What's new
What's new

User's report on solid state thyratron replacement

9100

Diamond
Joined
Nov 1, 2004
Location
Webster Groves, MO
For over ten years I have been playing ring around the rosy with thyratrons on a Modular 10EE. One of the problems has been that it is in a shop where any of a dozen people can use it and most of them think that once they have hit the stop buton, no further action is needed. Frequently the tyratrons would be lit with no one running it because the last user went off with it still energized. One tube finally burned out and we had pretty well used up the spares. We bought a pair of Scissio's solid state replacements and I installed them yesterday. They really are drop in (or more accurately bolt in) replacements. They ran right off the bat and a little tweaking of the max speed and IR Comp controls was all that was needed.

We will have to see how they hold up in service, but for now they get 5 stars.

Now, if someone would make a solid state replacement for the C3J, we could be rid of thyratrons.

Bill
 
Don't you miss that purple glow? Seriously, this is great news. Between these thyratron replacements and the modular box replacement, the DC drives ought to go another 100 years. How hard would it be to build a modular from scratch, using these components?
 
You didn't mention where you bought yours or what they cost, 10EE owners want to know. :-)
 
Ok, let me redefine "starting from scratch": suppose you have a WIAD machine as the starting point. That means you have aging insulation, components that are obsolete, a system that has probably been overheated for many years. By comparison a modular machine has components that are largely still available, it is more reliable, less prone to overheating. The DC panel is much simpler. The machine wiring is vastly simpler. However, and this is a crucial thing to keep in mind, the modular is a direct evolutionary step from the WIAD, it is not a total new species. There are more similarities than differences between modular and WIAD. Now we have two recent developments that take the evolution one step further, and they apply to both members of the family: solid state thyratrons and replacement diode boxes. But that doesn't solve all the problems, especially for WIAD. So what is the logical next evolutionary step, one that would be in reach of the avg 10EE tinkerer, since that by definition includes all present company that owns a running 10EE with a DC drive.

So as a starting point, you have all the transformers, they seldom go bad. The real question is, was I correct in the last paragraph? Are all the necessary components available to build a modular? Or even better, are there components available to simplify the design of the modular even further? One quandrant, four quadrant, breaking resistor - all that is food for discussion.
 
Don't you miss that purple glow?

Not a bit. When I was 20 years old, in 1955, I went to work at Greenleaf Manufacturing. My first assignment was sorting out nuts and bolts, then I was handed a schematic for a gyro temperature control that used a 2D21 miniature thyratron to control a relay. I got it working but when I leaned over it, blocking some light, the firing point changed. It changed with temperature, phases of the moon and the receptionist's menstrual cycle. It also got me some bad press from the vice president because it was such a mess from countless changes, trying to make it work logically.

Finally, I drew up a schematic using a 6BN6 tube to sense the output of a thermistor bridge. The 6BN6 was an FM discriminator that had almost snap action in this service. It worked so well that it became the standard control for our gyro test systems, my first commercial success.

After a lifetime of working with mercury vapor rectifiers, gas lasers, and other assorted argon, xenon, and other gas filled things, the glow from 10EE tubes just means having to keep changing them until I get a matched pair. Good riddance. If anyone wants something to light up, I have some big thyratrons I got in a package deal I bought for the ignitrons, I can make you a good deal.

The link Cal listed doesn't work right now, but it should be good in a while.

The replacements were for a customer's lathe and their purchasing agent bought them, so I don't know the exact price, but I think it was around $175. Don't hold me to that, just an imperfect memory.

Bill
 
The overall build quality looks very nice and the product seems well-designed. It also included an 8-page user manual.
The downside is while they do say it is likely to work in a WIAD lathe, there are no instructions for using it/them in one.
So, you modular guys have it easy and one of we intrepid WIAD owners will need to be a guinea-pig for this swap.
 
I would expect them to work the same. The tubes we were using came from a WIAD that the owner converted to an AC motor / VFD drive. The thyratrons were doing the same thing in both lathes. In any case, you get a warranty.

Yesterday one of the owners of the shop asked me when I was going to install the SCR units. I said they had been in there for two weeks. His reply "No wonder it ran so well."

Bill
 
I would expect them to work the same. The tubes we were using came from a WIAD that the owner converted to an AC motor / VFD drive. The thyratrons were doing the same thing in both lathes. In any case, you get a warranty.

Yesterday one of the owners of the shop asked me when I was going to install the SCR units. I said they had been in there for two weeks. His reply "No wonder it ran so well."

Bill

I think it will work the same but need to translate: "In this lathe, the two filament transformers ... For these transformers, connect the K lead to the bottom secondary terminal. The T6 and T7 transformer secondaries are phased in relation to the thyratron gate drive circuit and therefore it is important to connect the K lead only to the bottom secondary terminal as shown and not to the top one."
 
The AC from these transformers will have a peak either on the corresponding peak on the high voltage or the negative peak. The voltage from the transformers will be added or subtracted from the gate voltage that triggers the SCRs, so having them mismatched will cause uneven operation.

My first suggestion is to contact the maker but he may not have sold any units to a WIAD owner yet. To an extent, we are still beta testers because he does not have that many in the field.

In the absence of information from him, I have a modular lathe available and the works from a WIAD that I can test and compare them. If I was doing it on my own lathe, I would connect the return to the transformers' center taps and eliminate the problem, but he is selling to machinists, not electronic technicians so he tried to make it as simple as possible.

Bill
 
Last edited:
The AC from these transformers will have a peak either on the corresponding peak on the high voltage or the negative peak. the voltage from the transformers will be added or subtracted from the gate voltage that riggers the SCRs, so having them mismatched will cause uneven operation.

My first suggestion is to contact the maker but he may not have sold any units to a WIAD owner yet. To an extent, we are still beta testers because he does not have that many in the field.

In the absence of information from him, I have a modular lathe available and the works from a WIAD that I can test and compare them. If I was doing it on my own lathe, I would connect the return to the transformers' center taps and eliminate the problem, but he is selling to machinists, not electronic technicians so he tried to make it as simple as possible.

Bill
The other thing you could do is disconnect the primaries on the filaments transformers.
 
The other thing you could do is disconnect the primaries on the filaments transformers.

A good point I hadn't considered. We don't know if the approximately 1.7 volt peak is part of the system or simply tolerated and configured to affect both SCRs equally. I suspect the latter because if the high voltage and filament voltages are in phase, the instantaneous filament voltage would only have much effect in middle speed operation. Going very slowly or near maximum speed, the filament voltage would be near zero and not a factor. If the filament connections were opposite, the operation would be rougher in the midrange. It could be that if the filament voltage was on a negative peak during the high voltage positive half cycle, pulling the SCR cathode negative would have the same effect as driving the gate positive and that SCR would conduct through most of the half cycle.

Bill
 
I replaced the non-functioning module in my 1965 EE with a new unit from Tim. The build quality is excellent, the shipping was fast and I think the price was very reasonable ($345). This new module fixed all the problems I was having with this lathe and it now works great.

Had the C16J's been bad I would have replaced them with Tim's units as well.


Spindle Module Replacement
 
Thats pretty fantastic. I have a carefully packed and labeled set of C16Js acquired at high cost when I thought they were going full dinosaur.

Labeled in case the lathe finds another caretaker someday. Without the many labels my fear they would potentially go straight to the trash as some obsolete glass bottles.
 
Two months in, the machinists are delighted with the SCR replacements. The lathe runs smoother with the uniform firing, down to barely turning. I am equally delighted with being rid of hunting through tubes to find a balanced pair. I have one functional tube to dispose of. When I asked the owner what he wanted to do with it, he said "Take it with you. I never want to see it again."

Bill
 
That sounds very good.
One of the things the ee does very well is cutting screw threads. The ability of the machine is advanced over other machines because of the precision spindle control.
Basic, at threading speeds the carriage will stop at the end of the thread automatically within a couple of thou every time "half nuts always engaged".
Cutting threads with the half nuts always engaged has many advantages. The machine will easily cut threads at 200rpms in back gear at stop within .002" from a shoulder.
If things are not right with the drive, it can drop out, dynamic braking is lost.. The tool will crash into the work, if you don't have time to grab the half nut lever.
Pushing the vacuum tube module drive to the limit, because you have to make money, I found the drive will warn you when its about to wig out.
Anyway, I think the drive can be tested with an 8" chuck, machine in backgear at about 200rpms. Run the machine forward 4 seconds, let spindle stop total 4 seconds, run machine in reverse 4 seconds.
Repeat 20 times. My best module machine will do that, but if I try to speed up the switching to reverse by 2 seconds, the drive will overheat and kick out, yikes.
What I described above relates to metric and other precision threading always leaving the half nuts engaged.
I believe the machine is most stressed using that function, other then overloading the machine in other ways.

Did I say how nice it is to cut screw threads on a Monarch ee ?
 
The thing I would like to have is a puller like the Hardinge compound that would be automatically activated.

The other thing I would like to try would be some filter capacitance to average out the SCR pulses. My lathe is a Reliance MG, so it doesn't have the problem. I did try capacitors on a magamp drive I played with and promptly boiled them. There were some massive currents flowing in and out of them. A filter choke like is used on transmitter power supplies might be in order. I don't dare suggest that to the owners of the machine I converted to SCR. They refer to the lathe with one particular operator as "The money machine". Wanting to have more down time would get me beaten severely about the head and shoulders. In a shop with three Moris, a Hwacheon, and a Harison, the 10EE is the favorite.

Bill
 
With an inductor only, you need to be sure there is a current path to ground on the input side of it. Otherwise the energy is wasted on a negative pulse that goes nowhere.

Re sharp spikes, one of the things I like about saturable reactor/magamp systems is that the core does not saturate all at the same time so the top of the pulse is rounded, without the harmonics extending to many multiples of the AC frequency. I had a three phase reactor going with a full wave bridge rectifier, giving three times the pulses, which naturally ran a lot smoother and would not require as much filtering. All the cores and copper didn't come cheap, though.

Bill
 








 
Back
Top