Unboxing/repair of an HP 3468A DMM and HP 8662A Synthesized Signal Generator
Вставка
- Опубліковано 23 чер 2023
- I picked up some used HP gear for the workbench. One working, one not so much. Can't go to sleep without making an attempt to fix it :)
HP 3478A precision bench DMM
HP 8662A Monster Synthesized Signal Generator (not working)
Many manuals available for free download from Keysight Technology (AKA HP/Agilent)
www.keysight.com/us/en/lib/re... - Наука та технологія
The inside of the signal generator is a work of art, wow.
Back in the late 1990s I worked for about a year at a HP Service center near Schaumburg, IL. Back then each center specialized in certain types of equipment, the one I was at handled signal generators, spectrum analyzers, communication test sets and power supplies. I worked on signal generators, test sets and later power supplies. Many times we would start out connecting the equipment to a HP computer that ran a test routine from a mini-cassette tape! (Old test set-up for old sig gens!) We used rechargeable electric screwdrivers to open all those module covers. Hydraulic scissor jack carts to move the big stuff on and off of the benches. Rarely did we troubleshoot to the component level, just get a new module from stock. The frequency standard input jacks allowed several instruments to work off the same frequency standard in a set up for consistency. We did calibration also. I have a few "souvenirs" from those days, just little parts. HP test equipment did things the best way, not the easy way. I have great respect for it, but it usually is very heavy!
The more you use it the more familiar it will be.Good job.
Nice finds, both really nice instruments. I have no issues with secondhand equipment, especially with high quality instruments. The HP 3478 A and B are excellent bench meters. We have used a 3478B for years at work and the only issue to date we had is a broken handle. It and the reasonable used price of these units inspired me to purchase the 3478A I have in my shop. I have no doubt you will be well pleased with this unit. An FYI, for the interface I believe that a National Instruments GPIB card and IE488 interface cable will work with the HPIB interface since NI Labview has drivers for that instrument (also I believe the manual tells you how to write a BASIC program to access that instrument). But the card likely will cost as much as you paid for the instrument.
The 8662A is definitely a monster but once you resolve that clipping issue it will be a fantastic instrument for doing alignment.
I feel the same way about buying test equipment. I recently bought a HP 8656A on ebay for less than 200. The pictures on ebay showed that the digital function of it still appeared to be alive. They they advertised it as working.
When I received it all the functions of the controls showed they were intact, I ran through them. I found a complete manual for the unit online at no cost. The problem was that the unit produced no output signal.
Within half an hour I located the problem with the help of the schematic. It was a blown output fuse and I had the type in stock. Replacing it took care of the problem. I suspected that a user most likely had shorted the output in the past or connected it to some voltage that they should not have and the fuse merely did it's job.
I too enjoy fixing test equipment and as you say, it's usually a good value. I also pick up fantastic deals on test equipment as well as antique radios which are my thing at the ham swap meets.
We can just as easily apply our skills to repairing test equipment as well as the other things we work on in this hobby. No need to buy new unless there's something that's a fair price and we need that kind of functionality or we just want to treat ourselves to a new toy. I needed a PWM signal generator recently for testing a CNC machine axis motor driver board. I have a number of signal generators but none have the PWM feature. I could have bought one for around 100, but instead breadboarded a circuit to create that functionality using a standard square wave generator. All with parts I have laying around. Sure it took more time but I don't put time limits on my hobbies. It was also more satisfying to build the circuit I needed myself.
Great video. I love HP test equipment from the 70's and 80's. High end, top quality devices. Made with the best quality components from Resistors to transistors. Needless to say that electronic and mechanical design is the best. The modular construction was used by Wavetek as well and, at the time of service, is priceless. I have a few HP/Agilent equipment, some basic and some extremely complicated like the 8662. I always do a complete recap. before bringing them to my workbench, since a 40+ year old elect. capacitors will eventually fail. That way I won't have an unexpected surprise when I need it the most.
I ordered up a set of new caps today :)
That power supply is a nigthmare. I swap it in my 8663A with meanwell switchers two years ago. No problem at all. It is a versatile generator, all in one with phase mod, pulse, BPSK, AM, FM... just a perfect machine and wonderful to see them working. Cheers...
The AR488 open-source USB->GPIB project on the EEVBlog forum is a super cheap and easy way to play around with GPIB and it does work with old HP gear. Since the HP 3478A has calibration data stored in a battery backed memory, it's definitely worth getting a GPIB solution going to back that data up. Then if the battery dies, you can replace it and reload the calibration over GPIB in a few seconds. It worked great for my 3478A.
HP GPIB (General Purpose Interface Bus) was also standardized as the I.E.E.E. 488. It's an 8 bit, bi-directional, parallel, multimaster bus... low speed, compared to new instrument control solutions... but, it was extremely popular. You could have up to 15 devices on the bus. The Commodore PET even had a version of it. The Agilent, and National Instrument controllers (actually, System controllers) are very expensive, for what they are... if you count the parts, and the plastic; but, they work. I've seen Girlando's project... and most (if not all the other projects.) Many of them try to use a micro-controller to implement a system controller (Girlando's original project used an Uno) The GPIB was very well speced... and explained... much like all the HP doc, at the time... they wanted people to be able to use GPIB. So it's a mystery to me as to why people trying to implement these, haven't read the spec... and if they have it's a bigger mystery to me as to why they try to do it the way they do it. The Agilent, and National Inst., USB system controllers work... but from Keysight, or N.I. they cost around 500USD, last time I looked. HP, and N.I. made 8 bit controller for IBM PC, and 16 bit controllers for ISA bus, and 32 bit controllers for ISA... and they do pop up... but, they are not cheap either... you need and older computer (for the bus timing, and software)... and they don't always work (There was often software, and hardware config issues.) Anyway if you spend the bucks... you get a good instrument control solution that's pretty flexible, and robust... but, if you don't... you might end up with another... time consuming hobby. Ask me how I know.
The DMM brings back memories, I worked at the division that designed and built it, LID.
I totally agree with you. Whoever love vintage electronics can’t skip buying vintage HP’s. They are amazingly built and very precise, despite the age. I have both instruments you purchased, the multimeter was an easy fix. The 8662s ( I have 2 units) are still in need to be repaired. Be careful of the backup buttery pack behind the front panel, they do leaks and damage a lot of parts (don’t ask me how I know). 8662 has sweep features and up to 5 programmable markers, very handy for visual alignments as you know. With a fast XY scope you can build an excellent test setup. I love this staff, and your video 👍.
Thanks for the advice. I'll check out the battery when I remove the front panel to service the rotary encoder.
I agree totally with your choice to use older top-shelf test equipment instead of the new Chinese stuff. As a Test Engineering Manager in the 80's I had lots of experience with the things you are using today. Many times I have said that I would take my 40 year-old Tektronix scopes over my new off-brand one. Recently I went back to using an old 545B to work on a broadcast transmitter that I am restoring. It still works and is still accurate and a side benefit is that it helps heat my shop in the winter months.
Many people reading posts here got no idea that a 30 MHz Tek 545B is 17X12X24 inches at 64 pounds. It does have a nice display. The Tektronix 555 at 68 pounds for the scope, and 45 pounds for the power supply is my all time shop heater.... just over a kilowatt... and almost all that being heat.
Great scores Bob! The old HP stuff are often terrific buys. Their original MSRP can be eye watering. Fascinating just to study their build quality and attention to detail. In case you haven't come across it, those screws get buggered up because people try to use standard Phillip's bits. HP used Pozidriv screws back then. Great content as usual!
Thanks. Ah, that makes sense regarding the screws. Just order up some news power supply caps. You'll be seeing more on this device next weekend.
Great stuff. Another part of the hobby also is obvious: I could hear the breathless excitement as you got parts working. The power stayed on! The error lights went out!! We have sweep!!! We were right there with you as we get someone else's abandoned repair, and figure it out. Very exciting and satisfying! (yes, like many others, I got a high end but old piece of gear (HP3336B) cheap, have no idea how to use all it does, crazy more accurate than I will ever need, but I enjoy it, and that is the goal) Thanks
That is a robust peirce of equipment, very serviceable which is great. Would make a perfect addition to your bench
I worked as a Test Engineer for an analog IC company and we did lots of stuff for HP(and later Agilent) Sometimes it was our stock part but HP wanted there Internal P/N marked on it so no one knew what it was and had to go to HP. But we also tested to special specs HP wanted for the parts as well.
In the back of your manual should possibly be a break down of what UXXX actual HP P/N is. I've also found cross references online that would point to what the generic part is..like LM101 or 7904-15 regulator.
Sometimes even HP's own service manuals show the original part number alongside the HP order number in the parts list.
Ah yes, loads of HP house numbered parts. At least all the capacitors are clearly labeled :)
Nice Piece of gear!
That's a beast! Love your videos! Best on utube in my opinion!👍
Very cool. The one cap open started a lot. But noisy. I really like that✔️
Right with you. I think i enjoy the test equipment more than the actual testing, LOL. I was a metrology tech in another life. I work at an electronics parts store in Washington state, and we recently got the mother lode of older test gear, but nothing this high-end as that synthesizer. We did get one of the HP meters in the first part of your video. It sold quick.
That consider one of the best signal generators HP made to date just great very low phase noise and very high performance generator.
Definitely try to restore it fully it as well worth it. It also quite serviceable as nicely layout. The tantalum’s will be fine, I won’t waste money on replacing them, it excessive ripple and heat that kills them, I haven’t seen them fail in this model generator and pretty reliable. It the PSU I would of course put low ESR capacitors in. Also invest in a programmer and read the EEPROM they do fail. I always backup EEPROM and NVRAM on other equipment as they do fail in he NVRAM on Tektronix scopes as battery fails and hold firmware and even EEPROM as have seen them go bad as well and corrupt some of the data. Then you have a backup of it does happen.
That is an amazing unit.
Great deal on the HP 3478A with two manuals no less. Seems to be missing it's other leg but you can fabricate one from
a metal coat hangar quite easily. The HP 8662A synthesized signal generator is dead on arrival but that is why it was such
a cheap purchase and maybe will need minimal repairs to get it to work. The BAMA page shows a HP 8663A ( similar to yours) manual however the power supply board definitely needed looking at first. With that now repaired you must look
for the internal battery that may need replacing. Your STATUS button has a light blinking so you need a warm up time
of 30 minutes and it should go out. If not it shows an error code when pushed. Fun with new "toy"! Great video Bob!
I think it's there and I just didn't have it folded down. Full manuals are freely available online from Keysight. www.keysight.com/us/en/lib/resources/technical-specifications/operation-and-service-manuals-for-older-obsolete-rf--microwave-instruments-610466.html.
I have error code 10 that's gain and modulation. Not surprising since I only have coarse attenuation and modulation does not work. Otherwise, the frequency and sweep seem to be working fine.
What a trip. To get it to work by just changing 3 caps and then work out that multiple error lights are just caused by the internal clock ref being disabled by a button. So now the only error light is 10B which maybe something to do with why the output waveform is distorted. What a result. I wonder how much power it consumes? Might be a good idea to copy all those PALS and GALS if possible and any ROMS it has. Hope it serves you well.
Test equipment is just a tool !
The key is understand the spec , the function , its capability & shortcomings .
You can normally work around most problems with much lesser equipment .
Or just make some custom equipment ... it’s a home lab !
Diy low-end / low cost equipment is part of the fun ….
I agree in generally, but I'm running a side business as well where time is money. I need to have a core of reliable, robust equipment. That's why I've been upgrading lately.
I would love to see the HP 8662a in operation.
I usually buy used, did way before eBay, just got better after eBay. I recently bought one of those cheap component testors from eBay, might never use it now but sure would have been nice to have in the day. I also got a Fluke 179 meter because it was cheap for what it can do and I wanted a really nice tough meter. I think I paid 90 for it 5 or so years ago. Probably has a bunch of functions I'll never use but does great for checking batteries and whatever else I need to check. I have a dual trace techtronic scope at a car swap meet that was supposed to work, but I also got the service manual and someday I'll get bored and trace it out with the manual. I gave away a couple of simple scopes that were low bandwidth because I didn't really use them but decided I should have one of better bandwidth and more channels around. I only paid 20 bucks for it so no big loss.
What a beast that signal generator is. Don't bother replacing those axial tantalums, they are practically everlasting, unless exposed to overvoltage. And they are super expensive nowadays (they always were pricey, but today it's ridiculously expensive). Definitely worth replacing all the electrolytics in the PSU, and everywhere else, especially 105°C rated ones, because the early 105°C rated and/or low-ESR capacitors are very prone for leaking electrolyte. And I would really suggest to read out and archive all the EPROM contents, even if they are available online. There could be different hardware revisions that require different ROM versions, so it's safer to save the actual EPROMs you have in this. I have this exact trouble with a Philips laserdisc player: ROM bins are available online for my model, but only for different hardware revisions. Some EPROMs read fine after 50 years, but I have seen corrupted firmwares in EPROMs less than 20 years old.
Thanks for the advice. I've been warned about EPROMs in other pieces of gear I have as well. Time to buy a reader. Was surprised to see how inexpensive they are these days :)
@@bandersentv And someone below mentioned this signal generator has a battery pack behind the control panel, which tends to destroy boards by peeing on them. You should remove that immediately, but prior to that, investigate the service manual to understand what that battery does. It is possible that it keeps the calibration NVRAM content, so if you disconnect it, you may convert the instrument unusable, in need for a full factory recalibration.
If the battery is a primary lithium pack, it is entirely possible that it still has enough voltage in it after 38 years to keep the SRAM contents. In that case, you should connect an other battery parallel to it before removing the old one, to keep the cal data intact. And preferably put the new battery on long wires, place it far away from any spots where it can cause damage in the future.
Love your channel and agree with you in regards to older test equipment especially HP, got a bench full of them myself. Now, Where to get that Dumont Tee Shirt you have?
Several website have the Dumont shirt. I think I bought mine at Printerval. printerval.com/dumont-televison-network-dumont-television-t-shirt-p1486948
Just as a guess, I’d say that whatever stage is right before the output driver/attenuator is the problem. Clearly, the attenuator works, so if the problem was the driver transistor/IC, I would expect that reducing the output level would reduce/eliminate the clipping, which clearly was not happening, even at -33 dB. And yes, the innards do look a lot like the Apollo stuff CuriousMarc showed on his YT videos. 😊
The attenuator works, but strangely. It seems to ignore the fractional part and only change in large chunks. Repairing the power supply may fix the clipping. If not, there are several adjustments including an offset in the module with the error light.
The Hp8662a once calibrated (which is a really long painful process) it a fantastic unit.
LOL, I've gotten taste of the process from watching other videos. I have a frequency counter and the new, accurate DMM. I hope that is sufficient. I do not have a spectrum analyzer.
@@bandersentv That is one of the drawbacks of having such precise test equipment, you need top of the line spected equipment to adjust them!
Your frequency counter time base should come from a accurate standard like a GPS controlled oscillator. Your in the big leagues now. Getting your HP equipment calibrated so that it met it's specs wasn't cheep.
@@notimetolooz9655 True, but ballpark is good enough for my needs. If the frequency is within a few 10 Hz and I can get rid of that clipping, I'll be happy. Also means I can replace three items currently on my workbench :)
@@bandersentv I'm thinking the AGC error is the source of the clipping. An amplifier is maxed out and distorting. The amplitude control still works because the attenuator is after the source of distortion. I was just saying that you won't get six digit accuracy if it is only calibrated to four digits, etc.
The higher voltage caps will also have higher ESR (so it seems normal that’s the 450v replacement would read over 2 ohms where a 250v part would probably read lower)
Exactly which makes me wonder why the tech in the very detail repair tutorial uses 450 volt caps. He claims they have a lower ESR! He also misidentifies a tantalum cap. I'll show you all the caps I went with. There's a great new line of high temp, low ESR acial caps from Vishay
@@bandersentv Awesome! Not sure why my name is showing up as "user..." something in my comments. Just had to create a new Google account because my old one I forgot was tied to an old school Google account that got deleted, so lost all my subscriptions and everything. Your channel was one of the first ones I re-added. -Skyler Hagen
Never mind, fixed it. Apparently we have a "handle" now instead of a name.... what next
The TV CRT on your bench is from a TV set in the 60ies called a "Futura" (Sylvania ? )
1959 Philco Predicta
Very informative video. Thanks.
I’m curious of…
Though I’m not entirely sure what they are. But I have seen some videos where guys do alignments with “ spectrum analyzers”.
Why don’t you go down this route ?
Thanks
Very expensive and even more overkill than the gear I have now. Think of it as a combination sweep generator, wideband radio receiver and XY monitor in one. Really geared more for communications receiver work then televisions.
Oh ok thanks that makes sense. I really wasn’t to clear on what they were :-)
G.M. Can you recommend a specific RF signal generator that can sweep? I restore old, mostly general coverage AM receivers and already have a Knight 8650 but the distortion and harmonics are really bad. And it does not have a sweep function. John
Sorry I only limited experience with radio sweep gens. The HP 3336A is very nice but big, heavy and maybe overkill for your needs.
I've just started watching another one of your "test Equipment Gear" videos and I'm anxiously waiting to see you fix this $29,000, 66 lb, ultra low noise instrument. You didn't show us
how to remove the top cover. Can you provide me with a (3) or (4) Step process to remove cover on my 8662A that I just bought this past Sunday afternoon.
Yes, the 8662A cost $ 29,000 with no Options in 1981. Best Of LUCK getting this beautiful generator up-and-running !
I do at the 20 minute mark. Remove one screw in the middle. I also did mostly fix it. I have several more videos on it.
By the way I picked up an 8664A a few weeks after the 8662A. Newer, bigger, heavier and more expensive.
Yes. I know. I watched all of the 8662A and 8664A videos. I now want to remove the power supply cards in my 8662 and re-cap with 105C, low ESR caps. I removed the rear "center" screw but cover will not slide back ....whole thing feels like it's welded in place ....gudd dammit.
You brought the 8664 back to life. Use it as it is. It looks right at home in your lab !
@@garyspencer1711There is some metal mesh material packed in around the from for shielding. That could be what is holding you up. I think disassembly is also covered in the manual.
I have a hp8643a and was a little disappointed in its sweep capabilities. It has better specs for digitally stepped sweeping opposed to continuous sweep. In my limited experience I am used to continuous sweep and wondering if you ever use stepped sweeping.
No, I've only used continuous sweep. I can see how it might be useful though. I've observed that sweeping too fast can effect the response curve. Stepping might give the circuit more time to settle. Or just reduce the sweep rate (if you can) I think the 60Hz most older gens use is a little too fast
@@bandersentv Thanks for the reply. Was looking up specs for your 8662a and holy cow that things base price was 42 grand originally!
@@cgeorge6786 Yes, it's quite a beast. That's why I don't mind spending some time and money fixing mine.
@@cgeorge6786It is literally one of the best signal generators HP made to date performance wise. It one of my personal favorites as used in production many years ago and performance is top notch, definitely worth repairing to 100% and recalibrating.
@32:00 I realized this wouldn’t get fixed in this vid 😊
I got a lot further than I expected that's for sure.
It made me very nervous that you were running it with out screwing down the PSU cover. Looks like silicone/zinc oxide thermal conductive paste, so the cover also functions as a heat sink.
Appreciate your concern, but it's fine. I only ran it for a short period plus there is plenty of ventilation with the cover off.
Old HP, Tek, GR, and Keithley gear is well worth saving from the dumpster. With a few exceptions faulty gear from those companies is repairable.
Also, I referenced your video on Adapt-a-Cap at 17:55 in a video I just posted ua-cam.com/video/qUSdtCYZeRY/v-deo.html
Thank you. I just subscribed. Strange i haven't come across your channel before. That 3330B looks amazing.
Who throws this stuff into a dumpster? I find that hard to believe... and I worked in a large institution with a lot of this stuff. Somebody always took it. I'm not saying it doesn't happen; or it doesn't happen where you are...but, nobody does that here. Where have you seen 1980's HP gear, in a dumpster?
@@willthecat3861 I had a temp job at an authorized Sony repair center and threw a bunch of test equipment in a dumpster. No one was allowed to take anything and they smashed some gear with a hammer to ensure it couldn't be used.
@@willthecat3861 You might not find something like that in a dumpster these days but years ago it happened. HP would stop supporting the equipment after a certain numbers of years. It companies could not get a piece of equipment calibrated or repaired out it would go. Few people that worked with the equipment all day wanted to use it at home on their time off.
outstanding ! see ya. Bandersentv, 🤘
Shops looking a little better. you need a few things. a sencore lc75 capacitor analyzer. a signal tracer...ie ieco, heathkit etc. rigol 1054z scope.
Why? I have them all except the Rigol, but never use them. Even some of this gear I don't expect to keep out once I'm done with some alignments. Multimeter and good soldering iron is all I need most of the time. I'd much rather have the space. Just use a scope to trace a signal. I find injection to be much more useful anyway. I rarely waste my time testing old paper or electrolytic caps - just replace them.