EEVblog
Вставка
- Опубліковано 15 жов 2024
- PART 2: • EEVblog #538 - HP35670...
Dave repairs a HP 35670A Dynamic Signal Analyser. Well, kinda...
Forum: www.eevblog.com...
EEVblog Main Web Site:
www.eevblog.com
EEVblog Amazon Store:
astore.amazon.c...
Donations:
www.eevblog.com...
Projects:
www.eevblog.com...
Electronics Info Wiki:
www.eevblog.com...
Hi EEVBlog, I was the one , person in charge on the DSA PCB repairing and calibration.You can ask me
on the hardware thing.
I love these repair videos, it's really encouraging to see you analyze and troubleshoot electronics that 99.99% of people would just pitch. If your electronics don't work, it's not like 45 parts went out in it, it's probably 1 or 2 that failed. If you can hunt them down, you have a good chance of repairing it.
it is a real bummer that most people don't have the knowledge to repair electrical equipment and just decides to toss it in the dumpster
I own one of those - fully loaded with options 4Ch - sweep is great for diagnostic. Basic is more like HP version of Pascal. East to repair, I am getting an external microphone module. Those units fully optioned are like swiss pocket knife - your unit is more like a cheap version of an oscilloscope. To option it up - some is software but you will need a chip burner. Loads of statistical and system goodies but only if it is optioned up. Keep up the good work
Yes, it's an inductor, but is in parallel with the xtal only.
I've seen those grey PVC feet go melty - it isn't due to heat - it's a chemical thing, takes many years & makes a really sticky mess - probably something to do with plasticisers.
Dave-- Should see what the same -18V rail connects to physically, because that likely went out of range high and damaged the -12V regulator. Maybe it took out some dual-supply opamps, as your problem seems now analog in nature. Love these repair videos btw :)
Before you cleaned it up I thought that the goo might have capacitively loaded down the oscillator circuit...
These repairs that end "not fully functional" are great! This is how many of us end up! And its great to see your process and choices you make.
The end result is really secondary...
Yep, it's making up for stray capacitance in parallel with the xtal and it also affects the pullability of the xtal through the trim cap.
I love how you include debugging steps even if they didn't work.. super helpful.
The inductor in parallel probably makes up for unwanted parasitic capacitance in parallel with the xtal. Without it, your oscillator would probably oscillate slightly above the xtal's frequency. Short it out and measure the frequency. A parallel inductance can compensate for parallel stray capacitance. Pretty standard circuit actually...
Well that Goop use to be a Gray Rubber Foot that Melted..
Hi Dave I got one of those units fully optioned in 2018.
Channel 1 is source but it needs to be told what mode its running.
I repaired my unit added few goodies such as a link to PC.
Keyboard external had few weirdness but we fixed that.
Great unit running 2 x 68xxxx processors.
its a winner with chicken dinner! //Nickj
Is there any chance the inductor is actually electrically in series with the tuning capacitor? That would make more sense. Just with the tuning cap, the frequency of the oscillator would ALWAYS be above the xtal's label frequency. The inductor pulls it back down to the wanted frequency.
That goo on the top board looks to me like it's a melted bumper, like the other two nearby. Similar color and placement.
It's an inductor. So some sort of tank arrangement, maybe as a Butler oscillator on second thought.
Definitely a possibility. I hope it's that easy.
That Greg too is what's left of that foot after something like WD40 or some other solvent attacked it! Not heat! I have seen it happen before.
Yes, didn't think of that! How on earth would it get so hot as to melt like that?
I've seen some that seem to have melted for no apparent reason. Almost like the material just starts breaking down. Leaves a nasty, sticky mess.
Great video. I hope you take the time to continue repairing it.
pvc,rubber and things like that can do "funny" things over time, while repairing old reel2reel machines some of its rubber belts been melting away, some becomes sticky glue and others turn into hard plastic. it all depends in the chemical looks of the used material.
I really liked this video. If you do more troubleshooting on the unit, I hope you make more videos. It's super interesting to me tracking down faults and problems like this. Keep up the good work Dave!
Nothing changed. Because this thing is so big I had to troubleshoot on my teardown bench instead of my main bench, hence the different camera angle.
When I was in the Navy I got my hands on a 35Ghz Spectrum Analyzer. I had no idea what the test equipment cost at the time. But I kicked the shit out of it dragging it all over the ship.
Maybe, but how would it know unless there is another absolute reference in there to compare with?
I do hope it's something simple like that though.
Most of the time they're not chips but resistor arrays (like 16 resistors to supply) these are used to load the open collector outputs of the other ICs.
to use the mainboard with two bus to bus connectors is to use a ribbon cable connection to and from the same connections with all pins connected to the same pins on each connector. it allows for operation of the board sepret from the direct bus to bus connection.
read the schematics, I understand that on each board there is voltage regulator to change the +/-18 to other voltages, so if the psu fail and give to much voltage that regulator will die but the baord work.
Not necessarily due to heat. Some rubber just deteriorates over time. I have seen VCR pinch rollers which became as sticky as glue traps.
I'd bet that the reason the ADC gate array is failing is that the i2c bus isn't coming up. this often happens if an RTC clock's battery has died- i2c is open collector signalled, so a chip that expects to have a battery on it will tend to load the lines through its input protection diodes, crowbarring the bus.
Power supply: I don't want to power up this system anymore. Its faulty. I'm going to disable myself.
Dave: How bout I do anyway
Wow! A Trident Video Card! I haven't seen one of those since I had a 486.
I think it might be failing self-tests because it's detecting interference? Have you tried with the EMI shielding covers on?
It's also heat. Though you're probably right, in this case it was a chemical thing because the ones i've seen cooked were a different color. This one still had that grey.
Yes, must be. It would be a tank resonator in that case.
You referred to the cylindrical object near the leads of the crystal as a 1.5 ohm resistor. I think you will find that it is an inductor and most likely in parallel with the variable capacitor forming a tank circuit in the oscillator.
I have a HP 8562A SA with almost the same enclosure and the front cover also has a weird smell! I kind of think it's from the rubber gasket?
I haven't a clue even what this device does but a great video nonetheless :)
Jeri from technical illusions (CAST AR) kickstarter just passed funding, and she is looking to put together a design for manufacturing team! I know you do youtube full time which is why you would be perfect to help work on the project. Even though Jeri is in the USA with the internet today I'm sure you could work together. You could make videos about the process too and would be contributing what is sure to be a huge product!
Unbelievable how you narrowed down the fault to such a component!
Hey Dave,
Can't wait for part 2! Any chance you might still need to tweak that variable cap, by the Vreg you fixed, to get the freq right?
-Brooke
Looks like the LOCA adhesive has seen better days, you can repair this if you buy the correct Loca UV Screen bonding :)
you have to cure it with a UV lamp :)
Probably the power supply cooked it. I've seen something similar in a Panasonic tv.
Here's what i think happened. The foot unstuck itself, fell on the bottom where it got cooked (for whatever reason) then it dripped on the board. Also, that corrosion isn't rust, it's alu on metal corrosion via electricity. So a lot happened to this before that power supply went for a walk.
I must commend you on your patience, it's true the boon is big if it's fixed but most people would throw it out.
this may be ANOTHER power supply failure. My spectrum analyser didnt worked many of the options because a tantalum shorted on the +40vdc line
If the old PSU took out that 12v negative reg then could have done a whole lot more damage. There isn't any more of those regs elsewhere inside by chance?
Ohhhhh, I was to buy the exact one and I told him too that he got ripped off!!! So you were the only bidder for $500! Nice one. He said the PSU was disconnected since! Nice one Dave! At least its in good hands...
if it smells like WD40 that might be your problem, WD40 says cleans electrical contacts on it, but when you spray computer parts with it, or contacts, its slightly conductive, and screws up whatever you spray with it(I know this from experience on an old 660gt gfx card I ruined with it)
Perhaps some of the failures are caused by the clock being out of adjustment.
Pretty much if it's the custom chips.
wow, what an accomplishment. you truly are an inspiration to us all. why be first when you can be second?
I suspect the power supply failure knocked out the supplies for those parts that aren't working. If some of those custom chips are dead it's basically beyond repair right?
Dave Nice vid. I think someone tried to clean the screen and it got into the back of it!
Nicely done! Your troubleshooting skills are excellent!. Although this is unrelated do you know or anyone viewing this know the new cost of a Tektronix 2712 spectrum analyzer and a 2707 tracking generator was? Greeting from Apache Junction, Arizona.
Hopefully one of of those PLCC sockets I mentioned!
I really look forward to the next part of this... Excellent!
Yes, but it's not like I have those lying around!
Hiiiiiiiiiiiiii Dave, wow I faced the same problem with a CC Debugger. it was almost failing to recognised the chips and rarely showing the Green light.
so i tested a very small Voltage regulator that regulates 5.1V from its USB interface into 3.3 V, and couldn't find any similar regulators so i just did a hand made Zener with a resistor, it regulated only till 2.9 V and it worked.
it was able to erase and program chips. but it needed exactly 3.3 V to debug the connected chip.
Regards. & Thx
A melted rubber standoff is usually an indication of a massive failure of some part overheating?
I love it when dave got it working! I couldn't not understand a word out of his mouth haha
@ 25:00 Could you just bodge/hard wire a laptop power supply to the DC input the voltage range and amps draw makes me think it's feasable and cheap. How did the PSU end up costing $1600, was it built from scratch or a HP stock part, maybe they where dodgy supplies to begin with. Anyhows great upload and hows the Delorean clock ;D.
Yes, it's a gas plasma screen. I think I said LCD out of habit.
The last shot (other then it being out of focus) was really well framed - you should use it more.
ADC array sounds like they're using an interleaved digitizer.
Check the power supply to the ADC chips that are clustered together in a row or column.
Oh wow, something like that -- not a full computer -- with a disk drive on it? Weird! I have a MIDI keyboard with a floppy disk drive on it, and even that's weird! When video recorders and media players with disk drives came out, I thought they were weird too. But I'm used to DVRs now, so not as big a deal. Still weirder than just computers with them, though. What do you do with the little drive on this one?
+MaxxFordham I'm guessing it could be used to upload firmware over to it, in case of a software revision? There's a serial port on the back that could be used for the same purpose though, so maybe it can print a signal to a disk (i.e printscreen)?
Maybe, TripleDz.
I love these videos. They are supreme and informing.
The amount of failures makes me suspect that a voltage rail is kaput. This suspicion is not made smaller by the fact that the power supply is dodgy. Check all voltages, supply wise and local... :)
If we are luck just all the regulator on the +-18v are died and everythink is well.
Have you tried to put a signal in?
The ADC is strange that run at 18v, but the miexer and rf amplifier could be run at that voltage
By the way finnaly a video where you do some trouble shooting. Last time a displòay error solved imediantly. On the LeCroy hybrid ic died (I'm still thinking that the short could be under the shield cover of the time base, that you've never checked).
The DC power connector is a 2 pin LEMO by the way
Definitely continue this please! And, as always, be careful what you ask for, got a real good repair case now, don't ya!
Could at least some of the failures be caused by the oscillator being off (trimmer misadjusted)? Possible with tight tolerances for frequency such as with ADC gate arrays and such, I'm not sure.
could the test failures be related to that trimmer cap you played with?
Hello I wonder if you can help me. I am after a video magnifier and not no what will work for me. Some time ago I see you using one. Have you done a revue on this or could you tell me the one you use please.
All the best Richard
I like how all their older gear has that pouch on top
That gray mess, is that a molten foot, as there are already two are stuck on the printboard?
that melted rubber foot looks like it took a ton of heat, is there anything under it ? i couldnt tell on the video.
The problem with some of these (relatively) cheap power supplies is the lack of over voltage protection on ALL outputs.
unless you can find another model thats broken down for some other reason for cheap.
Can look for other voltage regs on other boards.
Start of an interesting series.. how often are you going to release new episodes? - cannot wait to see it fully working!
He was screw over on the power supply repair, they didn't even attempt to clean up the mess.
The color bars are confusing, I cant read the value of the gold bar in the middle.
Might be a capacitor as it is it is in parallel with a crystal.
things that old always have bad capacitors check for them with an ESR meter, many times it is the bad capacitors killing the supply. dont toss it.
Good luck with this one, I don't think PSU could kill so much, most of the time this PSUs are quite overprotected on the outputs so to get anything higher than rated voltage is quite hard on them even if a failure takes place. And also wow, they took so much money for fixing the power supply and it just died while on your bench :P
I watch a lot of your video's and wondering where you download your manuals from ? I have the hardest time finding good service manuals.
so the power got busted. Like the ac part. This could kill the regulator on the clock modul. So i think it killed other regulators too. The fails seemd to be all differnt , parts.
So check first all the other volatage regulators befor starting to look on other parts!
That resistor is a inductor, some funny markings on components ive seen on HP equipment,
Than the question would be what cause the psu to fail? Some short on the board pheraps.
He should have tried an external monitor first, to verify its not the screen that is faulty
Great video, Dave. Hey, are you affected by those fires down there?
I think if you do get it fixed and eventually decide you don't need it anymore and want to resell it, the guy who spent $1600 on the power supply repair should get first dibbs. ;-)
The issue with the LCD could be as simple as removing old polarizing film and applying new film.
So the power supply fried and took the tiny little chip with it i could have sworn it be a crusty wire or something
+crapper1 Thanks for spoiling a 42 minute video.
+crapper1 Perhaps the most pointless observation ever.
Did you tune the clock? I notice that you frobbed the tuning cap.
yes it is. these feet are notorious for melting over time.
Is that core broken??? @ 24:00
Slow zoom at 15 and a half minutes could use some dramatic music for effect
Well, I didn't want to mention that DFM is not exactly a fun job. Don't want to spoil it for the eager taker :->
I think maybe that component looking like a resistor is a inductor ?
I wonder if your power supply rails are clean or if there is a bunch of garbage and hash riding on them, maybe from another faulty part throwing some noise back into it. I still don't trust that power supply.
I only know a handful of words you say in this video, but it was exciting nonetheless!
FANTASTIC repair vid as ever. These are my favourites!!!
Look forward to more work on this blighter if you get round to it at some point.