HP 8116A More Help Needed

[Mick B]

Hello all, I was given this HP 8116A w/001 option and it has an issue. When turned on the led's and display all come on like always, but they never go out? There are no stuck keys- when holding down a key the stuck key error is displayed "0". It looks like the first segment on the display is brighter than the rest looks like a 1? I am also missing both 24v rails unplugging all the other boards & lifting the +0Ω W15 resister still no voltage, Lifting -0Ω W16 resister I have both 24v rails. Getting to the op-amp to test is going to be a bitch I could get to the pins from the bottom, but I like that idea less. Has anyone seen this problem before? maybe save me some time maybe the chip U30 is not doing a reset? and the other boards need to be in circuit. I have been up hours reading and tracing schematics I'm starting to see double. I'm feeling a little

[Mick B]

Good morning, Update, all boards installed. Went through the troubleshooter for 24v rails and on to 10.4 the Shaper and Output Amplifier. As always it starts with clearing the error code by pushing the LCL key all this does is display the stuck key error "o". The reset pin on U30 is at 5v grounding it for a moment does nothing. Checking the voltage at CR501 &2. it's not a full +/_ 15v. OK!  NEXT step set it up for XXX and check the output?
I need to just STOP. .... OK regrouping.
1. Yes, I have no 24v rails, (A problem on the -24 rail).
2. But it seems, I need to find out why it's not completing the self-test first. or is it?
3. I can't find the information to diagnose this.
4. Is it in the manual and I'm missing it?
5. Is this what it looks like when the RAM, (which is the next step in the boot process fails?
 Can I please get some help? I know there are many experienced HP warriors familiar with the 8116A in this group. I want to keep this. I really don't want to see this on E-Bay, Parts only.

[precaud]

I'd correct the power supply issues first, then check ALL of the 'lytics, especially down in the output stage. These things run very hot.

[Mick B]

Yes, VERY, VERY HOT! it's a wonder they last so long. I started on the power supply first, whatever is drawing to much current is on the main board A1 I checked all the electrolytics and as time permits, I'm working on the tantalums. And the current sense resistors and the driver transistors are OK. Got to start somewhere as there are just to many components on the main board and unfortunately, it's not a dead short so I can't inject a little current and then use my thermal camera. Oh ya I also pulled U200 A 24v op-amp. There are 13 or so 24v transistors in the output amp section that could be leaking and not to mention, a few diodes. WHAT A TIME SUCK! 
Thats why I wanted to try the display problem because it lights up without the 24v rails. I can stop the Op-amp from doing a reset and perhaps then it will throw a code. but with the display all lit up I'll never know. GOD help me!
I also pulled the 3 HP chips

[Mick B]

WTF?    417 people read this and no response as to the problem of all the display LEDs on. Come on guys, is it because the enthusiasm for something this old has come and gone? or maybe the more seasoned Techs (hobbyist) have left the group? I'm just trying to learn. This forum uses to be the GOLD Standard for information on electronics. I'll give it a few more days before I stop and visit fleabay. and another piece of classic HP goodness slips through my hands  and i will just have to keep using my Juntek8090.

[rhb]

Well, most people would ask the HPAK list on groups.io.  Lot's of old HPAK EEs.  The  purpose of the list is keeping HPAK gear working.

Reg

[bdunham7]

Come on guys, is it because the enthusiasm for something this old has come and gone? or maybe the more seasoned Techs (hobbyist) have left the group? I'm just trying to learn. This forum uses to be the GOLD Standard for information on electronics. I'll give it a few more days before I stop and visit fleabay. and another piece of classic HP goodness slips through my hands  and i will just have to keep using my Juntek8090.

I'm one of the 417.  I've repaired a few 8116A problems and they can be quite a challenge.  If I knew offhand what your issue is or had some idea, I'd certainly tell you, but I don't.  You'll have to track down all your power supply problems somehow before you go further; working on one of these with any known PS issues is madness even if you have a display coming up or whatever.  The other pitfall of these is that you may fix 10 things and then still not be able to calibrate it properly because one of the custom hybrids is bad.  So you've just picked a particularly difficult and arcane little monster to work on and no matter your skill level, you need a lot of patience.  Try the HPAK group if you like.

[Mick B]

Thanks to those that responded. I have taken your advice and posted this problem on HPAK. hopefully someone there can help. and for those that care. I found at least one output amp transistor shorted. -24v problem solved. Parts might become an issue.

[Mick B]

UPDATE! The 24V rail problem is resolved. There were a few shorted and open Shaper output amp Power transistors, and 2 bad resistors New Transistors on the way.
I could use some new heatsinks for those transistors I only got 2 off and destroyed one in the process. any advice as to how to remove these without killing them?
I still have the issue with the LEDs I could still use help with these to problems.

[pqass]

First of all, I don't own an 8116A.

However, looking at the schematics, I see that the LEDs and 7-seg. display are controlled by two 74LS374s and ICM7218A (U24, U23, U22 on page 3C).  If I saw LEDs stuck-on, I would check if the CPU (U1 on page 3A) was reset and is running properly; seeing if the logic power rail is 5V, that there is activity on the address pins (pins 9..25 for A0..A15) and clock (pin 37 for E) being 1/4th of 4Mhz (I think).

[Mick B]

First of all, I don't own an 8116A.

However, looking at the schematics, I see that the LEDs and 7-seg. display are controlled by two 74LS374s and ICM7218A (U24, U23, U22 on page 3C).  If I saw LEDs stuck-on, I would check if the CPU (U1 on page 3A) was reset and is running properly; seeing if the logic power rail is 5V, that there is activity on the address pins (pins 9..25 for A0..A15) and clock (pin 37 for E) being 1/4th of 4Mhz (I think).

Pqass, Thank you for responding, you have no idea how grateful I am to have you and others take an interest in my problem. After removing all 6 Shaper output amp power transistors, and confirming I had both 24v rails, I hooked up all 4 outputs of U7 to my scope. and all 4 did exactly what they were supposed to, no PDD. I reassembled it and turned it on, it performed the LED/Switch check and then "Displayed E-41" as expected.
 I was ecstatic. I'm thinking maybe because when there is a current draw on R35 or R36 it kills the 24v regulators but all the other voltage are still present keeping U1 in a constant state of shut this thing down. and sending a continuous PDD signal instead of a 1ms pulse, keeping the reset signal from being enabled in U27 locking everything up. this only a guess as I am a noob, and I'm sure there is a more logical explanation,for this madness.
Now I need to come up with some heatsinks for the power transistors on i'm dead in the water.

[Mick B]

 I was going to say I have hour's but it's days into this HP8116A, I have got it to the point of "E41" which is to be expected, I have removed all 6 power transistors, if someone knows how to remove the heatsinks without breaking them,  please tell me. Hopefully someone bought one of these FG's for parts. I also posted this in the sell & buy forum here.
 
-$ I'm in need of the heatsinks from an HP piece of HP test equipment or anything that has a heatsink like the picture. Thanks.

[MarkL]

You can search around for "TO5 heatsink" and try to locate something usable.

It looks like a direct replacement would be Wakefield 260-4TH5B, which has a BeO (beryllium oxide) insulator.  They are not readily available due to the high toxicity of BeO, but you could probably substitute 260-4TH5E, which is the epoxy insulated version.

Mouser has the latter in stock, but you're not going to like the price at $26/ea.

There's a listing on ebay for the studded epoxy version, if it will fit in the 8116A.  Or maybe you can remove the stud and drill and tap it to fit like the original.  $5/ea.

  https://www.ebay.com/itm/372640777684

You didn't say what broke on the original, but I'm guessing one of the sides separated from the BeO ring.  I've had this happen on several of these style heatsinks (on a Tek FG504) and I just reattached the two pieces with thermal epoxy.  Don't break or grind the BeO ring.  It's a hazmat; read up on it.

[factory]

The same heatsink is also used in the 8112A pulse gen, it's listed as "260-4TH5B-special thread", the manual fails to give any warning about BeO, some older oscilloscopes from Tek & HP also use BeO insulators.

It shouldn't be hard to find a dead 8112/8116 for parts, the hybrids are often missing or faulty.

David

[Wallace Gasiewicz]

I ran into a similar problem with a TO 5 power transistor in a DC to DC converter in a military radio.  The part was an unobtanium TO5. I substituted a TO 66. I had to cut off the "ears" however. The transistors in a TO 5 package are also made in TO66 packages with identical characteristics but higher dissipation. I think this is because of the better heat sinking. The transistors have different numbers though. I did this because I fortunately found a bag full of the flanged transistors at the radio shop. I have quite a few.
I am writing this in the hope that you may be able to mount the TO 66 package, The "ears" do require some extra room. In my application I could not fit them into the spot. Also the "ears" could not touch because of their polarity. 
If you can mount a flanged transistor, it will probably dissipate the heat more readily. 
I hope I got the numbers of the TO packages correct.
I recently repaired two HP 8165 "signal source" I do not think it was worth my time.

[Mick B]

Thanks Mark, Great information!...BUT your scared me, until I check and TO-5 & TO-39 are compatible, because the only transistors I could find were only in a TO-39 pkg.
How many I need will depend on how many I kill trying to get them off. I download the Wakefield-vette catalog I'm going to call them in the morning and see if I can find them somewhere cheaper than Mouser (WOW$ cha-ching) you could buy a working 8116 for that kind of money. Most likely I will end buying the ones off e-bay after my call. and seeing if I can make them work. Or maybe something else will come up.

[Mick B]

Thanks Dave, Hopefully I can find the 260-4TH5E, and the screws for those @ 4-40 UNC common I believe. Checking out e-bay for a dead one tonight. There is still the problem of getting them off of the old transistors without demolishing them.

[Mick B]

Thanks for the suggestions, I hope it doesn't come to that. This is definitely worth my time; the experience is priceless. This has been a joyfully frustrating adventure. as I'm teaching myself electronics with a little help here and there. And it was free.