I just dug out the 3582A after a move awhile ago. It powered up with no vertical baseline control for the text or scan. I re-seated the control and power supply boards (but not the hard to get at input boards.) That gladly restored vertical control and the unit functions as I remember. But the screen dot-matrix characters are quite distorted.
I got a service manual on order, but wonder if any members have performed service on these or seen similar faults. I've serviced Tek485 power supply faults and repaired HP display HV supplies as well as the logic in Tek2240, so I am not squeamish in regard to the 3582A innards.
I kinda know that this unit is a bit limited by modern standards, but it has its uses.
I'm still inclined to think this is a signal or grounding connection problem. The text distortion is essentially static and still readable, barely. And, interestingly, I think the distortion is the same for every repeated character. That is, all A's distort the same and all W's the same, but different than Z.s etc.
I am also checking out the Yahoo HP/Agilent group.
I got a service manual on order, but wonder if any members have performed service on these or seen similar faults. I've serviced Tek485 power supply faults and repaired HP display HV supplies as well as the logic in Tek2240, so I am not squeamish in regard to the 3582A innards.
I kinda know that this unit is a bit limited by modern standards, but it has its uses.
I'm still inclined to think this is a signal or grounding connection problem. The text distortion is essentially static and still readable, barely. And, interestingly, I think the distortion is the same for every repeated character. That is, all A's distort the same and all W's the same, but different than Z.s etc.
I am also checking out the Yahoo HP/Agilent group.
From pictures of the unit, it looks like it uses the HP Vector Display (HP 1745) like many other instruments of that era. You may have to get a separate service manual for that. Manuals Plus or Artek Media are your friends in need.
I tried calibrating one of those displays a while back and failed miserably. I would recommend only tweaking the pots for astigmatism, focus, trace rotation, and the like. On some models of that display unit, the pots are accessible at the rear panel of the instrument. On others, they're trimmers on the top board of the display unit.
The HP1745 display is a self-contained unit. It receives commands via a parallel bus, handles all the vector graphics, and drives the CRT. There's a jumper onboard that will make the display unit show a test image.
I've also found that many of the older HP instruments benefit from a thorough cleaning of all the board edge contacts and mating connectors. Deoxit works well for that.
Anyway... Good luck.
~Tom
I tried calibrating one of those displays a while back and failed miserably. I would recommend only tweaking the pots for astigmatism, focus, trace rotation, and the like. On some models of that display unit, the pots are accessible at the rear panel of the instrument. On others, they're trimmers on the top board of the display unit.
The HP1745 display is a self-contained unit. It receives commands via a parallel bus, handles all the vector graphics, and drives the CRT. There's a jumper onboard that will make the display unit show a test image.
I've also found that many of the older HP instruments benefit from a thorough cleaning of all the board edge contacts and mating connectors. Deoxit works well for that.
Anyway... Good luck.
~Tom
Last edited:
The Agilent group was not much help, but I reasoned that since the entire display was shrunk & distorted, the problem might be in the path to the CRT. I removed the CRT driver board (under the plastic) and reseated every connector I could get at, including those to the CRT tube. After careful reassembly, the 3582A powered up in perfect form with crisp, well formed characters located in the correct CRT positions!
I hope this info helps others with this system...
I hope this info helps others with this system...
I've got something similar going on with mine.
the text is clear and in focus, but not all of it shows up, (which i can live with)
the traces (channel a and b) do not show up, except for one fifth of a second or so after the crt is turned on. (which is long enough to verify nothing else is broken)
a zenar protecting the 100v supply failed shorted, as well as the 100v regulator, so the 100 v supply is now supplied by a 5K potentiometer, and the most text shows up at around 85 volts...
sounds like its a power supply issue...
the text is clear and in focus, but not all of it shows up, (which i can live with)
the traces (channel a and b) do not show up, except for one fifth of a second or so after the crt is turned on. (which is long enough to verify nothing else is broken)
a zenar protecting the 100v supply failed shorted, as well as the 100v regulator, so the 100 v supply is now supplied by a 5K potentiometer, and the most text shows up at around 85 volts...
sounds like its a power supply issue...
I would carefully unseat every PCB including the power modules and the boards under the metal plate to the right. Then, WITH THE POWER OFF, lift the plastic cover off the CRT electronics. Carefully disconnect and reconnect ALL PCB connectors including those at the CRT tube. Do em 1@ a time if you dont have a wiring diagram.
Do/did you have fault LEDs lit on the power supplies?
Let us know how you progress.
Do/did you have fault LEDs lit on the power supplies?
Let us know how you progress.
hi, thanks for replying.
One of the universal 15-7 volt supplies was bad, I already replaced it with a small offline flyback supply.
Yesterday I got it working.
All it needed was some adjustment to R109,112,6 and R46, via trial and error. As well as a replacement of Q7 and R28. not sure what those two do yet.
Also, not sure what i did but Q10 no longer gets too hot to touch, it may have been shorted out, looks like something pushed it over on its leads.
Anyhow, I'm in the process of scanning via a digital camera the complete repair and operation manual. I'll post a link to it here when i'm done, which might be a week.
Btw, the power demands are actually quite modest, overall it only consumes 80 watts in standby, and i'd say 20 or so of those watts are lost in the 12 and 5 volt linear regulators on the cards themselves... but if it ain't broke.. fix it till it is.
One of the universal 15-7 volt supplies was bad, I already replaced it with a small offline flyback supply.
Yesterday I got it working.
All it needed was some adjustment to R109,112,6 and R46, via trial and error. As well as a replacement of Q7 and R28. not sure what those two do yet.
Also, not sure what i did but Q10 no longer gets too hot to touch, it may have been shorted out, looks like something pushed it over on its leads.
Anyhow, I'm in the process of scanning via a digital camera the complete repair and operation manual. I'll post a link to it here when i'm done, which might be a week.
Btw, the power demands are actually quite modest, overall it only consumes 80 watts in standby, and i'd say 20 or so of those watts are lost in the 12 and 5 volt linear regulators on the cards themselves... but if it ain't broke.. fix it till it is.
http://www.johansense.com/tech_docs/hp3582a/lightweight.djvuhttp://www.johansense.com/tech_docs/hp3582a/hp3582A_L.djvuhttp://www.johansense.com/tech_docs/hp3582a/HP3582A_L.djvu
50MB djvu, the 500Kb JPGs are in the parent folder in case you want them.
50MB djvu, the 500Kb JPGs are in the parent folder in case you want them.
I have found that the Yahoo group Manual_Exchange
Manual_Exchange : Manual Exchange
often has a copy of the HP manual that I'm looking for. Download what you need - upload anything you have that they don't.
Manual_Exchange : Manual Exchange
often has a copy of the HP manual that I'm looking for. Download what you need - upload anything you have that they don't.
Power supply voltage - check
Look for bad caps in power supply and on the boards. Very common from that vintage is failed tantalum caps. Either shorted or opened. If shorted they get very hot and drag down the voltage. Open will result in noise on the supply. I have repaired quite a few HP and Tek instruments with this issue.
I had a 3582, relatively easy to work on.
paul
Look for bad caps in power supply and on the boards. Very common from that vintage is failed tantalum caps. Either shorted or opened. If shorted they get very hot and drag down the voltage. Open will result in noise on the supply. I have repaired quite a few HP and Tek instruments with this issue.
I had a 3582, relatively easy to work on.
paul
I 2nd that advice. I recently fixed my 6.5 digit DMM by replacing a bad tantalum (and its partner in the other supply). Originally I noticed that the -15 supply was low (0.5 V) and that the 79L05 reg was hot, so I replaced the reg, but that didn't help. I eventually traced it to the cap. A good lesson was learned.
Hey, I know this is an old thread, but its mine and for anyone who has or uses one of these analyzers, I have another tip. I dug this machine out again to do some AC/DC coupled measurements using 15-20VDC on the input. I found that after switching the AD/DC coupling switch several times, both channels got stuck on DC no matter the SW position.
If one looks at the A1 input PCB schematic, it is clear that this AC/DC function is made by closing a reed switch across the decoupling cap of 0.33uF. It is a computer controlled switch/relay, so either:
So this looks like a surprising design flaw by HP. Very unexpected!
The input DC range is 100V and the cap is 0.33uF with about 0.35 ohm ESR. At 100VDC when that reed closes, the current is a >250A impulse!
I lifted one end of the capacitor and added a 50ohm series resistor. So should have HP. Tested and fixed. (Here is one place where a high ESR is a good thing, as it protects the contact and is insignificant when compared to the 5K series input resistance already there.
If one looks at the A1 input PCB schematic, it is clear that this AC/DC function is made by closing a reed switch across the decoupling cap of 0.33uF. It is a computer controlled switch/relay, so either:
- computer cant read SW
- computer cant control relay
- capacitor is bad shorted
- relay is welded closed
So this looks like a surprising design flaw by HP. Very unexpected!
The input DC range is 100V and the cap is 0.33uF with about 0.35 ohm ESR. At 100VDC when that reed closes, the current is a >250A impulse!
I lifted one end of the capacitor and added a 50ohm series resistor. So should have HP. Tested and fixed. (Here is one place where a high ESR is a good thing, as it protects the contact and is insignificant when compared to the 5K series input resistance already there.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.