EEVblog
Вставка
- Опубліковано 15 жов 2024
- Dave continues the investigation into repairing the HP 35670A Dynamic Signal Analyser. This time using the Flir E8 thermal camera to check for any overheating components on the suspect A5 analog board.
Part 1: • EEVblog #536 - HP35670...
Part 2: • EEVblog #538 - HP35670...
Part 3: • EEVblog #540 - HP35670...
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...
It's barely 60 seconds to desolder and socket a DIP8 device, so personally I would pull out at least one of those OP AMPS and test it out of circuit
Hi Dave, read these comments, we are all dying to see you continue this fix! We realize it's a lot of work, but these challenges are the good stuff! Pretty please?
Seems like lots of people, myself included, would like to see you continue this one. You know you owe it to your fans. Bueller... Bueller... (please!).
What you need to check for is if the op-amp outputs are pinned to rail, and mathematically disregard what the inputs are at. I find bad op-amps all the time in audio amplifiers, and they will idle just fine. But they are pinned to rail, or otherwise not following the input voltages.VERY common issue.
I agree with you. I just had 5 2x800W power amps in here with their preamp stages damaged. Some were damaged only on one channel, some were working and than not working and some were just dead. None of them however were shorted or overheating. Even the voltage regulators were fine. When i fixed them i went to see how they were hooked up and discovered that the rack they used had no ground connection and a mixer with PWM powersupply. All the amps had isolation switches and when they were set to "input grounded" there was ~110V differece between the mixer and their input. We grounded the rack and everything returned to normal.
I'd probe the opamps in the input side to see if their output is near 0. If they are toward the supply rails, I'd suspect them.
The thermal info of 100C/w passed by quick. If that was resistance to the die, then you would expect the die to be that hot, not the case. Might be worth replacing one to see if it runs cooler.
I had a TEAC 24-channel mixer the - supply went out in a similar manner. We're talking hundreds of opamps. Luckily only 2-3 went out but they did not get real hot.
I agree, replace one of the hot opamps on the right side just to see what happens.
Exactly, just measure the difference between the opamp input pins!
It is a few years now since I have trouble-shooted anything like this, without a schematic, but I would inject a small (
This is exactly what I advised just before this message.
Nice video. Maybe you could try to desolder the hottest op-amp and put it in a test circuit to verify it is working. If the hottest one is working it is very likely the others are working and you can be quite confident they are not a part of the problem... Realy hope you get this!
Love all the new bids Dave! Glad your putting them out. I have a ravenous appetite for the Eevblog knowledge.
I'm usually not that kind of person that leaves negative commentary, but you have all sort of tools, desolder at least a few opamps and breadboard then doesn't take much time at all.
Keep going Dave, love seeing these types of videos.
Dave why don't you add 1khz singal op-amp inputs and see outputs if they work? I tested like that on one board and found dead op-amps. You have good test points where you can add your own +15 and -15 and then add some 1khz singal to op-amps. For testing the board I wondered why you use extenders as you can use your own DC power sources and add dc to test points and start poke the parts. And because you do not have clock signals you can input then your own signal to op-amp legs and see how they perform.
I want to see this DSA repaired, please go ahead!!
Take a look at that www.ebay.co.uk/itm/HP-Agilent-35670-66505-Analog-PCA-Board-for-use-in-35670A-/321282832323?pt=LH_DefaultDomain_0&hash=item4acdf2f7c3
Craft4Cube At USD$749.00 I don't think it's really an option.
Very sad you never finished this repair. I felt like you were very close.
Is there a video number 5 on this ?
The op amp that was replaced last time looks to be a lot cooler than the others, so perhaps a number of them are fried. I suppose it would be easy enough to take one from the group of nine and test it externally. Replace it with a working one and have a look at the board again under the flir?
Dave, what you think about making a fundamentals friday series on reading and interpreting datasheets? Starting with simple components like diodes, transistors (both BJT and FET's), to more complex ones like OPAmps. I think this would be a real help for beginners. The more detailed the better. Thank you.
I know it's now 2019 and generally I love your content but I really wish you would take these on as restoration projects and see them through to the end.
I know I would have love to see this working, older HP analyzer have great performance overall.
So did this unit get fixed in the end?
Love the videos Dave, keep them coming!
Just get a handful opamps start replacing them. I think you will spend way more time chasing faults, then you would if you were to just replace all opamps on the left side of the borard.(about 10 opamps, should be quick with your vacuum desoldering tool)
Hi Dave, great video mate. I am waiting for my E4 (will be converted to E8 soon!) to arrive in the post soon, these cameras are game changing. I'll be using it to check my current designs and hopefully confirm operating temperature of my PCBs powered up with time, so looking forward to this! Keep it up!
BTW repair videos are awesome, I know it may be frustrating for you but entertaining to see other people fighting a bad component until the end.
Love these videos and learn a ton from them. Keep them up!
cool instructions on how to fix such a complex instrument! love the FLIR!
Did you check supply voltages directly on the op-amps which were NOT warming up, in case of a local supply issue, also maybe it is an op amp which has gone open circuit, not short circuit, so checking the ones which don’t get warm would be an idea.
The saga continues, it will be interesting to see future videos..
Might make sense to replace one or two of the opamps currently running at similar temperatures and see if their temperature then differ from the rest?
I suggest taking a few suspect OP Amps out and testing them separately, put them through their paces, if anything is wrong with them surely you will find out.
Don't give up!
i know this is an old video now but i read you found the schematics, i can't find a part 5 and my ocd is kicking in, did you try to fix it? if you did, did you have any luck?
Probably HP wants you to send it in to them. We had a CRT monitor come in for repair the company wanted us to box it up and send it to them they wouldn't sell us parts or a service manual.
Try replacing the 4053 switch IC's. Quite sensitive devices as I remember from my circuit design days.
I noticed those boards used exactly the same levers to remove them as the hp4261a, which I modelled up and 3D printed new levers for as they were brittle and broke on my unit, you can download the lever (and my other designs) from: www.thingiverse.com/TheDefpom/designs
Do not give up.
That Sony extender card is for a BVW-70 series Betacam SP VTR.
This thing is turning out to be quite the bugger! I still want to see it working, if only for the process.
Will there be a part 5? Would really like to see it repaired ;)
It is highly unlikely that any of the op-amps survived if they are all rated for the usual 18v absolute maximum voltage. It certainly wouldn't be difficult or terribly time consuming, though, to desolder them all and install sockets, they're only 8 pins each for goodness sakes! That really shouldn't take more than an hour or two at most. How can it possibly take more than a couple minutes each to desolder some 8 pin chips?! :)
Obviously, the problem would be if there are other things on there on that rail that went to almost 20 volts that can't take 20v but I'd expect it to be mostly just the op amps that are on that -15v rail. You could put a meter in continuity mode and just run down all the pins of every chip to see what is connected to the -15 rail other than the op amps.
I'd say it is certainly still at least worth the attempt on a $20k piece of gear even if you just tried installing compatible-type but cheaper op-amps at first for testing.
Great video Dave!
There is one sure thing, this HP35670A unit when working would be worth much more if you went ahead and changed ALL the opamps, and Digital ICs, sometimes its worth getting all the opamps which are really cheap and digital circuits and just change the lot at the same time !, a matter of interest do you have any DC offset on the input BNC's? i am thinking of protection circuit?...
I really hoppe you can get this unit working, maybe we might see a Part 5?
Agree with others, Op-amps most likely gone puff. And that regulator looked very toastie.....??
Note to self: always put Crowbar Protection and Polyfuse on expensive designs.
keep up the good work Dave :)
Hmm, wouldn't the current stop at the bust packages that are sorted?? Seems like almost every device on that board is hot. I'm not really knowledgable in EE though.
It may be a dumb idea since there is a lot of thermal energy dissipation, but maybe trying to freeze the components and seeing which one "defreezes" faster and testing them one by one, checking for short or open components and resistance in them to see that they're not pinned to ground or something.
What about checking the voltage across the input of each opamp? If the opamps work the voltage is close to 0 volts, right?
Should find faulty ones with ease.
I love your videos, good fun but with your unique voice I keep picturing Steve Irwin and I'm waiting for you to say "Crikey I stuck me thumb up the CRO's bum and you should have seen it go" lol. Keep it up!
The Sony extender card is for certain Sony standard definition broadcast equipment, like BetaCamSP decks and linear editing equipment / deck controllers.
Do you think the E60 is better than the E8
great video, thanks Dave
Forgive my ignorance but why don't you simply remove the over heating voltage regulator and replace everything that does not heat up??
Because all of the devices were positve and negative voltage powered. If you remoove the positive or the negative rail the opamps will not be able to compensate for the absence of one rail and shit may happen on their outputs. You might get your DACs burned out if they were fed by a positive feedback opamp.
look to see which one warms up first,dont forget resistance checks
Which voltage regulator on there is getting the hottest? Do you have a current probe to see which chip is pulling the most juice.
Just watched the 2nd repair vid and that hottest voltage reg was the -15v so defiantly looks like more buggered op-amps. Shouldn't be getting that hot at idle, Even when out of the main unit.
Any progress on this?
No part 5 yet. Seems like Dave came a gutser.
Where is part 5?
He gave up
I should first watch the other 3 videos, because it's so long ago that I forgot the most!
Basically, he didn't manage to fix it in those either :-D
did you fix it ?
I have one of these. Great video. Thumbs up.
I just had a quick look at the Block diagram,
A10 Rear panel PCB J202 has a Source output,
I noticed the Mike connector power supplies, +8V on Pin 2 ?
There is a REF power supply on A5 PCB + 6.2V TP403 & - 6.2V TP413 looks near to a trim pot on page 8-19 of the Service guide PDF link I sent you before.
Perhaps protection diodes ? perhaps some good angel from Agilent might send you a full manual, hope so, it's not easy working without a diagram for such a instrument like this one.
In the old day's HP used to have everything in the service manual diagrams, full descriptions of the circuit, even the manufacture of the part & cage number used in the instrument, BUT these day's everything has changed, block diagrams, & a lot of headaches trying to repair TE as good as this one you have,
I admit I still really like the "Old HP equipment", but since HP want to Agilent & most of the HP / Agilent instruments are now made in Malaysia everything changed sadly :(
I think people should be able to sign a none disclosure agreement with Agilent - HP when they want to repair "their own obsolete TE" to obtain a full service manual with schematics & parts list, I do think that's fair, I am sure most people would say "YES" as Agilent - HP would not touch really obsolete TE. Agilent still have placed the old HP service manuals on the Agilent site including diagrams, I could argue the point if there able to do that, then why not supply the full service manuals for the Agilent new kit after say a number of years after the last one sold to be fair, think GREEN Agilent !, there is too much electronics scrap out there in the world, I love old HP kit :)
Keep up the good work, I really enjoy a heads into your videos as a engineer & hobbyist myself :)
I hope your able to fix this one...I keep on looking :)
Agilent have sadly said no way on the schematic.
+EEVblog Where is part 5?
in the unknown dimension....
Thanks Dave!
Whats the cost of replacing all these op amps ? I suspect less then junking the unit and that would eliminate them as an issue. Assumption they are common and cheap if not I will go off and hide in the corner and hope people forget i suggested that.
The cost isn't the issue; desoldering each DIP-8 package is.
DPortain
Dave seems to have some very nice toys and I think a rework unit that might make that not so hard. Poor people like me would cut the legs off and remove the package than heat and remove the legs. Solder wick or compressed air even a drinking straw would clear out the remain solder in the pin hole.
Just my 2 cents.
"stands out like a dogs hind leg" never heard that one before lol
Like to know what the meaty diodes do near the Source Output connector on A5, personally I would do a diode check with a Fluke DMM with power off to see if there are any obvious faults "shorts", should not take too long, after that power up and do a voltage check on the inputs and outputs to the OPAMPS, to see if any of them hit the rail voltage,
I see a Red COTO relay in the unit, does that switch from rear to front source by any chance?
Have you looked at LinkedIn for service chaps who work at Agilent - HP, maybe this is a common fault, perhaps there are protection diodes on the Source Output as your getting some low signal there.
I noted something is causing the unit to indicate OL on the front panel, sounds like the input to the main AD is pulled up to rail,? or maybe due to the gain being set high as source is very low this might cause a failure your getting, I see lots of CMOS switching,
I check my end to see if my contacts here can help.
Good luck to you, I know on the used market place a working unit is rather expensive !.
Best regards
Philip from UK.
could it be the expansion slot that it is on so if it is its on the main board (sorry about my bad english)
this video made me subscribe nice work keep it up always
Hey.
I am curious about the IDC ribbon cable extension.
Dave shows it at 4:14 mins.
Does anyone where i can buy an already assembled DIN 41612 extender cable?
Thx for help!
If you get the circuit diagram, then you can just follow the input until you come across the problematic IC or whatever it is.
What about the tantalum caps ??
There is a part nr 5 to hp35670 ?
I would build a probe from an 8-pin dil socket to easy connect and measure all the opamps in circuit. One after another with at least three channels on the oscilloscope.
I'm silly. I would have pulled all the op-amps out and built an external test circuit and tested each one on a breadboard. Set them up as a simple inverter with 2x gain and 'scope them out.
why bother? just put in new ones, they cost 10c each...
Eviltech
Not these op-amps. The op-amps used are probably $15.00 each, although it probably wouldn't hurt to just test it with a fistful of low end crap op-amps just to test.
Hey Dave you shogun make a part 5 explaining you gave up or something! Or even maybe call a minion from the audience to help you change the 30 zillion optamos and move forward with the project! Come on! ;-) and thanks for the content!
Hey dave is not worth it to try any more.For my experience when i had short V.R
i had also a short IC's connected to that power rail supplied from this V.R.
Is it just me or is the extender card on an angle?
do a tear down of that cheap camera you got in your mail bag please i know lots of people who would love to watch it
Awesome stuff :)
it was solved???
Welp. come a total gutser today Dave, no luck on any video i watched. U/L mate
can you update the play list
HP Agilent or whatever please send Dave a schematic so he can fix it.
Don't you mean "Keysight"? *eyeroll*
The FLIR was booted the whole time, as the OSD was visible, just had to remove the shutter. Oooops :D
Absence of knowledge manifests itself as ignorance.
Doom2pro Actually, the definition of ignorance is the absence of knowledge.
Thomas Fowler
It works either way, multidirectional :)
Do you plain to do another video
Probably most of the op amp can be get via free sample.
Keep on doing vides please!
Looks like Murphy got you with this one also? lol
After four attempts I think it's probably at the level of DMR = Doesn't Merit Repair
you're not a "quida" Dave!
And here we can see a wave of pointless comments every new video gets. It's got only 172 views for now. It actually gets 21 minutes after the video was uploaded.
Yeah fix that sucker
greate vidio +100500
Euhm Dave: V time I???? Its U times I ;)
Sigh, I wish I could drop 6k on a thermal camera like that (or have FLIR send me one free, guess that's the perks of being a famous youtuber :/)
666 video lol
i had an op-amp fail on my apple iigs (in the audio output) and it turned into a weak microphone. tapping on it or speaking near it was picked up in the audio output. but never ran hot. LM318 or LM308 DIP. weirdest thing.