I have repaired several of these on my channel, I even have a source for replacement switches linked in some of them. Bad switches can confuse the processor and cause odd behavior.
@@IMSAIGuy The correct parts are SKEGACA010 ...they fit perfectly & their "t" top fits the plastic caps without gouging them up. The price has gone up since they were discontinued 4/2019 paid $29 for 100, 4/2021 paid $29 for 100.
It's possible that only the two digits work on the display and that the counter may be working. I would certainly check the display and re-seat the socketed ics.
I have one of these counter and when I got it I had little issue with the small ic on the right side looking from the front panel. If I can recall correctly I had to change the ca3083 and the SN74ls138n. Anyway look carefully that area besides the switches on the panel....good luck!
Wow---I didn't realize the good condition 1992 I got at a hamfest for $50 was such a good deal :-) I need to pop mine open and see if it has an hour meter. I don't think mine has GPIB though.
Two comments. I replace caps with a bulging top, even if it's just the plastic bulging, because I figure it's bulging because the capacitor has been outgassing in some fashion. Second, you might want to put some tape over that exposed UVEPROM lid before it drops a few bits from the window being exposed to daylight. BTW, I love big old crystal ovens. Like old wines, the longer they have aged the better (more stable,) they are. .
@@jonka1, I suspect the problem is that it's already too late. I'm pretty sure the seller sent out a dead unit on purpose in a 'bait and switch' in order to dump it off for whatever they could get BECAUSE the processor EPROM was dead, had lost data. They do that with age. If you watch the video closely you will see that BOTH the GPIB UV-EPROM AND the processor UV-EPROM on the mother bd are missing their UV window covers. Unless both stickers are sitting in the chassis they were intentionally removed by someone trying to fix the unit.) A search of the web indicates 3 likely suspects for problems with these: A bad 10 MHz oscillator connection. (They use the 10 Meg reference as the processor clock,) especially on start-up, before it can switch over to an external reference. Bad UVEPROM data (they can be reprogrammed with copied data from a working one. The stuck keys that were mentioned in the video, which can usually be cleaned. I fix a lot of old gear. :)
In a surplus sale of an Tektronix 2335 I found a similar hours meter put inside. I guess this was part of the military contract it originaly came from. After more then 5000 hours the meter was consumed and the little coper piece is now down the tube to the bottom.
You already know this because you have already numbered the episodes from 1 to 4, but here is my guess anyway. The frequency counter may be in better condition than it appears at the moment. It is possible that the display is defective, which would explain why only the last digit appears, the first nine digits remain off although they have the correct value.
I love my Racal Dana 1992....one is broken....the other one got lost in New Mexico....it worked great...maybe mine only counts to 512 MHz...so I will pay attention...
They're nice counters, I have the 1998 which strangely has never had button problems. I think maybe you have a display problem, looks like your only showing least significant digit
Hello IMSAI, your test videos are quite useful to me, could you please do a video on phase shifters found on ebay, they are wideband 1 to 2.2 ghz, i wonder how they work.
I took a counter from an old energy meter that I now use to count the seconds of a transformer I'm using with my ZVS driver so I have a rough idea how long I've been running the driver without it breaking. It's not very accurate but it shows me the rough amount of seconds which is enough for the application I'm using it for.
I've clocked my 1998 with 04B OCXO from an atomic clock (I have GPSDO, Rubidium and Cesium references), the 04B is a damned fine OCXO and was less that a few nanoseconds off, so close it wasn't even worth breaking the 15yr old cal stickers on the adjustments.
MC146805 is the external bus version of the 6805 microcontroller with a whopping 8K address space. I'm designing an SBC around this chip because I was gifted a bunch of parts from the local maker space and I'm using it as an excuse to learn SBC design, PCB layout, etc.
I am kicking myself for not buying the one of these I saw for about £150 recently in working order and good condition. I don't actually need it, but seeing all the coolness, I wish I had bought it now!
I always enjoy watching the master...
I worked for Racal-Vadic, (Modem Division), in the the mid to late '80's. Located in San Jose, CA USA
I have repaired several of these on my channel, I even have a source for replacement switches linked in some of them.
Bad switches can confuse the processor and cause odd behavior.
I watched your repair. I saw those links and the sellers had no more switched. spoiler alert, I do find a solution
@@IMSAIGuy I had a few videos, they have different links, one was to digikey.
@@IMSAIGuy The correct parts are SKEGACA010 ...they fit perfectly & their "t" top fits the plastic caps without gouging them up. The price has gone up since they were discontinued 4/2019 paid $29 for 100, 4/2021 paid $29 for 100.
It's possible that only the two digits work on the display and that the counter may be working.
I would certainly check the display and re-seat the socketed ics.
I have one of these counter and when I got it I had little issue with the small ic on the right side looking from the front panel. If I can recall correctly I had to change the ca3083 and the SN74ls138n. Anyway look carefully that area besides the switches on the panel....good luck!
Wow---I didn't realize the good condition 1992 I got at a hamfest for $50 was such a good deal :-) I need to pop mine open and see if it has an hour meter. I don't think mine has GPIB though.
Two comments.
I replace caps with a bulging top, even if it's just the plastic bulging, because I figure it's bulging because the capacitor has been outgassing in some fashion. Second, you might want to put some tape over that exposed UVEPROM lid before it drops a few bits from the window being exposed to daylight.
BTW, I love big old crystal ovens. Like old wines, the longer they have aged the better (more stable,) they are.
.
I saw the missing tape and wondered why he said nothing. Hopefully it's not already too late.
@@jonka1, I suspect the problem is that it's already too late. I'm pretty sure the seller sent out a dead unit on purpose in a 'bait and switch' in order to dump it off for whatever they could get BECAUSE the processor EPROM was dead, had lost data. They do that with age. If you watch the video closely you will see that BOTH the GPIB UV-EPROM AND the processor UV-EPROM on the mother bd are missing their UV window covers. Unless both stickers are sitting in the chassis they were intentionally removed by someone trying to fix the unit.)
A search of the web indicates 3 likely suspects for problems with these:
A bad 10 MHz oscillator connection. (They use the 10 Meg reference as the processor clock,) especially on start-up, before it can switch over to an external reference.
Bad UVEPROM data (they can be reprogrammed with copied data from a working one.
The stuck keys that were mentioned in the video, which can usually be cleaned.
I fix a lot of old gear. :)
The BNC for Channel C looks like it's a Picofused BNC.
yes, I finally figured that out
In a surplus sale of an Tektronix 2335 I found a similar hours meter put inside. I guess this was part of the military contract it originaly came from. After more then 5000 hours the meter was consumed and the little coper piece is now down the tube to the bottom.
the little meter operates by platting metal with currect
Those RC's will be a cool restore! Good work
You already know this because you have already numbered the episodes from 1 to 4, but here is my guess anyway.
The frequency counter may be in better condition than it appears at the moment. It is possible that the display is defective, which would explain why only the last digit appears, the first nine digits remain off although they have the correct value.
That's what I was thinking!
Yes that is at least a very good possibility.
Nice. Roll on part 2. :)
I love my Racal Dana 1992....one is broken....the other one got lost in New Mexico....it worked great...maybe mine only counts to 512 MHz...so I will pay attention...
They're nice counters, I have the 1998 which strangely has never had button problems. I think maybe you have a display problem, looks like your only showing least significant digit
Hello IMSAI, your test videos are quite useful to me, could you please do a video on phase shifters found on ebay, they are wideband 1 to 2.2 ghz, i wonder how they work.
I put it on my list
I took a counter from an old energy meter that I now use to count the seconds of a transformer I'm using with my ZVS driver so I have a rough idea how long I've been running the driver without it breaking. It's not very accurate but it shows me the rough amount of seconds which is enough for the application I'm using it for.
At Rockwell...we used an atomic clock...for the time base on lots of instruments...
I've clocked my 1998 with 04B OCXO from an atomic clock (I have GPSDO, Rubidium and Cesium references), the 04B is a damned fine OCXO and was less that a few nanoseconds off, so close it wasn't even worth breaking the 15yr old cal stickers on the adjustments.
Nice...no wrist strap on an insulated bench....you just arced and buzzed the 74LS....not that much CMOS....
MC146805 is the external bus version of the 6805 microcontroller with a whopping 8K address space.
I'm designing an SBC around this chip because I was gifted a bunch of parts from the local maker space and I'm using it as an excuse to learn SBC design, PCB layout, etc.
The "time meter" is so fun! I need to find one.
I am kicking myself for not buying the one of these I saw for about £150 recently in working order and good condition. I don't actually need it, but seeing all the coolness, I wish I had bought it now!
is a great device 👍
Holy cow! I just got one for $20!
score!
The input C connector comes off because there is a fuse in it.
I figured that out eventually. funny thing is, my other counter is the same.
8651....an Intel
10 MHz oven oscillator...?
It is pretty odd. 5MHz oven, doubled to 10MHz.
SMB
So you're that guy that's been "watching" for several weeks. lol, good info, I love ebay, I've wasted most of my retirement time on it.