Arista switches with MLAG configured. Everything looks okay but the virtual-router isn’t pingable and/or working? Make sure you have a statement on both MLAG members of “ip virtual-router mac-address xx:xx:xx:xx:xx:xx” where the xx:xx:xx:xx:xx:xx is a mac address you make up. This isn’t on an interface, it is switch wide. Both switches carry the same mac address since it is used for fail over.
Some years ago I picked up what has always been my favorite hazer, a MBT-300. This machine has a lot of similar units sold under different names I believe, like Antari and such. HZ-300, HZ-350, etc.
After some time it developed an issue where it still worked but it was much much louder. Something was amiss.
Opened it up, and when running the compressor shook a lot. It sits on vibration isolation dampner things, which are basically a screw on each side of a rubber bit.
I looked up the specs from the Mfgr of the compressor, GAST I believe was the company. I got the info about what it should need as far as threading of the feed, but what I ordered would not work.
It turns out the proper thread is M6, not 1/4″. You can find M6 isolation things on eBay and the like. Cost was around $10 for 4, and this fixed the unit. Though it ran $20 since I bought the wrong ones at first 🙂
Super loads of thanks to the people on AtariAge.com for help on these!
Many moons ago I bought a bunch of Atari 8bit parts from the CHKD Thrift Store in Norfolk. I’m talking 15 years ago or more.
It turned out that all of the floppy drives were bad. I sat one them forever. Sold two for peanuts at the Vintage Computer Festival East, but kept back one that had a Happy 1050 enahancement, and one random one for spare parts just incsae.
So I pop open the Happy equipped drive, since that is cool and I figure that is the one I want to work. When I open the other drive and actually remove the shield around the CPU, I notice there is an ICD US Doubler enhancement installed! Whoot!
One drive had a bad WD IC, which I found by comparing chips. The other drive had an issue where it kept rebooting. I have already forgotten what I fixed. Phew.
The replacement WD chip required that I re-calibrate some things, once again documented on AtariAge.
Post on AtariAge:
Somehow, somewhere I achieved a 2nd Atari 130XE home computer. This is the more advanced version of the first computer that I started on – the Atari 800XL. I really liked the Atari 800XL, it was a sleek computer and looked better than the Commodore 64.
At some point I had a 65XE but can’t seem to find it, but was confused when I found the 130XE that didn’t work. I think maybe my buddy Cyberflux gave it to me — I’m not really sure. Anyways, we can’t have broken stuff taking up space so the idea was to fix it.
First things first, the machine started up straight into memory test. ROM passes, RAM fails. Most of the RAM fails. I look up videos on youtube, and theirs seemed to have the same crazy pattern with most of the RAM bad. I ordered some replacement chips, and tried a trick I read about of piggybacking a good IC on the bad, walk thru them and find the failed one. That didn’t work for me, I think there was two bad ICs. The chips on the left are the lower 64K so try to fix those first. I used a Hakko 808 to desolder the RAM chips 2 at a time and socket them. I put two new chips in, then moved the other two down. Eventually the last two chips brought the machine to life and it booted straight into basic.
I used a SALT cart to test the RAM. All is good.
But the keyboard….
The keyboard had something spilled in it. The machine looks great from the outside, but keyboard spill damaged the keyboard membrane. This mylar sheet has contacts that mate to the keyboard keys. The mylar looked good, but tested bad. At the edges of the contacts it deteriorated where it’s between two sheets. Repair would be daunting, since there was at least 15 bad spots on the thing:
I contacted BEST Electronics, but didn’t notice the email timely. I bought a replacement (mine is a mitsumi keyboard, however the mylar did not say mitsumi or have a part #?) The replacement was around $30 shipped. It feels less thick than the original which worries me long term, but it works 100%.
With these two things fixed, the 130XE is 100% working.
In the process I also ended up repairing two 1050 floppy drives and built a SDrive-MAX! More on that next posts.
ATGeek and I got to handle Audio and Video for B-Sides Charm 2019 in Towsen, MD. Great fun was had! Music acts were three artists well known to us: Inverse Phase, Mikal kHill and Wreck the System. The talks of the event were interesting, and overall it was a fun con. Some things are likely to change for next year as things are tweaked. Overall, a fun time!
The picture above is Inverse Phase on stage.
So a friend and myself volunteered to do lasers, sound and lighting for an upcoming IT security conference. Seems cool. We get up balls early to drive to Towsen MD to do a walk through on the hotel and try to get the ability for fire watch and haze for the lasers (denied by hotel. Sigh.)
On the way back we stopped by the MAGFest warehouse, another event we volunteer with. Right near the warehouse is this music store (drums/guitar/keyboard/perc/pa) that has a lot of new and used music equipment. I always like to peek at the keyboards and Rob wanted to check out the PA subs. I see this unit in the racks of random bits, an Elation AR-32RM. Maybe Rob spotted the DMX connectors. Not sure. I had been thinking about looking at the OLA software on Linux or Pi/Arduino for instant playback to some RGB LED wash lights I got from a deal back in the Norfolk VA days.
This thing is pretty broken. The welds that hold the front on were broken off, so someone used thin strips of gaffers tape to hold it together. No way, I think. Last time I was in this store drooling over some old Roland analog rack synth they quoted eBay fully working prices, but it was worth a shot. “Hey man, would you take $20 for this thing? It’s pretty broken but I might have a use.” “$20 if I never see it again, no warranty” “Oh yea, deal man.” $20 in the hand, $20 in his drawer, the next step we were out the door.
Two nice black cap head screws and nuts, some time on the drill press and some luck…. good to go! Looks pro! Getting the stupid gaff tape gum off was more work then fixing the damage, goo gone was no match. Magic eraser seemed to work best.
Last step…. the software!
The software is BROKEN!
So, after downloading the software distro and digging out some old laptops I found that the Windows 32bit drivers… are broken. Off to the 64bit drivers… which are… broken. After all the work to get the thing looking good, and with the DC Dive Show around the corner (where I plan to use it) the clock was ticking. I emailed Elation but didn’t catch a response. I decided to dig a bit.
The drivers portion of the software contain a FTDI driver, which is pretty common. But the vendor ID from the unit doesn’t match the driver ID in the INF files from the FTDI driver. By changing the vendor IDs in the driver files to match, it made it work. I was able to get the software to work on both 32bit Windows running on a physical laptop, and Windows 10 64bit running in a VM. The thing is, it might have broken FTDI support for other things, I’m not really sure about that part. I will try to post the driver bundles soon but WARNING … I fudged it to make it work, no idea how proper things are or what will happen to other hardware that uses FTDI. A friend thinks you can edit the device ID and vendor ID with some tool and just make the thing use a normal FTDI driver but not sure on that.
Overall I was able to overcome limitations and get the unit recording some DMX sequences from Freestyler/Entec OpenDMX. The unit works pretty well, but there is a timer that is global not per sequence which is really annoying. But happy with the unit, and it will come in handy in driving Color Kinetics lights without a sexy City Theatrical power supply but with the PDS-500e supplies that I have.
It’s possible that some of these units have a different set of IDs? I tried archive.org and got some earlier revisions of the software and the drivers from those failed to work also.
From the world of arcade repair… as the cold of winter hit a few of my pinball machines started acting up. On game nights with friends, it was creeping up to a 75% failure rate! Ouch!
This one took me for a bit of a ride. Twilight Zone was getting random resets. Sometimes, I was getting audio and graphics errors, as if the ribbon cables between CPU and Audio PCB needed reseating. But things weren’t adding up. It turns out that tapping the ASIC chip would cause all sorts of issues.
I attempted to do a sloppy job prying out the ASIC and cracked one corner of the socket. My goal was to carefully deoxit the legs of the IC and maybe pins of socket. Oh well. Found some sockets on eBay, ordered them and thanks to Hakko 808 desoldering tool it took about 15 minutes total to remove the old ASIC socket and solder in the new one.
So far, things seem good. I won’t believe the game reset issues are solved — that seems to be an issue with the AC input connector on the power driver board, but time will tell.
I’ve always had a huge respect and admiration for products from the Austin Texas based High End Systems / Lightwave Research. They developed earlier lighting fixtures like the Trackspot and Intellabeam. Their new hardware just pushes things even farther.
In the lighting software world there are free and cheap applications that can run shows, like QLC+ and Freestyler DMX. There are commercial programs like Martin PC, Chamsys Magic Q and Grand MA2. On the cheap side it’s not uncommon to use the software with the “DJ” controllers like the Novation Launchpad, Akai APC, APC II and APC Mini. The commercial lighting software usually has physical control surfaces available that are brand specific, and very expensive. On some of the commercial applications it may cost $600 to add midi functionality to be able to use the lower cost “DJ/live performance” style controllers.
I got to thinking about this old console that High End Systems made called the Status Cue. It’s a physical control surface that connected to a rackmount PC containing an ISA card. The ISA card supported two “universes” or networks of lights. It could do industry standard DMX-512, or the proprietary Lightwave Protocol that HES used on it’s fixtures before the DMX-512 standard came along.
I had Status Cue in my feeBay watch list forever. They just never come up. On reddit I had asked but they’re mostly gone. The PC ran Windows 95 or Windows 3.11 software, and it probably doesn’t do well (I don’t know) with modern fixtures. I have never used the system, I don’t know how easy it is to create new fixtures. Many of the features now might not have existed then so it wasn’t setup to control the now common features perhaps.
Well, I was heading to Florida so I finally just made a random offer of $50 on a console that was listed for $950. It was missing fader knobs, and had no computer or connecting cable.
My goal is to take the console, figure out what it says to and from the host computer that I don’t have, and build a widget to crap out standard midi commands.
I picked up the console. It’s heavy. It’s big. It’s not travel friendly, and I think there is something to be said for modern smaller Wings for controlling the modern software. This is more about the challenge than practicality.
But the buttons. The buttons are NICE. All the way home, my hand would slip to the back seat and I would clicky clicky on the buttons. See, the APC-40 are nice especially at their price point. But these consoles were like $20K in current dollars, so they’re very high end.
I looked on-line, no schematics. It shouldn’t be hard to figure out whats on the DB-25 that connects to the host. But just for sport, I shot an email asking for schematics. No way they’re going to have schematics for such an old piece of hardware. When I checked my email mid next day, there they were. HIGH END SYSTEMS YOU ARE AWESOME. When I asked Martin about schematics for the Roboscan 1220 to look at feasibility of converting to switchmode lamp ballast and power supply to shed lots of weight they told me only service centers get that. Nice people from the internet sent Roboscan schematics to me in no time, but why hide the info.
The rotary encoders are also silky smooth. The trackball is similar (as in really similar) to that which is in my Centipede machine.
Anyways, at this point I have the console powered up via a simple power supply. I have a RS-485 USB dongle on the way. The board uses a Motorola microcontroller (68331 or something) and I have no idea if the uart on that MCU will talk at a normal baud rate common with PCs. We will see as soon as the RS485 adapter arrives, since I seem to have lost or gave away the others I had.
Potential issues will be figuring out how I should deal with things like the rotary encoders and track ball to interact with the software. And what software am I targeting. Freestyler DMX, QLC+ and Pangolin Beyond would be my main ones (though I usually use Beyond with APC-40, and haven’t gotten into QLC+ heavily.)
More to come on this fun project!
No great pics of the deck during this post, maybe I will update.
So I swung by a Goodwill to donate some things, and did a quick check. I never find vintage computers, and most of the video games are picked over. I never find that copy of Space Quest III. But on this trip I did find a Sony TC RC70ES. I debated with myself over it for a little while. It didn’t have the wood ends, so it doesn’t really seem like a ES home component. But it is an ES deck. Less than a year ago I picked up the Nak 680 and have been enjoying it, so why not a little competition? And I tried to fix my childhood Sony dual tape deck and failed — so that still stung a bit. Should I? Yeeeaaa why not.
Brought it home right as I went on vacation. Diving in Jupiter with Eric, Chris and Mary! Chilling in Key West. South Beach Miami! Silverball Museum in Del Ray! Hanging with family. And pre-ordering some belts for this thing from eBay.
So when I got home from vacation the parts were there and ready to swap.
First, the old belt was nasty AF. It was goo, and the goo got on everything. Rubbing alcohol cleaned it off well, but I have no idea if that will hurt the new belts.
Removing the first layer to get to the flywheels was no problem. But the next layer of disassembly resulted in parts popping out all over. So service manual was found, so that is good. People care enough about this model for it to be out there.
Got it back together, annnnnd BAM. Sounded bad. I had to play with the speed adjustment on the motor, but I don’t have a real reference tape. I recorded 400hz from youtube on the Nak (which is not speed calibrated) and threw it in the Sony. It seemed to be moving all over the place.
But the Nakamichi didn’t just work. It seemed to come to life after playing a tape or two and loosening up. So I ran a tape on the ES.
After one side it started squealing. LOUDLY. I ran. I ran so far away. Hit power button. Did some quick google research, and as I remembered the pointer is to the drive motor that moves the capstans. So I put a drop or two of oil on it via a syringe type device used for purging air from printer lines. It cleared up pretty quickly. But not sure if mineral oil is ideal.
The speed still seems to be unreliable, and I don’t know that I really like the sound quality. The CrO2 tape that I recorded Capital Cities – A Tidal Wave of Mystery on the Nak when played back on the Sony just isn’t the same. First the deck detects the CrO2 and you loose high end I think. And there is more hiss. So far I don’t have much of a love relationship with the deck, but I still like the Nak 680 a lot.
Also, after working on the Sony ES deck I now know that I could have probably fixed that childhood Sony deck with motor lubrication. The mechanisms are still crazy complex, kudos to the Sony engineers that came up with auto reverse system and rotating head and all of that.
Time will tell on this deck. Next will be to get a real calibration tape. But I fear I might have to rebuild the motor as well, and I’m not sure of my motivation on it.
Years ago I picked up a handy unit called the MX-8. I think I found it for $20 on craigslist. The unit has 8 or so inputs and 8 or so outputs. You can software connect any input to any output. It can do data inspection and changes and other stuff, but all I use it for is basically a multi thru box.
I tend to leave my cheapest synthesizer on all the time, it’s a Roland JV-30. I bought that dead, swapped a diode on the power input and it was good to go. By leaving the keyboard, MX-8 and one MIDI module on all the time I am more tempted to stop and bang around on the keyboard a little.
All this uptime on the little MX-8 eventually resulted in it dying. Hit keyboard key, nothing happened. Figured power loss wiped out the current settings but nope, no power.
First I checked the fuse, but the fuse broke off in the fuse holder and wouldn’t come out. Fuses blow for a reason anyways.
De-racked unit and popped lid off. Was always curious what was inside.
CR-2032 (I think that is the size) lithium battery holds the memory. If your MX-8 forgets everything, this battery is easy to replace and this should fix your issue.
The power issue was a standard ?7805? voltage regulator. It died from heat. The heat is from being left on 24×7 for 10 years. When I replaced it, I re-spooged it with some fake artic silver heatsink compound. For sport, I added a 2nd heatsink behind the factory one so there is a bit more surface area ti dissipate heat.
Unit looks well made inside.
I think the hardest part of this entire ordeal was replacing the fuse holder I broke.
I did consider moving to a switch mode power supply, like from a phone charger. I didn’t have anything handy so I just kept it on the original linear regulator. I could save some heat and power by converting it. Perhaps next time.