HP 54501A Restoration and Calibration

[xrunner]

I should be receiving an HP 54501A oscilloscope this week for repair / restoration. I've wanted one for a while because we had them in my lab circa 1990's. I know there are other threads on it but I'm going to post a lot more pics of the insides / parts for other people that may benefit from them, as well as do some repairs such as replacing and socketing the NVRAM IC. I was also informed that one of the rear BNC jacks has been "dented" and so I'll be looking at how that can be fixed. I know it is a panel jack that goes to a push-on RF connector but I don't yet know what kind to order if I choose to make that complete new jumper assembly.

I also want to go through the full calibration including adjustments on the main board, but I do not have the "extender" that the manual says to use, so I'll be looking at how that might be done without it (the extender just moves the PSU out of the way).


Plus - I like GREEN !!! 

The actual scope coming is attached below.

[xrunner]

So it arrived this afternoon. Seems to work OK for the most part. It will complete all the self-test & self-cal routines but it will not store them, but I was aware of that problem and the part is on-order.

It also has a busted BNC jack on the back, I will be looking into a fix for that. Also found an original document in the case bag, looks like it was purchased by North Carolina State University.

[Wuerstchenhund]

Nice scope! We had them at work when they were new and I also had a 54510A as my own scope, and I love the single knob interface.

BTW, I'd strongly recommend to replace the Schaffner mains filter if that hasn't been done in the last 10 years as they tend to blow up and spread some corrosive gunk all over the mainboard. They are still made, usually not very expensive, and the new versions should be more reliable. Just make sure you don't buy "new - old stock" as that wouldn't really solve your problem.

[xrunner]

BTW, I'd strongly recommend to replace the Schaffner mains filter if that hasn't been done in the last 10 years as they tend to blow up and spread some corrosive gunk all over the mainboard. They are still made, usually not very expensive, and the new versions should be more reliable. Just make sure you don't buy "new - old stock" as that wouldn't really solve your problem.

Yea thanks - I've heard about that and I'm going to attempt to find a replacement that fits in the existing cut-out.

Here's a few more screen shots for today. I'm just posting these for anyone browsing the web looking for info that may need to see them.

So, I know that the NVRAM is bad (really the battery inside), so what happens when you run the self-test? On the back of the unit is a red switch that sets the calibration to either protected or unprotected. If that switch is in the unprotected mode, you see that it passes all the self tests (with a bad NVRAM). If it is in the protected setting, the self-tests do not pass the NVRAM, D/A converter, or A/D converter (with a bad NVRAM). And finally, you can see the screen at the completion of the self-calibration. However, once you turn it off it will all be lost with a bad NVRAM; but it does complete properly. 

Only other thing I've done is to pull out the power supply to see if it can be jury-rigged in such a manner as to allow one to tweak the tweakable adjustments underneath it without using the "extender cable" the manual talks about - I think it will be possible.

More to follow ... 

[Smith]

Better check the caps in the power supply too, they tend to leak. Maybe its better to replace them, before they go leak (and they will). Its verry common with HP scopes of that era. Mine had the same problem.

The mains filter is quite expensive, in europe at least. I did replace it, but it cost like 60 euro's. You could probavly find them cheaper on ebay.

Only thing I have to do on my 54502A is replace the fan, it's so loud! I already have a new EMB PAPST laying around.

[Wuerstchenhund]

So, I know that the NVRAM is bad (really the battery inside)

As to the bad NVRAM, I'd replace it with a FRAM IC instead another NVRAM module. Works the same and doesn't need a battery.

The mains filter is quite expensive, in europe at least. I did replace it, but it cost like 60 euro's. You could probavly find them cheaper on ebay.

I paid around £20 (around 25EUR) for a brand new one a couple of years ago.

[TAMHAN]

Redo the PSU _ASAP_. The Boeschert Cartridge WILL go "puber!" on you in the worst of times - happened with my 53310A MDO.

[rf-loop]

So, I know that the NVRAM is bad (really the battery inside)

As to the bad NVRAM, I'd replace it with a FRAM IC instead another NVRAM module. Works the same and doesn't need a battery.

The mains filter is quite expensive, in europe at least. I did replace it, but it cost like 60 euro's. You could probavly find them cheaper on ebay.

I paid around £20 (around 25EUR) for a brand new one a couple of years ago.

Do you know HP54xxx   memory handling /CE signal pass for using FRAM

FM16W08:

Users who are modifying existing designs to use FRAM
should examine the memory controller for
timing compatibility of address and control pins.
Each memory access must be qualified with a low
transition of /CE. In many cases, this is the only
change required. An example of the signal
relationships is shown in Figure 2 below. Also shown
is a common SRAM signal relationship that will not
work for the FM16W08.
The reason for /CE to strobe for each address is twofold:
it latches the new address and creates the
necessary precharge period while /CE is high.

I do not know exactly how HP bus do. So, I ask do you know? Or is this important?


Changing NVRAM is bullett proof operation. Ok there is battery inside chip  (or  if use NVRAM socket and low stby current RAM, or if use Maxim NVRAM controller chip), external battery).  Last time I have changed NVRAM's to HP545xx scopes they have been around or even over 20years without fail (But version HP used is bit better than usual). If NVRAM quality is same 20 year extra life for this scope is lot of.

[Wuerstchenhund]

So, I know that the NVRAM is bad (really the battery inside)

As to the bad NVRAM, I'd replace it with a FRAM IC instead another NVRAM module. Works the same and doesn't need a battery.

Do you know HP54xxx   memory handling /CE signal pass for using FRAM

If I remember right it's just a drop-in replacement for the 54500 Series scopes.

However, the later 54500 scopes (i.e. the ones with multiple knobs like the 54542A) have a clock (they use the clock in the Dallas module), which is lost when using FRAM. That's not a problem for the 54510A though, which doesn't have a clock.

[xrunner]

Thanks for all the responses. I have already ordered the NVRAM but will look into the other option. It won't be a big deal to swap the NVRAM after I mod the board with a socket for that IC. I also take note of the power supply capacitor replacement issue. I will probably take it apart today and clean the very fine black dirt out of it and take note of the caps I need.

One task I have is to repair the rear DC Calibrator output BNC, because it is so damaged I can't connect a BNC. I'm not sure why they used what is an RF-capable connection from the rear of the unit to the board for a DC signal. What I need to know is, what is the connector that they used to connect the cable to the board? Is it an SMB? I don't have any here to compare it to. Please advise. The BNC and SMB (if that's what it is) are crimp on connectors. I could buy those and re-produce the same jumper, for the sake of keeping it all original, but in theory, couldn't I just use a solder cup BNC chassis connector in place of the BNC crimp?

Thanks for the input!

[xrunner]

Proceeded to clean the power supply, crt driver board, and associated aluminum. Washed off with alcohol and dried. Re-assembled and tested to make sure I didn't jostle a component loose, still works. Completed the power supply adjustment and the CRT set-up.

Note the CRT made by Matsushita and CRT driver made by Hitachi.

Now need to pull the main board and remove the NVRAM ... 

[Smith]

You can just replace the BNC with some random BNC connector, it's nothing fancy. Be carefull not to shorten the original cable to the connector though, it's quite short if I remember correctly. I would recommend fastening the NVRAM module to the socket with a tie-wrap.

BTW, give it at least have an hour before selftest/selfcalibration. It will fail when the system havent had time to properly warm-up.

[xrunner]

You can just replace the BNC with some random BNC connector, it's nothing fancy. Be carefull not to shorten the original cable to the connector though, it's quite short if I remember correctly.

I may try to reproduce them, if not yea, I don't see what special needs are required for a DC signal there.

I would recommend fastening the NVRAM module to the socket with a tie-wrap.

I will take that under advisement.

BTW, give it at least have an hour before selftest/selfcalibration. It will fail when the system havent had time to properly warm-up.

Yep.

By the way - what's up with those front BNC nuts? I had to remove them all with long-nose pliers because I couldn't get any socket on them and an open end wrench kept slipping off. They appear to be 5/8" but the socket won't fit because of the space against the case, and a regular open end wrench seems to slip off. The flats on the nut are very small. Yuck.

Anyway - I'm pulling the main board now.

[AF6LJ]

This is off to a good start, nice score by the way.
I am looking forward to seeing your progress.

[xrunner]

This is off to a good start, nice score by the way.
I am looking forward to seeing your progress.

Thanks. Next up was the bad NVRAM IC. I removed the board and removed the old NVRAM. I then installed a 28 pin dip socket. Also cleaned the board as good as I could - it sure is a pretty board.

The whole thing is now basically stripped down. I want to remove as much of the black dust that I can, that dust that is attracted to HV areas. I'll probably just dip the case in the bathtub. I also plan to clean the CRT as much as possible.

[xrunner]

Gave the old CRT a thorough cleaning. Made a nice pile of dirty alcohol prep pads (I love those things). HV really attracts a lot of that fine black dust.

You can see the burn-in on the CRT - the ghosts of measurements past. I wonder what kinds of signals this scope has seen over the last 27 years - maybe if we look hard enough ...

But the cleaning went well and it looks almost like new.

[AF6LJ]

Wow!
I always turn down the intensity when the scope is not in use.
Especially on gear that I used daily at work.

[xrunner]

Wow!
I always turn down the intensity when the scope is not in use.
Especially on gear that I used daily at work.

Yep, good idea. No telling how many years it was left on all day, but really the screen seems bright and just fine.

Just cleaning the keypad assembly today. I'm posting these pictures for anyone in the future that may need to see the assembly, for whatever reason.

[xrunner]

Put the scope back together, but still waiting on the NVRAM. I wanted to test it but when I removed the old NVRAM two pins broke off the chip - pin 14 (GND) and pin 17 (DQ5). Just to see what the instrument would display, I powered it up without the NVRAM chip. You can see what the screen displays in the attachment. I'm posting that in case someone in the future needs to see what a non-functioning NVRAM chip causes (not one that simply has a bad battery).

What I did was to tack on two wires to the chip, where the small stubs were where the pins were - I was able to solder to them. I then ran the GND wire to ground, and the DQ5 wire to the socket under the chip, and inserted the chip in the socket. When I powered it up, it worked as it should (at least the same as before). 

So now I can move on to the board calibration, which doesn't require the NVRAM chip to be able to store data.

[xrunner]

I can't buy a new Schaffner filter like the old one. The only new part numbers available is this

FN 393-2.5-05-11

The original part is

FN 393-2.5-05-12

No supplier I went to had the original part number, they only sell the new part number (-11).

I eventually called the NA office of Schaffner, and the engineer said they changed the design in 1993. He gave me a copy of the change notice (please find attached). Here's the deal - the new one (the -11 model) would work when using it on 120 VAC, because those connections made by the removable switch block are the same (it connects windings in series). BUT - the 220 VAC connections are NOT the same as the old model. That's what the engineer alerted me to. Take a look at the PDF attached.

So, the question I want advice on is - do I buy the new model knowing that if somebody used it on 220 VAC something bad might happen, or do I get a used -12 model, knowing the connections are exactly the same but might have a lot of hours on it? What should I do?

 

[tautech]

By the way - what's up with those front BNC nuts? I had to remove them all with long-nose pliers because I couldn't get any socket on them and an open end wrench kept slipping off. They appear to be 5/8" but the socket won't fit because of the space against the case, and a regular open end wrench seems to slip off. The flats on the nut are very small. Yuck.

LOL
Machine a socket down or hunt for a quality thin wall model from say Snap-on or Stahlwille.

I can't buy a new Schaffner filter like the old one. The only new part numbers available is this

FN 393-2.5-05-11

The original part is

FN 393-2.5-05-12

No supplier I went to had the original part number, they only sell the new part number (-11).

I eventually called the NA office of Schaffner, and the engineer said they changed the design in 1993. He gave me a copy of the change notice (please find attached). Here's the deal - the new one (the -11 model) would work when using it on 120 VAC, because those connections made by the removable switch block are the same (it connects windings in series). BUT - the 220 VAC connections are NOT the same as the old model. That's what the engineer alerted me to. Take a look at the PDF attached.

So, the question I want advice on is - do I buy the new model knowing that if somebody used it on 220 VAC something bad might happen, or do I get a used -12 model, knowing the connections are exactly the same but might have a lot of hours on it? What should I do?

 

Google is your friend and there's some 230V units on eBay:
http://www.ebay.co.uk/itm/FN9222R-16-06-Schaffner-16A-250VAC-Mains-IEC-Filter-Socket-Panel-Mount-/182117819268
http://www.ebay.co.uk/itm/Bulgin-PS620-3A-Mains-Filter-IEC-Input-Socket-250VAC-3A-40C-OM387F-/391212205302

It is an IEC socket type right?

[xrunner]

It is an IEC socket type right?

Well yea but I'd rather get the new part and mod the socket somehow to prevent the 220 option from working than put in something that doesn't look original. I'm looking at some used original parts on Ebay also, there is a date code on the label at the bottom.

I'm also wondering if the switch block from an old model would fit in a new one and may lend itself to modification.

[tautech]

It is an IEC socket type right?

Well yea but I'd rather get the new part and mod the socket somehow to prevent the 220 option from working than put in something that doesn't look original. I'm looking at some used original parts on Ebay also, there is a date code on the label at the bottom.

I'm also wondering if the switch block from an old model would fit in a new one and may lend itself to modification.

OK but show us an image, is it switchable too? 

[xrunner]

OK but show us an image, is it switchable too? 

Yea it's switchable, give me a little bit and I'll post some pictures. I also want to hear from Wuerstchenund because he said they were still available. I sent him a PM and I'd like to know if he knows of any new old stock somewhere becasue that original part is not available as new.

Keep in mind - I know of nothing wrong with the part that's in the unit right now. I was just advised that they blow out by Wuerstchenund. Maybe it's not a big deal, how many blow out, what percent blow out? Any given part can fail. I mean the thing's been around for 27 years and the part is still working just fine and I can get the actual part (used) on Ebay right now ... 

How about this - buy a new part and epoxy the switch block in the 120 VAC position ... 

[xrunner]

I'm going to buy the new part and install it, and keep the original part. If I ever sell it I'll give the buyer the option of keeping it as-is, (it can't be used on 220), or I'll send them (or re-install) the original with the notice that some of the older parts may fail, and it can be used on 220 if the block is inserted that way.

I'm not really sure why they even changed the way the wiring works for 220, per the little wiring block. The industry must have decided power supplies were going to be designed differently at some point in time. The change notice is dated 1993.

Done deal.

Anyway, the USPS is out for delivery with my NVRAM chip so I've got bigger fish to fry for now. Thanks for the advice.

[xrunner]

The NVRAM is now installed and I went through the self-cal and logic trigger delay cal to make sure it would save the results. All passed. It now comes up with no error messages and the waveforms are displayed properly with respect to their amplitude and offsets, and the triggering works correctly.

Yippee! 

Next up is the manual adjustments on the main board, after which I'll do another self-cal

[xrunner]

Proceeding with the manual system board calibration. Section 4-14 Track and Hold Offset.

This is what the signals at TP1 and TP2 look like when the adjustments are complete using R274 and R74.

Refer to Service Manual pg. 4-9

[xrunner]

I found an archive picture on the internet of the lab console and equipment where I worked - the AFEWES lab (Air Force Electronic Warfare Evaluation Simulator). There is an HP 54501A there in situ.

Can you see the 54501A in the rack?

[xrunner]

Several of the adjustments cannot be accessed without the extender cable set because the PSU is in the way, specifically, some trimmer caps near the front of the scope. I was able to adjust many of the pots toward the back with the PSU in-place (because my hand is small), but the ones up front just cannot be done.

I'm going to re-arrange the PSU so that the scope will still operate and so that I can get to the caps. I found that it can be operated when connected in the geometry shown in the picture

It looks a little scary but I'll be OK. 

[AF6LJ]

Whatever you do BE SAFE

[xrunner]

Whatever you do BE SAFE

Thanks Sue - no problem. It looks hokey but it's very safe. 

[AF6LJ]

Whatever you do BE SAFE

Thanks Sue - no problem. It looks hokey but it's very safe. 

Doesn't look that bad to me, I have done similar.

[xrunner]

Well I'm still alive. 

I had to move the PSU as shown (because I did not have the "extender cable" for this scope) to complete the Attenuator Compensation Adjustments - Section 4-15. It states to use a 100 Hz square wave, 300 mV pp, into a 50 ohm termination at the scope. However, using the given parameters of the square wave they stated, I found that adjusting the compensation caps produced absolutely Zero observable change in the response to the square wave on any of the four channels. I decided to change the freq. of the square wave to 10 kHz and that allowed me to see and make the adjustments. Don't know if the manual has a mistake, but it seems so.

So having completed that part of the cal, I put the PSU back the way it's supposed to be.

[rf-loop]

Well I'm still alive. 

I had to move the PSU as shown (because I did not have the "extender cable" for this scope) to complete the Attenuator Compensation Adjustments - Section 4-15. It states to use a 100 Hz square wave, 300 mV pp, into a 50 ohm termination at the scope. However, using the given parameters of the square wave they stated, I found that adjusting the compensation caps produced absolutely Zero observable change in the response to the square wave on any of the four channels. I decided to change the freq. of the square wave to 10 kHz and that allowed me to see and make the adjustments. Don't know if the manual has a mistake, but it seems so.

So having completed that part of the cal, I put the PSU back the way it's supposed to be.

And what is your square wave rise time and other signal charateristics. Adjustments need do with equipments what are listed in service manual or comparable. In this case HP8116A programmable pulse fenerator. If not available, then need look carefully HP8116A specifications and use something what is as near as possible (because in service manual, text is valid for named equipments)  There is quaranteed rise time and overshoot etc.
Every step in HP service manual is just perfectly right.  This is not made in china.

Also it is good practice to follow paragraph 4 order starting from 1, every step in order. This is how it is designed to do. (if know exatly what is doing then also know where can do different and how to execute own procedure)

 If on your board things do not go as in manual, there is two reasons, device is somehow failed or you do something wrong. One problem where things do not go as in manual may be wrong instruments used for job.

[xrunner]

And what is your square wave rise time and other signal charateristics. Adjustments need do with equipments what are listed in service manual or comparable. In this case HP8116A programmable pulse fenerator. If not available, then need look carefully HP8116A specifications and use something what is as near as possible (because in service manual, text is valid for named equipments)  There is quaranteed rise time and overshoot etc.
Every step in HP service manual is just perfectly right.  This is not made in china.

Uh - yea, I know it's not made in China, it's made in the U.S.A. 

I do not have the specified pulse generator, so I'm using the best I have - a Rigol DG1022 which has the following specifications for the square wave -

Rise/Fall Time < 20 ns (10% to 90%), (Typical, 1kHz, 1 VPP)
Overshoot < 7.5% (Typical, 1kHz 1Vpp)

It may not be fast enough rise time, but I'm not going to go out and buy a specific pulse generator just for one alignment on one scope. I'm dong a restoration the best can do with what I've got at the moment.

Besides, I was able to do a decent job by just using a higher frequency square wave which clearly showed me the adjustment changes. If the adjustment for the 100 Hz square wave (using the generator specified in the manual) were to have been in a different position than it is now (meaning the position I adjusted it to for the 10 kHz F), then the higher frequency square wave would have looked terrible. So you see, the adjustment now has to be very, very close to what it should be anyway, or the higher frequency would not look flat. There is not a different adjustment compensation capacitor for different frequencies - only one. It can only be in one position when you are done. Do you understand that?

Also it is good practice to follow paragraph 4 order starting from 1, every step in order. This is how it is designed to do. (if know exatly what is doing then also know where can do different and how to execute own procedure)

Yea, yea, I did follow it in order ... Give me a little credit LOL.

If on your board things do not go as in manual, there is two reasons, device is somehow failed or you do something wrong. One problem where things do not go as in manual may be wrong instruments used for job.

Manuals have errors from time to time, I'm not talking about HP specifically now but in general. But I want you to look at an attachment I scanned in from the manual for my HP 5334B, and read the text right under ABOUT THIS SUPPLEMENT and tell me what it says there ... remember this is an HP manual. 

I've seen errors in manuals before. So to claim it doesn't happen is just factually incorrect. Things can go wrong because of three reasons - device is somehow failed, you do something wrong, or there is an error in the manual procedure.

And as I said, I'm not going to buy the exact generator or a new generator that has the exact characteristics of the one specified in the service manual just for this one specific alignment (if that's what the problem is). If I get a better one (waveform generator) someday I'll go back and see if what you say is true, but for now it's as good as it's going to get, which is pretty damn good.

There are several things I need to do now. I need to finish the alignment checks, the only one left is the Oscillator Output Check. Then I will be making two new cable assemblies for the Ac and DC calibrator outputs on the back. Even though only one is bad, I'm just going to make two new ones so they match.

[rf-loop]

I know.  I have lot of HP/Agilent Tektronix manuals etc errata and correction sheets. Some of them also very severe.

Also I know how to repair calibrate and adjust HP545xx scopes. I have done these more than "lot". Just boring work for salary. In some storage I think I have still couple of these (severe failed) for recycling.

Usually (mostly) these models do not need readjust exept if example mobo is changed. Mostly they pass specs for cal without adjustment. Main principle for adjustment is that first need be sure that something need really adjust. But, this particular case is so easy that it can do also nearly how ever. But if do not think just this case, I have seen it so many times over years that with wrong equipments someones have made "adjustments" and things have gone only more bad than before adjustment. (I do not mean you do it so)  This scope is easy piece of  bisquit because it is only 100MHz BW and there is only very extremely limited things what can adjust at all.

[xrunner]

Usually (mostly) these models do not need readjust exept if example mobo is changed. Mostly they pass specs for cal without adjustment. Main principle for adjustment is that first need be sure that something need really adjust.

Yea but I like to adjust things - even if they may not need it. 

I paid for this ticket and I want the whole roller coaster ride. This is my entertainment don't you see? 

[Wuerstchenhund]

I also want to hear from Wuerstchenund because he said they were still available. I sent him a PM and I'd like to know if he knows of any new old stock somewhere becasue that original part is not available as new.

Sorry for only replying now but I was down with a cold.

When I said that replacement for that Schaffner mains filter is available I did not mean the original part, which is no longer made (the whole line has been redesigned, also because of the tendency to explode). But Schaffner has other models that fit the cutout and allow 120V/240V switchover. I can't remember what the exact model was I used in my 54510A but I remember that I had to modify the mains voltage switching logic for the new module slightly. I decided against hard-wiring the mains voltage to 240V, which would have created some potential issues for people that owned the scope after me.

In any case, Schaffner should be able to tell you what the suitable replacement models are.

Keep in mind - I know of nothing wrong with the part that's in the unit right now. I was just advised that they blow out by Wuerstchenund. Maybe it's not a big deal, how many blow out, what percent blow out? Any given part can fail. I mean the thing's been around for 27 years and the part is still working just fine and I can get the actual part (used) on Ebay right now ... 

The question is not *if* your old filter will explode, the only question is *when*. All these old Schaffner filters will blow up eventually, as due to their age the caps in them become dry, heat up, and due some design issue this is pretty much guranteed to end up in a more or less violent disintegration. There are no external signs telling you if your filter is going to explode tomorrow or in a year, and no warnings before it happens. The fact that your scope's filter is still working is probably down to being operated at 120V mains, as the speed of aging is also dependent on the mains voltage (so filter operated at 240V will age faster to some extend). But that doesn't mean that your filter is fine, it has passed its best-by date a very long time ago, like every single filter of that series that was made.

There's absolutely no point in doing any restorative work on the scope while leaving the original filter in place. When it pops it will blow a large load of dark, corrosive gunk on your nicely restored scope, which in the best case is just a hell of a pain to clean up and in the worst case will damage your scope's PCB beyond repair.

This should also explain why looking for "new old stock" would be a silly thing to do, because no matter if used or not, this "new old stock" will have degraded as has every other filter of that series, and you're essentially just replace one time bomb for another.

[rf-loop]

Usually (mostly) these models do not need readjust exept if example mobo is changed. Mostly they pass specs for cal without adjustment. Main principle for adjustment is that first need be sure that something need really adjust.

Yea but I like to adjust things - even if they may not need it. 

I paid for this ticket and I want the whole roller coaster ride. This is my entertainment don't you see? 

Oh, in this case all can understand and yes it may be fun but also useful. Same here when I do things in hobby side... curious and experimental. Still one Collins Receiver is waiting if I like start adjust it - there is totally out of order this bit complex mechanical variable IF tuning system.

[xrunner]

Sorry for only replying now but I was down with a cold.

No problem.

When I said that replacement for that Schaffner mains filter is available I did not mean the original part, which is no longer made (the whole line has been redesigned, also because of the tendency to explode). But Schaffner has other models that fit the cutout and allow 120V/240V switchover. I can't remember what the exact model was I used in my 54510A but I remember that I had to modify the mains voltage switching logic for the new module slightly. I decided against hard-wiring the mains voltage to 240V, which would have created some potential issues for people that owned the scope after me.

In any case, Schaffner should be able to tell you what the suitable replacement models are.

Yes, as I learned (I called Schaffner USA), and I'll attach the revision again for you if you're interested, the new model which fits the cutout exactly and has the same connector, was re-designed. It switches internally differently for 220V than the old model. I have no idea why this change had to be made. But - the new design (-11 see attachment) is no different in the 120V switch block position, which means I could install it and use it just fine here.

So, what I'd like to do is buy a brand new model - the FN 393-2.5-05-11, and mark out or black out the lettering on the reversible switch block so there is no 220v option indicated to the user, and also label the rear so as to indicate it is only for 120V. As you can see from the second attachment, even if I put in a different model filter and re-wired the internal wiring, there is still lettering on the back that indicates 120V/230V in two places. I mean, unless I scrape that off also, I can't prevent a person from doing something dumb and connecting it to 220V, if it ever left the US. So do you think it's acceptable to replace with the new design and black out the lettering on the reversible block so that 220V is not indicated? Is that acceptable?

The question is not *if* your old filter will explode, the only question is *when*. All these old Schaffner filters will blow up eventually, as due to their age the caps in them become dry, heat up, and due some design issue this is pretty much guranteed to end up in a more or less violent disintegration ...

Yes I agree with all of that statement ^^^

[Wuerstchenhund]

So, what I'd like to do is buy a brand new model - the FN 393-2.5-05-11,

Probably no bad idea. I think that's the model I bought when I replaced the one in my old 54510A.

and mark out or black out the lettering on the reversible switch block so there is no 220v option indicated to the user, and also label the rear so as to indicate it is only for 120V. As you can see from the second attachment, even if I put in a different model filter and re-wired the internal wiring, there is still lettering on the back that indicates 120V/230V in two places. I mean, unless I scrape that off also, I can't prevent a person from doing something dumb and connecting it to 220V, if it ever left the US. So do you think it's acceptable to replace with the new design and black out the lettering on the reversible block so that 220V is not indicated? Is that acceptable?

Frankly, I'd do it properly and keep the device 120V/240V switchable. Aside from that disabling the voltage switching is more of a hack job (in which case it would be easier and look more professional to just buy a FN390 Series filter without the voltage switch instead of black-felting a switchable filter), disabling the voltage switch facility could well have other implications. Fact is that the 54500A Series has been specc'd by the mfg for 240V compatibility, and it's not unreasonable for people to rely on that spec. So if you disable the 240V mode, and someone buys your scope (i.e. as a replacement for an existing ATE), powers it up at a 240V mains line and the PSU blows up (potentially causing further damage beyond the scope), you might well end up being liable, and I wouldn't count on that simply blackening/scraping out 230V markings would cover your ass. Chances of this happening are probably slim but it's enough if that ever happens once with the scope you sold, which has some potential to ruin your day.

My recommendation is to do the work properly. Especially since we're talking about mains voltages here, which aren't exactly insignificant.

[xrunner]

Frankly, I'd do it properly and keep the device 120V/240V switchable. ...
My recommendation is to do the work properly. Especially since we're talking about mains voltages here, which aren't exactly insignificant.

I really appreciate your help Wuerstchenhund - but how? Did you look at the change notice from Schaffner? I'm going to post the relevant schematics one above the other in this post.

They changed the way the wiring is switched inside the thing, and you know you cannot get into it to modify it. Look at the schematic. In the old one (the bottom one) and the new one (top) the 120V wiring is essentially the same as each other - the hot 120V goes into the PSU, out and right back in again, and then out back to the neutral.

They changed the way the 220V is switched though. In the old one, that "middle" loop path is simply opened and the 220V is simply across the two input terminals to the PSU. In the new way, it ain't like that. You can see the difference in the top schematic.

I do not see any way to change what they did internally on the outside of the filter. If you say buy a new filter, so it won't blow out my nice restoration, then I agree and it will live happily ever after at 120V. But you also say it's best to make it still work at 220V. How? If the person that gets it from me changes the little block so they think it's going to work at 220V, then how is it possible to keep that functionality? Like I said, what they did cannot be accessed - it's buried inside a potted case. And I see no way to reverse that on the outside to keep the same 220V option with this PSU. There is no way to "unconnect" what the switch block connects inside the filter. There is no wiring change you can make outside to reverse what happens inside when the block is set to 220V.

Does anyone else see the problem here? 

I have another idea but I want to first exhaust this issue.

[Messtechniker]

What about a big red honkin' label saying "For 120 V mains only!"
as a last resort on the unit itself on two or three places to get
you out of any liability.

Yours - Messtechniker.

[xrunner]

What about a big red honkin' label saying "For 120 V mains only!"
as a last resort on the unit itself on two or three places to get
you out of any liability.

Well, I can't have my cake and eat it too right?

It's an old piece of kit and the original part is not available brand new - they are all used old stock. It's impossible in any practical sense to buy a new one and cut it open to change the way it works. So I have two choices -

• Keep the original part that allows it to work on 220v but risk blowing black snot all over it
• Buy the brand new better part which switches the wiring differently for 220V but works on 120V.

But here's what I would like to do, to both be safe and allow the new part to be used. Keep in mind the thing would have to be sold by someone someday to a country that used 220V (I won't be the one that does that though). What I propose is to buy the new part. I will then epoxy the two top middle insertion slots in the back of the opening where the switch block goes in to prevent it from being inserted for 220V operation. As you see in the pic, the two middle prongs are what do the wiring change between 120V and 220V. They are only one one side of the block. The block is flipped over to change the wiring.

So if someone even tried to change the block, it won't go into the slot when it's turned over. It can only be inserted one way after I'm done. In the manual it states -

"Before connecting the instrument to an ac power source, ensure that the line selector module is installed for the correct voltage"

So with my epoxy mod, this will be impossible. They will have to re-install an original used part to do that, and thus the whole problem will start over for them, but at that point it's their decision to risk the filter blowing out. Really, isn't what I'm doing the safer approach? I mean, if it's common knowledge that the filter blows up, what I'm doing is the safer choice anyway.

I will also label the back nicely so that it states it's for 120V only, covering up the existing labeling on the back, and completely remove the lettering for the 220V selection on the switch block.

What else can I do given only two options?

I sure wish I knew why the stupid part was re-designed. It must have been to comply with some sort of new industry requirement.


Please find attached the pic of the switch block that is used to change the mains voltage input.

[Nuno_pt]

xrunner, it sounds like an save approach.

[tautech]

What else can I do given only two options?

Keep a mod book for the unit that stays with it and add a sticker inside and outside explaining such.
Original and modded schematic too, then only an idiot will FUBAR it.

[xrunner]

Keep a mod book for the unit that stays with it and add a sticker inside and outside explaining such.
Original and modded schematic too, then only an idiot will FUBAR it.

I'm working on that today. I have the new Schaffner filter on order.

After I get done, only a complete fool, who would not have paid attention to the original instructions and notices anyway, will be able to damage it. A fool such as I'm talking about, would simply plug a new original scope like this straight into 220V when it was set for 120V, paying no attention to the warnings on the back.

Nothing can be totally idiot-proofed. Not even HP was able to achieve such a goal. 

[tautech]

Another thought xrunner.....I've come across manuals that in the first page or so there's been a card overlay stuck in place containing country specific information, often it's the first thing you see when opening the manual.

[xrunner]

Another thought xrunner.....I've come across manuals that in the first page or so there's been a card overlay stuck in place containing country specific information, often it's the first thing you see when opening the manual.

Another thing I can incorporate is an actual warning notice from HP regarding the filter, which came with this scope in the pouch that attaches to the top of the case. You can read the attachment; I will add to the documentation, to be on the safe side, that

• The filter was changed to the modern version to avoid any problems with the issue in the notice
• The filter was changed to avoid a known defect in the original filters that cause them to catastrophically fail

However, the modern part does not allow this model scope to work with 220V, so that option has been disabled.

Yay.

Boy, this project has been interesting so far.

[tautech]

 

3rd sentence 3rd paragraph.....HP will replace the mains filter free of charge.

Wonder if they'll honour that? 

[xrunner]

3rd sentence 3rd paragraph.....HP will replace the mains filter free of charge.

Wonder if they'll honour that? 

Gee I dunno. Hey I know - I'll drop the confirmation card in the mail (see attachment) and see if it gets returned. Maybe there's a crew of engineers there that were never notified of the company change. If it doesn't come back, we only have a PO box to go on ... somebody needs to find out whose getting their mail. 

[xrunner]

Next up is making two replacement cables for the DC & AC calibrator connections. Only one is damaged but just so they look the same I'm going to make two new ones. Have the SMB parts and RG-174, waiting on the BNC parts.

[TAMHAN]

3rd sentence 3rd paragraph.....HP will replace the mains filter free of charge.

Wonder if they'll honour that? 

My memory is fading a bit now. But I can remember that Intel still honors the guarantee for the Pentium I according to the German c't newspaper...

[xrunner]

My memory is fading a bit now. But I can remember that Intel still honors the guarantee for the Pentium I according to the German c't newspaper...

Maybe I should try to send the cope back LOL. The safety notice states that if the ground connection doesn't pass the test to send it back. OK it's 26 years later, but since it's a safety issue and the part isn't available now, maybe they will send me a brand new Keysight scope! 

[xrunner]

Got the brand new Schaffner line filter today. There was still one thing about it I couldn't tell from any pictures. The new switch block that it comes with - as labelled - is wrong for the 120V position. The 220V position, is the correct one for the old way the PSU was wired, back in the day. 

So, I can easily make it work, but I will use the old switch block in the new filter. Welcome to restoring old equipment.

Also got the BNC connectors I need - so I should have all the parts now.

[tautech]

Got the brand new Schaffner line filter today. There was still one thing about it I couldn't tell from any pictures. The new switch block that it comes with - as labelled - is wrong for the 120V position. The 220V position, is the correct one for the old way the PSU was wired, back in the day. 

So, I can easily make it work, but I will use the old switch block in the new filter. Welcome to restoring old equipment.

Also got the BNC connectors I need - so I should have all the parts now.

Quadruple check that, sketch it out and be sure.
Is there no way the wiring can be reconfigured to make the switching correct?

[xrunner]

Quadruple check that, sketch it out and be sure.

I did, I octuple (is that a word) checked it, I just tested it with the scope and it works OK.  If that thing had blown my project I'd never have been able to face the forum again. 

Is there no way the wiring can be reconfigured to make the switching correct?

Nope. Not in any practical sense of the word. The problem lies inside the new design of the Schaffner filter. Something, for some reason, caused them to change the design back in 1993. For some reason an industry standard changed and they don't make these parts with wiring like the old ones. It's not possible to open it and change the wiring because it's all buried in potting compound. I wish there were a way. I've stared at that schematic for hours and there is just no way to undo what is done inside the thing with extra wiring outside. There is no way to undo a "short" with more wiring, something like that concept.

But, all is well. I will now drip some epoxy into two of the ports back inside where the switch block fits, to prevent it from inserted upside down, and call it done.

[xrunner]

Is there no way the wiring can be reconfigured to make the switching correct?

Well, it's going to be OK, and work as originally manufactured. Not being one to give up easily and look things over again, I took another look at exactly what these diagrams meant in terms of what's inside the filter and what the switch block does. It's a little confusing, because the diagram doesn't clearly show what part of the circuit is the switch block and what isn't. What changes are hardwired and what aren't.

Also, I had in my mind what the Schaffner engineer told me over the phone, that it would work on 120 but not 220 as compared to the old model. But, it turns out, if you use the old switch block in the new filter, it will make the changes that are required the same way as the old model. That's one thing the Schaffner guy didn't take into account, and probably wouldn't have realized, is that I have an old one with a different switch block that I could use in the new one.

Anyway - gah - I won't have to make any label changes to tell people now. It sometimes takes me a while, but I get to the end of the yellow brick road.

[xrunner]

Made up the new replacement cables that connect the DC and AC calibrator outputs to the rear panel. Installed new Schaffner filter. Back panel finished - good as new.

[xrunner]

It's done! 

Works well, for me it's a nice unit for my collection too.

Got my eye on another used piece of test equipment, may have another thread coming ... 

[tautech]

Nice.
Looks really sharp, thanks for sharing. 

[AF6LJ]

It's done! 

Works well, for me it's a nice unit for my collection too.

Got my eye on another used piece of test equipment, may have another thread coming ... 

This was a great thread, your scope came out really nice.
I am looking forward to your next installment.

[xrunner]

Thanks to you and everyone that made suggestions. That Schaffner filter was a pain! 

[rf-loop]

Nice work!

Also nice to see peoples want restore these old well designed and made beauties.

I hope also thät today chinese designers looks and learn how carefully these equipments user interfaces are designed.
There is difficult to find any bug or design error. UI ergonomy, useability is really designed by peoples who really know how equipments are used in real work.  (Of course it is old and performance is technically obsolete but this is other thing)

[xrunner]

Nice work!

Also nice to see peoples want restore these old well designed and made beauties.

Thank you.

I enjoy doing it, and I hope the threads I post (and others that are doing outstanding restorations) will be available for years to come, so that others wanting to do the same thing for these great vintage instruments will be helped in some manner. Things in the future will never be made the same way, and it's important to preserve the quality of some of these instruments.

And many thanks to the EEVBlog for hosting these threads!

[roberto]

Hi, gentlemen and thank for the support.

I bought one used HP 54501A and I LOVE IT but.

It gave me problems when I use it for a while, I had a bad smell (non electronic blowing components) and the oscilloscope cause differential relay action.

In my opinion is the power supply or something similar, do you think can be the Schaffner filter?

I'll change it immediately.

Tanks for support, cheers RA

[Tim88]

BTW, I'd strongly recommend to replace the Schaffner mains filter if that hasn't been done in the last 10 years as they tend to blow up and spread some corrosive gunk all over the mainboard. They are still made, usually not very expensive, and the new versions should be more reliable. Just make sure you don't buy "new - old stock" as that wouldn't really solve your problem.

Yea thanks - I've heard about that and I'm going to attempt to find a replacement that fits in the existing cut-out.

Here's a few more screen shots for today. I'm just posting these for anyone browsing the web looking for info that may need to see them.

So, I know that the NVRAM is bad (really the battery inside), so what happens when you run the self-test? On the back of the unit is a red switch that sets the calibration to either protected or unprotected. If that switch is in the unprotected mode, you see that it passes all the self tests (with a bad NVRAM). If it is in the protected setting, the self-tests do not pass the NVRAM, D/A converter, or A/D converter (with a bad NVRAM). And finally, you can see the screen at the completion of the self-calibration. However, once you turn it off it will all be lost with a bad NVRAM; but it does complete properly. 

Only other thing I've done is to pull out the power supply to see if it can be jury-rigged in such a manner as to allow one to tweak the tweakable adjustments underneath it without using the "extender cable" the manual talks about - I think it will be possible.

More to follow ... 

Man, sure am glad I stumbled across this post. I picked up a 54502A a while back that didn't want to power up. When probing for voltages, something sparked and it powered up. Was planning on eventually replacing it not knowing the darn thing can explode black gunk all over the place. Thanks!

[xrunner]

Man, sure am glad I stumbled across this post. I picked up a 54502A a while back that didn't want to power up. When probing for voltages, something sparked and it powered up. Was planning on eventually replacing it not knowing the darn thing can explode black gunk all over the place. Thanks!

Glad you found it. If you have any pics of your work post them here.

[midtex]

Hello there. New member here. I recently bought an HP 54501A for cheap, knowing that the NVRAM issue was what was causing the self-test failures. I replaced the Dallas chip with genuine item from Mouser. The unit passes all self tests and I have gone through several of the training exercises detailed in the Quick Start guide. Everything works exactly as it should - except......

The vertical axis seems to have a small amount of ripple in the signal. All voltages whether sine or square or DC all "wiggle" for lack of a better word. If I tell the scope to display real time voltage, the number is not steady, but jumping all around by several tenths. The background video - scales, menus, labels, etc are all rock steady, so this is not a CRT issue. It is only measured signals and it does this whether using the rear DC or AC reference ports, or when using an external signal generated by a function generator.

I was wondering if anyone has any suggestions of where to start? I'm guessing it is possibly a reference voltage that is not regulated properly? Would this likely be a power supply issue or main board? I do not have a schematic, but do have the service manual which does provide information on some calibration.

Thanks in advance for any advice!

[Tim88]

Mine had a bad power entry module that was bad, it is a common failure I believe. Its the module that the line cord plugs into. They don't make them any more but there is a replacement available. I am not at home so don't have my notes handy but there is info on it on the interweb.

[cjuried]

Hello all,

I have one of these units that passes all selftests except for a failed AD Converter. Any thoughts on suspects I should keep my eye out for? HAs this been a common issue with these units?

Thank you,

Chris