What's new
What's new

C16J and Spindle Module Replacement Project

Timothy Jones

Plastic
Joined
May 11, 2007
Location
Central Massachusetts
I've decided to tackle the bull by the horns and come up with solutions for several components on the 10ee that I've seen fail. I have a tube type machine from about 1972 (not a WIAD). My list of what fails (based on experience) are the following:

1. Contactors and relays in the starter compartment
2. C16J thyristor tubes
3. In the spindle module: The rectifier diodes and in general the soldering which can have cold joints. I've not seen one fail, but I suspect the selenium suppressors may fail with age. Also the potentiometers can go out of spec or just fail.
4. The speed control potentiometer

Going down this list, I don't see much of a problem with the contactors and relays. There are enough equivalent components out there that replacing one is probably not so hard. They are basically mechanical and there is also always a possibility of actually fixing one. The C16J tubes are a different story. If you can find one, the current asking price seems to be exhorbitant. As for the spindle modules, I know from personal experience that they can lead you around the barn and back. I kind of like Monarch's advice of keeping a spare handy. I suppose Monarch might be able to sell you one, but I'd hate to contemplate the cost. Lastly the speed pot. That is real problem. It is 2 gang with each half being non-linear. Basically there is no replacement.

You might ask why not just go with a VFD or modern DC drive like so many others. My answer is that I think the tube drive is way cool. I love the purple glow. I want to keep my machine original (or at least in the spirit of original). There may be more practical reasons also. Doing a conversion is expensive and time consuming. I can run my lathe direct drive (no back gear) at 10 rpm with smooth unstoppable torque. I don't know, can a VFD do that? The bottom line is that I think Monarch did an excellent job with that drive and I don't want to change it with a lot of work for something less. I'm sure there are others of you out there who also want to keep your lathe unmodified but are running up against the brick wall of irreplaceable parts.

So I'm starting this thread to cover a project to first create a modern SCR based drop in (affordable) replacement for the C16J, and second to create an (again affordable) replacement for the spindle module. (If this works, I'll leave the speed control pot for a different thread). I think some people have already suggested the SCR idea but I'm not sure anybody has actually tried it.
 
I have several brand new ones in the box.

I just checked Ebay and you are correct, they are nuts for what they asking.

PM me if you are interested.
 
So I'm starting this thread to cover a project to first create a modern SCR based drop in (affordable) replacement for the C16J, and second to create an (again affordable) replacement for the spindle module. (If this works, I'll leave the speed control pot for a different thread). I think some people have already suggested the SCR idea but I'm not sure anybody has actually tried it.

I have looked into this a bit. It shouldn't be too difficult. I have also experimented with magnetic amplifiers. I have had a MG 10EE running on them, both single and three phase. The nice thing about them is that they are intrinsically current limiting. You can set one up to deliver the the maximum rated current to the motor and save the brushes et al from surges. If you want to correspond about this, email me at [email protected].

Bill
 
Attached is a schematic of an SCR approach to the C16J. The big problem in doing this is that the tube grid drive is high impedance (240kohms series resistance) and not sufficient to drive the gate of a beefy SCR. So I've added a MOSFET Q1 to drive SCR D1 gate.

The tube requires 20 amps at 2.5 volts to heat the filament. The SCR has no filament, so I've added a pair of back to back diodes D2 and D3 which basically leaves the filament transformer secondary open circuit while preserving the cathode current path out through the transformer center tap. This is quite nice because it eliminates the filament current entirely. (Both filament transformers are unnecessary and, if you wanted, you could connect D2 and D3 cathodes directly to monarch circuit 51. There is no harm in connection through the transformer secondaries as with the tubes, and for direct pin for pin replacement that is what you should do).

The remaining components are for protection of the MOSFET and SCR.

The tubes can switch up to 16 amps, so one worry that I have is keeping D1, D2, and D3 cool. If you figure a typical diode drop is 1 volt that means the each device may need to dissipate as much as 16 watts which is quite a bit.

The C16J tube starts conducting at a grid voltage right around 0. The SCR turns on at around a volt or so. The MOSFET however won't start to conduct until the gate/source voltage is upwards of several volts. I worried about this at first but then decided that the exact gate turn on threshold was probably not very critical. I've built a prototype and measured it at just around 5 volts. This is a difference from the tube, but given the simplicity of an uncomplicated FET buffer I felt it was worth trying it out before getting all fancy.
 

Attachments

  • c16j1.pdf
    16.3 KB · Views: 251
It isn't clear what the status of your lathe is. Are these parts failing, or are you saying that they are the most likely to fail in the future? You wouldn't want to end up like me with a lathe that runs perfectly with boxes and boxes of parts stacked up everywhere that you never use.

Are you trying to fix something that isn't broken?

I could try to argue with you and say some 16Js last 40 years, they are a pain to pay for, but how often is anyone changing these?

The rectifier diodes are really trivial. They are easily replaced. I bet I could replace all of them for less than 5 bucks with new parts that you don't even have to fit in the holders (just solder them in).

Potentiometers are easy to source. I think too many 10ee owners think they have to buy OEM parts when there are improved and cheap replacements available if you look for them.

John
 
It isn't clear what the status of your lathe is. Are these parts failing, or are you saying that they are the most likely to fail in the future?

Are you trying to fix something that isn't broken?

John

My motive is this: I have one lathe working nicely with good tubes. I have 4 others that aren't working. I thought that perhaps I could solve my own problem while at the same time offer some solutions of interest.
 
Read the archives of this forum. There are at present 153 pages. In fact if you go to page 153 and look at the bottom of the page, the very first post is " Monarch 10ee drive system - repair or replace ?" that discussed SCR replacement. Start there and read upward.
 
John - I should probably go back and exhaustively read the archive. Thanks for the suggestion. I have followed these discussions sporadically for several years but may have missed some. I've have managed to produce some replacements that appear to work. I apologize, however, if I am covering old ground. I also apologize if I am somewhat inept in using the forum - I don't post much.

I decided the easiest way to test the C16J replacement was to build a printed circuit board (I hope I've already posted the circuit schematic) and mount it in a tube in a way somewhat similar to the vacuum tube. At first I thought paper phenolic which is a good insulator would be good, but then balked when I found out how much it costs. I finally settled on polycarbonate. I liked this because it sort of looks like the glass on the C16J.

I mentioned in my last post that I was worried about keeping the parts cool. The tube is 9 inches long and mostly filled with copper heat sink. I'm not happy yet with how the heatsinks are attached because it won't work well with thermal expansion. For these prototypes, the pcbs are also held only by friction which is inadequate but good enough for testing.

The first picture shows the pcb. The second shows it assembled with heatsinks. The third shows the mounting tube, finally, voila, a pair of fully assembled C16J vacuum tube replacements.
 

Attachments

  • IMG_0316.jpg
    IMG_0316.jpg
    24.2 KB · Views: 281
  • IMG_0317 1.jpg
    IMG_0317 1.jpg
    19.6 KB · Views: 223
  • IMG_0320 1.JPG
    IMG_0320 1.JPG
    94.9 KB · Views: 303
  • IMG_0318 1.JPG
    IMG_0318 1.JPG
    74.9 KB · Views: 224
Potentiometers are easy to source. I think too many 10ee owners think they have to buy OEM parts when there are improved and cheap replacements available if you look for them.

While I agree with your other points I have to diagree here - the pots are not easy to source. Using a WiaD drive as an example the pots are/were custom with opposing halves of the resistance shorted in the armature pot and the field pot. Further, they were a 4 watt wire wound high spec pot (Clarostat 58C1) that is not longer available except from old stock.

Just for kicks I spent the time with someone here to make a replacement pot - it turned out to be possible but was far from simple requiring several pots for spare parts (at least the parts from a tandem assembly and a couple of pots each with double the needed final resistance) and some PC trace repair paint. In the end it worked well and I hope for a long time, I never heard of any failure.
 
Attached is a schematic of an SCR approach to the C16J. The big problem in doing this is that the tube grid drive is high impedance (240kohms series resistance) and not sufficient to drive the gate of a beefy SCR. So I've added a MOSFET Q1 to drive SCR D1 gate.

It needs a driver to run some neon tubes in the WiaD cabinet so those of us can still enjoy the light show.
 
That's very interesting. I'd be interested in how you did that. I'd like to go into my own non-linear pot solution in another thread when I've had a chance to work on it and try a few things. My thought, however, was to get two pots and gear them 2:1 to a master drive gear with half the teeth cut off. Then the 1st half turn of the master gear would engage 1 pot for one turn and then the 2nd half turn would engage the 2nd pot for one turn.
 
That's very interesting. I'd be interested in how you did that. I'd like to go into my own non-linear pot solution in another thread when I've had a chance to work on it and try a few things. My thought, however, was to get two pots and gear them 2:1 to a master drive gear with half the teeth cut off. Then the 1st half turn of the master gear would engage 1 pot for one turn and then the 2nd half turn would engage the 2nd pot for one turn.

If we're talking about the pot modifications: all I really did was to remove the insulation from half of the pot travel (the second half for armature control, the first half for field) using paint remover followed by MEK and some really careful scraping. Thinking that I was unlikely to get the resistance wire tinned I used silver bearing trace repair paint on the exposed wires to short the resistance in the windings. It seemed to work but like I said I don't have long term results.
 
Timothy,

Amazingly quick progress to already have a PC card ready to go!

It needs a driver to run some neon tubes in the WiaD cabinet so those of us can still enjoy the light show.
LOL. How about using some of the LED chips that are used for "designer" light bulbs to provide the glow?

Cal
 
I started working on this in the beginning of November. I see that I may have given an impression that I was only starting this effort when in fact I am in the middle of it. I should have said that right up front. I only decided to post yesterday when I began testing, and I've only just installed the finished tubes in the photo below this morning. I've also done work on the spindle module which I would like to share as well. I am still trying to get used to the medium of the forum which I know is interactive and which I am still not quite familiar with. So once again, sorry for any confusion.

I plugged my new tubes in yesterday but the first attempt ended in failure. They started working, I got immediately excited and began to push load and rpm. That lasted about 30 seconds and then the lathe lost speed control below about 1500 rpm. 1500 was the lowest it would go. Also it sounded wrong. I was worried that I had fried one or both SCRs which could imply that there was something seriously wrong since they are quite robust and not really prone to failure unless you really abuse them.

After taking them out and retesting them on the bench, I found that one was still working and the other had a broken MOSFET. I felt this was a very good sign since it is easy to blow one by exceeding the maximum gate/source voltage. After contemplating the circuit for a bit, I decided that since there was no protection against positive swings of gate voltage (a situation that initially I thought was not needed) that maybe the gate voltage protection diode D4 needed to be bidirectional instead of unidirectional. I put a 24 volt transient voltage protection diode in place of the regular diode I had there and tried again. That seems to have worked.

The green traces in the pictures below show the anode to cathode voltages across first, a C16J tube, and second, an scr replacement, under identical speed and load conditions. As you can see the two traces are almost identical. (Note that the probes were AC coupled so what looks like 200 volts on the scope is actually 0 volts across the devices. AC coupling moves the average value of the wave form to the zero point on the screen and in this case the average value is about -200 volts).

I do notice a slight asymmetry in the conduction intervals of the SCR replacements while the C16Js seem to be quite symetric. I suspect that this could be due to the higher gate turn on thresholds of the replacements and also variability between them. However, I didn't notice that this seemed to have any significant affect on the way the spindle felt and my perception of the operation of the lathe. Everything was smooth, very stable, and seemed to behave equivalently to changes in torque. The one exception to this was at very very low RPM, say less than 10 rpm, I felt the SCR tubes exhibited a very slight delay in response to torque changes although I might be imagining that.

I've run the lathe for an hour or two on the replacements. I am, of course, still uncertain about reliability and longevity, but I think it is a start.
 

Attachments

  • IMG_0324.jpg
    IMG_0324.jpg
    61.9 KB · Views: 348
  • IMG_0320.jpg
    IMG_0320.jpg
    51.9 KB · Views: 207
  • IMG_0319.jpg
    IMG_0319.jpg
    49.2 KB · Views: 200
Timothy,

LOL. How about using some of the LED chips that are used for "designer" light bulbs to provide the glow?

Cal

If you open the door it is very reassuring to see the glow. It tells you that everything is working, and I saw one post in which imbalance appeared to indicate a problem, so maybe there is some function. I wonder if led intensity could track current flow...hmmm.
 
If you open the door it is very reassuring to see the glow. It tells you that everything is working, and I saw one post in which imbalance appeared to indicate a problem, so maybe there is some function. I wonder if led intensity could track current flow...hmmm.

If you were to bury the LED into the polycarbonate tube and frost the tube a bit...

Anyway, seriously cool work. Congratulations!
 
There are actually several separate "projects" involved here;

1) The "light show":

Argon glow lamps are sold AS sources of illumination for special needs. No need to expect power Thyristors to produce that particular wavelength of light merely as a byproduct of operation. I'd actually suggest the often insanely cheap screens surplused as automakers changed backup camera displays, or those sold for "aftermarket" and DIY. Sizes down to 3" or so, round or rectangular, 65K+ colours, and you could also display camera shots of "whatever".

Let's send that project a coupla doors down the hall, 'coz there' no limit to it but imagination. And.. it hasn't anything to do with making chips, no way, no how.


2) RPM control, better-yet, possibly improved regulation under varying load:

Replacing the control pots is not hard. One simply shorts half - or near-as-dammit half - of the turns of a common wirewound. Nor are they unobtanium. Plenty of them around from VFD or DC Drive conversions. ISTR I have three sets of the toilet-seat sized ones from MG conversions. The challenge there isn't the half-range deal. It is simply the cost of very high Wattage units.

If you want to use a "mechanical" approach to crossover-point or even shape of the curve, no need for weird gearing. Just use "slide" type potentiometers and levers, push-pull rod, weed-wacker cord and spool/hub, or plate-cams to control resistance. Or... use a packaged linear regulator and power pass-element to get the same effect.

Mind - these are not needed for either VFD or Solid-State SCR conversions anyway. Nor should they be needed for a DIY hollow-state project.

Wiser Engineers with larger budgets have long used dynamic "Field REGULATOR" circuits that auto-sense the changes in load and supply the needfuls to BOTH Field and Armature requirements when operating in the Field Weakened range. A single 'vanilla' potentiometer controls the lot, and one gets far, far better stability.

The various 10EE drives were built to tighter budgets for simpler needs than the high-grade DC elevator drives they were derived from is all. A machine operator could "set up" for a give cut, even a complex one. An Elevator has to take whatever load gets on or off at each stop. "Experience" wasn't much help, even when there was still a human operator. Which was being phased out as controls got better.

3) The DC control panel:

It is time - perhaps past the time - the collective "we" came up with more easily supported counterparts to these.

Challenges?

The OEM FWD-OFF-REV power contactors have BOTH a mechanical interlock to prevent powering both directions at once, AND have similar switch and relay logic in the selector controls. Solid-state contactors, nor even Mercury-Displacement ones, would not have that mechanical interlock.

BFD.

Well-engineered controls with fuse or circuit-breaker protection are usually more than good enough to prevent asking for FWD and REV at the same time.

The "sensitive" relays used to control braking, acceleration, and monitor for field loss (or field simply too damned weak), are not hard to built counterparts to. Solid state not an absolute requirement. Stock "sensitive" relays can be biased to work. Common approach "back in the day" and the originals were exactly that.

IMNSHO, replacing the DC control panel with more common/modern components is a worthwhile exercise.

I'd personally use some of my stash of Mercury-Displacement contactors for the FWD-OFF-REV main. No external arcs or sparks, the "contact point" material is self-renewing. Downside is the 10EE would do longer operate reliably when hung from the overhead and run upside-down. I think I can live with that limitation.
:)

Main power contactors aside, the rest - Field sensing, braking, acceleration - is easily done in solid state - analog OR digital.

Why do we not see more work in this area? Simple lack of need.

OEM panels are seriously long-lived, generally repairable. VFD or Solid-State DC Drive conversions don't use ANY PART of them at all.


4) Hollow-State to Solid-State Thyristor conversion:

Not wise. Not a direct conversion. The two races have the same function, but different operating environment and characteristics.

By the time one has taken all that into account, the SCR-class DC Drive has been "re-invented", and not very well.

Download the manual for a Eurotherm/Parker-SSD 512C-16. See how much of that Analog complexity you feel up to duplicating.

Be aware that decent used ones only cost $200 to $400, and even ones bought brand-new with warranty are "affordable".

"Bottom Line"?

I could see merit in coming up with a newer, smaller "DC Control Panel" that could run any of MG, WiAD, or Modular Drive 10EE with simple option strapping.

Otherwise?

Either do the VFD or DC Drive conversion.. or preserve the MG, WiAD, or Modular drive as it was originally built.

If nothing else, Monarch and their suppliers / contract development partners (Reliance Electric and Engineering was a bleeding LEGEND, already for that, and long before Monarch's 10EE....) had waaaay more invested in testing and tuning and responding to field-experience feedback that any of us 'ere are likely to live long enough to replicate.

2CW
 
The dual control pots are easy. Apply twice the desired output voltage to one section and clamp the voltage at half that with a voltage divider and a diode or a zener and do the same on the other with it clamped to the high end (diode turned around). By adjusting the reference voltages you can change the crossover point.

I worked with thyratrons when they were the only game in town. They are more tolerant of brief overloads than SCRs, but that is about the only good I have to say about them. They are fitful beasts that age. I recently did the Easter egg routine with the Modular I maintain, changing tubes until you get even glows. In the mid 50s a company I worked for had thyratron powered 400 cycle inverters we used to test military equipment. When they unbalanced we would trade them from the spare box until we got equal glows. The extracted ones went in the box for next time. We never bought a new tube, just played musical chairs with them.

Remember that we are running on the last remaining C16Js and no one in his right mind would tool up to make more. Timothy Jones is on the right track. Personally, I would not try to look like tubes and just mount a heat sink there, but he obviously knows what he is doing so I leave all that to his judgement.

Re diodes in the control board, I had to replace a couple because the field timer stuck on. The lathe is in a job shop where anyone can use it, no dedicated operator. They were leaving the power turned on all night, which put field current on the whole time. The head would get so hot they complained about the effect on accuracy and I think the heat contributed to the diode failure. Fortunately I had replacement diodes and I replaced the time with a modern one.

I do appreciate the light show. My first high powered ham rig used a 304TH for the final amplifier, two 304TLs in AB1 for modulators, powered by a pole pig with four 866 mercury vapor rectifiers in a full wave bridge, all mounted just behind their panels with windows so I could watch. At night I would turn out the lights for dramatic effect. With 375 watts of filament heat and approaching 900 watts of plate dissipation in an attic room with no air conditioning, only a 17 year old would have endured it just to talk to people.

Bill
 
Bill - That sounds like a good idea with the zeners. I have assumed that the pots are a real problem, but as both you and rkep[ler seem to have workable solutions, maybe not quite as bad as I thought - thanks.

Going back to rep[ler's idea to recreate the glow, at first I was thinking "well that is a fun idea but probably a bit frivolous". Well maybe not. I think my SCR replacements are unbalanced and that might be a good way to rectify that. If there were a pot in place of R1 and R2, one might be able to tune the rectifiers into balance with the help of a glowing indicator.

The polycarbonate tube could make an effective light pipe. Machining a bevel on the end of the tube makes a nice frosted finish, increases the surface area and emits light at a pleasing viewing angle. I tried it with a flash light and it works nicely. I'm thinking of plastering a flex circuit with several side mounted backlight type leds on it around the inside of the tube under the aluminum mounting where you can't see them. I would have the leds projecting into holes drilled in the tube so they can inject light into the cross section of the plastic.

I'm toying with several simple led drive circuits that would be constant current, but modulated based on the thyristor on time. I think this might give the appropriate light modulation but I will have to try it.

One final twist is that the machined bevel could be cut as two circular grooves, some distance down the tube from the end, one on the inside and one on the outside, overlapping, but only partially through, which would have the effect of extracting 100% of the light at a spot other than from the end. This spot could the same as on the original C16J.

Maybe this illumination thing is an idiotic and unnecessary waste of time, but I am seriously thinking about it. If you have an opinion and care to share it, let me know.
 

Attachments

  • IMG_0322 1.jpg
    IMG_0322 1.jpg
    32.3 KB · Views: 115
Lights would appeal to what I call the magpie nature of the human animal, attraction to shiny things. One time I was chrome plating some rifle parts like the trigger guard, etc. I was playing with the temperature and current to achieve a finish that was starting toward satin chrome but only slightly, sort of like you had blown your breath and fogged them a little rather than the usual glaring finish. I showed them to a friend, who was unimpressed. Then I showed him a part that had gone to full bright like most decorative chrome. His response was that I was finally getting it right. Oh, well.

Bill
 








 
Back
Top