attachments you selected? Do you just pull them trough the hole stack? Please report back if they are strong enough to hold everyting together.

[bson]

No idea how things will go together.  I'll see when it comes here, but I have an assortment of long M4 screws, nuts, spacers, O-rings, etc, so even if the fasteners don't work I'll see what I can figure out.   I can also mill a bracket from acrylic if needed.

[sixtimesseven]

No idea how things will go together.  I'll see when it comes here, but I have an assortment of long M4 screws, nuts, spacers, O-rings, etc, so even if the fasteners don't work I'll see what I can figure out.   I can also mill a bracket from acrylic if needed.

Nice!

The hottest part I could find so far was R103 on the power board. At 28 ambient I get about 110-115degC with the original fan @28degC ambient and about the same for the Sanyo with 28V.

[precaud]

I think it would be better to measure the temperature on one of the pcb's in the card nest that runs warm. That's where the effects of fan pressure changes would be greatest, and most consequential. Exposed power components near the fan aren't going to be affected much by it.

[sixtimesseven]

50dBa ! That's one noisey fan! The fan noise and crt are what keeps the 3577A (and 3562A) off my bench. Lovely instruments crippled with noisey fans and hard-to-read crts.

I probably just install a quiet 12V PC fan with a 24V converter and test with the thermal cam if the 80 CFM are enough without the CRT heating the instrument.

The Achilles heel of the 3577A appears to center around the input relays. That was the major design change between A and B versions, a redesigned front end using different relays.

Do you already have some with that issue? I'm still waiting on Artek to get me the service manual so I have no idea yet what these relays are, but are they replacable at all?

Teledyne 412 series apear to be still available at mouser: https://www.mouser.ch/Teledyne-Relays/Electromechanical/Relays/High-Frequency-RF-Relays/412-Series/_/N-5g33?P=1yzsh13Z1z0zl30

Rather expensive if you have to replace all of them! 

IDK, but maybe you can find the HP part number for the relay 0490-1404 in the service manuals of other instruments?

[precaud]

IDK, but maybe you can find the HP part number for the relay 0490-1404 in the service manuals of other instruments?

Yeah, these things are pricey, that alone would have been good reason for HP to move away from them.

I've found a couple other HP instruments from that era that use the same type relay, but not the exact same model. I found a couple 432-series in a DSO (forgot to write down the model number). And the 16531A digitizer plugin for 16500A logic analyzers has a couple of 'em, but I haven't pulled them yet to see their model number. Should be able to pick up those boards for cheap, noone is using them anymore.

EDIT : The 432-series relays came from an HP 54200A, a 50MHz, 200Msa/S 6-bit DSO. The main board has three of 'em.

[sixtimesseven]

IDK, but maybe you can find the HP part number for the relay 0490-1404 in the service manuals of other instruments?

Yeah, these things are pricey, that alone would have been good reason for HP to move away from them.

I've found a couple other HP instruments from that era that use the same type relay, but not the exact same model. I found a couple 432-series in a DSO (forgot to write down the model number). And the 16531A digitizer plugin for 16500A logic analyzers has a couple of 'em, but I haven't pulled them yet to see their model number. Should be able to pick up those boards for cheap, noone is using them anymore.

EDIT : The 432-series relays came from an HP 54200A, a 50MHz, 200Msa/S 6-bit DSO. The main board has three of 'em.

Does using "old" relays make much sense though? I mean they could be at or near their end of life and it would be a shame to go trough the hole adjustment routine just to switch out relays again...

[precaud]

Does using "old" relays make much sense though?

I don't have a problem with it.

I mean they could be at or near their end of life and it would be a shame to go trough the hole adjustment routine just to switch out relays again...

They could be, but probably are not (near their end of life). They've had a very easy life - the 54200A (and 54201x) 's I have here have been sitting around doin' nothin' for more than 15 years. Ditto for the digitizer plugins.

Plus, I tested the instruments before parting them out and the relays were fine.

[precaud]

I'll experiment as well.  Ordered a relatively inexpensive fan and dampeners:

https://www.digikey.com/product-detail/en/G1238L24B1-FSR-EM/1570-1785-ND/8120198/?itemSeq=268069836
https://www.digikey.com/product-detail/en/LZ550/381-2709-ND/5043883/?itemSeq=268069888

Any progress on your fan replacement?

[bson]

I'll experiment as well.  Ordered a relatively inexpensive fan and dampeners:

https://www.digikey.com/product-detail/en/G1238L24B1-FSR-EM/1570-1785-ND/8120198/?itemSeq=268069836
https://www.digikey.com/product-detail/en/LZ550/381-2709-ND/5043883/?itemSeq=268069888

Any progress on your fan replacement?

I have the fan but haven't taken the time to remove the instrument from where it sits to open it up yet...  Given its size and weight that's a 2 person job. :/

[Berni]

And i have fixed the clicky snagging switches on mine with a quicker fix:
https://www.eevblog.com/forum/testgear/hp-4145b-semiconductor-analyzer-repair-with-photos-(display-garbage-dead-rom)/msg1743731/#msg1743731

They all work perfectly now

[precaud]

Any update from folks who have replaced the 3577A fan with a quieter one?

[Berni]

These fans often tend to be pretty high performance fans with a ton of airflow. So most equivalent fans that move the same amount of air are not going to be significantly quieter.

But in a lot of cases these fans are overkill in the first place. They are designed to keep the thing cool when its stuffed snugly into a big rack of equipment that is pumping out heat and perhaps the rack placed into a industrial environment as part of a automated test setup. So what i did instead is slow the fans down by inserting a resistor in series with them, a lot of these fans actually get quite quiet when ran a bit slower, but when doing this i always check the temperature of the hottest components to make sure im not causing anything to overheat by slowing it down too much.

[precaud]

Those are good points, Berni.

I've used series resistance to lower the operating voltage and fan speeds before. But with that approach, pressure would scale down proportional to the speed reduction. Not generally a problem if circulating air in a large, open enclosure. But after looking at the various cooling paths in the 3577, I'm thinking that isn't a good thing in this case. The power supply components probably wouldn't be affected by it; that air path is direct and open. But pressure is needed to push air through the small holes in the bottom of the card nest, and out the holes in the top panel of the cards.

That's why the Sanyo fan looks attractive; it generates the same pressure as the stock fan, with only a little less airflow.

Probably the best move to reduce cooling requirements is to swap the crt out for one of those nice LCD displays. Hard for me to justify that cost though, as long as the crt is still working...

[Berni]

Well the static air pressure can also have a big effect on the noise of a fan too. In my homebuilt x86 NAS project i was trying to cool it all by drawing air trough the small case using a 120mm fan, but no matter what fan i used i could never get it to run near silent while also pushing the required amount of air trough it. Eventually i ended up grabbing a dremmel and cutting out the hole pattern where the fan mounts and blows trough so it was now a wide open hole, this finally solved it and got the noise level down to where i wanted it, now the harddrives are louder than the fan when all of them spin up.

So the restrictive air path could also be one of the things that is making the fan loud, or making turbulent noise as air makes its way trough the restrictions.

Oh and looking at the datasheet for the HP 3577A it claims it uses 450VA max, so that's indeed quite a power hungry instrument, ofrcouse the real power use in watts is likley quite a bit lower, but i think its going to be quite difficult to cool without making a lot of noise. The big ass HP 8566B Spectrum analyzer i have has a power consumption of 650 VA, sounds like a hair drier with the powerful fans and yet still gets pretty hot.

[precaud]

Yes, you may be right. I've had similar experience.

That's why I was asking for feedback from someone who has successfully reduced the noise by replacing the fan.

[bson]

I'm trying to install a NewScope-5 in mine, but can't for the life of me get the CRT out.  The instructions say "on the HP3577A there is one additional screw on the bottom right that needs to be removed" but I have no idea where this screw is.  I've removed every screw that I think can hold it, but it's not budging.  There are two screws on the bottom back, one on the right, one on the left that attach to a captive bracket, but that's two screws and the left one is inaccessible without pulling the assembly partway out.  Just removing the right one won't help.  So I'm at a loss what f*cking screw I need to remove...  I don't care to keep the CRT, it's going in the trash anyway, so at some point I'll just start ripping stuff out from the top until I can disassemble enough to pull the tube out from the front.  But I'd rather not make a mess.  Has anyone else done this already and know where this one additional screw is?

Edit: nvm, it's not at the bottom right but at the top right and looks like it's only holding the top PCB onto the cage.  It's not.  There are also two screws under the right side adhesive cover holding the display by the bezel.  After removing those it slides out.

[precaud]

The crt ass'y slides in on a "shelf". Look down at the bottom of the rear of the crt unit, you'll see one large-ish screw that secures it to the shelf. I bet that's the one.

Did you swap out your fan yet?