I’ve known about these devices since they were introduced a decade ago, but never really had access to one. I decided to get one to see how it works. There’s a seller on eBay that has both the tube and a PCB for a buffer amp, and I decided to pick one of those up. Here’s the good stuff:
The device itself is kind of interesting - if you said that it looks like a vacuum-fluorescent display, you’d be correct. That’s exactly what it is.
What is this, exactly?
A vacuum-fluorescent display, or VFD, is a type of display that uses a filament, grid, and plate - just like a normal tube. The plate in this case is a painted substrate that is flooded with electrons, and this flood is either allowed or prohibited due to the charge on the grid. It’s and on-off device, but it still fulfils the basic requirements of a triode tube.
The Nutube takes that idea one step further. Instead of a painted substrate that’s a numeral or symbol, the plate on this one is just that - a rectangular plate that is flooded with electrons from a filament, modulated with a grid in the middle. It does glow in operation, but that’s not the point - this is a flat-pack tube that can amplify audio. It was designed, AFAIK, for Korg’s musical instruments because “tubes!” and because this device doesn’t need the high voltages that something like a regular 12Ax7 device would need - even though those higher voltages are trivially easy to generate these days.
The tube’s numbering doesn’t really follow standard conventions, but whatever.
The vendor provides a BOM for this device’s board, but it’s out of date. I’ve created a new one and will post it as I build this device. In the meantime, I’ve ordered parts and hopefully will have time to assemble this device in the next few weeks:
Stay tuned for an updated BOM and assembly thoughts.
I had a request in regards to a Simpson 260 multimeter. I was asked if I knew what the handle bolt sizes are - I do not, but I decided to take a quick look to determine if there was anything of use easily visible. I have a series 8 available, so I said I’d take a look.
The first thing of interest is the handle itself. The handle is held on by two large hex bolts with a thin head. These bolts pass through the handle. Inside the handle itself is a piece of metal. This metal pivots with the handle, but doesn’t necessarily appear to be part of the handle.
The bolt then passes on through the case.
Unfortunately, I can’t get this apart to measure the end of that, I suspect these have to come out before the meter assembly pulls from the case back.
That’s all - this was just a check of some parts on the unit. Stay tuned, more good junk on the way!
This device was an easy fix. Some wire, a couple of probes, some grommets - and we’re done. Cable is even color-matched to the unit!
I scrubbed the case down, and gave the switch a spray of Deoxit as a preventative, even though whatever the OEM used was not really tarnished like other switch rings I’ve seen.
So. Does it work?
Sure does! The new battery reads as a new battery.
That’s all for this device, it goes on display until needed. Stay tuned for more EICO 950A goodness!
Working with this unit has been interesting, especially with the original problem of being a shorted speaker. But, for now at least, the unit is operational but still has some issues to solve later on.
What’s going on?
After the speaker replacement, there was simply a lot of noise. half a volt of noise, to be exact. This was unacceptable, and there were two potential causes - the transistors, which most say to replace with the capacitors, or the carbon composite resistors.
The transistors helped considerably, but only Q3 and Q4 - this reduced the noise levels to about 1/4 of the original, as received value. That’s pretty good, Q4 was noisy and replacing it with a (not really much newer) transistor cleaned that up. The rest, unfortunately, it probably coming from resistors and I’ve read posts indicating that metal film resistors here help considerably. That will be another project later in the year.
Beyond that, I changed the meter drive level. The meter on the front is simply a mechanical version of an eye tube. That is, it’s not calibrated in any way, it’s just deflection. You really had to drive the thing hard to make it move, and I wanted that to happen at a lower input. Paralleling a resistor with the input resistor has allowed that to happen, and it’s placed as such it’s going to be easy to remove should I or a future owner want to bring it back to spec.
Anything else?
There’s nothing to say other than to pay attention to your replacement transistors. Leads aren’t always the same from part to part!
In the last part of this series, I put some new transistors in the device. And had to replace them once again because they were inserted incorrectly.
With that being fixed, it’s now time to check the unit for actual operation, make one final mod, and see what we have. Here’s the data:
What the manual says:
What the old transistors said:
What the new transistors said:
And finally, the noise levels:
The noise is very similar to what we saw after Q3/Q4 replacement, so the previous transistors, while providing the noise, don’t seem to be causing it. Some commentors on the youtube videos suggested the resistors are probably doing it, and I tend to agree. Old carbons are noisy.
The transistors also show some improvement in voltages, especially Q1 and Q4. Q1 should be off when measured in this configuration, and it is. Q4 is now closer to the Vdd rail, as it should be. Current draw on Q4 has also reduced, but is still high - probably because it’s amplifying noise. We’ll see what resistor levels do to that.
The final piece is to change the meter level. You really have to drive this thing hard to make it deflect, and I don’t want that. Calculations made during a forthcoming video suggested something around 3.1kΩ would make the meter happy. I accomplished this by paralleling a resistor with the 10kΩ input, and that resistor turned out to be around 4.7kΩ. It was stuck into the holes in the terminal strip so a future owner can easily remove if needed.
With that, this project is finished. I’ll probably start a new one later to swap out those old carbon resistors - but for now, that’s all and the device is working well. There’s one more post for the wrapup, and please check out the YouTube playlist for this device.
This is an interesting piece of radio history. Originally seen at the Findlay Hamfest in 2024, it showed up again in 2025, at which point I made the vendor an offer and took it home. Silvertone was the house imprint of Sears & Roebuck, once the “everything by mail” store in the USA. Sears expanded into physical stores, and devices like this unit would be sat on a counter at the radio department so you could test your batteries. Sears, of course, would be happy to sell you new “Powermaster” brand (another house imprint) batteries if yours were depleted.
The case itself is in ok condition, but it appears to have had some water ingress at some point. You can’t really see it in the image, but there’s some rust near the front of the slope. Note the two fuse clips near the top, those would have held the probes.
This is nothing more than a simple load tester, and is full of resistors to divide the voltage down for the meter and provide some load testing for the battery you’ve connected to it. It’s essentially a specialized voltmeter.
Before we go any further, does it work?
Sure does, this new battery reads decent on the unit. Probably will work a little better once probes are re-attached.
Of interest here is this knob. It appears to have been made in Chicago, and may have been sourced just for this device by Sears. After some research on this, the company, Davies Molding, appears to still be in business!
I have some cable and probes for the unit. Both of those were probably purchased at Mendelson’s Surplus many years ago, for far less than you’d get them today…I do miss that place.
It took a hot minute to figure out where the leads went, but after some investigation I determined that the negative goes to a big mass of resistors, and the positive goes to a terminal that has a second piece of phenolic on the switch. That one became obvious when I examined it with a magnifying glass and noticed it seemed to be a bit shinier than the rest, indicating it had been disturbed at some point in the recent past. Makes me wonder, did this thing have probes on it until recently and someone wanted those vintage pig pokers? Don’t know…
There’s still a bit left to do, I have some grommets on order for the holes and still need to solder the probes on, but this is what it will look like - eventually…
Stay tuned, this thing is going to work again soon!
Last night, I sat down with the EICO 150 and started to do the final wrapups on the unit.
Apply a signal.
Nothing.
Excuse me, what?
I focused on the function switch, but it seemed to be working properly after some study. I was able to put a signal in after Q1, and have it work, so that really threw me. I was able to bypass the input blocking cap on Q1 and get more signal, but highly distorted. So what’s going on here?
I decided to try the transistor(s) I pulled out for Q1 and Q2, starting with Q1. Removed the new Q1 and just tacked the old one back in. Immediate signal improvement. A check on the new Q1, and it’s just a diode now.
I get a new transistor for Q1 and check it. It’s good…and there it is.
Leads. The pinout on the transistor is different than the OEM unit.
I replace Q1 and Q2 (which is also a diode now) and I have output loud enough to make your ears bleed.
Rookie mistake. That’s on me. Always check your parts, and pay attention to what your tester is telling you.
Next post is actually doing the final tests I wanted to do, and to try and change the sensitivity of the output meter a little. Stay tuned!
I’m not sure why I didn’t run that lead behind the other one, instead of in front. Who knows.
This device sat in my rack for many years, both providing power and USB services for various things. It quit back in March, the power supply giving out like switchers do. You can find that post at the bottom of this one.
I was going to simply toss the device in the recycle bin, but…high current switchers are so cheap these days, there’s no reason not to fix it. So, let’s open it back up:
I wasn’t able to get an exact replacement for the supply, so I got a MeanWell equivalent unit. It won’t exactly fit inside, so we’ll make do.
For this task, I drilled a hole in the rear of the chassis and connected the new supply lines directly to the input of the hub using some good 16GA wire. I probably could have used a single strand here, as the internal wiring was somewhat lacking in size - but I used two, because just in case. They’re soldered together, and covered with two layers of shrink tubing.
The other ends were stripped, terminated with an aglet, and screwed into the supply, which now sits on top of the hub chassis. Open frame devices like this aren’t something that I’d put in a place where others could touch, but for me - it’s fine. The hub is back online and working.
There’s a youtube video for this device: Coming soon.
I’m happy that a piece of what could be e-waste is now back in service.
The EICO 150 was back on the bench tonight for some transistors. I started by adding a new terminal for Q4’s output, meter tie point, and transformer input.
I replaced Q4 with a NOS NTE152, new grease, and a new mica insulator of the proper size. Yeah, I got some shiners on the wire there, I tried to be careful but even my orange stick went through the insulation. I’ll get some nail polish and touch those later to coat the wire.
The part I used was an EGC152, which is the same thing as an NTE number, as NTE purchased the ECG line from Philips way back when, who in turn had purchased the line from Sylvania.
Q3 also was replaced at this time, with a generic 2N3906.
I did some noise testing to see what was going on now:
Noise levels had calmed down substantially:
There was still too much noise, so I went ahead and replaced the other two transistors.
Noise levels calmed down to what I would consider just background nosie. So are modern transistors that much better, or were these simply dying? Well, I replaced Q4 with a device that’s barely 10 years older than the one in there, so it’s certainly not “modern” - I’d have to guess that the transistors were simply breaking down from years of abuse. The Internet seems to think that these are a “replace with capacitors” item for this unit, so maybe there’s some truth in there.
There’s a little more to do, I want to make some measurements for posterity, and adjust the meter drive a little - you really have to crank the gain to get meter deflection, much like an eye tube unit. I want to dial that back some so a comfortable listening level gives noticeable deflection.
Back to the transistors, here’s what I pulled out of the unit:
Q1:
Q1 was originally an EICO number, but the previous owner replaced it with an SK3124, twice. I assume that’s because they blew it out by connecting input to B+ on a tube device. I used a 2N3391A as it’s replacement.
This part would have been an EICO part as well, but it had a Motorola HEP 726 instead. This is a generic AF transistor, ala 2N3904, etc. I used a 2N3391A as it’s replacement.
Here we have our first EICO part. This is the complement to the output transistor, and as such is PNP. Almost any PNP AF transistor will work, but may have to be selected to reduce or prevent oscillations. I used a 2N3906 which seemed to work well.
Here’s the big boy. It’s some EICO number, un-cross-referenceable. It’s an NPN power transistor, almost anything with a few watts of dissipation and an hFe of 100 or so will work here. I used an NTE/ECG152 on the suggestion of the Internets. It recieved a new mica insulator and new grease as well. The old insulator was just a little round piece of mica and the previous owner delicately balanced it under the transistor.
I’m happy with the noise levels here, all I can really hear now is the hum from the power supply. There’s a few things left to do, I want to measure voltages for posterity, and adjust that meter level some. One more part, and then final thoughts. Stay tuned!
There are only a few more parts to do on the range switch, so let’s get them done. There are two capacitors, a ?? type and a mica type. This is what they looked like:
And here’s what they ended up looking like:
I also took the time to replace some of the … more interesting jumper routing on switch terminals, and clean up the terrible solder job. Last thing to take care of on this unit is the power supply area, and I need to decide how I want to lay this one out. Stay tuned!
