You said it was developed in Germany, HP had Production facilities in Böblingen, the units made there had a Sticker "contains imperial and metric hardware " inside. So maximum fun getting them reassembled is guaranteed :-)
I have two of the Pulse Generator models of this instrument and have spent quite a bit of time inside them, nursing them back to health. Regarding the hybrids, I think they actually use the same hybrid in a couple of different positions in the signal chain which is interesting! Thank you again, your videos are excellent!
Used a couple of those when I worked at HP MPG (Medical Products Group). The frequency counter was off on mine as well. And you're right, it was designed at Hewlett-Packard Böblingen, Germany.
I have one of these devices which went wrong when the instrument was knocked in transit. Fortunately I have the original service manual from HP and I set about fault finding it. The fault turned out to be due to one of the computer type power supply connector pin being open circuit. I took out the pin from the connector and it looked totally fine, but when I put my soldering iron on it, it fell into many pieces.
Had a couple of these and I hated them with passion. Pulse generating ASICs run very hot and if you buy them for repair on ebay, you mostly get those with dead ASICs, so not much to repair. Also no PLL, everything is analog and controlled with DACs from microcontroller, so like 5% frequency accuracy. Used the dead ones to harvest parts and eventually got a few working ones which I then sold for good money on ebay, though ;-)
That +5 vs. -5.25 V issue may warrant further investigation via the schematic. If it is the kind of arrangement I'm thinking of, a resistor may be off there. Not that it's likely to be too critical.
I cut my professional teeth on ECL. Yes, it was fast as hell but consumed a ton of power to do it. Eventually, CMOS killed it. If you followed ECL recommendations, there should also be a -2.0V supply in there, too.
I just got a couple of these without understanding the main flaw in these is the unrepairable ASICs. Both had E31 and E53 errors. Both seem related to the VCO/Slope IC. It's a real shame that ASICs seem to be the main cause of fault symptoms because I really like the ease of use of these FGs. Question regarding calibration, do you calibrate all your equipment in-house? If so how? I mean you'd have to send out at least some standards for external calibration correct? Also, if you ever get one of these with a noisy fan, removing them is a nightmare. Usually HP is #1 for ease of service (I love how you can prop the PCBs up on those hooks), but they really complicated that fan assembly.
Odd that with a pot for every darn thing in that instrument, HP cheaped out and didn't design the +5 and -5.25 supplies to include individual adjustments.
Lifting one of these things is a suprise. You expect weight on par with something similarly-sized (438A, etc.) but it's just like a solid rock. Always catches me when I take it down off the shelf.
This "display amplitude as into 50 ohm load" always gets me when I use HP function gens. First thing I do in let's say HP 33120A is go to the menu and change output termination to "High Z" - it doesn't change output impedance but shows open circuit voltage. It's one of more confusing features - maybe it's just me but when I set 1V peak to peak on function generator, I want 1V peak to peak with 50 ohms output impedance on the connector (measured by "ideal" voltmeter), not 2V.
If you're generating fast rise/fall times, 50 ohm termination is a must. Without it, you'll get ringing at the receiving end of your cable! See that overshoot and undershoot at 8:12? Notice how it's gone when Marc applies a 50 ohm terminator at the end of his cable. Transmission line theory works!
Yes, of course. As I said, these function gens ALWAYS have 50 ohm output. The "output termination" option literally multiplies numbers on the display by factor of two (so you enter the value you'll get at output if it's properly terminated) ;)
12:23 Isn't using T splitter with one output not terminated a bad thing to do? Doesn't it negatively affect signal by inserting reflections? Or is reflection impact negligibly low in this case? I'm not a specialist, but will be thankful for correcting my small long-line transmission knowledge.
I'm normally an avid watcher of your videos but having spent the last 20 years working in a cal lab working on these exact generators (amongst hundreds of other things) I'll give this one a miss, lol.
Personally I much prefer a 3325A synthesized function generator. It only goes up to 21 MHz, but it also goes down to something like 0.00001 Hz And it is a digital divider and you can externally clock it from GPS, though I seem to recall the external clock is annoyingly a 5 MHz input. Since I'm much more commonly working with audio and power frequencies than RF, being able to get accurate audio band frequencies and sweeps is right nice. It is also quite repairable unless the microprocessor dies, but that seems to be rare. What does screw up is the fancy relays, but they can be disassembled and cleaned.
The pulse generator from which I believe this was derived can generate really short pulses (a few nanoseconds wide), so at the time it may have been more economical to trigger fast logic based on a ramping waveform than to try to build a digitally clocked solution.
How are its distortion specs? Analog VCOs, for a lot of decades, were more pure than any synthesized sinewave generator. You should dig up an old HP 200 and check it out.
Keysight are still located in Böblingen near Stuttgart - "founded in 1959 as the first HP factory in Europe", they say. Rohde & Schwarz - Munich (and some other places in Bavaria) Hameg - Cologne, if memory serves
Yes mismatch reflections... but, why two attenuators? Is it to make a 50 ohm through BNC to BNC connection?... because each attenuator is BNC to N... or something?
Because I only had 10 dB attenuators on hand, and wanted 20 dB for easy 1/10 reduction. Note to self: buy 20 dB attenuators to calibrate HP instruments...
Hello Do you have the service manual with schematics and/or the pcb layout. I have the same instrument but the battery leaked and damaged some traces on the top pcb. What kind of battery you have in the instrument, is it rechargable or alkaline? From the number on the battery i have i found that the batteri is 2.5v alkaline but on some forum i people are talking about rechargeable battery.
The service manual is available for a few bucks from Artek Media. [edit] I got the battery wrong in my first answer, mixed up with another instrument. The manual says it's a NiCAD battery. I vaguely remember I changed it, but don't remember it as a NiCAD. The manual also implies there was a revision with the battery circuit, so maybe some have rechargeable NiCADs and some have non rechargeable, maybe long life Lithium batteries?
@@CuriousMarc Does the service manual from Artek Media has schematic or pcb layout? I downloaded the service manual from Keysight but it only has the programing, testing and adjustment procedures no schematics or pcb.
Wow, an all-analog function generator under digital front panel control. That's simultaneously beautiful and horrifying. Thanks for the peek inside.
Calibrating old test equipment brings a good sense of structure and wholesomeness into this time of turmoil . Thanks for the vid, Marc.
You said it was developed in Germany, HP had Production facilities in Böblingen, the units made there had a Sticker
"contains imperial and metric hardware " inside. So maximum fun getting them reassembled is guaranteed :-)
I've used on of those in the USAF. We used it as a reference timing source measuring Frequency Envelope Delay Distortion on S3 Communication lines.
-5.2 volts. The unmistakable indication of ECL!
Man old hp gear is the Best!
Thanks CuriousMarc!
Oh I know Mr. Shoehorn. He is always a busy fellow.
Thanks for another great video - that's quite some calibration process, keep them coming Marc.
I have two of the Pulse Generator models of this instrument and have spent quite a bit of time inside them, nursing them back to health.
Regarding the hybrids, I think they actually use the same hybrid in a couple of different positions in the signal chain which is interesting!
Thank you again, your videos are excellent!
Used a couple of those when I worked at HP MPG (Medical Products Group). The frequency counter was off on mine as well. And you're right, it was designed at Hewlett-Packard Böblingen, Germany.
That instrument has crazy capabilities. We used them for the TOW2 circuit card screening.
I have one of these devices which went wrong when the instrument was knocked in transit. Fortunately I have the original service manual from HP and I set about fault finding it. The fault turned out to be due to one of the computer type power supply connector pin being open circuit. I took out the pin from the connector and it looked totally fine, but when I put my soldering iron on it, it fell into many pieces.
Hi Marc, where do you keep finding these treasures? Keep up the good work!
Had a couple of these and I hated them with passion.
Pulse generating ASICs run very hot and if you buy them for repair on ebay, you mostly get those with dead ASICs, so not much to repair.
Also no PLL, everything is analog and controlled with DACs from microcontroller, so like 5% frequency accuracy.
Used the dead ones to harvest parts and eventually got a few working ones which I then sold for good money on ebay, though ;-)
I love fixing up old test gear.
I thought the frequency was off by design on this model :) Mine is particularly noisy too. Thanks for sharing
Yes!
My Prophet REV2 keyboard music synthesizer also has most of those waveforms, but it does not go quite as high as 50 MHz... 🤪
There's a world of difference between digitally synthesizing a 20 kHz sinewave vs. a 50 MHz sinewave....
@@demef758
You obviously didn't get the joke... BTW, the REV2 is DCO based (analog oscillators digitally controlled).
Thanks:)
That +5 vs. -5.25 V issue may warrant further investigation via the schematic. If it is the kind of arrangement I'm thinking of, a resistor may be off there. Not that it's likely to be too critical.
-5.2 Volts is used for emitter-coupled logic (ECL), much faster than TTL or CMOS for the era.
I cut my professional teeth on ECL. Yes, it was fast as hell but consumed a ton of power to do it. Eventually, CMOS killed it. If you followed ECL recommendations, there should also be a -2.0V supply in there, too.
Yes, start by looking at the specs ;-) Freq. is 5% precision.
Where did it all go wrong for HP?
Looks like you could key it into a dipole antenna and transmit Morse code! 16V at 50ohms is 5W... QRP! Enough to talk around the world.
Yes that's its main quality I think. It's seriously powerful. It will come in very handy pretty soon, as you'll see.
I'm dusting off my old valve transmitter right now, CW is coming back to save the world.
It would be an good idea to get FPGA/CPLD versions of those trouble maker chips.
I just got a couple of these without understanding the main flaw in these is the unrepairable ASICs. Both had E31 and E53 errors. Both seem related to the VCO/Slope IC. It's a real shame that ASICs seem to be the main cause of fault symptoms because I really like the ease of use of these FGs.
Question regarding calibration, do you calibrate all your equipment in-house? If so how? I mean you'd have to send out at least some standards for external calibration correct?
Also, if you ever get one of these with a noisy fan, removing them is a nightmare. Usually HP is #1 for ease of service (I love how you can prop the PCBs up on those hooks), but they really complicated that fan assembly.
Odd that with a pot for every darn thing in that instrument, HP cheaped out and didn't design the +5 and -5.25 supplies to include individual adjustments.
It may not have been possible to have separate adjustments.
They could perhaps be in a tracking configuration and all you are really doing is trimming the zero level about which both supplies sit.
Makes me wonder why they didn't design in some basic readback of frequency. But then I guess they'd have been tempted to close the loop.
You called it compact and it looked compact.. Until you rotated it and it was 3 times as long as wide.
Lifting one of these things is a suprise. You expect weight on par with something similarly-sized (438A, etc.) but it's just like a solid rock. Always catches me when I take it down off the shelf.
This "display amplitude as into 50 ohm load" always gets me when I use HP function gens. First thing I do in let's say HP 33120A is go to the menu and change output termination to "High Z" - it doesn't change output impedance but shows open circuit voltage. It's one of more confusing features - maybe it's just me but when I set 1V peak to peak on function generator, I want 1V peak to peak with 50 ohms output impedance on the connector (measured by "ideal" voltmeter), not 2V.
If you're generating fast rise/fall times, 50 ohm termination is a must. Without it, you'll get ringing at the receiving end of your cable! See that overshoot and undershoot at 8:12? Notice how it's gone when Marc applies a 50 ohm terminator at the end of his cable. Transmission line theory works!
Yes, of course. As I said, these function gens ALWAYS have 50 ohm output. The "output termination" option literally multiplies numbers on the display by factor of two (so you enter the value you'll get at output if it's properly terminated) ;)
12:23 Isn't using T splitter with one output not terminated a bad thing to do? Doesn't it negatively affect signal by inserting reflections? Or is reflection impact negligibly low in this case?
I'm not a specialist, but will be thankful for correcting my small long-line transmission knowledge.
You've been tweaking reactive circuits and you just now realize it's an analogue system with digital counters and controls? Still cool as hell.
I guess you don't have a 5335A as they call for in the manual?
I'm normally an avid watcher of your videos but having spent the last 20 years working in a cal lab working on these exact generators (amongst hundreds of other things) I'll give this one a miss, lol.
Finally
Personally I much prefer a 3325A synthesized function generator. It only goes up to 21 MHz, but it also goes down to something like 0.00001 Hz And it is a digital divider and you can externally clock it from GPS, though I seem to recall the external clock is annoyingly a 5 MHz input. Since I'm much more commonly working with audio and power frequencies than RF, being able to get accurate audio band frequencies and sweeps is right nice. It is also quite repairable unless the microprocessor dies, but that seems to be rare. What does screw up is the fancy relays, but they can be disassembled and cleaned.
Amen to the 3325A.
The pulse generator from which I believe this was derived can generate really short pulses (a few nanoseconds wide), so at the time it may have been more economical to trigger fast logic based on a ramping waveform than to try to build a digitally clocked solution.
Stay tuned for the next episode... Or is that, Stay calibrated?
the 8116 is peak HP
Mmmm. A nice bowl or porridge and Marc for breakfast. BTW, how many zeros after the decimal point in a gadgillion ? Lol...
How are its distortion specs? Analog VCOs, for a lot of decades, were more pure than any synthesized sinewave generator. You should dig up an old HP 200 and check it out.
Not bad. As you can see on the FFT, third harmonic is a good 30 dB down. Although I would trust my real spectrum analyzer more.
Do some calibrations..... Nice Instrument i didn't know that there was a hp factory here in Germany! Or was it made by Rhode und Schwarz or Hameg?
Not sure where it was made, but quite a lot of the fancy HP higher-frequency stuff from that era was designed in Germany.
Keysight are still located in Böblingen near Stuttgart - "founded in 1959 as the first HP factory in Europe", they say.
Rohde & Schwarz - Munich (and some other places in Bavaria)
Hameg - Cologne, if memory serves
Yes mismatch reflections... but, why two attenuators? Is it to make a 50 ohm through BNC to BNC connection?... because each attenuator is BNC to N... or something?
Because I only had 10 dB attenuators on hand, and wanted 20 dB for easy 1/10 reduction. Note to self: buy 20 dB attenuators to calibrate HP instruments...
and here I thought a DLP projector was crammed for space
My HP 8116A front panel is dead. While the fan is running normally, what do I need to check if the front panel is off.
That could be so many things. I’d check power supplies first of course.
Hello Do you have the service manual with schematics and/or the pcb layout. I have the same instrument but the battery leaked and damaged some traces on the top pcb. What kind of battery you have in the instrument, is it rechargable or alkaline? From the number on the battery i have i found that the batteri is 2.5v alkaline but on some forum i people are talking about rechargeable battery.
The service manual is available for a few bucks from Artek Media. [edit] I got the battery wrong in my first answer, mixed up with another instrument. The manual says it's a NiCAD battery. I vaguely remember I changed it, but don't remember it as a NiCAD. The manual also implies there was a revision with the battery circuit, so maybe some have rechargeable NiCADs and some have non rechargeable, maybe long life Lithium batteries?
@@CuriousMarc Does the service manual from Artek Media has schematic or pcb layout? I downloaded the service manual from Keysight but it only has the programing, testing and adjustment procedures no schematics or pcb.
@@vanekolev3005 Yes it has the schematics, the PCB placement (where the components are) but not the trace layout, and the parts list.
@@CuriousMarc Thank you for the information, i will get the schematics from Artek Media.
I think Dave (EEV) would say it's all about the tongue angle... Btw. I need a new FN, not buying this model. Thanks for the warning Marc ;)
What about that battery?
Gee.... No Option 001? ;)
So.... What is the purpose of the Panasonic 2/3 AA Lithium battery?
HP = Has Problems : )
high priced
@@vap0rland High Performance.
Another Tektronix buddy ? ;-)
I have 4 Tek scopes.
I have one HP5335A though.
I have one of these. It's garbage for anything precision.
Looks infuriating