So nice to see someone fault-finding in a rational manner, instead of just swapping bits in and out. Of course it helps if you have (a) the manuals and (b) a nice pile of spares for when you do find something broken. Looking forward to part two!
Hi Marc, I have two of the 5245L, both are working more or less fine (one has two boards with very marginal neon-decoder-plate that I will be replacing with an updated silicon replacement). One note regarding the oven heater control and the transformer for it: It is unfused on primary side, so if anything goes south and shorts the secondary, the transformer will be burnt. A fuse is very easily installed on one side of the sheet metal besides the transformer and provides safety.
@@HainjeDAF Well, this is not my invention, hp installed a replacement board in later revisions itself. Its schematic can be found in the service manual of those revisions, of course with hp part numbers. Replacement chip can be CD4514 4bit decoder and ULN2003 for nixie driving instead of lots of transistors. Look for change 11/1912A (for 1224 logic) or 13/1932A (for 1248 logic)
Great job, Marc. Nice to see somebody figuring one of these beasts out! I rescued a 5247M from the physics lab dumpster many years ago (back when it was still publicly accessible!). This (budget?) model is considerably rarer and difficult to find schematics for. After establishing the counter didn't budge, I ended up buying the service manual and started poking around inside. I soon realised that it, too, had previously been monkeyed with. Infact, it was in pretty bad shape overall. One of the Nixies had failed as it apparently took air. The 5254C 3 GHz (!) plugin it came with had completely seized up, and no amount of solvent I tried (from mild to aggressive) could free it up for more than a day or so. After several weeks I realised the project was gonna take more of my time -- and bench space -- than I had to spare, so gave up and sadly flogged the thing for parts. I really do regret that, but after seeing the difficulty even you had in sourcing replacements trannies, it's clear to me I wouldn't have had a chance in hell of getting it working again, or at least not up to spec.
I started my career as an electronic technician back in the mid-70s. I was an Air Force metrology technician and had used and repaired lots of those counters along with their plug-ins.
Just got a 5245L with 5255A down converter from my friend, and I'm more than jealous about your plugin piles! BTW the unique nixie driver circuit on this machine gives the tubes slightly fade-in fade-out transition, which is amazingly satisfying. Wish you good luck repairing your unit (I'm sure you are), and thank you for showing this beautiful piece of electronic engineering.
I’d say my electronics knowledge is at the level where I have a pretty good understanding of Ohm’s Law. Nevertheless I find these repair-a-then videos completely fascinating.
I also have one of those, repaired it a while ago, main fault was a replaced transistor, someone put a silicon transistor in it, and they have problems with more than 5V reverse on the B-E. With a germanium transistor, it came back to live. I totally loved the neon lamp based BCD decoders. Wonderful machine. It came from a local large radio station. One day I want to get some new nixies because the Russian replacements are ugly, they used the same digit for 5 and 2 which simply isn't pretta. The original nixies had several problems, so I had to replace them, but I'm not happy with the current state.
Wow! Maximum compressed data file with no bit memory-consumption. Maybe the search function depends on the quality of the glasses for checking the HP-FICHE! 😁🙃 Nice Video, as always. Cheers!
One day I will find to read the Horowitz Hill standing on the board behind me ;-) Helped me quite a lot for some minor OP Amp stuff I had to do during my chemistry/laser spectroscopy PhD
Marc, in case you have not see this: A Uniquely Complicated Nixie Tube Clock: HP 5245L Electronic Counter & ERA EasySynth++ ruclips.net/video/ri_uOuomBks/видео.html
The Q2 failure on board A35 failure described is a common one (no 10MHz); the most likely root cause is a faulty plugin (I have an example here of a 5254A which due to rough handling had a short on the plugin connector). And don't forget the ferrite bead, if you do (or HP did) or you will find a helpful parasitic on the Freq Reference output somewhere in the 85MHz region. I would then refer people to the classic article "Prevent Emitter Follower Oscillation" in Electronic design Vol 24#13
The most vulnerable part of a frequency counter is the input circuitry. It's very easy to push a to high of a voltage signal in. By the time you've realized what you've done, it's to late.
Crazy to me that there was a time when it made sense to pack a piece of test equipment full of transistors but build memory elements out of neon lamps, instead of using those very same transistors to simply make some discrete flip-flops...
I always assumed that it was a transistor breakdown voltage issue. The neon nixie's required more voltage than the transistors of the day could handle, so they had to come up with a very creative way of accomplishing it without the use of transistors.
The display boards aren't exactly BCD...they're something close but not quite. That is what I thought of at first when I tore down a Dymec 2401 a long time ago. I still have one or two of the display boards. BTW that meter actually had tubes in the power supply, including a couple of Nuvistors which I unfortunately no longer have. I did salvage an enormous amount of good germanium transistors out of it.
You are right. Actually there were three or four options you could get, one would give you "modern" 8421 BCD, other 2421, etc... I think most of the early HP instruments that output in BCD didn't have 8421 by default. We'll wrangle with the unusually weighted BCD counting in the next episode.
All switches on all equipment should be worked vigorously from time to time as a matter of practice. A friend of mine had a Kenwood TS 830 ham radio that would flutter in frequency. He took apart the radio, re-soldered connections, tightened screws and many other things, There's a button on the front marked "fix" that was only used if you had a crystal for one "fixed" frequency instead of the dial to tune the frequency. Obviously no one ever used this feature. Then fix was pushing the fix button repeatedly. It never had a problem again!
6:48 the signals on the back are more than convenient. Zero beat the 10MHz output against the 10MHz broadcast from WWV and bingo bango, you have calibration traceable to NIST.
Those things was cutting-edge electronic wizardry for that era. The best one could do with the available technology. Just curious to know how much does it cost in current dollars.
Hey Marc! Purely out of curiosity, do these HP machines and test equipment have functions, and things they do, that may or isn't found on modern equipment? Thank you Marc for your videos and time as always!
I only know about this sort of electrical signal stuff from my modular synth so maybe this is a completely normal thing but I thought it was interesting that the waveshape seemed to be different for each clock division. I would have expected it would be square waves throughout.
did anyone else notice at 17:29 (and again at 19:18) it was counting and showing 1khz? thinking maybe it just takes time for it to count and average the signal?
Very cool video. This goes way above my head.. but I've come across this hardware which is going to be scrapped. Would it be feasible to remove the nixie tubes to repurpose them for a clock? Is there anything I should know before attempting that?
HP and TEK were so awesome in their heydays. It's sad how in the 1990's, HP got typecast as a consumer electronics company, selling ink-jet printers at below-cost while price-gouging on the ink.. and mid-tier generic PC clones. If I had a time machine, I'd go to early- 1960's Silicon Valley and buy a home.
The instrument division and true HP heir is Keysight. They still make very beautiful, very expensive instruments, particularly top end RF and optical communications.
if anyone knows the music marc uses in these intros/outros, i would appreciate the information. ive tried to shazam it etc, but it comes up mua roi by le thu, there are several songs, but i cant find any song by that name that sounds even close
I used some Soviet instruments and frequency meters from the 80s. I am surprised that many Soviet devices have 90% similarity in design and internal structure. I think that the electronic industry of the USSR just decided to copy Western samples.
The biggest issue I have with this channel is Marc’s voice is so calming I fall asleep and have to watch the video multiple times 😅
So nice to see someone fault-finding in a rational manner, instead of just swapping bits in and out. Of course it helps if you have (a) the manuals and (b) a nice pile of spares for when you do find something broken. Looking forward to part two!
Not to mention c) the knowledge necessary to actually troubleshoot these circuits.
And d) the talent of explaining complex ancient electronics in an easy and fascinating way.
And e) having the right test instruments.
f) that is why Right to Repair is important
Swapping bits....good one
I think I have the metal can version of the transistor you need . I’m willing to donate it to the cause.
Hi Marc, I have two of the 5245L, both are working more or less fine (one has two boards with very marginal neon-decoder-plate that I will be replacing with an updated silicon replacement). One note regarding the oven heater control and the transformer for it: It is unfused on primary side, so if anything goes south and shorts the secondary, the transformer will be burnt. A fuse is very easily installed on one side of the sheet metal besides the transformer and provides safety.
Hi, I'm curious to your silicon solution
@@HainjeDAF Well, this is not my invention, hp installed a replacement board in later revisions itself. Its schematic can be found in the service manual of those revisions, of course with hp part numbers. Replacement chip can be CD4514 4bit decoder and ULN2003 for nixie driving instead of lots of transistors. Look for change 11/1912A (for 1224 logic) or 13/1932A (for 1248 logic)
I think Marc won't be satisfied until he's got one of everything in HP's catalog...
It's like Pokemon for adults. 😂
I could watch this for hours!
8:56 you know it’s a serious vintage repair-a-thon when you end up with two cans of deoxit on the bench
Great job, Marc. Nice to see somebody figuring one of these beasts out!
I rescued a 5247M from the physics lab dumpster many years ago (back when it was still publicly accessible!). This (budget?) model is considerably rarer and difficult to find schematics for. After establishing the counter didn't budge, I ended up buying the service manual and started poking around inside. I soon realised that it, too, had previously been monkeyed with. Infact, it was in pretty bad shape overall. One of the Nixies had failed as it apparently took air. The 5254C 3 GHz (!) plugin it came with had completely seized up, and no amount of solvent I tried (from mild to aggressive) could free it up for more than a day or so.
After several weeks I realised the project was gonna take more of my time -- and bench space -- than I had to spare, so gave up and sadly flogged the thing for parts. I really do regret that, but after seeing the difficulty even you had in sourcing replacements trannies, it's clear to me I wouldn't have had a chance in hell of getting it working again, or at least not up to spec.
I always enjoy a Marc Repair-a-thon video! Excellent work as always, sir!
Over 18GHz in the sixties, wow!
It is a happy day when we get a new video from you Mr Marc.
Reminds me of the Tandy catalogue, was always a joy to see what was “new for 1983” 😂
I started my career as an electronic technician back in the mid-70s. I was an Air Force metrology technician and had used and repaired lots of those counters along with their plug-ins.
Wow😮. Fault-finding and repair, in conjunction with comprehension. Such good material! 👍
Just got a 5245L with 5255A down converter from my friend, and I'm more than jealous about your plugin piles!
BTW the unique nixie driver circuit on this machine gives the tubes slightly fade-in fade-out transition, which is amazingly satisfying. Wish you good luck repairing your unit (I'm sure you are), and thank you for showing this beautiful piece of electronic engineering.
HP was/is one of the best instrument makers I've ever had the pleasure to work with. I remember tuning R-390s in the Navy with one of them (HP 8657B).
CuriousMarc + an ailing HP 5245L = quality content
I’d say my electronics knowledge is at the level where I have a pretty good understanding of Ohm’s Law. Nevertheless I find these repair-a-then videos completely fascinating.
What a series😮 I can watch that every day 🤩Looking forward to next part hopefully not so long 😃
I also have one of those, repaired it a while ago, main fault was a replaced transistor, someone put a silicon transistor in it, and they have problems with more than 5V reverse on the B-E. With a germanium transistor, it came back to live. I totally loved the neon lamp based BCD decoders. Wonderful machine. It came from a local large radio station. One day I want to get some new nixies because the Russian replacements are ugly, they used the same digit for 5 and 2 which simply isn't pretta. The original nixies had several problems, so I had to replace them, but I'm not happy with the current state.
Love the photo of the sat dish installation -- really beautiful location!
God, I love Nixie tubes!
one the best channel on you tube ,tnx for informative subjects
Analog devices will always be repairable!
My 5245L is still going with no issues after 60 years. Had some 2N708s and 709s fail in my 5360A computing counter.
Enjoying the repair a lot! What a magnificent unit :)
Wow! Maximum compressed data file with no bit memory-consumption. Maybe the search function depends on the quality of the glasses for checking the HP-FICHE! 😁🙃 Nice Video, as always. Cheers!
The rm503 was my very first scope I ever had. Got it when I was in high school. Loved that scope.
I used this in the Air Force in testing Radar LRU's (Line Replicable Units) from the F111D
One day I will find to read the Horowitz Hill standing on the board behind me ;-) Helped me quite a lot for some minor OP Amp stuff I had to do during my chemistry/laser spectroscopy PhD
04:40 I didn't expect to see Antoine in your basement.
Excited for the next episode, you all have the coolest gear to work on, and you're good at it :)
Cheers, using my 5248L still to this day..
I have this exact counter fully functional! Mine just doesn't have the accessory module. Mine still works and yes I use it!
WOW that's a nice counter
Back then the frequency spacing was .030 MHz
Amazing work, and most difficult in providing an explanation. Very well done, Marc.
Nice! Used many times in AF cal lab. Ca'ld too.
Most things made in 1960 were top-notch. Present company included.
Good work.
Marc, in case you have not see this:
A Uniquely Complicated Nixie Tube Clock: HP 5245L Electronic Counter & ERA EasySynth++
ruclips.net/video/ri_uOuomBks/видео.html
Nice to see you here! Nicely done!
The Q2 failure on board A35 failure described is a common one (no 10MHz); the most likely root cause is a faulty plugin (I have an example here of a 5254A which due to rough handling had a short on the plugin connector). And don't forget the ferrite bead, if you do (or HP did) or you will find a helpful parasitic on the Freq Reference output somewhere in the 85MHz region. I would then refer people to the classic article "Prevent Emitter Follower Oscillation" in Electronic design Vol 24#13
Can't wait for the next one 😁
Its like an adventure puzzle :-)
The most vulnerable part of a frequency counter is the input circuitry. It's very easy to push a to high of a voltage signal in. By the time you've realized what you've done, it's to late.
Crazy to me that there was a time when it made sense to pack a piece of test equipment full of transistors but build memory elements out of neon lamps, instead of using those very same transistors to simply make some discrete flip-flops...
I always assumed that it was a transistor breakdown voltage issue. The neon nixie's required more voltage than the transistors of the day could handle, so they had to come up with a very creative way of accomplishing it without the use of transistors.
The display boards aren't exactly BCD...they're something close but not quite. That is what I thought of at first when I tore down a Dymec 2401 a long time ago. I still have one or two of the display boards. BTW that meter actually had tubes in the power supply, including a couple of Nuvistors which I unfortunately no longer have. I did salvage an enormous amount of good germanium transistors out of it.
You are right. Actually there were three or four options you could get, one would give you "modern" 8421 BCD, other 2421, etc... I think most of the early HP instruments that output in BCD didn't have 8421 by default. We'll wrangle with the unusually weighted BCD counting in the next episode.
Hooray! Back to HP bench gear AND Apollo. If you gotta go Tek, the RM503 isn't a bad choice either.
Nice. Mine has similar problems due to the STORAGE toggle switch on the rear panel occasionally making a poor connection.
All switches on all equipment should be worked vigorously from time to time as a matter of practice. A friend of mine had a Kenwood TS 830 ham radio that would flutter in frequency. He took apart the radio, re-soldered connections, tightened screws and many other things, There's a button on the front marked "fix" that was only used if you had a crystal for one "fixed" frequency instead of the dial to tune the frequency. Obviously no one ever used this feature. Then fix was pushing the fix button repeatedly. It never had a problem again!
Sweet first. Can’t wait for the ups and downs of this pretty piece of 60s kit
4:00 Oh yes, back in the sixties you still needed to hand-crank your lab equipment to start
6:48 the signals on the back are more than convenient. Zero beat the 10MHz output against the 10MHz broadcast from WWV and bingo bango, you have calibration traceable to NIST.
@AvE “Full Chooch” mention at 11:58
What are you talking about?
Very heavy and also usually the fans are quite loud. But they work nicely if you can accomodate one.
Yep it's a loud boy.
Still waiting for Marc to make up names for those engineers on the picture based on their clothing...or missing of some.
As always very interesting to follow you doing the debugging and analysing of such a piece of old tech. Thanks Marc.
Those things was cutting-edge electronic wizardry for that era. The best one could do with the available technology. Just curious to know how much does it cost in current dollars.
Hey Marc! Purely out of curiosity, do these HP machines and test equipment have functions, and things they do, that may or isn't found on modern equipment? Thank you Marc for your videos and time as always!
In my 5245L, I changed the fan to a more modern more quiet fan.
That might be a worthwhile mod. Which one did you use?
@@CuriousMarc I used a 12 volt one of the same physical size. It had almost the same air movement but was considerably quieter!
I only know about this sort of electrical signal stuff from my modular synth so maybe this is a completely normal thing but I thought it was interesting that the waveshape seemed to be different for each clock division. I would have expected it would be square waves throughout.
From a time before HP turned into the printer cartridge mafia.
It's funny how those old school test equipments are much more repairable compared to modern iPhones.
The components, and performance aren't the same 😉
@@Irilia_neko the trillion dolar gigacorp thanks you for coming in their defense.
What repairability and performance have to do with each other?
One was built with repairability in mind and one was designed to be as unserviceable as possible.
did anyone else notice at 17:29 (and again at 19:18) it was counting and showing 1khz? thinking maybe it just takes time for it to count and average the signal?
Looks like it had non-zero values on the display at 13:36 as well.
Or maybe a not good connection somewhere
It isn’t counting. The gate light is off. It comes up with random digits on powerup until you clear the counters.
I got CH3-34 9digit 200MHz nixie frequency counter with s^^t load of transistors in it tunnel diode logic circuits
Very cool video. This goes way above my head.. but I've come across this hardware which is going to be scrapped. Would it be feasible to remove the nixie tubes to repurpose them for a clock? Is there anything I should know before attempting that?
What?! Thou shalt not pilfer an HP 5245L for its Nixie tubes! You should repair it or preserve all its parts preciously as spares.
HP and TEK were so awesome in their heydays. It's sad how in the 1990's, HP got typecast as a consumer electronics company, selling ink-jet printers at below-cost while price-gouging on the ink.. and mid-tier generic PC clones. If I had a time machine, I'd go to early- 1960's Silicon Valley and buy a home.
Cheeky Roland Juno-6 or 60 behind the transistor tester.
A Juno 6 or 60 in the backgroud. Is this a new project, Marc?
That's my trusty Juno 6 that I bought new in the 1980s. Still works today!
@@CuriousMarc I had no idea you played keys too!
01:09 is that a solder blob on one of those IC's in the counter?
Has one of the first Corvus 5 meg drives,,,in packacking..with EVERTYHING. Interested?
What happen to HP now?
Computer Division?
Instrument Division - Agilent?
Are they still in operation right now?
The instrument division and true HP heir is Keysight. They still make very beautiful, very expensive instruments, particularly top end RF and optical communications.
My internal 1 Mhz oscillator could not be adjusted to 1Mhz.
Have an external 1Mhz oscillator that is lock to the color burst frequency?
if anyone knows the music marc uses in these intros/outros, i would appreciate the information.
ive tried to shazam it etc, but it comes up mua roi by le thu, there are several songs, but i cant find any song by that name that sounds even close
From the doodly-doo on the Quindar Tones video: Music Credit: Crinoline Dreams by Kevin MacLeod (edit: wait, that's just the elevator music)
It's Festive Dinner by Pony Music.
18 GHz, this was before Ku band was used
Yaaaaay!! New video!!! I love them!
I don't know if they were even using C-band yet
I think it was not till the 70s they started using 3 GHz to 5 GHz for satellite
What about a BF244 N Channel FET.
Hi is there anyone ho can help me to a photoresistor board to control 1 nixie? I laid my hand son a 5245 but one photoresistor board is broken.
How come know one has commented on the Roland Juno 6???
How much are they?
👏👏👏👏🔝🔝🔝🥇
Silicone on Sapphire.
I used some Soviet instruments and frequency meters from the 80s. I am surprised that many Soviet devices have 90% similarity in design and internal structure. I think that the electronic industry of the USSR just decided to copy Western samples.
im so early this video is 360P
Congratulations?
It actually counted at random times...
It wasn’t counting, the registers just power up with random values.