I am using auduo test for power supply capacitors. I listen to them making their own sound (capacitor to the ear) and passing sound to good quality headphones. Motor run capacitors are not all the same. The newer ones (metallized polypropylene vegetable oil) may indeed sound terrible, but the older ones may be quite good. The best of all are antique paper in oil with foil electrodes.
I believe that the way to improving current capability of power supply is oversized transformer, swinging choke, and low DCR components, not increasing the size of capacitor bank.
I believe that the way to improving current capability of power supply is oversized transformer, swinging choke, and low DCR components, not increasing the size of capacitor bank.
I agree, unless faulty a rectifier has no sound.
I built a valve mixer with 1n4xxx series and the switching spikes were very noisy. Changed to Schottky and problem went away.
I've used the Panasonics that Eli linked. Is that type generally referred to as DC Link, regardless of brand? Or is DC Link a specific brand?
I've heard people mention MKP and I always assumed that they were referring to a series of caps made by Wima, which are labelled MKP. But your comments make it seem like it refers to type of construction, perhaps metallized polyester?
I have some Wima FKP types, which I believe are polypropylene film and foil types but, again, I thought the designation referred to a series of Wima caps.
What's the difference between the MKPs, the DC Link types and the Panasonics Eli linked?
I have no way of measuring but the Panasonics seem to work well and they are reasonably priced, although they are about 4x more than a similar Nichicon electrolytic. They are larger, of course, and more difficult to mount.
I have no idea what they were "designed for" or how they're constructed, though I assume they're not considered "audio caps". If they were they'd be labelled with metallic gold script and cost much more.🙄
I understand that the temp rating of electrolytics is, in itself, not critical but I've always figured that the hour rating (10,000 hours or more) should be a good indicator of general quality and reliability. Caps with those specs also seem to have other specs, like ESR, that are pretty decent.
Thanks for suggesting the ALC/ALS caps. Digi-Key lists them with the cheapest (and lowest value, 330uf) being about $12, which is the about the same as a 50uf Panasonic film. I rarely use anything over 100uf, though, and it looks like 330uf is the lowest size they stock.
Other than the obvious difference in capacity, how do you think they compare with the Panasonic film?
Are those of you who use SS rectification unconcerned about inrush current and the possibility that the tube plates will be hit with significant voltage before the heaters have a chance to fire up?
I know thermistors can be used but my impression is that they are of limited value. Better than nothing but probably no more effective (perhaps even just a bit less) than a directly heated tube rectifier and much less effective than an indirectly heated tube rectifier. Maybe I'm misinformed about their relative effectiveness?
I sometimes use a thermistor on the primary to help ease the inrush current to the heater winding, even though the amp is tube rectified.
Again, it seems to me that an indirectly heated tube rectifier would be the best choice for a preamp or lower power SE amp where the current is more stable than a high powered PP amp.
I know thermistors can be used but my impression is that they are of limited value. Better than nothing but probably no more effective (perhaps even just a bit less) than a directly heated tube rectifier and much less effective than an indirectly heated tube rectifier. Maybe I'm misinformed about their relative effectiveness?
I sometimes use a thermistor on the primary to help ease the inrush current to the heater winding, even though the amp is tube rectified.
Again, it seems to me that an indirectly heated tube rectifier would be the best choice for a preamp or lower power SE amp where the current is more stable than a high powered PP amp.
Until you get above 500 V., cathode stripping is of minimal concern. Also, the slight delay provided by a well selected NTC inrush current limiting thermistor in the PSU filter wiring is sufficient, when an "instantaneous" on, SS rectified, C- (bias) supply is employed. The negative voltage applied to the O/P tube(s)' control grid(s) electrostatically shields the cathode(s), before B+ "rises".
That's good to know. Is cathode stripping the only concern when the B+ comes up before the heaters? For some reason I thought there were other issues.
So, if a SS rectified bias supply is powered from the same PT as the B+, the correct placement of a thermistor would be on the center tap of the HT supply, not on the primary so that the C- is unaffected. Correct? I suppose the filament windings would also get their own thermistors on the secondary side too.
I've been using battery grid bias on some builds and the battery is not switched so there is always negative voltage on the grids, even when the amp or preamp is turned off. Is that a problem? If not, does that mean it would be best to put the thermistor on the primary where they are more typically used?
Stage amps have an "on" switch and a "standby" switch for the HT.
Pretty simple solution. Useful for mesh rectifiers as well since they can spark if they're not warmed up.
Pretty simple solution. Useful for mesh rectifiers as well since they can spark if they're not warmed up.
Last edited:
I look at it this way. Tons of tube gear has been made with silicon diodes and no b+ delay. You'll see it even on luxury brands like McIntosh.
When unswitched battery grid bias is employed, electrostatic protection of the cathode is always present.
Power should be applied to heaters immediately. Get emission going. If you want heater thermistors, use a separate filament trafo that's energized before the rectifier feed is energized.
In a FWCT or bridge setup, the place for the thermistor is in the "hot" line, between the rectifiers and the PSU filter. In a "full wave" doubler PSU, use twin thermistors as shown in "El Cheapo" or a single thermistor in the line between power trafo and doubler cap. stack junction.
Attachments
I thought the main issues with tube rectifiers were:
* large voltage drop / sag
* didn’t like driving large capacitance smoothing caps.
The slow start can be nonthermistor emulated with a Dirac/triac, or simply use he LT3080 and/or LT3081 soft start. With heaters you can parallel LT3080s for current and get a steady voltage with soft start, if you want corrent limiting then simply add a LT3081 after it and set the Ilim pin.
* large voltage drop / sag
* didn’t like driving large capacitance smoothing caps.
The slow start can be nonthermistor emulated with a Dirac/triac, or simply use he LT3080 and/or LT3081 soft start. With heaters you can parallel LT3080s for current and get a steady voltage with soft start, if you want corrent limiting then simply add a LT3081 after it and set the Ilim pin.
How exactly do you use a 35V LDO voltage regulator to soft start a 500V power supply?
I’ve never seen a soft start circuit like that for heaters before. I can hardly believe this could be necessary. Never used it either.
Regards, Gerrit
Regards, Gerrit
LOL... Nick, you need to express yourself a little more clearly and honestly.
In #54 you said you can get soft start by "simply using the LT3080".
Tom's regulator is a very nice piece and I own a couple myself. But to call that "simply using the LT3080" and using it as a soft start when you don't need regulation is absurd. And in most situations you don't want to regulate B+, just the screens.
In #54 you said you can get soft start by "simply using the LT3080".
Tom's regulator is a very nice piece and I own a couple myself. But to call that "simply using the LT3080" and using it as a soft start when you don't need regulation is absurd. And in most situations you don't want to regulate B+, just the screens.
Ahh you didn't say all that 😉 😛
It's very clear with SS vs Tube, from my humble playing during research with modelling, is just how hard SS starts are compared to ye-olde tubes. One interesting concern that Carlson (Carlsons Lab YouTube channel) demonstrated is the impact on heater failure rates on some of the less reliable euro-tubes with SS power.
Yeah, I have a few vintage pieces that use SS rectification with no delay and I've never really worried about it much. But thermistors are cheap and they certainly have benefits. So when I'm restoring something I often add them.
Voltage drop varies but it's no big deal. Just take it into account when the PS is designed and you'll end up with the B+ you need. And, interestingly, indirectly heated rectifiers that provide a slow start typically drop less than directly heated ones.
Sag, as I understand it, is more of an issue evident in PP amps, especially if they're run hard. Current doesn't vary nearly as much in a preamp or SE amp.
The cap size limitations apply to the first cap in an C input filter. Subsequent caps can be large, which helps deal with sag issues if the amp has them. And it's easy enough to add an extra RC or LC if sag is still a concern. Of course, choke input filters can also be used.
I'm totally ignorant about all the SS devices that seem to be so popular with so many people here on the "Tube" forum. So I don't know how they compare in terms of their ability to slow the startup. I know that indirectly heated tube rectifiers are simple and provide a slow gentle startup.
Yeah, not really necessary - if it was then you would see lots of heaters burning out.
But take a look at the failure mode of an incandescent light bulb. The only time they seem to fail is when they're first turned on, and if you have them on a dimmer, they last much longer. Obviously, their filaments are much thinner and more fragile than those found in tubes, which is why they emit so much light and a tube's filament only glows dimly.
So, while it may not be absolutely necessary, a slower start will reduce the chances of failure. If you're using tubes that are either more vulnerable or rare and expensive, it's a little added insurance. And easily implemented using a cheap thermistor. No need for complex circuitry.
Last edited: