HP 54645A Scope Problem

[UtahMikey]

Hi Folks,
I just bought a HP 54645A Scope on eBay (Oct 2022).  It said it powered up, and looked good condition.   One reason I bought it is it has the IEEE-488 plug in option and had free shipping.
So it was $150 US + tax.  I have the version D with Logic Analyzer I bought back in the 90's New $5200 !!!  What a nice scope you can get now days for that!
After taking a sticker off the unit Looks NEW!   It came from a university per a sticker.  Inside not one bit of dust or dirt!  I was so happy!   It turns on and works fine, lots of CRT intensity.
Passes all self tests and calibration.   .....BUT,  DC balance is bad.   I set the trace levels, and the move when I change the GAIN?   I ground out the Battery in the Memory Chip, and replaced it
with an external 3.3V CR2032 battery.   I checked voltages per user manual, they are within spec.  Still have the problem.   I am using some almost unreadable schematics found on this site.
I think they are for another model.   These scopes are easy to use and have deep memory.   My favorite scope!
For now I am Stuck!  Too bad, this scope looks absolutely new!  It probably had this problem for a long time, so never got used.  No Scratches or dirt.
Would like any suggestions and front end schematic and component locations?
Thanks,  Mike
I am a Super Senior now, but was an  EE engineer years ago.

[ingalopez]

Hi UtahMikey.
Sorry not much to say about that fine piece of equipment!.
Looks like it need a recalibration. Have you looked for service manual?

Good luck

[edavid]

The CLIP is available here: http://archive.org/7/items/HP54645NCLIPImgs

Note the whole zip file is 400MB.

Unfortunately the schematic sheets are chopped up, but they are readable.

[xavier60]

I have a small collection of  used HP 54645A  also.
One of them developed what seems like a similar problem in one channel.
After needlessly replacing the PGRM preamp IC, I found a resistor which had drifted high.
All I remember at the moment is that the resistor was green, one of a pair in each channel. 

[xavier60]

This is a snap I took at the time.

[UtahMikey]

Thanks Guys for the Help, I need a lot of help.   Sorry for taking so long getting to thank you all.
The one good link to the documents was a great start.  I have spent several days printing out the documentation.. I made a schematic by taping together all the pages together.  HP did a crappy job withe the schematics.  They did not take time to annotate the designators in any order.    It has taken me hours to find some of the parts.  You might find resistor 101 in parallel with R411, or such nonsense.  SO it is Hard to find part locations.  I have tried measuring for those precision resistors that were mentioned.   I adjusted the vert position to 3V and measured the voltage division of the 4K and 20K resistors.  They are very close per divider calculations and with a 4 1/2 digit Fluke.  I would have to remove some parts to get rid of some parallel loading.  I still don't know exactly how the gain and offsets get processed?  I DID find a point where Gain AND Position affect the voltage???
AND (Big Hint?) when I move the position control to ZERO (front screed annotation pops up with 0.00Volts), THE GAIN Knob leaves the trace at Zero.  In OTHER words, the channels work if you Zero Out the vertical position.  As soon as you offset the Vert Position,  changing Gain makes the trace jump?   So the Scope works fine with Vertical positions set to zero, both channels on top of each other!
Maybe some one with a better understanding can figure something out from this CLUE?  I have added a exterior CR2032 battery and the scope calibrates fine.  The protection switch works too.
Thanks, Still a struggling 77 Yr. Old Senior! 
Mike, WA6ISP Ham in 1961!

[Miek]

This sounds like it might be a feature rather than a bug... in the menus under Utility -> System Config -> Expand Vertical you can select either Center or Ground. If it's set to Center then the view of the trace will zoom relative to the center of the screen when adjusting Volts/Div. If it's set to ground then it'll zoom relative to the ground reference.

[xavier60]

This sounds like it might be a feature rather than a bug... in the menus under Utility -> System Config -> Expand Vertical you can select either Center or Ground. If it's set to Center then the view of the trace will zoom relative to the center of the screen when adjusting Volts/Div. If it's set to ground then it'll zoom relative to the ground reference.

I have just had a play with that feature on one of mine. It allows up to 20VDC to be zeroed out while allowing the channel to be set to max sensitivity.
It's going to be very useful to me.