[Sc]

Hi
Probably this discussion happened before, large capacitors were expensive and hum will decrease the higher the better ..... - yes probably true. But - does it make sense, as e.g. seen on many recreations of vintage studio equipment or classic guitar amp's to put up far more than ten times the values ?? To my experience - rather not. With DC - heaters (where large values make sense) and proper grounding design hum is usually not a big issue. You can easily overload tube rectifier designs even if the first cap is a suitable one. Tone of tube rectified guitar amp's (sag) to my experience changes with cap values and personally I like the values close to the original. The same with vintage studio gear - where these had maybe a simple tube rectifier and some 16 to 50uf or on larger ones tube stabilized and regulated circuits, we now see solid state regulated PSU's with 1000nds uF over the full B+ line. Any tonal or functional experience's on these more critical studio applications ? Given hum is under control - just vodoo to stick more to original values  ?

[gyraf]

agree, this shouldn't be overdone - specially not when "recreating" classics

[rock soderstrom]

I agree. Especially with guitar amps, this does not change the sound/character for the better in my view. In studio equipment it is wasted money and space after a certain point. In addition, it also stresses the rectifiers unnecessarily.

[abbey road d enfer]

It depends on the intention.
If the goal is to recreate a faithful avatar of a coveted piece of gear, one should stay close to the original, which is not always possible, since recent capacitors have very different characteristics than period one. New ones have less losses, smaller DCR and are more robust.
If the idea is to create a modern version of a respected model, it is certainly desirable to take advantage of the improved performance of modern components.
For studio gear, increasing the nominal value of smoothing caps makes sense, as long as the rectifier and xfmr can handle it, just like using solid-state regulators for voltages where stability is essential to the operation (Vari-MU compressors are an example).
For guitar amps, I agree that increasing the PSU caps on a Reverberocket would ruin its sonic signature.

[EmRR]

Agreed.  Have seen plenty of recreations with 10x the filtering which don’t behave like originals with minimal filtering.  Nor have I found originals with practical hum problems, maybe pre-1950 but even then I have 1930’s equipment with hum specs that are hard to beat. 
Most hum problems in classics come from power and audio transformer placement in my experience, in those cases no additional filtering ever helps. 

[Tubetec]

I thought Id add that tube rectifiers come with a recommended maximum value of following capacitor , as mentioned by Rock theres a chance of damage to tube rectifiers by exceeding that value . Guitar amps where the output stages are often driven into audible distortion the time constants in the power section do have an impact on tone . The modern Chinese AC30 has a switch on the back panel allowing selection of 22 or 50 uf , its a subtle enough effect , it makes the bass end a little more solid at higher levels . The other thing I wanted to add relates to choke fed screen supplies on guitar amps , ok you can replace it with an equivalent resistor but I strongly disagree with those who say it makes no difference to the tone , again the differences are more pronounced as you crank up the volume past clipping .

I have to admit to overdoing it a bit in terms of large caps and chokes in some of my preamp supplies with the aim of making a smooth overload characteristic .All the above aside there is still a place in my heart for that old style slightly fuzzy on the peaks sound like you find on an Aretha Franklin record, no doubt a mix of the gear and the tape of the time , you just dont get that much in records nowadays .

[Rusan]

Just thinking out loud here.

Another negative effect that increasing cap values can have is on the PSU's stability.  I know we're dealing with Class A small signal circuits and this really comes into play with Class AB operation tube power amps, but I thought it still worth mentioning because it's yet another reason why blindly increasing the filter capacitance in vintage tube gear just for the sake of increasing it can be a bad idea.  I realize that not a lot of vintage push pull amps have CLC filters before the power tubes, but some do (and I've DIY'ed several) and some of the below principles do apply to Class A in a certain measure as well.

Consider a sudden, high amplitude transient at low frequencies, where the CLC pi filter must instantaneously deliver a lot of current.  A small value input filter cap naturally has more voltage sag but recovers from transients quickly and smoothly, while a large value input cap exhibits better regulation, but upon recovery its output voltage can significantly overshoot the nominal rail voltage and then settle back down with ringing oscillations.  The choke's DCR will help damp the ringing, but of course Class AB amps with CLC filters don't ever (or rather shouldn't) have high DCR chokes because of the resulting poor voltage regulation.  Unless of course, it's a guitar amp where serious voltage sag is what you're actually looking for.

Increasing the value of the CLC output cap reduces ripple, improves regulation, damps oscillations, and lowers the PSU's impedance at low frequencies.  But, it can be a two-edged sword due to the bigger cap's typically higher ESR, which means higher PSU impedance at high frequencies, especially as the cap ages. Ringing will also be pushed to a lower frequency because the larger cap changes the time constant of the RLC lowpass, which depending on the amplifier topology and AF frequencies of interest might or might not have a negative effect.  Lots of stuff to consider.... 

As several have mentioned, large value caps after tube rectifiers can send them to an early grave due to the increase in RMS current value. Abbey noted that it also stresses power transformers, and not only does it increase the I²R heat gain in the secondary, but if it actually saturates the core on transient peaks you'll then have harmonics of the AC line frequency in your power rail.

FWIW, the many times I've tried greatly increasing the capacitance value of a C filter or the output of a CLC beyond "normal" values (in both Class A and Class AB circuits) has not once yielded positive sonic results, at least not to my particular set of ears.  The almost universal negative artifact has been a recessed, lackluster midrange, and bypassing the big caps with small value film caps to cut the ESR has always introduced other sonic anomalies.   

Just some food for thought, so to speak. 

Rusan

[Sc]

Thank's a lot for your opinions - grateful not to be the only one to doubt the sometimes extreme increase practices. I believe that especially the fact, that a PSU has to be fit to load large capacity is not very well known out there. Another issue to the valve rectifiers is the low impedance / DCR of modern power trannies. Recently burnt a "holy grail" GZ34       

[EmRR]

Another issue to the valve rectifiers is the low impedance / DCR of modern power trannies. Recently burnt a "holy grail" GZ34     

I haven't thought about that, it's certainly the case that audio transformers now versus then use larger wire with lower DCR, and it makes things like tube compressors behave and sound differently than the originals they 'clone'.

[Sc]

I haven't thought about that, it's certainly the case that audio transformers now versus then use larger wire with lower DCR, and it makes things like tube compressors behave and sound differently than the originals they 'clone'.

Just started a new topic on that - thanks - very interesting.

JS Transformers ( Jørgen Schou) 1:40, 20ohm-30kohm, recomended applications?

Started by motordisco « 1 2 » The Lab

Experience with beyer BV/TR 310 003 002? 220ohm-2K ? Applications?

Started by motordisco The Lab

7963 submini applications?

Started by ihscoutlvr74 Drawing Board

PNP Transistor Gain Higher than NPN?

Started by CJ Drawing Board