Film Vs Ceramic Capacitors for Tube Unit (Film in AC application?)

The 1N914 diode is a signal diode. It will explode with a loud BANG! if you try to use it as an AC rectifier. Pick some nice low Qrr diodes so there's less noise. FREDs are good, so are Schottky. Make sure you have enough PIV rating.

You're fusing the transformer secondary before the rectifier, so fuse is placed between the winding and the rectifier.

An Amphenol CL-80 is 3 amps max, resistance @ 25 degrees C is 47 Ω, and a CL-90 is 2 amps max, resistance @ 25 degrees C is 120 Ω. Arts prefers Ametherm so suggest you research that. That's for the primary side. You'll need to size the thermistor for the load. If it isn't enough the thermistor never gets hot enough to drop in resistance. So figure out what the load will be.

Would this diode be a good choice? Also I don't think I have the skill to estimate the load before repairs are complete, so I will be ordering the thermistor at a later time. Like I said earlier, I know next to nothing about inrush limiters. Would it be useless (or even dangerous) to order and install the CL-80 before the thermistor is ordered?
 
Would this diode be a good choice? Also I don't think I have the skill to estimate the load before repairs are complete, so I will be ordering the thermistor at a later time. Like I said earlier, I know next to nothing about inrush limiters. Would it be useless (or even dangerous) to order and install the CL-80 before the thermistor is ordered?

Diode is the wrong type. That is 600 V @ 3 A. You need a higher voltage rating, not higher current. Get something around 1,000 V and you'll be fine. Standard package is fine.

Your inrush limiter must be hot enough to drop in resistance.
 
Diode is the wrong type. That is 600 V @ 3 A. You need a higher voltage rating, not higher current. Get something around 1,000 V and you'll be fine. Standard package is fine.

Your inrush limiter must be hot enough to drop in resistance.
I'm not afraid to admit I'v run into a wall, I'm having trouble searching for low Qrr diodes. Digikey does not let me sort by that value, and I can not even find mention of that value in some data sheets. Can I use the 1N4007 diode used in that schematic? Also would I be right to assume that moving the leads "over" one pin and bridging with diodes is a solution only for the Dynaco, or should I do an identical modification to the 6X4 rectifier?

Also, This is the capacitor I have selected for the death caps.

Thank you for all of your help with this project, this is only my second project working with tubes.
 
The types with lowest Qrr are Schottky Barrier Diode or Schottky Silicon. The magic phrase on the datasheets are “Reverse Recovery Charge” or “Reverse Recovery Current", expressed as Qrr but sometimes as Irec.

The 1N series is not a good choice. 500 nC Qrr vs 20 for the SBD. Cost to get a better diode is very low, maybe a few dollars.

Yes, you should do the same type of modification of adding diodes to take up the PIV stress on the 6X4. The Dynaco modification does this by putting diodes in series with the rectifier's cathode-plate diodes. You'll need to figure out the pinouts to see how this works. If the 6X4 shorts cathode-to-plate then you've got solid-state rectification with no delay. If it shorts heater-to-cathode you lose the transformer unless you have a very small fuse at the transformer's B+ secondary. The latter is far more likely than the former.
 
I usually use UF4007 as a standard replacement rectifier for this sort of thing. With the tube rectifier in there as well I suspect the SS diode's properties are not as crucial anyway.
 
I usually use UF4007 as a standard replacement rectifier for this sort of thing. With the tube rectifier in there as well I suspect the SS diode's properties are not as crucial anyway.

The issue is that the Qrr creates AM band noise which can cause transformer stimulation and ringing, and tubes are high impedance so that noise can cause ringing and oscillation.

Cleaner power is always a good idea.
 
With the high impedance of that tube rectifier I'd expect any ringing to be damped out quite nicely. Maybe bypass them with a small value cap if its still an issue.
 
With the high impedance of that tube rectifier I'd expect any ringing to be damped out quite nicely. Maybe bypass them with a small value cap if its still an issue.

Qrr creates an AM carrier wave modulated by mains harmonics. Not as readily damped as ordinary low-frequency DC ripple.

The snubber is a good idea for diodes and ransformers. Using SBD reduces Qrr to vanishingly small levels, like 20 nC compared to regular silicon. So the snubber is less critical. Stll a good idea, though.
 
The types with lowest Qrr are Schottky Barrier Diode or Schottky Silicon. The magic phrase on the datasheets are “Reverse Recovery Charge” or “Reverse Recovery Current", expressed as Qrr but sometimes as Irec.

The 1N series is not a good choice. 500 nC Qrr vs 20 for the SBD. Cost to get a better diode is very low, maybe a few dollars.

Yes, you should do the same type of modification of adding diodes to take up the PIV stress on the 6X4. The Dynaco modification does this by putting diodes in series with the rectifier's cathode-plate diodes. You'll need to figure out the pinouts to see how this works. If the 6X4 shorts cathode-to-plate then you've got solid-state rectification with no delay. If it shorts heater-to-cathode you lose the transformer unless you have a very small fuse at the transformer's B+ secondary. The latter is far more likely than the former.

This is the candidate I found, which claims to have "no" reverse recovery charge. Also I found this 80mA fuse to put before the rectifier.
 
Last edited:
Ok, you have confused recovery time with recovery charge. Separate concepts.

Those diodes have 14 nC, as per the datasheet, and in keeping with the 20 nC typical value I cited.
 
Ok, you have confused recovery time with recovery charge. Separate concepts.

Those diodes have 14 nC, as per the datasheet, and in keeping with the 20 nC typical value I cited.
Thank you for all of your help, I will order the parts list I have created immediately. Now I need to figure out how I am going to get at all these pins...
 
Ok, you have confused recovery time with recovery charge. Separate concepts.

Those diodes have 14 nC, as per the datasheet, and in keeping with the 20 nC typical value I cited.

Ok, so according to this pinout:
6x4.png

Pins 3 and 4 are the heaters, pin 7 is the cathode, pin 1 and 6 are the anodes, and pins 2 and 5 are not connected.

Here is a diagram of how I understand to insert the fuses (80mA, 250V) and diodes:
vub6yrc.png


Although like the Dynaco, it seems I could wire the fuses (80mA, 250V) for the anodes to pins 2 and 5, before bridging to pins 1 and 6 with the diodes.
 
You get an A+ for doing the homework assignment and posting diagrams.

Yes, solid-state diodes go in series with the tube rectifier sections. That way the solid-state takes all the PIV stress, not the tube which really can't do it well. Would not hurt to add some snubbers to the diodes to remove the Qrr. Here's Jim Hagerman's paper on that: http://www.hagtech.com/pdf/snubber.pdf

Smart work on putting fuses on the tube's output to protect the tube, not merely the transformer, rather than have a tube fail from a short and then blow the secondary fuses. Spend a buck on fuses and never need to worry about losing a rectifier tube in a catastrophic way from overcurrent.

I suggest adding:
(1) A fuse to the secondary B+ center-tap because it has the current from both sides passing through it. That should cover all eventualities. Some people only fuse the center-tap but this doesn't take into account an alternate path for the B+ current through the heater circuit, if grounded, or through a wiring path.

(2) Fuses to the heater secondary, same arrangement—both legs and the center-tap—for the same reason: if a heater shorts it could conceivably cook the transformer. Heater fusing is tricky because of the massive current during startup, but the inrush limiter keeps that current down to the point that a fuse, likely a slo-blow, will work while it would not work in normal situation without a slo-blow that was drastically over-rated to withstand startup. But the inrush limiter makes it work.​

The B+ fuses should not be 80 mA as that is at or above the 6X4's capacity. Figure out what the circuit is using. Probably few tens of mA. Your supply voltage is below 250 V so the fuse rating is honored? When the filter capacitors are empty the charging current can be quite high, but the fuse should tolerate this.

Don't forget to fuse the transformer primary as well as the secondaries, in case something shorts in the transformer. This is to prevent relying upon your mains fuse/breaker as the safety device. Again, unlikely but this is how good design practices work.

You are still planning to add an inrush limiter for (a) heater secondary and (b) rectifier secondary, since cold-start is 5-10x normal current load. You will need to size this thermistor for the amount of current being drawn.

That will be a very protected supply and A+ work.

Another idea to consider as long as you're doing this work is to add a TRIAC to the power switch. This way the power switch, which is usually a hard to find component (hasn't been made in 60+ years), doesn't move any current, so the contacts forever last. An AK thread explains how to make this modification.

Again, good work on doing the homework and posting the diagram. More people should do this.
 
You get an A+ for doing the homework assignment and posting diagrams.

Yes, solid-state diodes go in series with the tube rectifier sections. That way the solid-state takes all the PIV stress, not the tube which really can't do it well. Would not hurt to add some snubbers to the diodes to remove the Qrr. Here's Jim Hagerman's paper on that: http://www.hagtech.com/pdf/snubber.pdf

Smart work on putting fuses on the tube's output to protect the tube, not merely the transformer, rather than have a tube fail from a short and then blow the secondary fuses. Spend a buck on fuses and never need to worry about losing a rectifier tube in a catastrophic way from overcurrent.

I suggest adding:
...

By B+ center tap do you mean the cathode on pin 7, or the path to ground for the transformer secondaries? Also, Can I place the same 80mA fuses (will have to order more) i am using on the two heater legs (and the rectifier secondary path to ground if that was what you were referring to) and path to ground? Would the same fuses work for the transformer primary fuses you recomend installing? If I understand the terminology correctly, the transformer primaries are the two wires going off the inductor symbol on the left side of T1.
Lastly do you recommend any start point for reading about current estimation?
 
Using phone, so a bit terse.

Look up heater current in each tube's datasheet. That is hot. For cold, multiply by 5 to 10. So 80 mA is far, far too low for heater supply. Each tube is about four times that hot. Also, add in bulb current which can be as high as a heater. Why I suggested and LED.

Again, 80 mA is MAXIMUM for recifier, at point damage occurs. See how much you are drawing. Look at bias setting.

Primary fuse is total AC current drawn from mains, so sum B+ and heaters. Heaters will dominate, of course.
 
Using phone, so a bit terse.

Look up heater current in each tube's datasheet. That is hot. For cold, multiply by 5 to 10. So 80 mA is far, far too low for heater supply. Each tube is about four times that hot.

So I shouldn't use the 80ma fuses for the heaters on the rectifier as previously planned right? If I am reading the schematic/pinout I posted above correctly and this site, the heater current for the 6x4 is .6A right?
 
Sounds about right. But the rectifier heater has its own winding. You need heaters on the other winding plus the bulb.

80 mA is B+, but that is too high for tube output, so that doesn't save 6X4 from short in B+ circuit. If rectifier shorts, a few times 80 mA is likely what transformer can withstand until fuse flows, but best figure out what actual current is.Two different values.

Two different currents here for B+, one between transformer and rectifier guards against rectifier short, which takes out transformer,one for output guards against short in B+ which takes out rectifier and transformer.

Messy explanation, but too hard to edit on phone. Sorry.
 
Also, This is the capacitor I have selected for the death caps.
NO, use only X/Y rated caps. Mouser has them. They are usually rated at 275vac, which covers any line voltage up to 240v (high line 264v, max).
These caps. will open circuit if they become shorted. Thus preventing the "death" part of the scenario.
 
NO, use only X/Y rated caps. Mouser has them. They are usually rated at 275vac, which covers any line voltage up to 240v (high line 264v, max).
These caps. will open circuit if they become shorted. Thus preventing the "death" part of the scenario.
According to the listing, they are Y2 rated. Isn't Y rated the type you want for this usage?
 
Back
Top Bottom