 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
203.129.50.139
In Reply to: RE: US made NOS are good, too posted by Garg0yle on August 28, 2015 at 08:58:02
Fast, soft recovery SS diodes are reliable, quiet and have no difficulty filling up large capacitors. Also, reliable and inexpensive.
Stiff power supplies are a good thing.
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Follow Ups:
I agree, usually I do build my amps with a SS supply.
As Eli mentioned it does have it's merits for slow turn on, although might not be necessary for receiving class tubes.
By the nature of the design, I don't think I will hear an improvement in bass, but I suppose I could try by making something similar to a copper cap.
Infamous sockpuppet
But it is an important feature for most valves.
You can simply put a thermistor in the primary / AC mains side of the power transformer and leave it at that. And the heaters will come on first.
Or add another in the B+ / HT secondary on the other side. IIRC that is what we did for the LEAKs.
Or you can use a 555? timer circuit to delay the B+, set to the time your circuit takes to get all its filaments nice and hot.
Or ..... with Eli's suggestion using both VRs and SR diodes.
I simply saw the space where the GZ34/5AR4 socket was and wanted it for a filter cap. ;-) For the best sonic reasons.
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Tim I'm not sure if you are familiar with my amps, but do you think I would still see a bass improvement with these 1 watt class A amplifiers?
Even though they have valve rectifiers, the B+ is rock solid in this design.
(As well as the B+ on the 6J5)
It is a direct coupled design with the B+ for the driver coming off the cathode of the 2A3 similar in topology to the Free Lunch and BB Proof etc, but much better performing.
Fortunately they are mono blocs so I could in theory do an A-B comparison, it's just that with this design I would have to reconfigure all of the operating points to account for the increased B+.
It's doable, just tricky, I would have to be reasonably confident before I break out the soldering iron.
While I have not modeled the SS diodes in spice to look for an improvement, it is hard to imagine they would be better then what is there now, from a practical point of view.
Right now a 20Hz full blast sine fluctuates the B+ ~0.8 volts.
a 1000Hz full blast sine fluctuates the B+ ~0.01 volts.
If I may pat myself on the back, they are phenomenal.
Are there other areas I should look for improvements that I am not considering?
Infamous sockpuppet
I am NOT an expert on valve circuits nor on PSU's but I do like to absorb general principles, that are supported by empirical evidence.Viz? A good high-storage PSU sounds better than a good but VR-limited PSU. Ripple will be a whole lot lower, for a start.
And, that SS diodes offer you the possibility to substantially increase the energy storage in a valve amplifier, which VRs don't like. If we square 50 volts we get a factor of 2,500 if we square 300 V ( 6 times the volts) we get a factor of 90,000. Divided by 2,500 we get 36 times the storage.
In other words a quite small valve amp - in WPC terms - can have a far stiffer PSU than most BIG SS amps. And IMO&E far better bass. Less dry and tight perhaps but lots of slam. If it's a PP NFB amp then the advantage is considerable.
That is I am not claiming that good SS diodes - SRs / Schottkys - snubbed or not sound better than VRs, but pointing out that VR's do limit the stiffness we can achieve in a PSU.
Peak Inverse Voltage is the limitation VRs have.
I don't see how a VR can to do anything special to the sound a PSU gives us, but I know that a VR will limit the volts we can get and severely limit the amount of storage we can have. I also know that their use in guitar amps is because of the sag they are able to give. No Hi-fi amp needs sag.
So I ask, why do it?
Back to your amps?
It may, as you think possible, be overkill to increase the storage of your 1 watt amps. I don't know.
While it is quite possible to have too much storage which can pulse the PT too much, I haven't heard any negative consequences from these big PSUs in the LEAK St20s. The PT in these is probably a bit stressed in hot climates, so it could have been an issue.
I don't have an OEM LEAK St 20 circuit diagram to hand, but they had IIRC, three? or four 32uf? (small ones anyway) capacitors, circuit-value wise. Two in one can I think.
We used 4 x 470uf 400V dark-brown Nichicon caps in the bass-duty amp and 3 in the mid-treble amp. The last two in each amp are bypassed with MKP and MKS caps.
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Edits: 08/30/15
OK just to update this thread, I am going to use 5V4 rectifiers instead.
They have a little more drop then the 5AR4, but modeling suggests that I won't have to change the amp around.
I suppose because I am not drawing near the rated current through either tube, the voltage drop in practical use is pretty low and comparable between the two.
At full rating the 5AR4 drops 17 volts while the 5V4 drops 25, for all intents and purposes they are within a percent or two of each other in my application.
Thanks for all the suggestions, I'm not sure why I over looked the 5V4 the first time around, I guess I just missed it reading incomplete lists.
So I have a nice pair coming from the 'bay.
The price is SO MUCH better for the 5V4 as well.
I think I paid around $24 USD for the tested and matched pair, that is more like it!
(Coke-bottle glass, black-plate RCAs, what's not to like!)
The 5AR4 is off my list completely in the future, I won't back myself into that corner again by choosing them. I will let the other turkeys fight over them and the crappy new production. Screw that noise.
△ᴉʇɐuᴉɯnllI oᴉpn∀△
Tim,
All PN junction diodes exhibit a reverse recovery spike (switching noise). In modern UFnnnn types, the spike is small and easily snubbed.
The somewhat recently introduced high PIV silicon carbide Schottky diodes are noise free. The only item of advantage left for vacuum diodes is the slow start of types with cathode sleeves. Several good methods are available to "soften" the start of SS rectified PSUs.
Given forward drop considerations, old units built with vacuum rectifiers are frequently best left that way.
BTW, the great vintage amps known for superior bass performance (McIntosh 2nn, H/K Cit. 2, Marantz 8B) employ SS rectified PSUs. :> D
Eli D.
More power too, though we didn't exploit all the extra volts. The real benefit was a much wider full power bandwidth, which is more important to fidelity.
The bass is better than from the Class_A STAX DA80 I had for back-up, and it should be, with so much more energy storage.
There are a few Rs in the PSU to hold back the HT volts a bit. IF I wanted to, I could reduce the resistance in their PSUs, and put in a choke, and still perhaps get a bit more power.
But, when we started we weren't fully aware of the EL84M option.
Learnt a lot. .........
Get rid of all high mu / low GM small-signal tubes and you can use a lot less NFB in the gain stage. (LEAK power amps have a lot of gain, lost some.)
Add a CCS to the splitter-driver stage, and you can use less NFB around the following OPT stage.
Use NFB locally, and far less overall loop NFB. (Figure out what is enough loop NFB for your room/spkrs.)
Use best possible R's and Cs for each NFB section, local or loop.
Have a big *STIFF PSU - because you now can - and by-pass the last two caps.
*How much energy we can store in a PSU is given by the voltage squared and valve amp PSUs have 100s of volts.
IME valve amps can have far better bass IME&O than SS amps can. Warm expressive nuanced tube bass, but with SLAM.
IF I find I need more than 20watts Pentode (bass) and 16W UL (170 Hz up) for the being-built Manger based speakers? We can lose those Rs and use those extra volts! OR change the PT!!
I do agree that for some classics it is best to stay at the same power. Less stress on the original iron / output tubes, where the design runs them hard already.
I tried to stay with the GZ34 VR but they weren't cheap ..... and watching a Chinese one fail was not fun. Lesson one.
In short, I don't see any audible / lifetime / MTBFailure advantage to VRs, given that we can slow the inrush in other simple or more complex ways. Dropping them increases power-bandwidth and hugely improves the bass.
Who needs 'sag' in a hi-fi amplifier, eh?
We used the GZ34 octal sized hole for another PSU cap. ;-)
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
