Profile cover photo
Profile photo
Michael Kukat
Communities and Collections
View all

Post has attachment
Bought cheap BNC connectors for my WordClock cables several years ago. What could possibly go wrong? They are just BNC connectors.

This was my spare cable. i already noticed this effect on another cable and replaced it (crimped more of them for a larger setup i had back then). I didn't use any nuclear bombs on it, i even didn't roll over it with my car. Okay, i just slammed the right one on the table to see what happens after it already fall apart. Not too much to avoid damage to the table. This crap breaks like glass. Maybe even worse.

There is nothing on this planet that can't be totally botched.

Post has attachment
Inspecting the inspection camera

Yesterday, i have shown you a small part of this Sony DXC-3000A i recently bought, the optical filter block that sits in front of the color divider, composed of an IR cut filter and an optical low-pass, and some third thing i don't know.

Today, i wanted to have a look at the CCD block's capacitors because this old Sony stuff is full of ELNA capacitors and i was already impressed that those on the other boards don't leak. Sometimes i guess Sony bought non-leaking ELNA capacitors but got leaking Nichicon caps while all the others got the rest :)

So i didn't find any issues inside the camera but took some photos to show you how this beauty looks inside. This camera is extremely maintainable, a very clean construction, nearly as nice as the JVC KY-17 i recently took apart. I forgot some important photos for the KY-17 so no teardown post of it yet. Back to the DXC-3000A, which not even is a broadcast camera, it was more targeted to the prosumer and industry, just look at that ridiculous built-in microphone, this is so totally consumer-style :) But at least from the construction and quality of the guts, i don't see a difference to comparable Sony cameras around that time. The DXC-327, which is slightly newer, even has much more problems with leaking capacitors. On the other hand, i think the whole DXC series was more for semiprofessional use, the broadcast video stuff was more the BV* line.

You can take the camera apart in a matter of minutes and you don't need much longer to put it back together. The connectors to the backplane are labeled, nothing can go wrong here, the board order is printed on the backplane as well as on an aluminum tab next to the battery compartment (which sits on top of the card cage), the whole CCD block can be easily removed by undoing 4 screws and pulling some connectors from the backplane.

On the electronics side, this camera already is microprocessor controlled (they were since a while), but has fully analog video processing. There is an on screen display in the viewfinder, the settings are stored using a supercap for a couple of hours, so no leaking battery for this. I think there are some more permanent settings that can be stored, maybe in an EEPROM, i didn't inspect it that deep yet. And then there are some internal switches for basic configuration like switching the output between Y/C and CVBS.

All in all a nice camera. I don't have a complete "normal" version, i bought it because i was curious about some PCBs that were mounted where the shoulder pad usually sits. Well, nothing spectacular, but the idea of having zoom presets was some inspiration for my soldering microscope, which i still have in the queue. And having a CCU connector is never wrong - i could convert one of my DXC-M3s to a camcorder by adding a Raspberry Pi now :)

So this camera was built into a machine or whatever to inspect something. I have no idea where exactly it was installed, but the changes/additions i found are:

- No carrying handle, no viewfinder, no shoulder pad, no battery pack holder
- The lens connector (carrying control and feedback signals between camera and lens) is removed
- The lens is modified, all the electronics are removed and the motors are directly interfaced to a connector which within the machine likely goes to the 2 additional PCBs
- An additional sensing device is installed in the lens, likely sensing the teeth of the zoom gear for feedback of the current zoom position
- An additional motor to control focus is added
- An achromatic lens and color filter has been added for a magnified view. The color filter is the same as the daylight filter you can dial in on the camera, i have no idea why they always add those filters, maybe they are of better quality than the internal ones. You get natural colors with the 3200K pre-set white balance of the camera, the filter wheel set to 3200K, this color filter in front of the achromatic lens and a cold light source like flourescent lamp - or for today, LED.
- 2 PCBs are added, connecting to the VIDEO OUT and CCU connectors, plus having several other connectors, likely going to the lens and the host system. Besides this "Zoom preset" board, there is nothing spectacular about them, the second board just seems to be an interface

I found a brand name for this conversion within the lens handle, but i already forgot it again. But i googled a bit and they still produce gear for QA purposes like inspection cameras and much more.

Not sure what i will do with this camera, but at least it fits into my collection of professional video cameras around the time when there was a switch from tubes to CCD sensors. This one was made in 1989, i think.

The last 3 photos are the view through the whole chain, the image quality is pretty impressive for a 30 year old camera with a new sensor technology back then, especially through a CVBS output. It might even get a bit better using the Y/C or even RGB outputs.
16 Photos - View album

Post has attachment
Fixing lens delamination - first try

I mentioned it from time to time that the Fujinon video lenses i got with several vintage video cameras all suffer from one problem - one of the laminated lens elemens delaminates, resulting in a blurred image with low contrast.

Interesting enough, there was not a single exception yet. The last one i got first had fungus, i didn't see the delamination, but after cleaning the fungus problem, the slightly milky look through the lens revealed that it also is no exception.

Beside all those Fujinons, i also have 2 Canon and 1 Schneider video zooms, they don't have those issues. Very interesting. So a while ago, i decided to give relaminating a try on one of the Fujinons i don't really need. It's a hard challenge consisting of first separating the elements, which is not too hard with the old canada balsam, but most people will tell you it's nearly impossible with the more modern epoxy based cements. If you separated and cleaned the lenses and the glass is okay, the next challenge is cementing it back together. Most people use canada balsam, some use the modern epoxy based stuff which is even harder to get for normal people, but one out there uses simple consumer glass glue. This is what i decided to give a try and ordered a while ago.

Now, with another new camera, i had the chance for a first try. The filter block in front of the color separator looked very ugly. Too bad i didn't make a photo of this, but i started working on this too spontaneuously yesterday evening, so i forgot. But meanwhile, i have some progress i need to share with you.

Yesterday evening, i have thrown this 3-element filter into an acetone bath. Half an hour later, i was able to easily remove the thickest element from the stack and with the help of some more acetone and a guitar pick, it wasn't too hard to remove the remaining epoxy from it. But i wasn't able to separate the remaining 2 elements yesterday evening, so i let it soak over night. Stil no success this morning, so i carefully applied a bit more force, there was not much to lose. The problem was the blue disc, which is the IR filter and this one is very thin, so it can easily break. But everything went fine, after patiently pushing around the seam with the fingernails it suddenly came off.
Same procedure with acetone and the guitar pick to clean all the surfaces and then a closer inspection. The IR filter definitely had some damage, but at least the center was mostly clean. I inspected a bit with the microscope, i wasn't too happy with what i have seen, but still better than before with the bubbles in the dissolving cement.

So let's go for re-cementing it. I never worked with this glue, so i was a little bit nervous. Cleaned all surfaces with the LensPen, compressed air to blow off remaining dust, a drop of the glue, pushed the first 2 elements together and got a horrible mess. Sure. That was too much glue. If i knew how evil this stuff stinks and what mess it leaves, well, everywhere, i would have use less :) But okay, after aligning everything and washing my hands about 5 times during this, cleaning the desk and so, the 2 pieces were together. I couldn't judge the quality because everything was contaminated with this glue. I didn't want to risk damaging the only important area - the area between the two pieces. They say 5 minutes until it's a little bit stable.

And they said 24h in daylight to harden. Well. Daylight. Maybe visible again in March? Okay, let's see how it works out tomorrow...

1.5h later, i decided to clean the outer surfaces and noticed that really nothing changed, the glue was still as soft as it was 1.5h before. And tweaking the whole thing to clean it didn't seem to damage the joint between the glass surfaces. So i cleaned it and decided to put the third piece in place also, looks like the glue is not so critical. And it really isn't. This time it was not enough glue, i added some with the finger, shifting the elements around, i did absolutely everything you would never do to get a good result :)

Then there still is this 24h daylight problem. How to i stabilize the whole thing without creating another mess with the plastic frame where the whole thing later goes? I wrapped some adhesive tape around it.

And then i had another idea. Daylight. 24h. What if i use an UV source? UHU doesn't specify the required wavelength, so i decided for the EPROM eraser, which looked plausible. To erase an EPROM under sunlight, you might need weeks. This erase needs minutes. So i gave it 5 minutes and voila - the glue hardened enough that it was impossible to shift the elements around to fix a minor misalignment i noticed. I removed the tape, cleaned the surfaces and threw it back into the eraser for another 15 minutes, which should complete the task.

Really bad that i don't have the before-photos, but the after-photos are impressive, even on closer inspection, i don't see the defects of the IR filter, maybe the glue filled enough small gaps that even this problem is gone.

Even if the separation with Acetone might have been a matter of luck, i'm very positive using the not so important Fujinon lens as my next victim for this :)
5 Photos - View album

Post has attachment
Capacitors at work

Not a full repair story today. I somehow lost my motivation to spend lots of time to shoot all those photos and write long articles when everything will be gone anyway in the near future.

But i'll still throw some nice photos at you from time to time :)

Yesterday, i got 2 JVC BR-S411E, professional dockable S-VHS recorders. They should fit a camera i bought before, but i'm still waiting for it. All in all, i noticed that i don't have JVC broadcast gear and i was curious about their built quality, since Sony messed it up with leaking capacitors and only Panasonic was really great so far. I need some Ikegami gear for a last comparison.

To make it short: JVC has the capacitor problem of Sony, bundled with the unmaintainability of an incredibly messy wiring harness, something i already noticed with their consumer gear. Those recorders look like a piece of shit inside compared to the clean construction of the Panasonic devices. Not that they have bad specs or are ugly or so, it's really just the guts that totally demotivates one to maintain them.

But build quality in general isn't my real problem. I noticed plenty of ELNA capacitors. At least in the late 1980s and early 1990s they might have been the worst capacitors every made because i always stumble across them when looking for leaking caps. The recorders mostly looked clean, but with some audio issues on one of them and some video issues on the other one, i just removed the audio board from the first one for a closer inspection.

And i found something on this audio board which might be part of the reason it doesn't behave correctly. I'll just drop here what i've seen :)

Those small ELNA caps were used a lot in miniaturized gear and they are nearly always leaking. But there are other types also, i already scrapped noce gear for leaking brown large ELNA caps and recently, i repaired a Roland XV-5080 with bulging ELNA caps in the audio outputs, one of them was the reason for the output being silent.

Retro electronic gear. It's like eating too much chili. It's painful as hell, but it's also so much fun.
2 Photos - View album

Post has attachment
Music tip

If you like some chillig electronic music and the sound of the good old Roland x0x machines, check out this great release of a friend:

If you like it, please buy it, that price is really ridiculous for the great work and she gets ripped off way too much already.

I hope you like it as much as i do!

Post has attachment
Novation BassStation repair - again

A while ago, i bought this nice little thing for not that much money - defective, as usual. As i just collect synths and rarely use them, the challenge to repair them is an additional use :)

Back then, there were no schematics of this synth available, so it was a little bit harder to find the problem, but poking around with the scope probe revealed that the filter is totally closed, independent of the adjusted cutoff frequency. It didn't take too long to find the exponential converter not working correctly and at the end, it was a 33KΩ resistor being open. Replaced it, worked fine again.

That doesn't happen very often that you find a broken resistor in such a modern piece of electronics.

Recently, i noticed that there is another problem with the filter or so again and today, i decided to have a look at it. Didn't take too long to find a 33KΩ resistor between the OSC mixer and the VCF being open. Hm. Wait. 33KΩ. I heard that a while ago. Okay, replaced it, and it was better. But what's that strange behavior of OSC 2? It's fluctuating like through a noisy potentiometer, but there is no potentiometer inbetween, 2 VCAs form the mixer. So i measured a bit more and found a half-dead resistor. Guess the value. Yes. 33KΩ.

Another test, now the VCO mixing is fine, but the VCF is silent in 24dB mode. Had a look at the schematics (meanwhile, someone reverse engineered the synth and posted the schematics:, suspected R61, had a look at it, 33KΩ, replaced it, partial success. Again the sound fluctuated. Followed the chain, let's have a look at R67. Guess the value. Replaced it, and voila, everything works again.

So no complaint about capacitors today, resistors also have the right to be evil! Especially the 333 ones. Because 66MΩ is uncommon. So they are just half-evil and i just replaced half of them meanwhile. Not sure if i wait for the next problem or if i replace the rest right now.
Hm. I think it's more interesting to wait for the next problem :) And i want to play a bit with it now, followed by the Model D, i recalled that i recently bought one and didn't play a lot with it yet. Maybe i'll add the Microbrute and set up some couch pseudo modular setup.
2 Photos - View album

Post has attachment
Tinkering with the ELV RS 500 room climate station

I have this nice station since nearly a year meanwhile, but triggered by a discussion with +Gerhard Torges, i connected it to USB for the first time today.

Based on some basic research found here: i dug out WireShark and had a look at what happens with some actions performed in EasyTemp running in a Windoze VM.

And this is what happened:

Interesting to mention that the station uses the USB VID of ST microelectronics and the PID is also used by some other gadgets out there. This is what happens if the USB-IF demands $5000 for an USB VID and even doesn't allow reselling PID blocks to individuals to share the cost for smaller developments.
I do the same for my hobby projects, but it appears that such products also make it into the field.
So this is how something so well-defined like USB suffers from totally stupid regulations for the required numbers.

Post has attachment
Soldering microscope, pro edition

Another thing i was thinking about since some time - and worked on a little bit during the weekend, was a way to make use of all those old professional video cameras that pile up here.
A while ago, i had the idea to use old surveillance camers for a soldering microscope. Well, it was mostly to have a reason to buy a sixpack of them. This didn't look bad with some tests, but i needed to get quite close to the PCB and needed lots of light, so it might not be really comfortable to work with.
On the other hand, my process at the moment looks as follows:

- Draw the PCB layout. With 0603 components, because hand soldering. During this procedure, get upset about the incredibly big size of those 0603 resistors, they are in the way everywhere around the 0.5mm LQFP micro.
- Print layout, exposure, development, etching, all in all half an hour for that, including cleaning the workspace afterwards.
- Drill the PCB (even with SMD, i still usually have through-holes to drill).
- Solder the stuff together. Depending on the complexity of the PCB, this can take several hours. I work with a magnifying lamp (you see it in this setup also), sometimes i need another magnifier for accurate positioning of components, all in all it's pretty guaranteed that i need a break after 30-45 minutes because after an hour, i definitely get a headache with my head pointing down, close to the magnifier, plus the required concentration all the time.

But it works. Built lots of stuff this way.

But there's always room for improvement. Recently, i ran across 1 Sony DXC-325P and 2 Sony DXC-327P, all with Fujinon lenses, plus 2 Canon lenses. All 5 lenses are equipped with a Heliopan 4/250 achromatic lens. Simply said, a 4 dioptre close-up lens with a maximum distance to the object of 250mm. The cameras were in industrial use, so no handle, no shoulder pad, but technically, they are good 1/2" triple CCD broadcast cameras.

After some calculations and thinking about the whole setup, i drilled some holes in some pieces of junk i had around to get the horizontal mounting bars for a provisoric test. Before constructing the whole stand, i wanted to test this and found a DEC 3000/300 and a DEC Personal DECstation 5000/25 plus some DEC TK50 tapes giving the right height for this test. Okay, choosing DEC for everything was by chance because i dug through some DEC stuff the last days and those were still around. I could also have used 2 boxes of beer and some NetWare documentation, but i just have one box of beer around at the moment. Due to the offset between the front and rear horizontal bar, i added an RS-232 adapter on something similar on the other side.

It's just about having a test setup to see if one can work at all with this. The video monitor is also provisorically attached, this is close to the location where i think it would make sense.

Well, all in all this looks very promising. The resistor on the screen is a 1206 type, so 0603 should still be very comfortable and maybe i reduce my favorite resistor size soon, if all this works out as intended. This is the maximum zoom, by the way, with a 16x zoom, there is some room to adjust it for the job. And due to the large lens, even moderate amounts of light allow to close the iris enough to get a good depth of focus. But i'll add 2 bright MR16 LED spots left and right of the camera, so illumination will also be included.

And as a final goodie - focus, zoom and iris are motorized and have feedback potentiometers, so i can add some foot pedals or so to adjust them while working. In theory, it would also be possible to add auto-focus using a RPi and some image processing, automatic iris control is already supported by the camera, i would just need to add the motor driver.

All those features are the reason why i decided against the small surveillance camera. Even if the final mechanical setup will be much more work, the results simply are much better than with smaller, simpler cameras. And it looks incredibly cool :)

By the way - i paid 12€ for this camera including lens. Okay, they went out for different prices, the most expensive one was 26€, interesting range for cameras with an identical description, but that's eBay. I had to replace some capacitors and the lens also is no longer usable. Yes, what you see here is an unusable lens. One of the elements of the zoom delaminates, so even if the image is still sharp, it has very low contrast. You won't believe your eyes when i switch this setup to the Canon lens, which will make it into the final setup :) But while working on this and risk dropping something, i'll use the components where an incident like this doesn't really hurt a lot. I hope i don't drop it on my feet, that stuff has some weight.

...To be continued.
5 Photos - View album

Post has attachment
VAXstation 4000 VLC maintenance

And this one also gets a new home. It has some history here, being the first computer beyond the home computers and PCs i had until then. A customer of our computer shop back then came in and asked if we can upgrade those machines (there was a second one, a VAXstation 3100 model 76), i just knew "VAX" being something completely different than the standard household appliance PCs. So i had to disappoint the customer, we can't upgrade them and they will not run Windows 95, but i would buy them. I forgot what i paid, but they were cheap, i also got a 19" fixed frequency monitor with the package.
So i had 2 machines without any clue about them. This was the start of a long and interesting journey into many exotic workstations and servers that were incredibly expensive when new but mostly cheap to obtain after they were 10 years or older.

After i couldn't do a lot with those 2 VAXstations back then, one day i found NetBSD, which ran on the 3100, this was the starting point to port it to the 4000 VLC also. I got some information about CPU and I/O registers on a somewhat strange way, which helped a lot to make this work. At the end, i also ported NetBSD to some other VAX 4000 machines, large fridge-sized boxes i had back then at the peaks of my collection. I loved those beasts, built like tanks and unbreakable. And totally different from all the rest, closer to mainframes than to "normal" computers.

Besides the history, there is not much great to tell about this one. It's a very nice looking pizza box, but "VLC" stands for very low cost, and this is what it is. It has a VAX CPU, yes. But no cache, maximum of 24MB RAM (which is installed), the frame buffer is not much more than just a frame buffer, i once ran Netscape or so on OpenVMS on it and it not just took minutes to start, it crashed after opening the second tab or so due to lack of RAM.

But on the other hand - no defects. PSU looks good, the NVRAM needed a battery replacement. It's a DS1287, which in fact is a DS1285 with battery and xtal on top, so you just need to remove some plastic around pins 16-20, cut the connections there to disconnect the original battery and add a new one somehow. I decided for a CR2032 socket glued to the chip, so the battery is easily replacable in the future. Inside the machine, there was just a rubber standoff below the board getting close to the point where it transforms to goo, so i replaced it.

The LK401 keyboard needed lots of cleaning and some more maintenance, one of the feet turned to goo and left a mess all over itself and some other keyboard i stored together with it, so i removed the 2 front feet and replaced them by felt feet. The numeric block 0 key was half-stuck, the reason was the rubber dome staying in the compressed position, maybe i stored the keyboard with 0 permanently depressed somehow. I removed the rubber dome mat, pushed out this dome, applied some hot air and it cured a little bit. I hope this will get better with the time, but at least the key still works fine, it might just hang a bit lower while the dome is not yet healed completely.

There was an OpenVMS 7.1 installed, last login October 2003. Didn't use it for some days :) But now, everything is tested and works great. OpenVMS 6.1 didn't like the Seagate hard drive (which i totally understand :), 7.2 wasn't that picky and could install onto it. So everything tested and ready to get used again.
24 Photos - View album

Post has attachment
VAX 4000 105A maintenance

My big VAX gets a new home, so time to check if everything is okay and play some last minutes with it. The 4000 105A is one of the later VAXen and was intended as an office server. No graphics, just text terminal logins, usually via network (whopping 10MBit/s), but for this, running VMS, it could serve hundreds of users. According to the person i got it from, it was used by 1500 or 1800 employees before being retired.

With 128MB RAM and CD-ROM, this is a really nice machine, it even has the dual-bus DSSI controller installed. On the rear, you see the large SCSI-1 connector, 3 DSSI connectors (the 2 right ones are the same bus) and 2 large 50 pin connectors between them. Using those, you can add an external enclosure to add some QBus cards, the standard of the larger VAXen of that time for expansions. The 4 covers in the left on the rear are for multiport serial devices, which can be installed within the machine.

As expected, not much work was necessary to bring it back to perfect shape. No leaking capacitors, i removed the leaking backup battery a while ago and replaced it with 2 new NiMH cells now, packaged in a small plastic bag, just in case they start leaking in the future. It got new rubber feet because they were already missing, and the lock mechanism of the front cover got some lubrication. That's all, the machine simply works like more than 3 decades ago, when it was built. I did a quick OpenVMS 6.1 test installation without any problems.

This machine is built incredibly maintainable, some easily reachable screws, most of them can't be easily lost, the main board can be removed by just pushing 2 tabs labelled "PUSH" and pulling it to the front. That's the good old DEC quality.

Okay, the CD-ROM took a bit more work, but at least no leaking capacitors yet, so no need to repair something.

I'm sure the new owner will love it and have lots of fun with it :)
22 Photos - View album
Wait while more posts are being loaded