EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: aroby on June 27, 2014, 03:29:20 am
-
I just picked one of these up for $45 on eBay, which I thought was a decent price for this equipment. It doesn't work (!), so I plan on fixing it or at the least learning something. I know there are a couple of other threads on the forum about these meters - they may be relevant, since the symptoms on this one are display related.
Before I start on my fixit adventure, I thought I'd ask here to see if anyone has any quick thoughts on the problem. The symptoms are that when the meter is powered on, the display continually flashes and the relay just behind the display keeps cycling. I assume that this is due to a mechanical problem, since moving the meter around either stops the flashing (display turns off) or changes what is happening. But it never changes to be a stable display. The meter also has the AC option which changes the behavior slightly when disconnected.
I'm going to check out the solder joints on the transformer, since that seems to be a common cause. The display seems to work fine - it's as bright as a I would expect. I don't see anything obvious on the board, although one of EEPROMs has it's sticker missing. This meter apparently has a military background, but cosmetically looks fine, just a bit grubby.
If anyone has ideas on places to start for this, I'd be grateful for any insight. I'm very much an amateur on this stuff, but given the manual states that the "8840a is designed to be easily maintained and repaired", I thought I'd give it a go.
Anthony
-
Definitely check for bad solder joints, particularly on heavy parts.
-
After checking for cracked joints etcand loose connectors, then make sure all power rails are correct and no excessive ripple (indicating dried electrolytics).
Also the PCB needs to be kept clean, any signs that others have been in there before you !. A good clean with IPA and tooth brush (nice new one please !! ;) ).
Certainly sounds mechanical in nature, with some luck thats all there is wrong with it.
Run the self test routines when you find and fix the mechanical failure.
good luck.
-
The meter also has the AC option which changes the behavior slightly when disconnected.
Let me know if you want to sell that AC option board. I've just repaired an 8840A that I plan to give to a friend as a gift. However, it doesn't have the Option-09 board, and I'd really like to give him a meter that has A/C measurements.
-
Oh, and probably you already checked this - but if not, check the voltage setting switches on the back panel. My 8840A was labeled as "won't stay on", and (one of) the issue turned out to be the idiots had the thing set for 100V mains and were connecting it to 120V. All of the DC regulators in the 8840A have thermal overload protection, so if you operate them in an overvolted input condition, you get exactly this symptom - meter appears to shut itself off. It's not inconceivable that this might cause a symptom like you are seeing.
Also, I assume you saw that Fluke has a crappy scan of the manual up on their web site.
-
Let me know if you want to sell that AC option board. I've just repaired an 8840A that I plan to give to a friend as a gift. However, it doesn't have the Option-09 board, and I'd really like to give him a meter that has A/C measurements.
Assuming I can get this working, I will be keeping the option board. If I fail, I'll let you know
Anthony
-
Also the PCB needs to be kept clean, any signs that others have been in there before you !. A good clean with IPA and tooth brush (nice new one please !! ;) ).
The board does look very clean, not even any noticeable dust on it. I thought it was amusing in the manual to see that the boards "can be cleaned with water-based commercial cleaning systems, such as dishwashers" and then left to air dry for 2 days. Not sure I'll be trying that!
The only thing I have seen that indicates that it has previously been opened up is a missing sticker on one of the EEPROMs. Maybe that has been upgraded at some time in the past. It looks a bit odd - the EEPROM is piggybacked on top of another larger IC. The calibration stickers were in place over the screws, and it seems to have been calibrated last in about 1990. The main board is from 1983 and the AC option board has 1984 printed on it.
Anthony
-
Oh, and probably you already checked this - but if not, check the voltage setting switches on the back panel. My 8840A was labeled as "won't stay on", and (one of) the issue turned out to be the idiots had the thing set for 100V mains and were connecting it to 120V. All of the DC regulators in the 8840A have thermal overload protection, so if you operate them in an overvolted input condition, you get exactly this symptom - meter appears to shut itself off. It's not inconceivable that this might cause a symptom like you are seeing.
Also, I assume you saw that Fluke has a crappy scan of the manual up on their web site.
I will double check that when I get home tonight. I did see the switches and they looked in the correct places, however 100V and 120V are just a small click apart, so that does warrant another look.
I did find the manual - what a great example of comprehensive documentation.
Anthony
-
I did find the manual - what a great example of comprehensive documentation.
Yes, it is fantastic. I wish the scan was better, though. I have trouble reading some things. I ordered a used paper copy off eBay, but ended up paying close to what I paid for the meter for it...
-
Also the PCB needs to be kept clean, any signs that others have been in there before you !. A good clean with IPA and tooth brush (nice new one please !! ;) ).
The board does look very clean, not even any noticeable dust on it. I thought it was amusing in the manual to see that the boards "can be cleaned with water-based commercial cleaning systems, such as dishwashers" and then left to air dry for 2 days. Not sure I'll be trying that!
The only thing I have seen that indicates that it has previously been opened up is a missing sticker on one of the EEPROMs. Maybe that has been upgraded at some time in the past. It looks a bit odd - the EEPROM is piggybacked on top of another larger IC. The calibration stickers were in place over the screws, and it seems to have been calibrated last in about 1990. The main board is from 1983 and the AC option board has 1984 printed on it.
Anthony
The larger IC is the Z8 CPU. The EPROM (not EEPROM) on top is the program ROM.
-
Sounds like a problem with U202.
-
Sounds like a problem with U202.
Can you elaborate on your thinking here? From the schematics, U202 appears to be a Z8611 microcontroller.
Thanks
Anthony
-
A while back, I bought 10 of these as parts units. After swapping and replacing some components, I was able to to get 6 fully functioning meters.
One of those I units I recall had the same problem and if I recall correctly, it was the U202 that was one of the issues as that initiates everything.
-
A while back, I bought 10 of these as parts units. After swapping and replacing some components, I was able to to get 6 fully functioning meters.
One of those I units I recall had the same problem and if I recall correctly, it was the U202 that was one of the issues as that initiates everything.
That's good to know. Hopefully not the case here, but I will see.
Anthony
-
A while back, I bought 10 of these as parts units. After swapping and replacing some components, I was able to to get 6 fully functioning meters.
One of those I units I recall had the same problem and if I recall correctly, it was the U202 that was one of the issues as that initiates everything.
That's good to know. Hopefully not the case here, but I will see.
Anthony
U202 is the microprocessor I think. This outputs control signals that, if they aren't correct, can cause all sorts of behavior. There is quite a bit in the manual's troubleshooting section on that.
But before I went there, I would spend some time reading the maintenance/troubleshooting section in the manual. And probably the first thing I would do, and which is what the manual suggests, is to measure the various DC voltages with another meter or an oscilloscope. There are clearly labeled test points on the PCB where you can measure these, and you don't need to remove anything other than the back bracket and the outer shell - 3 screws - to get to that.
-
My second 8840A arrived (remember, I'm trying to scrape together enough parts from 2 to make one fully-optioned, working meter). Although the eBay listing said it was failing the A/C tests (tests 1, 2, and 3), when I ran the self test it passed. I pulled the A/C board out of this meter and stuck it in the one that has the IEEE-488 option, and everything seems to be working good. I don't have the ability to test all the A/C ranges, but I did test from 200mv up to 10V using my function generator, and the A/C readings look correct within 3 significant digits.
-
This is turning out to be a bit anti-climatic. I started tonight by pressing on various parts of the main board and found if I put pressure on the microprocessor with the piggy back EPROM that the meter kicked into action. All the self tests passed (although I haven't tried them with the AC option installed) and I measured the voltage using my cheap Chinese voltage reference and that all passed (so either the Chinese reference is pretty good, or I have a massive coincidence on 4 voltages!). I have the front panel off right now to clean things up, so I'll see if it's all still operational when I put it back together.
I'm rather pleased given the price I paid!
Anthony
-
DO NOT RUB THOSE KEY PADS ON THE DISPLAY BOARD WITH ALCOHOL! IT WILL REMOVE THE CONDUCTIVE COATING!
-
DO NOT RUB THOSE KEY PADS ON THE DISPLAY BOARD WITH ALCOHOL! IT WILL REMOVE THE CONDUCTIVE COATING!
Thanks for the warning. I just used a moist (water) Q tip which worked fine.
Unlike many of the photos I have seen, the plastic on mine is a khaki green, vs. a grey. I assume this is because it was used by the military?
Anthony
-
If pressing makes the microprocessor work then you either have a broken socket or dirty pins, or a cracked joint on the main board. Clean and check, you will probably have to unsolder the socket on the mainboard if it is faulty, though I have bodged a new socket onto the old one before by cutting it off and soldering a new turned pin one onto the stubs, no way was I going to try to undo all the soldered wire wrap spaghetti under the original board, especially as it was about 10 layers thick there. peeled the plastic housing off and left the socket pins bare for easy access to solder.
-
This is turning out to be a bit anti-climatic.
LOL, that's exactly how I am feeling. I bought two "broken" meters, and was looking forward to the fun of investigating the root cause. But the only thing I've found "wrong" with them seems to be that someone had the voltage setting on the back wrong.
-
If pressing makes the microprocessor work then you either have a broken socket or dirty pins, or a cracked joint on the main board. Clean and check, you will probably have to unsolder the socket on the mainboard if it is faulty, though I have bodged a new socket onto the old one before by cutting it off and soldering a new turned pin one onto the stubs, no way was I going to try to undo all the soldered wire wrap spaghetti under the original board, especially as it was about 10 layers thick there. peeled the plastic housing off and left the socket pins bare for easy access to solder.
Now it's working I think I'll leave it alone. I couldn't see anything underneath the board that obviously looked wrong, and I don't have an IC extractor to pull the chips out so I can clean the pins and socket. This piggy back arrangement seems a bit fragile.
Anthony
-
Oh, those stamped pin sockets..... They likely have one broken tab on a pin somewhere, if it gives you more trouble remove the ceramic chip ( not the rom from the microprocessor but the whole 2 part package, the socket in the microprocessor is a very reliable unit and rarely fails unless it has had hundreds of insert extract cycles on it) and replace the socket on the board with a turned pin 40 pin socket, soldered to the board top and bottom. Inspect the board when the socket is out, and likely you will find one broken pin in unsoldering it.
-
The 8840 notoriously gets bad solder connections involving the power transformer. Don't bother checking them visually, as the solder connections may appear fine and still be bad.
Resolder them all and see if that fixes your problem.
-
Oh, those stamped pin sockets..... They likely have one broken tab on a pin somewhere, if it gives you more trouble remove the ceramic chip ( not the rom from the microprocessor but the whole 2 part package, the socket in the microprocessor is a very reliable unit and rarely fails unless it has had hundreds of insert extract cycles on it) and replace the socket on the board with a turned pin 40 pin socket, soldered to the board top and bottom. Inspect the board when the socket is out, and likely you will find one broken pin in unsoldering it.
Thanks for the tips!
-
If pressing makes the microprocessor work then you either have a broken socket or dirty pins, or a cracked joint on the main board. Clean and check, you will probably have to unsolder the socket on the mainboard if it is faulty, though I have bodged a new socket onto the old one before by cutting it off and soldering a new turned pin one onto the stubs, no way was I going to try to undo all the soldered wire wrap spaghetti under the original board, especially as it was about 10 layers thick there. peeled the plastic housing off and left the socket pins bare for easy access to solder.
Now it's working I think I'll leave it alone. I couldn't see anything underneath the board that obviously looked wrong, and I don't have an IC extractor to pull the chips out so I can clean the pins and socket. This piggy back arrangement seems a bit fragile.
Anthony
Of the 10 8840a's and a/af's that I have worked on, that's a first to see that eprom piggy backing the U202 chip. All of mine have that eprom in the socket labeled as U222, and, it appears that your main board has a chip in that socket also.
Wonder why?
-
Either they ran out of the mask rom chips, and looking at remaining product life and the cost of having a minimum run of 100k parts fabbed from the original mask set ( minimum order and likely the fab would have a 6 month lead time to make the basic wafers as well so delivery would be in a year or more), they decided to simply do the last few thousands using a production item of the development kits microprocessors, as they were likely still available, and even if they were 20 times the price of the mask version they would work out cheaper than buying in the mask rom for under 10k units.
-
Judging by the date codes (at least a year earlier than my own non-piggyback model), I think this a very early unit rather than a 'last run' unit.
-
Same reason then, the one year or so lead time for the mask product, and then they likely went through a few firmware revisions developing it before deciding to commit to a final mask layout.
-
Did not dawn on me till a few minutes ago that it looks they were just stacking memory there.
-
Judging by the date codes (at least a year earlier than my own non-piggyback model), I think this a very early unit rather than a 'last run' unit.
On the back of the main board it says (c) 1983 and board Rev D.
Anthony
-
Hi Anthony :)
In the picture I can see that CPU's firmware is V 2.5
In my 8840 there is a damaged processor. ( Z6211 with ROM mask)
I have built CPU's adapter on U8840 with external EPROM.
Unfortunately, I have found only v2.3 firmware on the net, but it doesn't work correctly
I need a version 2.5.
Could you make the dump of EPROM for CPU ( U202) ?
-
If you want to message me your email adx, I'll send you the 2.5 bin file. Not sure if it's allowed to be attached here.
-
If you want to message me your email adx, I'll send you the 2.5 bin file. Not sure if it's allowed to be attached here.
Thanks for the file, but unfortunately it is not a file for CPU U202 :(
I need a version 2.5 for U202 ( not U222 )
-
In case anyone is interested, the attached photo shows how the two meters turned out after minor repairs, cleaning up the stickers, and repainting the main chassis cover.
-
If you want to message me your email adx, I'll send you the 2.5 bin file. Not sure if it's allowed to be attached here.
Thanks for the file, but unfortunately it is not a file for CPU U202 :(
I need a version 2.5 for U202 ( not U222 )
My bad, I thought you needed the bin from 222.
Let me see if I can rip it from u202. Not sure if I can with the G540 programmer.
-
In case anyone is interested, the attached photo shows how the two meters turned out after minor repairs, cleaning up the stickers, and repainting the main chassis cover.
Nice. Mine is a muddy green color. I guess it was ex-Army.
-
Hi Anthony :)
In the picture I can see that CPU's firmware is V 2.5
In my 8840 there is a damaged processor. ( Z6211 with ROM mask)
I have built CPU's adapter on U8840 with external EPROM.
Unfortunately, I have found only v2.3 firmware on the net, but it doesn't work correctly
I need a version 2.5.
Could you make the dump of EPROM for CPU ( U202) ?
Sorry, I only just saw this. I don't have anything with which to dump the EPROM. Not sure that I want to pull this out of my unit - it looks like it would be a nightmare to get it plugged back in. If you have any suggestions how to do the dump without pulling the EPROM out, let me know and I'll have a go.
Anthony
-
Hi Guys,
Not sure if anyone is still watching this thread,
But I had a similar intermittent problem, due to a dry joint inside the piggy back socket.
I've fixed it and posted it here: https://www.eevblog.com/forum/testgear/fluke-8840a-intermittent-relay-clicking-resetting/ (https://www.eevblog.com/forum/testgear/fluke-8840a-intermittent-relay-clicking-resetting/)
Thanks to everyone for pointing me in the right direction!
-
The larger IC is the Z8 CPU. The EPROM (not EEPROM) on top is the program ROM.
Wow, they used a Piggyback Z8??? I built a few boards with it in the 80s and it was the neatest thing ever. Never saw it in any other products though.