Author Topic: Is this scope repairable? (Read 22009 times)

Mediarocker · « **on:** October 15, 2012, 10:40:48 pm »

So I ended up finding a steal of a deal on eBay (uh oh... here we go.) It's an old HP Agilent 54503A 500MHz Digitizing Oscilloscope. I managed to snag it at a cost I could manage, and figured since it's a 500MHz scope, it'd be a beauty to have on the bench.

The seller listed it as in working condition.

Well after waiting a few weeks for it to get here, I finally got it today, and powered her up. Everything appeared to be fine... So then I decided to test it and put it to work on a PWM circuit. Nothing. Not even a flinch on the screen.. wary, I ended up running a self test:

So it failed some pretty important stuff. I'd imagine the D/A and A/D Converter is kinda important, as well as the NVRAM and Analogue Trigger.

So I'm assuming that is why I can't use the dang thing.

So I figured maybe, it just needs to recal, so I dipped into the menu and boom:

So wondering if it was an easy fix, I gave Agilent a call. They told me that the Cal Ram can be typically write protected through a switch, which may be inside or on the back of the scope. I don't have anything on the back of the scope, so I figure I'm going to have to pop the cover.

Here's the problem. I'm not going to open it, unless I know it can be fixed. I'm thinking I'm just going to send it back to the seller... It is DOA and not in working condition. The device powers up but is a useless paperweight to me in this condition...

Anyone care to enlighten me or lead me towards a possible solution. I really don't want to send it back if it can be repaired..

If I need to pop a few easy to get to IC's I figure I'll keep it. But if it's a PITA or too much work, she's going back. I could use that money on something else.

tom66 · « **Reply #1 on:** October 15, 2012, 10:52:23 pm »

Yeah, it's fixable.

Common issue with the 54xxx scopes.

However it involves removing the NVRAM module and replacing it, which involves desoldering it... Then you need to calibrate it (which is best done by a cal lab, so it's expensive.)

Personally I'd just send it back.

I had the HP 54501A, which is a nice scope, but the single knob interface can become a little difficult to adapt to. But the 54501A was my first real scope so I didn't need to adapt.

ablacon64 · « **Reply #2 on:** October 15, 2012, 10:59:30 pm »

I believe it can be repaired but without opening you'll never know if the repair will be worth or not. It's a risk. Things may be worse on the inside.

I understand you don't want to miss the opportunity because of the price, but have you give it a thought that such a good equipment would not have been repaired by its original owner if the fix was trivial?

That problem indicates that the unit was not good before it was sent to you, I mean, can that problem be triggered by a drop during transportation? I don't think so.

Mediarocker · « **Reply #3 on:** October 15, 2012, 11:12:45 pm »

Quote from: tom66 on October 15, 2012, 10:52:23 pm

Yeah, it's fixable.

Common issue with the 54xxx scopes.

However it involves removing the NVRAM module and replacing it, which involves desoldering it... Then you need to calibrate it (which is best done by a cal lab, so it's expensive.)

Personally I'd just send it back.

I had the HP 54501A, which is a nice scope, but the single knob interface can become a little difficult to adapt to. But the 54501A was my first real scope so I didn't need to adapt.

So the NVRAM has to do with the A/D D/A converters as well as the Analogue Triggers? Wow such a simple part can really mess up a scope

I'm willing to solder bits, it's no biggie to me. Just I wanna make sure that if I so chose to keep it that I'll be able to fix all the issues. I have a friend I can turn to for recal.

Quote from: ablacon64 on October 15, 2012, 10:59:30 pm

I believe it can be repaired but without opening you'll never know if the repair will be worth or not. It's a risk. Things may be worse on the inside.

I understand you don't want to miss the opportunity because of the price, but have you give it a thought that such a good equipment would not have been repaired by its original owner if the fix was trivial?

That problem indicates that the unit was not good before it was sent to you, I mean, can that problem be triggered by a drop during transportation? I don't think so.

I completely understand. I think the reason why he was selling it was he had no use for it and got it surplus or something.

Though I count this as a DOA. It was advertised as working. It is not. It mearly powers up and cycles through the menus.

ablacon64 · « **Reply #4 on:** October 15, 2012, 11:21:34 pm »

Quote from: Mediarocker on October 15, 2012, 11:12:45 pm

I completely understand. I think the reason why he was selling it was he had no use for it and got it surplus or something.

Though I count this as a DOA. It was advertised as working. It is not. It mearly powers up and cycles through the menus.

Maybe the seller is not lying to you, maybe he just don't know what he had to sell, anyways, it's not an easy fix. Maybe you could quote for the parts and the calibration, then you decide if it's worth it or not to keep it.

tom66 · « **Reply #5 on:** October 15, 2012, 11:51:42 pm »

Yeah, the NVRAM stores the cal factors, so you need to replace it and re-do the calibration.

I have a feeling it's possible to get away with no cal data on the 54501A, it'll just be less accurate (default cal settings.) But the 54503A probably refuses flat out to run at all.

The problem with the NVRAM is the battery in it dies after about 20 years.

free_electron · « **Reply #6 on:** October 16, 2012, 12:07:22 am »

swap the nvram ( Dallas DS1251 i believe , can be replaced with an ST one) these things are getting harder and harder to get a hold or.
worst case , dremel open the top of the sram ( its a thick 24 or 28 pin DIL package ) to expose the battery ( a coin cell 3 volt lithium ) , snip off the coin legs and hotwire a new one ine )

after that run the self-cal. ( the scope has a cal output in the back ) . follow the procedure in the service manual. its a matter of connecting a shor tbnc cable between the self ca output and the inputs in sequence. the scope will guide you through it.

after that : good as new .

Mediarocker · « **Reply #7 on:** October 16, 2012, 06:10:44 am »

So all I gotta do is resolder a new SRAM/NVRAM module and it fixes everything? Even if Channel 1 and 2 are wore out, I still have 2 other channels to use. Well hell I think it's well worth it. I think I found it. It says Dallas DS1251YW-120 Nonvolitile SRAM and is located directly under the CRT.

Big question though, it also failed the Analogue Trigger test... Will this fix that as well?

Gall · « **Reply #8 on:** October 16, 2012, 09:21:56 am »

Yes. Just a dead battery inside that black box. But should be replaced as a whole.

DS1235 is basically just a SRAM with 15 address lines (256K=32Kx8) that works at 5V. It has a built-in battery so that it continues to work even without external power. This is hard-to find part but it should be quite easy to replace it with a small board containing an ordinary SRAM chip, battery and two diodes.

nixxon · « **Reply #9 on:** October 16, 2012, 01:01:10 pm »

Inform the seller about the scope's status and include those screenshots.

The seller will probably have no interest in the product in return, as is.

Ask if the seller want's the scope back for a full refund, or if he/she will compensate for the item not being as described (bricked). You may then keep it as a not-so-expensive repair project. You should get at least 50% of your money back.

And - all will be happy

ablacon64 · « **Reply #10 on:** October 16, 2012, 01:14:18 pm »

That's an excellent idea. I've had a similar problem with a product bought on ebay and they sent me a full refund (as a matter of fact they gave me credit).

Lawsen · « **Reply #11 on:** October 16, 2012, 06:05:53 pm »

Yes, this HP 545xxA is very much repairable. I have a similar vintage 200 MSamples/s oscilloscope. NVRAM is replace able/use a Dremel tool or screw gently ply off the cover and add jumper wires to a small clock battery holder to power it. Desoldering the NVRAM clock chip is the only challenge. Becareful of the high voltages in the fly back transformer and green phosphor cathode ray tube display. I love this oscilloscope more than the Rigol. This one, the Philips electronics controlled analog oscilloscope, and Yokogawa DL1502 with the thermal printer on the top and amber orange phosphor cathode ray tube in the shape of a small Macintosh classic. Yes is slow at 200 MSa/s, but the software has no bugs, not need to wonder about the next firmware upgrade. The HP-54501A just works and the HP-54504A has an even better input amplifier up to 400 MHz. Without the calibration table, often the analog to digital for sampling and measurement and digital to analog converter for the display goes into error computer wise. Digital oscilloscope is a computer. HP made this in Loveland, Colorado, U.S.A. Sadly, there are no oscilloscope made in the U.S.A. Electronics jobs left the U.S.A. like clothing making mostly left the U.S.A., too. Rarely, the analog amplifiers are burned by over voltage from the previous owners and tech/engineer. Look for burn marks on the traces in the input circuit board. You can repair this digital oscilloscope and use it. Calibration is not very costly, just find an independent shop to do it. I would like to learn how to calibrate it, but I do not know how.

Mediarocker · « **Reply #12 on:** October 16, 2012, 09:57:15 pm »

Magic smoke was released at some point. Possibly the PSU was replaced:

Or it might be the result of what caused of this really dodgy cap replacement....

Board looks fine:

Switch was either replaced or touched up here:

And some other touch ups here:

tom66 · « **Reply #13 on:** October 16, 2012, 10:33:48 pm »

No, my 54501A was like that.

It's just dust and heat, no magic smoke leaked out of my scope, it was all new build quality (original HP cal sticker intact before I went to repair a single solder joint on the PSU.)

Mediarocker · « **Reply #14 on:** October 18, 2012, 05:52:05 am »

So I'm assuming this scope is worth keeping... Let's do this. I'll be contacting the sell to see if he can give me a partial refund... Let's see how this goes.

mrflibble · « **Reply #15 on:** October 18, 2012, 02:46:59 pm »

Quote from: tom66 on October 15, 2012, 11:51:42 pm

Yeah, the NVRAM stores the cal factors, so you need to replace it and re-do the calibration.

I have a feeling it's possible to get away with no cal data on the 54501A, it'll just be less accurate (default cal settings.) But the 54503A probably refuses flat out to run at all.

The problem with the NVRAM is the battery in it dies after about 20 years.

Mmmh, this thread reminds me... I've got a 54720D here, and was wondering if it has the same sort of issue.

Quote from: TFM

Mainframe calibration allows the oscilloscope to establish the calibration factors for each slot independent of the plug-ins. These factors are stored in the oscilloscope’s nonvolatile RAM.

A quick google tells me that there are people with the 54720D + nvram problems as well. I've been meaning to check this out before it's too late, but you know how it goes. "Well, it's working just fiiiiiine. I'll do that later". But later never comes. Since I do not own a 54717A cal plugin, I think I'd better make sure I do something before the nvram goes *poof*.

Anyone with experience on this one? I don't mind investing some time in this to make sure this thing keeps running for years to come. I would however mind messing up a currently working machine. So any pointers are appreciated.

Mediarocker · « **Reply #16 on:** October 18, 2012, 03:56:49 pm »

I pulled the back off and popped the Cal Ram Write protect switch into the unprotected position. I booted it up and BAM. The scope works fully. A little bit of noise, but that's probably the old Eleneco probe I got.

ablacon64 · « **Reply #17 on:** October 18, 2012, 04:46:45 pm »

Lucky you! It's good when everything ends well! Congratulations for the buy!

Mediarocker · « **Reply #18 on:** October 18, 2012, 05:44:40 pm »

Quote from: ablacon64 on October 18, 2012, 04:46:45 pm

Lucky you! It's good when everything ends well! Congratulations for the buy!

Well I still have to replace the DS1235, but at least I know that the scope will work perfectly when I do so.

MetraCollector · « **Reply #19 on:** October 18, 2012, 09:38:57 pm »

Hello !

I bought a HP 54502A scope during August this year for 70€. I had the same issue.

I replaced the NVRAM chip and it works well again. You can buy DS1230Y equivalent on eBay, this is the suitable one:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110894147524
Marked as "-120"

You can desolder it easily with only chinese sucker, vias have great diameter and it is not soldered using lead-free vomit.

Dirty marks of brown flux on the board are normal, they are from manufactory.

My topic is here:
https://www.eevblog.com/forum/product-reviews-photos-and-discussion/hp-54502a-nearly-vintage-digitizing-scope/

Good luck and congratulations to this instrument. It is some kind of attraction in my dormitory

ablacon64 · « **Reply #20 on:** October 18, 2012, 10:17:13 pm »

Great topic, excelent pictures! I wonder how we missed that....

Mediarocker · « **Reply #21 on:** October 19, 2012, 03:14:32 am »

Metra, Nice thread!

Yeah I cant believe I missed that. I should have probably searched a little

Anyway, I will be ordering that replacement ASAP.

How did you clean your board? I would like to clean all the dust and grime from my board. It's not really grimy, but the dust is on it pretty good.

MetraCollector · « **Reply #22 on:** October 19, 2012, 04:18:31 pm »

I used vacuum cleaner and brush.

I always use it.
Sometimes tissue. In difficult situations a little bit isopropanol.

Mediarocker · « **Reply #23 on:** October 28, 2012, 07:10:03 am »

I replaced the NVRAM and now I'm having this occur during calibration...

Appears to be an issue with the AC calibrator? I have no idea.

MetraCollector · « **Reply #24 on:** October 28, 2012, 10:51:16 am »

Try loading default calibration. But make sure you switched the switch on the back of the instrument.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Is this scope repairable? (Read 22009 times)

Share me