*UPDATE* Ruhmkorff Coil & Crookes Railway Cathode Tube

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

The SCR is a TYN408G , are there any equivalents?, that one is £9 on fleabay!

 

[Sigurthr]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Ugh, man that is a disappointing driver. I thought you had a legit relay one, or at least a beefy modern silicon unit. Yeah, you're def not getting any meaningful x-ray exposure from that.

IMO, try a 100k resistor (or rather, I'd place a 1/4W or 1/2W 500k rheostat in its place and fiddle as needed) and see if it perks up as Cyp suggests. Beyond that I'd look into replicating an authentic relay drive, or even a hybrid variable relay/SCR driver (simple comparator to watch cap voltage + cap charger + relay trigger). Then you can upgrade the energy storage/pulse cap for more output.

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Feel free to post any suitable circuits Sig .

Can't find any 500K Rheo's, I'll drop in a 100K and wind it up on the variac...as soon as I find an equiv for the SCR.

Ped

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Guy sent me a pic...looks like a 20K to me?

Just got to find an equiv for the TYN408G .

 

[Cyparagon]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

So from the photos, this is what I'm reading:

The cap is charged on one half-cycle, then added to mains and discharged through the coil the next cycle. WHERE in the next cycle depends on the pot setting.

The rotary switch is clearly to switch the polarity of the coil, but I see no point in that :can:

(?) I'm not sure about that one diode. It might make more sense for that to be a diac.

(??) Yep, 20k metal film +/- 1%. But that's a strange selection, since it just sets the lower limit for the pot. A more standard 22k +/-5% would be fine for this (red red orange gold), but given that it has blown once, I would use a 1-watt or a half-watt as a replacement instead. Or if you've only got 1/4W, a group of four in a 2s2p arrangement of 22k 1/4W will give you 22k 1W.

This is of course assuming nothing else blew up.

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Thanks Cyp.

Yeah I was going to parallel 4 100K resistors, however, I think the SCR is suspect..the resistor died a violent, sudden death..no browning of the PCB showing prolonged stress...indicating something went wrong suddenly, overloading the resistor.
The pot is a (presumably linear) 220K.

Agreed the "polarity" switch is about as much use as mens tits. it has zero effect on the Crookes railway tube.

Just need to find an equiv for the TYN408G .

Thanks

 

[WizardG]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

The TYN408G is a 400V 8A SCR. Almost any 400V+ SCR with the requisite current rating should do the trick in this circuit.

 

[Sigurthr]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Looking over the original operation photos and seeing the build/design quality more clearly now, I'm not so sure improving the output power is a good idea on second thought. I'd replace that resistor and increase the heatsink on the SCR and call it a day. Any of the improvements I mentioned earlier are likely to cause secondary side flashover.

Regarding that polarity switch... it should produce a polarity change on the first peak of the output pulse, but if it is underdamped or undamped there will be oscillation counteracting the effect of the polarity change. So on the very high impedance loads you're using you won't see a polarity change, but there should be one for low impedance loads. The SCR only passes current in one direction and there's an antiparallel blocking diode, so the cap charging current pretty much only goes through the resistor string. The coil only operates every other half cycle, making the pulse discharge across the primary DC, even though there is a small reverse polarity current on the "off" half cycles.

For a new SCR try: TYN816RG STMicroelectronics | Mouser

or

TYN608RG STMicroelectronics | Mouser

You know, one improvement I would do, is place an antiparallel flywheel diode across the primary to catch the back EMF. In this case you'd have to place it across the polarity switch. It has to be antiparallel to the current flowing through the SCR, so anode on the junction of switch lead and capacitor lead (left side on Cyp's drawing). It's possible the resistor and scr didn't fail from only thermal overload, but rather that back-EMF induced overvoltage played a role. The funny thing is that's a horrible place for switch as it is subject to high voltage induction, so not only does it pose a failure point but user safety hazard. If the SCR or antiparallel diode ever failed-short it would place mains across the switch with the only current limiting being the reactance presented by the charging cap.

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Thanks for that.

I wasn't looking to improve or increase the output , just get the darn thing working .

Immediately after it failed, I ripped it open, evacuated the smoke and felt the heatsink on the SCR , it was cold, the SCR certainly didn't seem overloaded, I'm leaning towards back EMF like yourself.

 

[Cyparagon]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Do you have a bench power supply? Or even two regular supplies? I can give instructions on how to test the SCR if you do.

Immediately after it failed, I ripped it open...

Be careful. That cap has no bleeder that I can see, and can possibly remain at lethal charge for days.

 

[Sigurthr]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Do you have a bench power supply? Or even two regular supplies? I can give instructions on how to test the SCR if you do.



Be careful. That cap has no bleeder that I can see, and can possibly remain at lethal charge for days.

Good point! Do take care. Once you replace the blown resistor the resistor string will work as a bleeder if you short the mains input terminals - good for future maintenance checks.

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Yeah I'm fully aware of the retained charge, after building TC's for years..the lack of feeling in two of my fingertips are testament to that :crackup: .

And yeah I have a few variable PSU's .

Ped

 

[Cyparagon]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

You can probably test the SCR in-circuit (ie. just connect to it as it is) in this case. If it fails, just desolder it and test it again to be sure.

Connect a bench PSU across anode and cathode. + to anode and - to cathode. Current limit should be 0.5-2A (not critical) and voltage should be 5-10V (not critical).

The device should be drawing no current at this point. If it does, it's bad.

Simulate a gate pulse by connecting a ~10ohm resistor (not critical) to anode on one end and briefly poking the gate with the other end.

The device should now be drawing enough current for the PSU to current-limit. If it doesn't, it's bad.

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Right, PSU connected across A/K =no current drawn.
Pulse the gate and the SCR conducted and current limit kicked in, and kept conducting (even with the gate not being energized) , it latched like a switch. If i reset the PSU, it again drew no current, until the gate was energized.

 

[DashApple]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

The SCR will latch with a DC supply , its how they work , on a normal SCR they turn off when the holding current drops below the minimum required to hold it on , Hence with AC they turn off every half cycle till triggered again

The SCR looks to be good if no current is fowing before the gate is pinged , once it has been the SCR will latch on till you turn the PSU stopping the current flow .

 

[ped]

Re: Ruhmkorff Coil & Crookes Railway Cathode Tube

Great, I'll whack it back in circuit, along with my franken-resistor and see what happens.