5U4, 5R4, 5V4, 5AR4 & Solid State Rectifier Performance in Vacuum Tube Amplifiers 

this is more of an issue with high power radio transmitters. Everything I've come across regarding the tubes used in guitar amps says this is a non-issue. you're more likely to wear out the filament from turning the amp on and off many times that you are to wear out the cathode because of some imagined cathode stripping. In many tubes you can damage the cathode by running them with no B+ voltage for long periods of time. this was a big problem for people using tubes as active devices in digital logic circuits and others where they were run for prolonged periods with the control grid biased to cut off all plate current. eventually manufacturers figured out ways to make the cathodes less sensitive to this problem, and special grade computer tubes were produced that could handle that type of operating condition.

sporadic -Z Thank you, why is that?

2 things affect that. 1, the equivalent series resistance of your power supply. 2. the idle current of your tubes, the bulk of which, is the bias current on your output tubes. This is NOT indicative of trouble, but merely a characteristic of THAT amplifier.

Yep. It seems the internal resistance goes down as the rectifier model's letter (R,U,V,AR) and the current capability go up. The 5R4 has the highest resistance by far. It's natural therefore that the power output will increase for the same amount of distortion as the biasing goes up (both current and voltage) which results into increased headroom. It also possibly moves the amplifying tubes' operation into a more linear region of operation. In a way, one could perform a similar test by attaching a variac at the mains input and increase the voltage by the same amount and then use resistors to simulate the rectifiers' internal resistance. Perhaps it might also be interesting to compare the amp's distortion levels instead, holding the output power constant, and confirm with the distortion spectra although I would not expect any surprises there. Having the amp connected to its intended load could be also useful. There are people who regularly perform tube rectifier rolling in their amps and claim sonic benefits with a particular rectifier. They then go on to attribute these benefits to the rectifier in a manner that alludes that the rectifier has some internal sonic advantage. I treat these claims with scepticism. Each rectifier is designed to do a particular job, so swapping rectifiers and comparing performance on that basis alone is not tantamount to comparing like with like. There is the possibility that the rectifier choice simply alters the amplifier operation characteristics, i.e. alters the design. It would be better, I think, to have different versions of the amplifier appropriately designed to maximize performance for each particular rectifier. Again, it would not be comparing like with like, as we would be dealing with slightly altered designs, but it would make sense if one strives to achieve the best distortion figure at a set power level (or the converse).

@@Stelios.Posantzis That is absolutely correct. And most guitarists do not understand it. Changing rectifier usually changes B+ and screen voltages in a way that output tube bias, both anode and screen currents change. If you do not rebias output tubes after rectifier change (by adjusting fixed bias voltage or, in case of automatic bias, cathode resistor value(s)), the observed sound change will be caused mainly because of different bias and operating point of output tubes. (Of course, sound and THD of preamp an PI stage tubes changes slightly as well if anode voltages change.) BTW using variac to change mains voltage with a fixed bias amp causes the bias supply voltage to change a bit, which compensates slightly the B+ change. When you change only B+ rectifier, not the mains voltage, fixed bias bias supply voltage practically does not change.

@@Stelios.Posantzis I've never noticed that before! Thank you for pointing that out!

The 5R4 is intended for higher voltage power supplies (for example a power supply that has an output of 800-900 volts or so) than the 5AR4 and the 5U4 which generally are used in power supplies usually not over 500-550 volts or so). The 5R4 will work but your voltage will be a lower, probably by 40-50 volts so your power output will be lower also. The 5AR4 is the next thing to solid state diodes you are likely to get in a vacuum rectifier. Only mercury vapor (MV) rectifiers will have a lower voltage drop across the tube. usually around 15 volts. The more current drawn thru a vacuum tube rectifier (not the MV rectifiers) like the 5AR4/5U4/5R4/etc the higher the voltage drop across the tube resulting in lower voltage to the amplifier tubes. Sometimes guitar musicians in a musical creative style, use the voltage drop that is inherent in the vacuum tube rectifier, like the 5U4 as an example, as the voltage drop increases the distortion as the volume (and resulting power supply current) is increased and the power supply voltage drops which may enhance the music they are creating by giving it more crunch or over-drive or however they refer to it.

Thanks for your nice comments and encouragement. There were several examples in this video that I could have done better but sometimes what not-to-do can be as important as what-to-do. As for the varistors, I have found that in a warmed up amplifier, jumpering around the variable resistor makes an insignificant amount of voltage difference in the DC output of the rectifier - like a volt or two. The FFT software that does an excellent job at low frequencies and allows you to very easily adjust filament hum balance pots (60Hz) and examine the 120 Hz ripple of power supplies is the SpectraPlus. I have tried and still have, three USB sound cards but now I have settled on the one channel Scarlett Solo. It is 24 bit/192 KHz and does a good job. Is it really any better than the two channel Scarlett or the Behringer... ? probably not but the Solo is a bit smaller and I like it. The AD2 does not do a good job at these low frequencies but if you limit the dynamic range to 80 or 90 dB in the Digilent AD2, you will get a nicer, and I believe more accurate, harmonic profile at the higher audio frequencies (above 10 KHz) than you will with the Scarlett USB and SpectraPlus SW - I assume because the AD2 input channels are flat beyond 10 MHz as they are also the scope channels - so there is a trade-off there.

Thanks for your prompt reply. I think I'm going to swap the 83 rectifier out of my Hickock 539as a first SS mod. I have a Focusrite Scarlett 6I6 for a soundcard to try. I will look at the SpectraPlus trial and probably buy the Digilent AD2 for distortion measurements. Thanks for the info and keep em glowing!

For small value caps I use Orange Drops and for power supply capacitors I use garden-variety electrolytic capacitors I get off Ebay. If the voltage is above 450 volts, I simply put two 450 volt electrolytic capacitors in series, and shunt each capacitor with a 470K, 1 watt resistor. This value of resistor will not drain the capacitors quickly when the power supply is turned off but it will equalize the voltage across each capacitor and over several minutes drain the voltage off the capacitors. Hope this helps.

@@ElPasoTubeAmps That's how the device is built. Two 450 vdc caps in series, but their leaking. The strange thing is, someone used the on/off switch to shunt the caps with a large resistor when turned off. Thanks for the information. I'll check e bay.

The amplifier with only 10 watts with the SS rectifier or 5AR4 is absolutely amazing sounding on a pair of Paradigm Focus speakers in my little shop.. $100,000 of McIntosh equipment couldn't outdo it. I don't know what it is but something about those old 1940 era 6B4G triodes charm me. The highs are clean and the bass is tight but full range. It must be love.

David. keep in mind that it didn't overshoot because the 6B4G tubes are directly heated and warm-up instantly like the 5U4 and 5R4.

ElPaso TubeAmps t

What would be a better test would be the tube rectifier with it's 40uf max capacitance compared to the sound of UF4007 diodes with 500uf of capacitance. That is where much of the solid state rectifier slam comes from IMO. That and the lack of voltage drop under load with SS diodes.

HUGO FUGUSEV could you expand on that? I'm currently using a solid state rectifier on my VTA ST 70 " modern beefed up Dynaco St 70" would I see any benefits upgrading the capacitance of my cap?

Voltage doublers are actually very common. The McIntosh MC40 uses a full wave voltage doubler and so does the Collins 30L-1 linear amplifier. It it a perfectly good way to get high voltage from a modest voltage level transformer and it makes the transformer so much more versatile and valuable to use it as a, CT full wave, or a bridge, or a voltage doubler. That is three perfectly good ways to use the same transformer for a power source.

You are a musician and I am not so you hear things I am not aware of. On the other hand, I try to describe your, "resonance-complex resonance . . vs a clearer-sterile resonance" type of sound using harmonic profiles from a spectrum analyzer. I am confident we like the artifacts created by vacuum tube amplifiers which adds to the complex harmonic profile of the original music. Harmonic distortion typically rounds off softly at moderate clipping levels in a vacuum tube amplifier and can add pleasingly to the music but SS amps can have transitions at the clipping level that are abrupt and have harmonic profiles that we simply do not like. Surely our ears are ultra sensitive to these harmonic profiles or we would not be able to recognize individual voices. Also, moderate distortion levels add to the volume of the music so a 15 watt amplifier can sound as loud as a 50 watt amplifier in some cases and may be preferred. it is so complex it is difficult to put into works. I hope this helps and Thank You very much for your comments.

Hi David, I did remove the SS diodes and change it from a bridge to just a tube full wave and basically cut the voltage in half. I think this amp would do about 15-16 watts at the high voltage. Actually, it was way too high with the bridge and of course it did drop the power considerably. You are right, it is a miracle that the tubes didn't fail at the higher voltage. A good bit of difference in output power can be observed going from the 5U4 to a 5AR4. If I remember correctly from about 6 watts to 10 watts as the rectifier stands now without the SS devices to make it a bridge. I wouldn't worry to much about the DCR of the PS choke and I think the best Acrosound/Williamson amps are the Heathkit WM-5 (I believe...) where they either run a pair of 6SN7 or 12AU7 drivers. The older designs with the 6SL7 is very high gain (I believe about 100 mV for full output) and distortion is much higher and the gain is not needed - or desirable - in my opinion. Running the tube plates as high as you can and keep your electrolytic capacitors save is, again, in my measurements, the best way to get the most amount of power and the least amount of distortion. 480 volts is good if your PS capacitors are rated at 525 V. In your modeling program I bet you can see how the initial turn-on of a PS way over shoots the DC value of the electrolytic capacitors especially on the input capacitor of a PI network. Not so much so on a choke input filter. Even though this amp is at most 10 watts now, I have a real appreciation for its simplicity and sound and possibly the fact that it is triodes in the output (?) I don't know... and I am sure in a blind test I couldn't tell it from any other amp ranging from $5 to $50,000 if it was running at 3 watts into the speaker... :-)

I've been in a bit of a quandary since James Audio went out of business but, yesterday I stumbled upon Toroidy's toroidal outputs from Poland they even have Cathode feedback windings and they're about $126 a piece and rated at 80 watts PP . So I think I'm going to give them a shot only trouble is now I'm not sure how to incorporate the CFB winding into the Acrosound/Williamson circuits. I want to build something with a high power rating since it's going to end up driving my Magnepans. Any thoughts / suggestions would be greatly appreciated! :-)

As for the separate NFB winding that sounds like a nice transformer. I would like a little information on them so I can look into buying one (or two).. I determine the value of the NFB components by starting with the point of about 16 dB NFB. The way I measure it is to disconnect all FB and set the amplifier to maybe 5 watts - some level well below clipping and distortion. Then start adding in a FB resistor and see what the power drops to. Calculate dB loss with 10log(p1/p2) and see what you get in NFB then run a series of power and THD tests to see how it turns out. If that is not good, raise or lower the NFB resistor value and measure again. Start with maybe 2K ohms. On the amplifiers we tend to build, around 16 dB NFB seems to turn out OK. At 20 dB NFB things starts to get weird at the LF end. I "tune" the NFB resistor and capacitor for maximum power output which coincides with minimum THD and make sure it is stable and not trying to oscillate at LF (20 Hz and below). I do see amplifiers that tend to oscillate at maybe 1/2 to 1/4 Hz. Very slow rise and fall of the sine wave even when driven at 1 KHz. It doesn't seem to be a real problem but it does make the THD meter act weird and probably should back off the NFB a little bit. As for the parallel capacitor around the NFB resistor - that one is more tricky. I personally use a GR variable capacitor that will go up to (I think) 1100 pF and usually find a point around ... maybe... 200 pF that works well. Got to watch the top and bottom of the frequency range when playing with the C. Sometimes I just leave the C off. Hope this helps.

I think the answers to a lot of your questions lie in the average current each rectifier can handle. The 5V4 is 175 mA, the 5AR4 is 250 mA and the 5U4 is 275 mA. I am sure these are CCS ratings (Continuous Commercial Service) and can be exceeded in dynamic music conditions found in a musical amplifier without any harm to the rectifier. The next parameter is the internal impedance of each rectifier tube and without getting into specific numbers we know that the voltage drop across the 5AR4 is much less than the 5U4. So, as the recommendation to not use a 5V4 in place of the 5AR4 is probably a good recommendation as the current rating of the 5V4 is a bit low. That would be kind of like putting a 5Y3 in place of a 5U4 for something like that. The fact that you can use SS rectifiers says the power supply capacitors can handle the extra voltage so surely you can use any rectifier rated at enough current. I am sure you know that even the smallest power supply SS rectifiers are rated at least 1A or more and the voltage drop across SS diodes is only about 0.7 volts. This is going to give you the highest voltage, the most available current and the least "sag" when playing the amplifier at full volume. Using two 5U4' is a serious power supply. That is going to give you the ability to draw twice the current of one tube, if the load needs it, That is a pretty impressive power supply setup. The last parameter is the PIV of each rectifier which is 1550 for the 5U4. This would relate, generally, to a 800 volt CT power supply power transformer primary voltage. With 400 volts each side of the power supply transformer secondary to ground this will give you approximately 565 volts DC with no load or about 480 volts under full load (using SS rectifiers). I think these voltages are pretty close to what you are measuring. To be absolutely certain you can "roll" different rectifiers in each amplifier, I would want to study the schematic of each before making this recommendation. On the other hand, if your amplifier says you can run either vacuum tube rectifiers (i.e. 5U4) or SS rectifiers, I can't see a problem there using any rectifier that can handle the current of the least of the recommended rectifiers. As my memory serves me, the only amplifier I remember seeing a 5V4 in is an old (circa 1960) Fisher 70-A which is rated at 20 watts so I would stay away from the 5V4 in a large amplifier. I hope this helps.

That is correct. From the 5R4 to the solid state rectifiers there is about 50 volts difference in DC output of the power supply depending on the current drawn through the vacuum rectifier. Part of the point of the video is that you may or may not (probably not) be able to safely swap your old 5U4 type rectifier out for more modern day vacuum rectifiers or solid state rectifiers without surpassing the voltage rating of the electrolytic capacitors in the power supply. Also, my examples were not the best at showing over-shoot but SS rectifiers will provide an instant and higher voltage across the capacitors until the vacuum tubes warm up and draw down the voltage on the power supply. I forgot about the surge resistors I had in the primary circuit of the power transformer just for that reason, to limit inrush current so that surge voltages are minimized as the tubes warm up.

Great demo. Thanx.

Some people have actually made solid state devices to replace tubes. I remember seeing SS replacements for many of the tubes in vintage Collins Radio, HAM radio equipment, lots of years ago. I never saw any Collins equipment with these SS devices so I have no idea how well they worked.

I don't recall running across any specific schematics for the 6A5G tube but looking it up, it is the 6.3 volt filament version of the 2A3 which is the same specs as the 6B4G. You can find schematics for those all over the Internet so you just make sure you have the pin connects right and you should be able to find lots of schematics. vinylsavor.blogspot.com/2019/12/tube-of-month-6a5.html

They put diodes in series with the 5U4? What do they claim that does for the amp?

Yep the diodes are placed in series before the 5U4, I've heard a few reasons that I can remember 1) the tube gives a delay, so it's better than just going with diodes 2) the diodes increase the life of the tube / reduces the strain on the tube. I think there were other explanations too, I remember reading one person claimed better low frequency response but, I've never seen anyone post before and after measurements to back up these claims. I would be curious to find out what effects it actually does have on an amp.

I can't imagine it having any effect at all except lowering the HV by another 0.7 volts. I think of these types of changes like polishing the chrome lug nuts on a car wheel. If you make the lug nuts really shinny and put a coat of wax on them, it will lower the wind resistance and make the car slide through the air with less resistance and provide not only a smoother, faster ride but will increase the gas mileage. Combine that with a beer (or two) and an overall good day at work and home and some favorite music makes everything better. No disrespect meant to anyone as we are all influenced by our emotions and beliefs.

Speaking of various changes because of varying ideas, I've heard this one in regards to power supplies before (I had an amp with a dropping resistor because the HV was slightly too high, and a 10H choke) and was wondering what your take is. It's along the lines of "Big caps and chokes = dampened dynamics and a current starved amplifier. Low DCR power transformer with low value chokes and caps = dynamics." I got that response and was yelled at because I was running my tube at 100% dissipation for class A single ended. So I'm I really drastically shortening the life of my output tube by running it at its rated plate dissipation?

I think there is more than one discussion item here. A transformer with low DCR (DC resistance) implying higher current levels; low value chokes, I assume again low DCR (?) or low L (?) I assume still leaning towards higher current and then low caps (?). Does low caps means low C values (?) that doesn't add up to higher levels of current so that doesn't seem to follow. As for running your SE class A amp at 100% dissipation, it probably improves performance, i.e. power output and lowers THD so that is a good thing. Class A draws current through all 360 degrees of the cycle so actually putting a signal through the stage does not increase the plate current very much. Bottom line is there is always a trade-off. If you want your tube to last forever, run it at very low plate current and deal with the poor performance or if you like it performing best, run it they way you do and have a spare tube. I prefer doing it your way.

SCR's are different from a diode and can't be simply put in place as a rectifier. Actually, I made a video on SCR's some time back that turned out pretty popular that will show you how SCR's work and ate tested. Here is the link. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-S4To05OtDxs.html

Probably not. It is rare but half of a rectifier can fail and then we get only half wave rectification. I had that happen to me about a year ago with a 5AR4 and it was the devil to figure out. I think I would swap out the rectifier first and see if anything changes. I posted a video some years back on how a lose ground on a canned electrolytic power supply filter capacitor can cause all kinds of hum problems. Hope you get it fixed.

@@ElPasoTubeAmps Thank you Very much for your prompt answer. I have noticed that it hums even more with other tubes . So it has a lot to do with different types. I have 5 of them .I noticed Also that transformer right next to that tube kinda vibrates. This vibration diminishes if I put something heavy on top of it .I read about a heavy contraction called " The Brick " for the ridiculous sum og 100 USD. So instead I made my own out of brass covered with real leather. It seems to Work. I bought also a high end power cable with ground. But when I opened the amp , it has only + and -- .no ground. So that was a waste of cash. Thanks again. 😂😁

It's math. For AC to have the same heat power (as in you connect it to a heater) as DC it has to have it's amplitude square root of 2 time bigger then the DC voltage. After you rectify and filter it you will get a DC voltage of AC * sqrt(2), eg. 110VAC will give you 155VDC.

So, 110VAC RMS is like 155V peak to peak, therefore, the big caps charge up to 155VDC. But the 110VAC RMS would be more like 110VDC on a heater, i.e. RMS being more like a comparable DC voltage. So his 400VAC is likely RMS, like the primary 120VAC is RMS. But the ecaps would be more like PtoP. If the rectifier B+ without caps reading was had, it should be more like 400VDC, since RMS should be more like DC. So what you're saying is RMS and PtP are the reason, as I suspected. The transformer can only supply so much, even when saturated. 110RMS or 155 PP? So it's really about charged ecaps?

110 VAC RMS would be 155 V peak or 310 VAC peak-to-peak but when we rectify it full-wave, we turn both halves of the AC waveform (positive and negative) to positive pulses and end up with with a DC voltage that is a 120 cycle ripple voltage of 155 V peak from the original 110 V RMS source. If there is no load on the filter capacitors, then this capacitor(s) will charge up to 155 volts DC, the peak value that is feeding the capacitors. Once we start drawing current through our amplifier, this DC voltage will start to drop, depending on how much current we draw, down to a value that is typically around 1.25 times the original 110 V RMS which would be about 137.5 volts. While all of this has been said correctly, the actual DC voltage you end up with on your amplifier will vary according to the voltage drop across the particular vacuum tube rectifier you use, i.e. 5U4 or 5AR4, will produce different voltages under load. The 5AR4 will give you more voltage output than the 5U4, for example. The more current you draw from the power supply, the lower the voltage will be, down to a level that blows the fuse or burns up the tube and/or transformer. Musicians like this voltage "sag" as it gives them more creative power in playing their music. From the strict sense of high-performance power in the area of stable output, the solid state rectifier will win ever time. As can be seen in the video, using a 5R4 and using solid state rectifiers can result in an output voltage difference of as much (or more) as 50 volts more with the solid state rectifiers. To complicate it even further, much will depend on the type of filter network that is used on the output of the rectifier. I hope this helps.

One way to lower voltage on the transmitter might be to use a choke input filter on the power supply. I assume he has either a simple capacitor filter (no chokes) or a Pi filter. In either case you will end up with more DC output than the RMS value of the AC voltage. Connecting the output of the rectifier straight into a choke (with no capacitor to ground at that point) and only putting a capacitor on the output of the choke, will, if the choke is operated at close to the correct current rating, will give about 90% of the AC RMS voltage. Maybe this will help. If there are unused secondary windings on the power transformer like unused 6 volt or 5 volt windings, they can be put in series with the primary to lower the output voltage of the secondary that HV is derived from (got to check (experiment with) the phase in connecting them in series with the primary).

pneumatic00 Great comparison chart! Very useful at a glance. Thanks!

There's a good diode trick using 1N4007s to improve the reliability of the newer 5AR4s.

The 5AR4 has less voltage drop across the tube and will deliver a higher voltage output than the other vacuum tube rectifiers. Not quite as good as the SS rectifier but close. Also, the 5AR4 will not "sag" under load like the 5U4 The 5R4 - the big cream colored one, is rated at a higher voltage but a lower current that the 5U4 so the big 5R4 will be used in a power supply that, for example, needs to supply 800-1000 volts DC output whereas the 5AR4 and 5U4 will be used in power supplies typically under 600 volts DC output. An important factor in a vacuum tube rectifier is the PIV rating - see pasted below: "For rectifier applications, peak inverse voltage (PIV) or peak reverse voltage (PRV) is the maximum value of reverse voltage which occurs at the peak of the input cycle when the diode is reverse-biased. The portion of the sinusoidal waveform which repeats or duplicates itself is known as the cycle."

@@ElPasoTubeAmps Hi man, thks for the video . When you say that the 5AR4 doesn't sag , what would be then the difference with a solid state rectifier ? thks !

@@soundsokok The 5AR4 would sag less than the 5U4 under the same load. SS rectifiers have a constant voltage drop across them of about 0.7 volts and this voltage drop does not increase as the current increases so the SS rectifier does not sag. A 5U4 with no load to full load will likely drop (sag) 40 volts. This is what makes the 5U4 and other tube rectifiers get so hot. With all this said about the rectifiers, be them tube or SS, the voltage will sag within the power transformer to some extent even the wall outlet to some extent. It is normal and the way everything works. I would not want sag in my stereo amplifier but I might want it in a guitar amplifier for increased distortion or "crunch" or whatever it is called. Hope this makes sense.