Arduino Based Low Voltage Vacuum Tube Tester 

Thank you! It's really specialized for my needs, but the design as it is actually works fine at higher voltages too. I actually tested it out with an anode voltage of 100V, an anode resistor of 100k and a 10M/474k voltage divider. These were the results from that: i.postimg.cc/904ngMxs/unknown.png But it still has the problem of a set anode resistor and not giving us the real characteristics of the tube. We've been tossing around some ideas for V.2 to give proper datasheet style information, so hopefully I can expand on that and get something neat going!

Thank you so much for checking the channel out! There's just something so special about vacuum tubes, it's hard to put my finger on it. I think the scale of them helps because, even though the circuits end up massive, it's easy to visualize exactly how stuff is moving around the circuit. Our little bun is the most important member of the team here! Though I do love seeing some wild rabbits hop around the yard too.

Thanks for checking the video out! The old Accurate Instruments tester is not the best, but it does help to check how weak some tubes are as well as check for shorts. In the future I definitely intend to build a proper tube curve tracer that gives current traces for a variety of plate voltages. The primary reason I'm using tubes at low voltages is because I didn't start tube stuff until recently. Like most people, I was used to really low 5V circuits and had never touched anything with voltages higher than 100V. For a lot of people, myself included, the high voltages of tubes can often seem like gatekeeping. There are a lot of really smart enthusiasts out there that want to get into tubes, but don't. So, I started playing with them at low voltages, and they work surprisingly well. This opened the door to the hobby for me, and now that I'm more comfortable with tubes and how they work, I'm ready to start working on circuits with higher voltages. I would like to think that I've also opened the door for other enthusiasts to experiment with tubes as well that were previously afraid of the high voltages. The more people we have playing with tubes the better!

Thank you so much!

And one more thing - you could go one step further and be able to vary both anode voltage and grid voltage, then measure cathode current - this way you could generate the characteristic curves of the tubes or other components (I/V curves). EDIT: just watched to the end and you are talking about just the thing ;-)

Also there is a better way to reference the op amp to a negative voltage, I drawn up a schematic. imgur com/ a / unHHLEl you need to add the dot and remove spaces yourself. The filter needs around half a second to settle, the resulting voltage is rock solid ;-)

Thank you! And there's tons of room for improvement here! The idea to do it 16 times and accumulate the results is a great idea! Although, the biggest bottleneck in the test isn't from the baud rate. In the main loop, when testing, I do the AnalogWrite to set the grid voltage and then I have a delay(100) command. I found that if I didn't let the tube settle in to the new grid voltage, I had a few issues with getting reliable reads. I can bring that delay down to about 20ms, but any lower than that and I started getting problems. I way overshot it with a 100ms delay, but then I know the tube is nice and stabilized for my read. I totally forgot that I could set pullups on the Arudino! That would have some some code and simplified the start button a bit. I'm definitely a brute force type programmer, there's like zero elegance to my code, haha. As for the OpAmp, I used the 741 because surprisingly, it was the only OpAmp I had laying around. This entire thing was actually built from scraps I had floating around the lab (well, the breadboard version was at least, I ordered some new buck regulators for the PCB version so I wouldn't lose my only two). And filtering the buffering the output from the tube through an OpAmp filter is an excellent idea, but the primary reason I didn't do it here is I was initially building this to run at 100V (I even did some testing with it at 100V: i.postimg.cc/904ngMxs/unknown.png ), but I didn't have anything that would be able to buffer a 100V output like that short of using another tube. For the low voltage version, that definitely would have made more sense though! Thank you for the schematic on the OpAmp to a negative voltage control, I really didn't have a clue what I was doing and just used the first schematic I came across on the internet that I could get working on the breadboard, haha. It seems to work decently, but there's definitely better ways to tackle that. I'll probably leverage your more robust design for V.2 where we start building proper Datasheet style curves!

@@UsagiElectric I'm happy you find my sugestions useful :-) If you'd like to bounce some ideas about circuits just hit me ;-) About buffering the tube output - you still can use a large resistance value divider, just put a voltage follower after it (opamp with '-' input and output connected together, your input signal goes into '+' input, you take your output from the output pin) - this way you give your ADC a nice low impedance source to load its sample and hold capacitor. Having the circuit settle is crucial anyway with PWM used for analog output, it takes a moment for the voltage to get where it needs to be. And it's not a race but a piece of test equipment, way better to have accurate results one minute later, than garbage in 2 seconds ;-)

Thank you so much! It means a ton to hear that you're enjoying the videos and getting something out of them. That's epic that you're building an ALU from OpAmps, that's a really unconventional project, but I love unconventional projects! And, I think I may actually found your project on Hackaday.io: hackaday.io/project/181803-1-bit-cpu-useing-open-loop-op-amps I'll definitely be following along and am looking forward to more updates!

@@UsagiElectric that's it! :) I finally started drawing up my control register, it's going to be considerably more complex than the alu -- and might just be the most complex part of the whole CPU. Thank you for checking it out and linking it 😁 I don't know what kind of pace I'll make but I'm having fun all the same.

@@lindoran Awesome, I'm looking forward to seeing more on the control register too! Even if progress is slow, as long as it's fun, that's all that matters!

Thank you! And I eagerly await it too, haha. I have some ideas, but I still have a lot of testing and designing to do on it!

Thank you for checking the channel out! I'm hoping to start getting into the design of the cheap curve tracer sometime in the next few weeks. I've got a few projects on the breadboard I gotta clear off first, but it's next on the docket (hopefully). It'll probably take a while to get it working alright, but I'm hopeful it'll go smoothly! The PDP11 is a very cool machine! A friend of mine just picked one up recently and I'm excited to see him getting it going. The future is indeed built on the past, but I'm constantly surprised at how little things have changed! The fundamental designs are pretty much all the same, things have just gotten so much smaller!

Thank you so much! I'm curious how it will evolve too, haha.

Thank you so much!

Thank you very much!

She's performing the most important step: quality inspection! She may have been bribed with treats though, haha.

Thank you!

Thank you! I've been working through a bunch of ideas and I think I've got one that will work, time to start prototyping it and see where it takes me!

@@UsagiElectric :) really ! a open source one will help a lot of people, basically it will disrupt the market :)) good luck!

It would be better to test these with the voltage the audio circuits use.

Thank you very much! And a big greeting from Texas to Argentina!

I agree, it would improve things! Also datasheet claims that analog inputs are optimized vor 10K impedance sources, so your voltage divider might be a little high in resistance. Might cause errors at higher speed.

Thank you! That's definitely a good idea! The primary reason I didn't end up buffering the output was because initially I was planning on building this thing at 100V to hopefully make it a little more appealing to more tube users. I actually did some initial testing using a 100k anode resistor and a 10M/474k voltage divider. These were the results from the initial testing set up like that: i.postimg.cc/904ngMxs/unknown.png But, I didn't have anything in my collection that could buffer a 100V output (short of another tube), so I just went with a massive voltage divider and it seemed to work decently well. Once I decided to split the project into two (one being this low voltage version and another being a proper high voltage version that gets I/V curves like the datasheet), I just changed the resistor values a bit and dropped the anode voltage to 24V. Other than that, it's actually the exact same design. For V.2 we'll get a little more serious with the buffering outputs and making sure we're getting proper readings!

​@@UsagiElectric For V.2: Why don’t you go for buffering voltage after the divider, using an OP amp? So you can keep your high resistive divider, which is a good thing to have in order to keep load low! On the other hand the voltage follower doesn’t need to be capable of dealing with high voltages. Pretty nice project, looking forward to see version no. 2! :) Might gonna build one too...

@@schmitzvonschmitzen2870 Interesting! I was so focused on trying to keep the load on the tube to a minimum, I totally didn't think about the issue being on the Arduino side. In that case, a little OpAmp buffer after the divider makes a ton of sense! For V.2 however, there's a couple designs I want to test out, but one in particular would keep things pretty simple and not require buffering of any high voltage. I've got a bit of real-life stuff going on at the moment, but I hope to dig into and start testing out ideas for V.2 in the next few weeks!

@@UsagiElectric A quick cheat version of this is to put a small valued capacitor, like 1nF between the ADC input pin and GND. This provides a momentary rush of current for the ADC sample. You must also add a small delay before ADC samples so the cap voltage settles. It should make readings more repeatable without adding an opamp. The opamp will do better though.

Thank you so much! I don't have anything recent on the hardware and workflow, but I definitely plan to make a video covering the entire process sometime in the future! For quick reference though, I design all my PCBs in DesignSpark PCB using 1mm trace widths. Once happy, I export the Gerbers, and then open them in FlatCAM. From there, I get all the settings right for the tools I'm using (60* V-bit, 0.9mm drill, 1.5mm end mill) and export it as G-Code. Then, I stick a 2x4 in the clamp on our Bridgeport EZTrak CNC mill, use the biggest end mill I have to flatten the surface of the 2x4 with respect to the mill head, and finally just stick the PCB to the wood with carpet tape and hit go after zero-ing everything out. I've gotten into a pretty good flow and can go from design on the PC to actual PCB in my hands in under two hours, which is excellent for how impatient I am when working on projects, haha.

That old Morkrum Kleinschmidt is pretty cool, isn't it! It's actually pretty close to being good to go. The send portion works with no problems, and the receive portion operates, it just needs some serious adjustment. My biggest holdup on it right now is that I need to build a 120V 60mA current loop for it, and some kind of level shifter that lets me hook it into modern RS-232 interfaces. Here's a short video of the send portion putting some Baudot code into my scope: imgur.com/gallery/dxnTJJZ

@@UsagiElectric Maybe you should ask @CuriousMarc he has done a lot with Teletype current loops. 😊

Thank you, but I certainly wouldn't say super genius! Mostly, I'm just really, really stubborn and refuse to give up until I got a result I'm happy with, haha.

Haha, thank you! Unfortunately, I'm going to have to shave the beard for a work thing next week, but I think I'll let the hair grow out a little longer before giving it a trim!

@@UsagiElectric - That's sad. What about attending the get-together remotely and just leaving your video off? ;)

Thank you! My ultimate goal is to build something that will test pretty much any dual triode, tetrode or pentode you throw at it. Rectifier tubes and diodes may be a bit more difficult, but still possible? At any rate, the ultimate goal is to keep the total cost under $100, which means there will have to be some compromises in pinpoint accuracy, but I think I've got a design that will work! One of my goals early next year is to get building on it!

@@UsagiElectric Thanks for your prompt reply! I'd love to build something that I can learn from at the same time. I'm glad there are people like you who take the time to make videos. I still work on old guitar amps etc. but I have to use a friend's tube B&K tube tester which does simple emission, shorts & leakage etc. Was also thinking of building something like the Etracer kit. Now, I think I'll wait to see how your project turns out. Thanks again.

Oh man, I'm right there with you, that would be awesome! Unfortunately, there's just way too many variations that I don't think it would really be possible. Hmmm, I mean, you could test continuity across each of the pins and potentially determine the heater layout from there. But determining the voltage and current of the heater from just that would be... beyond my capabilities. But, if you could figure out the heater, you could then use low voltages applied to the pins to start determining what the layout of the tube is. Like, put 6V through a large current limiting resistor on two random pins, connect a random third pin to ground, and see if any emission happens. Through this you could maybe figure out where the cathode, grid and plate are. But, for tetrodes and pentodes that adds a whole other level of complexity. Alright, well, now my brain is going to be chewing on this for next few months, haha. First, I'll focus on building a low cost curve tracer for tubes, then maybe I'll revisit this idea!

Thank you so much! Interestingly, I initially designed it at 100V using a 100k anode resistor and a 10M/474k voltage divider. These were the results from the initial testing set up like that: i.postimg.cc/904ngMxs/unknown.png The upshot of this is that the tube is operating much closer to the voltage it's designed for. Though it still has the same drawback of just measuring the output with an anode resistor in place, which is great if you have a fixed circuit circuit environment you're working with, but doesn't really give results like you would expect to see in a datasheet. But, it definitely works, and you could even replace the anode resistor with a potentiometer to have a variable anode resistor for a little more flexibility in testing (though, it would also need like an additional 1k resistor to prevent the tube from pulling massive current)!

Thank you so much! I've heard conflicting ideas on what low voltage does to tubes. The biggest worry seems to be something called cathode poisoning. The cathode is nice and hot and boiling off electrons, but those electrons don't have a strong enough B+ to be attracted to, which can cause a chemical reaction between the cathode base material and the oxide coating on it. This can act like a high-resistance, which can start to mess with stuff. Now, everything I've read regarding cathode poisoning says it's when there is no anode voltage at all, so no current flow through the tube. So, 24V, while not much, is still something, so that may prevent cathode poisoning altogether. Honestly though, I have no idea, there's just not enough information out there about it. So far though, I haven't had any problems!

Thank you so much!

Thank you! And that's actually a really good idea, I never even though about that! A tiny, portable tester that just needs a 24V wall wart and can give genuine results on the LCD would be awesome.

Thank you!

Thank you!

oh you are your using a exel sheet for that

Ultimately, I would like to have V.2 have a database of tubes stored in an application that can then cross reference on the go, without needing to copy the data into an Excel sheet. For my purposes, V.1 here is great and will definitely get some use, but something that's a little more in-depth I think will help a lot more people!

Thank you!

Indeed!

Thank you! I really wanted this to be something that was dirt cheap and could be made by anyone anywhere in the World. I actually added it up, and in parts cost on the finished version, I have less than $15 in it! My goal for V.2 is to have something that gives proper curve tracer data for less than $30. I don't know if I'll be able to achieve that since I'll need a lot more complexity, but if I can, I think it'll be something that'll be useful to a lot of people!

Thank you! I've managed some decent progress on V2, but got tied up with other projects which put it on hold. I'm hoping to get back to it a little later this month!

That’s great news!! I’ve been sourcing a 100-200Vdc supply in anticipation!!!

If it can happen, it will happen!

Sure thing! They're just Harwin H3161 1mm PCB pin headers. I buy them from Mouser, but Digikey also carries them.

Oh, uh, I just kind of used what I had laying around, haha. As for the tube sockets, I use little Harwin 1mm PCB pin headers with a custom footprint on the PCB: www.mouser.com/ProductDetail/Harwin/H3161-01?qs=WS5Jv%252B%252Bx1qUxGQ9n1naX6A%3D%3D

Thank you!

Your video on the Roetest was definitely a bit of inspiration for this! So, nice to send some of that inspiration back your ways, haha. Hmm, now I'm thinking I need to make a new Gerber design file for more traditional sockets and that can be made by the regular PCB shops!

That's entirely possible! Also, sometimes the material that is coated on the plates can flake off and cause shorts too. If you're lucky, sometimes you can get away with just shaking the tube to break the short loose.

Unfortunately, I don't think something like the 8417 could work. There's a few issues that we would have to overcome. The fact that it's octal would require a whole new plug-in module that's a little too large for the space available. But the bigger issue is that the 8417 heater draws 1.6A of current, which means its in-rush current when cold is probably closer to 3A, which might be too much for the little buck converter I'm using to generate 6V for the heaters.

Thank you! And thank you for the offer, but I'm afraid I must ask not to add it to your compilation video. I would prefer if people could watch the whole video. There's a lot of context as to why it was built the way it was that I fear will be lost if the video is made into a shorter form for a compilation.

@@UsagiElectric We respect your decision and we will not feature this video. But if you have some projects that you might want to share, feel free to contact us. Keep up the good work.😊

That's interesting that the Aviation band tube radios ran such low anode voltages! I run low anode voltages because I'm a total klutz and I wanted my vacuum tube computer to be safe for an idiot who likes to stick his fingers in live circuits (ie. me). It turns out, if you're building logic circuits, some tubes work beautifully at low voltages. The 6AU6 in particular is a great low voltage logic tube, with sharp transitions and nice voltages at saturation and cutoff. It also helps that the 6AU6 was probably one of the most ubiquitous tubes ever made and is pretty cheap, haha.

@@UsagiElectricWell a master knows how to use their tools ;) In your case you are just looking for a Hi-Lo signal (digital) so you are using them as a switch vs an amplifier ;) So 100% correct to run them cold. Why waste the power. BTW I have a little typo in the above should be 12/14 or 24/28 to be perfectly correct. The first being the battery the next the working (alternator) voltage. It is the old engineers joke you can 2 of the 3, 'low voltage', 'medium power', 'no distortion'. In the case of General Aviation tube radios, they choose the first two. The real rub of 'Cold' tubes is some modern 'hybrid' tube amps that have a 12AX7 in them with 40v on the plate, they have to use an op-amp or two to cover up the distortion patterns., (snake oil really).

It should work at 12V with no problems at all! The resistor divider network for the arduino will need to be updated, but other than that, it should be pretty straight forward!

Try taking a look at an old 555 timer datasheet, there were many application circuits for all kinds of interesting circuits, the voltage inverter being one of them. Another interesting one in the same ballpark is a buck or boost converter with full feedback loop based on the 555.

@@gammaleader96 It just seems nuts, but thats one of the things great about electronics; the 'intended use' (be it intended or just taken for granted in my head like that) is often just one part of a components capabilities; the options for use or abuse are so many ...

The 555 is a truly great chip, people have made all sorts of wild stuff with it! It may not be the best negative voltage generator out there, but I really only needed a tiny amount of current, so I figured it'd get the job done!

I just found this and now feel the same.

Me too!

Unfortunately, the closest electronics recycling place is about two hours drive away! The joys of living in the middle of nowhere, haha.

@@UsagiElectric Stack up all your aluminum cans and steel cans and electronics and make it a little outing for day Bonus is buying the electronics they have