Really useful guide mate. I just got one of these from eBay with one channel out - feeling a lot more confident about digging about inside and giving it a service after watching you. Cheers, subscribed.
NFM cheers for the tip - I went through the installation guide and it turns out there’s nothing wrong - the seller and I both misunderstood what the Low output was. Getting similar distortion readout to yours.
If you still have the information, I am interested in the part numbers for the Vishay caps used on board A3, the Notch Filter Board. Specifically caps C2 and C3 which you listed as 50V, Radial, 20uF capacitors. I'm finding -10/+75% but no joy on +/-20% ones. It could be the ones you used have been discontinued. Anyway, with the part numbers I can more easily search for an equivalent. Cheers.
The MC68B21 is a PIA chip. A peripheral interface adapter chip that was commonly used along with 6800 and 6502 series processors to add input and output pins or ports. The standard 6820/6821/6822 type I/O chips were 1Mhz clock speed. The B on this one indicates 2Mhz clock speed capability if I recall correctly?
It won’t let me edit or delete the previous comment, so I’m replying to add to it. Anyway you later mentioned a MC68B09, and that’s a 2Mhz 6809 processor chip. Not sure whoever actually used those very much? The only place I can think of was in the Commodore SUPER PET computers, of which were weren’t many, and those were mainly sold to schools back in the early 1980’s. Actually the original Macintosh computer prototype was designed around a 6809 processor chip, then once Steve Jobs hijacked the project it was switched to an 8Mhz 68000 before it was released.
Wet-slug Tantalums. Big differences between cheaper dipped and wet-slug. Check out the current prices to replace a wet slug with mil-spec reliability and you might just have a myocardial infarction. Ever heard of Coltan?
You seem to skip over the amplitude settings. In fact, you go from shoving a board back in the machine straight to showing measurements without showing us what you did to get there. The % distortion is variable with amplitude so are you measuring distortion at 3V?... because no one has a clue since you didn't show any of that.
@@NearFarMedia Thanks, now that I've had it a few days, I realize I don't really need any additional filters. It has the 80kHz and 30kHz low pass which is all I need for audio work.
Maybe, I just used what I could get at the time that would fit. Good caps mean better performance and longer life, so no problems except maybe price. :)
Not a fan of arbitrary replacement of electrolytics, it's usually done by folk who don't know what they're doing. Were the old ones defective? Upping the capacitance can cause problems, HP designed the PSU to have those values, PSU design is complex, I'd be wary. Yes all electrolytics have a tolerance, usually -10/+50% but in reality are usually very near stated value. Not a good idea to touch the traces on the back of the board, the grease off your fingers can change the inductance and capacitance between these traces. This is a very advanced bit of test gear, with respect, not to be mucked about with if you don't know what you're doing. Nice bit of test gear anyroad, have fun.
Don't worry, I know what I'm doing. :) Been doing this stuff for nearly 20 years now (geez I feel old...) The PCB's were cleaned before they were replaced, so no fingerprints remained. The capacitors were measured and were showing a higher ESR that optimum, so that's why I decided to swap them out. They are just bulk smoothing caps and the values aren't super critical (unlike the tantalums on the other boards which I didn't touch, too finicky to match and no need anyway), If there was a problem with going higher value on those big PSU caps, I doubt the HP engineers would have chosen caps with such wide tolerances. The caps I installed were within the tolerance of the original caps anyway. Basically, the more smoothing, the less noise, the better performance as displayed by my test results. :)
He properly explained the choice he made for changing the caps. He actually showed the ones in the unit were labelled +75%, yes most will be close to the actual but HP designers were well aware of the stated tolerance. I'm not a fan of letting electrolytic cap sit in old equipment until they fail and take out pc boards and other components. I've been in the business 30years+ and do not agree with your statement. His video was excellent and the comment I will make is on the solderin technique. Solder should be applied to heated trace and lead not to soldering iron tip. Otherwise excellent, thanks for sharing.
*YES YES "ALL THE OPTIONS"... (NOBODY HAVE)* *THE GPIB ..IS NO "OPTION" ..IS IN "ALL OF HP STUFF" ..* *IF ..NOT !! "NOBODY" BUY THE "STUFF"" 100 YEARS AGO ...* *LIKE THE "DISPLAY" OR "PSU" .. OR "CABINET" ...* SO NO NEED TO "TALK" ABOUT "ALL" OPTIONS .. YOU NEVER WILL #BUY / GET /SHOW
Nice bit of kit. I have the A model that needed a cap replaced to get the distortion down.
Really useful guide mate. I just got one of these from eBay with one channel out - feeling a lot more confident about digging about inside and giving it a service after watching you. Cheers, subscribed.
They are a good unit, and they have a good service manual available, so following the repair guide should narrow down the problem relatively easily.
NFM nice one. I downloaded the PDFs, quite a lot to work through but it’ll be worth it. Cheers for the encouragement.
NFM cheers for the tip - I went through the installation guide and it turns out there’s nothing wrong - the seller and I both misunderstood what the Low output was. Getting similar distortion readout to yours.
If you still have the information, I am interested in the part numbers for the Vishay caps used on board A3, the Notch Filter Board. Specifically caps C2 and C3 which you listed as 50V, Radial, 20uF capacitors. I'm finding -10/+75% but no joy on +/-20% ones. It could be the ones you used have been discontinued. Anyway, with the part numbers I can more easily search for an equivalent. Cheers.
The MC68B21 is a PIA chip. A peripheral interface adapter chip that was commonly used along with 6800 and 6502 series processors to add input and output pins or ports. The standard 6820/6821/6822 type I/O chips were 1Mhz clock speed. The B on this one indicates 2Mhz clock speed capability if I recall correctly?
It won’t let me edit or delete the previous comment, so I’m replying to add to it. Anyway you later mentioned a MC68B09, and that’s a 2Mhz 6809 processor chip. Not sure whoever actually used those very much? The only place I can think of was in the Commodore SUPER PET computers, of which were weren’t many, and those were mainly sold to schools back in the early 1980’s. Actually the original Macintosh computer prototype was designed around a 6809 processor chip, then once Steve Jobs hijacked the project it was switched to an 8Mhz 68000 before it was released.
23:34 Nice EEVblog reference!
Cool HP Engineers. Color-coded levers. Borad 7 is purple ... :-)
Wet-slug Tantalums. Big differences between cheaper dipped and wet-slug. Check out the current prices to replace a wet slug with mil-spec reliability and you might just have a myocardial infarction. Ever heard of Coltan?
Yeah, they have a nice pricetag. Coltan ain't cheap for certain 'political' reasons...
You seem to skip over the amplitude settings. In fact, you go from shoving a board back in the machine straight to showing measurements without showing us what you did to get there. The % distortion is variable with amplitude so are you measuring distortion at 3V?... because no one has a clue since you didn't show any of that.
tantalum capacitors ARE electrolytic capacitors
I think these gray capacitors are mica type.
Hi, you mentioned that you would list down below the source of the new filters that someone was selling online.
Unfortunately the website has been down for a while now, so I don't think they are available anymore.
www.deckardaudio.com/products/HP8903Bfilter.php
@@NearFarMedia Thanks, now that I've had it a few days, I realize I don't really need any additional filters. It has the 80kHz and 30kHz low pass which is all I need for audio work.
BTW I swapped out the same six caps you did and my signal/noise went from ~94db to 104db at 1kHz, 1Vrms. Noice!
Oh nice work! The extra power supply filtering from the new caps is definitely working. :)
👍
I think it's overkill to use Simens Sikorel caps here.
Maybe, I just used what I could get at the time that would fit. Good caps mean better performance and longer life, so no problems except maybe price. :)
@@NearFarMedia If you want to keep it, sikorel is a good choice. It will last for decades.
Not a fan of arbitrary replacement of electrolytics, it's usually done by folk who don't know what they're doing. Were the old ones defective? Upping the capacitance can cause problems, HP designed the PSU to have those values, PSU design is complex, I'd be wary. Yes all electrolytics have a tolerance, usually -10/+50% but in reality are usually very near stated value.
Not a good idea to touch the traces on the back of the board, the grease off your fingers can change the inductance and capacitance between these traces. This is a very advanced bit of test gear, with respect, not to be mucked about with if you don't know what you're doing.
Nice bit of test gear anyroad, have fun.
Don't worry, I know what I'm doing. :) Been doing this stuff for nearly 20 years now (geez I feel old...)
The PCB's were cleaned before they were replaced, so no fingerprints remained.
The capacitors were measured and were showing a higher ESR that optimum, so that's why I decided to swap them out.
They are just bulk smoothing caps and the values aren't super critical (unlike the tantalums on the other boards which I didn't touch, too finicky to match and no need anyway), If there was a problem with going higher value on those big PSU caps, I doubt the HP engineers would have chosen caps with such wide tolerances. The caps I installed were within the tolerance of the original caps anyway.
Basically, the more smoothing, the less noise, the better performance as displayed by my test results. :)
He properly explained the choice he made for changing the caps. He actually showed the ones in the unit were labelled +75%, yes most will be close to the actual but HP designers were well aware of the stated tolerance. I'm not a fan of letting electrolytic cap sit in old equipment until they fail and take out pc boards and other components. I've been in the business 30years+ and do not agree with your statement. His video was excellent and the comment I will make is on the solderin technique. Solder should be applied to heated trace and lead not to soldering iron tip. Otherwise excellent, thanks for sharing.
*YES YES "ALL THE OPTIONS"... (NOBODY HAVE)*
*THE GPIB ..IS NO "OPTION" ..IS IN "ALL OF HP STUFF" ..*
*IF ..NOT !! "NOBODY" BUY THE "STUFF"" 100 YEARS AGO ...*
*LIKE THE "DISPLAY" OR "PSU" .. OR "CABINET" ...*
SO NO NEED TO "TALK" ABOUT "ALL" OPTIONS .. YOU NEVER WILL #BUY / GET /SHOW
Triggered much?