High B+ using SS rectifier

[abbey road d enfer]

according to schematic it is a transformer with 375VAC secondary, but this P-TF22772 is rated different, see attachment.

Indeed! I misread on the first pic.

 

[abbey road d enfer]

The peak voltage from an AC signal is V_rms*sqrt(2)
The V_rms at the wall may be somewhat higher or lower than 120V, but if it were 120v exactly, the V_rms of the secondary would be 325v. The V_peak then would be 460v. That should be the highest voltage you could see in your circuit (and you'd want all your HV caps to be a safety factor above that). I'm not sure how you are seeing 550v in this situation.

Remember that the rated voltage is for full load, so unloaded, it's somewhat higher. Typically the open voltage is about 10% above loaded voltage. The measured difference is much higher than 10%, so there is something wrong somewhere.

 

[abbey road d enfer]

I'll put a thermistor in there for sure, primarily to slow down the voltage on startup.

The main purpose of a thermistor is to limit the current surge at turn-on. It allows reducing the nominal value of the fuse, resulting in better protection.
It doesn't change much the voltage build-up, but also protects the reservoir caps.

 

[Cqwet Dbdfte]

You could use a HV MOS FET and a power resistor to shunt the +B to ground while the tubes warm up.
This is a tiny power amp, so a series HV FET+LT317 regulator could also work.
As it looks like a guitar amp, fidelity is not much of an issue, you may like all the harmonics produced by the amp.
I see no reason to use tube rectifiers, they can't handle much current, limiting filter capacitance severely. I used ~300 ohm in series with each silicon diode and then some 220uF 385V caps, followed by the above series regulator in a Hp 200 tone generator (puts out ~60V pp into 600 ohms)
HV caps (of same type) can of course be stacked, using balancing resistors.
I prefer 1V pp ripple over ~20 in my power supplies. Less work for regulator with a few mV pp noise.
Also, an aluminum chassis make great heatsinks for power resistors and (full pack) MOS FETs. Else a piece of aluminum sheet can be fitted inside a steel box.
Relieving the circuit from a tube rectifier reduces the heat, of the rectifier tube, and load on transfomer.
A CL-90 or similar can be used on the primary to allow much higher energy storage caps. I use 1500uF/660V in power amps with 480V +B.

 

[expat]

Hi Cqwet Dbdfte,
Thanks a lot for all the info. I would love to better understand what you're suggesting, do you happen to have any schematics that illustrate what you're describing?

 

[Whoops]

Reduce the B+ voltage with this circuit, it provides voltage reduction without extra sagging,
the Zener voltage rating you choose determines the voltage drop.







https://el34world.com/Forum/index.php?topic=21987.0

https://ampgarage.com/forum/viewtopic.php?t=27035

 

[expat]

Hi Whoops, thanks for that info, schematics and video link - I greatly appreciate your time and learning this technique. (And I love that cat's YouTube channel, by the way! Subscribed to him recently).

What would you and Cqwet Dbdfte have to say about the actual sound of the SS rectifier on an otherwise tube-built amp? How would you describe its sonic attributes?

 

[Cqwet Dbdfte]

Does a flat line DC voltage have a "sound"? If you want a soft power supply that changes under load and provides signal mixing and intermod products then we are not talking about a power supply but some form of signal alteration circuit.
Most LT/LM317 datasheets provide info on using the part to control an external HV transistor. FETs are easier to use than bipolar due to higher current gain.

 

[musipol]

It's no big secret that on a guitar amp, a B+ that sags when the amp is driven hard will give you a somewhat different sound (the feel while playing the guitar is definitely different). Personally, I think SS rectifiers, and an appropriately sized resistor, gets you close enough to a tube rectifier for my taste. In the case of this amp, you have way too much voltage and need to go further to reduce it, using some of the alternatives described previously.

 

[Whoops]

What would you and Cqwet Dbdfte have to say about the actual sound of the SS rectifier on an otherwise tube-built amp? How would you describe its sonic attributes?

There’s plenty of info and discussions online about that topic from people that know much more than me, just search Google and will have some good readings

As far as I remember people saying, it’s seem with a tube rectifier you will have sag, and with a SS rectifier you will get no sag at all.
So with SS the transientes will be faster, the amp will respond quickly, and will be more punchy.
With Tube rectifier and sag, the sound will be “more dynamic”, slower transients and response when you struck the strings, and more mellow.
Some people prefer SS rect others prefer tube rectifier.

That’s what I remember, I maybe wrong in my descriptions so Google this topic for other opinions and descriptions

 

[abbey road d enfer]

With a SS rectifier, there is moderate sag, resulting from the DC resistance of the Power transformer.
With a tube rectifier, there is more sag, because of the diode characteristics of the valve.
With the "power zener" circuit, there is as much sag as with the SS rectifier.
With a MOSFET used as a B+ regulator, there is no sag at all.

 

[Cqwet Dbdfte]

Power supplies are not designed on purpose to sag under load, this would be tolerable, or not, -side effect.
If a power supply sag provides some benefit then this should be quantified, as this could be too much, or too little.
It is easy to make a "soft" and saggy power supply, just add some series resistance.
Beware there is a possibility to get positive feedback with soft regulation.
Classic tube circuits had just "good enough" regulation, as it was not economically feasible to have good regulation, now super easy to accomplish.
The vacuum rectifier was the major culprit in this drama.

 

[ccaudle]

Some people prefer SS rect others prefer tube rectifier

Some modern amps have a resistor in the power supply circuit with a switch to bypass so you can choose sag or no sag as you deem appropriate.

 

[expat]

Excellent input everyone, thank you.
Its been pointed out that SS is 'more efficient' and won't burn out the PT as quick. But amps from 70 years ago with a tube rectifier still work great, and I'll be dead as hell by then. I figured I'd try the SS rectifier because of the punchiness - it sounds pretty cool (and super percussive) on my converted Phillips turntable from 1962 (selenium rectifier that i'm about to update to silicon diodes, 12ax7, el95). There's a great schematic out there for a Ampeg B-15 that combines 3 different versions, and you can go from SS to Tube rectification on a switch. I am feeling some feature-creep going on...
I have to think on this a little.
Greatly appreciate everyone's take on this.

 

[merlin]

Its been pointed out that SS is 'more efficient' and won't burn out the PT as quick.

SS is not 'kinder' on the transformer (if anything a tube recto is kinder because of its resistance and slow warm up). But SS gives you more freedom: you can use lots of capacitance if you want to, or not, you can add resistance if you want more sag*, or not, you can add a soft start circuit if you want to, or not.
A tube recto is just a cool-looking millstone around your design neck.

*You can even use an SS recto to do the heavy lifting, then add a tube recto in the DC path to give the appearance and effect of a tube recto, but with less of the design limitations!

 

[expat]

hey Merlin, thanks for the input.
So yeah, this thing is built in a baking pan, so i couldn't care less about how things look. I just really like having options, and as you point out, you can apparently milk a SS recto anyway you want.
I'm the guy who put 6 different value vintage caps on a rotary switch in my DIY Princeton's tone circuit (total waste, fyi). So, give me some switches and I will go there, just to find out the hard way for myself.

 

[Wavelength]

Using the 6V6 above 300V is going to cause problems. Originally when RCA helped Fender design amps they had the 6V6 at like 280V. When Fender built the amps the supply went up to 400V. This caused a lot of failures in the field and the design was beefed up. But not to 500V. The data sheet calls for max voltage of 320V. At 500V you have a few options like EL34 or 6550/KT88 or some of the newer variants of those. But watch the wattage as that can be a problem and lower your expectations. Remember it's about diminishing returns when it comes to amps. The balance of current and voltage and output transformer design really makes the difference.

 

[Cqwet Dbdfte]

AR's "Acoustic suspension" type speakers was probably the worst thing that happened to audio in the 60's. Being grossly inefficient meant much more power was needed to drive them, along came higher power amps with less performance. A ten dB less efficient speaker needed 10dB more power, the old 15W now needed 150W, hopefully.
If you want a small poweramp you also need a very efficient speaker. Obviously, or not, music production has a different set of criteria than reproduction.

 

[expat]

My buddy brought his variac and fancy DVM over and we brought it up to 120v and I got 504v on pin 3 (plate) of the 6v6 socket with no tubes installed.
I'm thinking now to just go for the series resistance and keep the SS recto to hear how it sounds.
Given that it's showing 504v, what would I need in terms of resistor values?

 

[Cqwet Dbdfte]

Depends on the current drawn by all the tubes. A voltage regulator eliminates this problem.