SimpleSE: C1 value

[squiffiness]

To all:

Looks like Mouser has discontinued their 47 uF, 500V rated electrolytic caps needed for C1, unless you buy 475 of them. Mouser is my source (without going into the why's of my choice, it needs to be Mouser).

The datasheet for the 5AR4 lists allowable values of the first capacitor (Cap Input Filter) as 4 to 60 uF. So, in theory C1 could be any standard value from 4 to 60 uF. 500V or better rating is needed, clearly for the SimpleSE.

However, can anyone convince me whether I should upsize/downsize to the next standard values, being 33 uF of 56 uF.

Neither seems to have much impact on the final B+ voltage (PSUDII sims), but the 56 uF seems to create a little more start up voltage overshoot than the 33 uF.

Danke

[rknize]

I would not go up in value. The larger the C1, the more stress it puts on the 5AR4 during the power-on surge. New production tubes are more prone to arching over. I would go with 33uF.

[akimmet]

I recommend going smaller as well. Most power transformers don't have enough winding resistance to make a 5ar4 happy. That 60uF figure usually requires extra 150ohm plate resistors to increase the impedance of the power supply enough to prevent arcing over. 47uF is already too much for some modern 5ar4 versions.

See rating chart 3 in the GE 5ar4 data sheet.
http://www.drtube.com/datasheets/5ar4-ge1959.pdf

[squiffiness]

Originally Posted by rknize

I would not go up in value. The larger the C1, the more stress it puts on the 5AR4 during the power-on surge. New production tubes are more prone to arching over. I would go with 33uF.

Originally Posted by akimmet

I recommend going smaller as well. Most power transformers don't have enough winding resistance to make a 5ar4 happy. That 60uF figure usually requires extra 150ohm plate resistors to increase the impedance of the power supply enough to prevent arcing over. 47uF is already too much for some modern 5ar4 versions.

See rating chart 3 in the GE 5ar4 data sheet.
http://www.drtube.com/datasheets/5ar4-ge1959.pdf

Thanks guys.

Any other thoughts on the stock values in the other section of the PS CLC filter (i.e. the selection of PS Choke 5H/10H and C2/C.aux) that follow?

My current BOM selections are a 5H (ESR 58R) Edcor unit (cost driven), with a 120uF electrolytic for C2 and an ASC 40uF/440VAC motor run cap for Caux.

[rknize]

Those selections are probably fine. 5H is a little low, especially given the low 58 ohm DCR (I assume that's that you meant). Some additional series resistance might be warranted to get closer to the 150 ohms of the original design. Folks typically run a bigger motor run cap (70-100uF range), but it's not that critical. I went with a 220uF C2, since I had it on hand and it fit the PCB.

[rmyauck]

Adding SS diodes, can help 5AR4/GZ34 life too without hurting sonics.
See post #18 and discussion:

Does the type of 5AR4 influence the sound? - Page 2 - AudioKarma.org Home Audio Stereo Discussion Forums

Randy

[squiffiness]

Originally Posted by rmyauck

Adding SS diodes, can help 5AR4/GZ34 life too without hurting sonics.
See post #18 and discussion:

Does the type of 5AR4 influence the sound? - Page 2 - AudioKarma.org Home Audio Stereo Discussion Forums

Randy

Not sure this would work with the printed circuit board mounted octal socket. Pins 5 and 7 go to nowhere and you'd have to jumper pins 2 and 8 to allow the socket to be rotated to make it work according to your sketch.

I believe that since the diodes are in series with the HV windings, they could be added to the bare winding leads and the diodes inserted into the terminal strip at the board edge in the design.

While I must be hallucinating, I seem to remember a thread somewhere on DIYAudio that mentioned this. Can't find it now.

Since I'm a novice at the tube stuff (OK with soldering), I am leaning toward an essentially stock build. This thread started based of the preferred value for C1 being unavailable at a preferred supplier AND limitations on my budget due to a limited WAF. Currently what the amp will eventually look like leads to uncertain WAF. If I spend too much, the WAF will tend toward zero.



There is another thread somewhere about wiring in diodes

[rmyauck]

Yes, that would work to just put them in series from the PS leads to the circuit board , paying attention to polarity. You could use heat shrink tubing to make it safe and sturdy.

That drawing is for point to point wiring saving using a terminal strip as there are convenient unused pins handy.

Anyway it will help the tubes last for years for a few cents as UF4007 diodes are at the most 35 cents each.

Randy

[rknize]

The subject has come up for amps like the Dynaco ST-70, which push the 5AR4 more than some other designs do. The modern tubes don't like it.

Dropping C1 to 33uF would normally have a negligible effect on ripple. The relatively low DCR and inductance of your choke makes me suggest that you make up for it anyway with a bit more C in C2. It's probably not going to matter, but it's something to think about.

[boywonder]

Squiffy, your choices look fine...and as Russ mentioned, the low DCR of your choke may result in a bit higher B+ than anticipated. You can always swap in a 5U4; that usually drops about +/- 30V than a 5AR4, depending on current draw. The 5U4 comes to life a lot faster than a 5AR4, so you lose some of the benefits of a soft start. Well, you still get a soft start but you'll get some B+ overshoot since the 5U4 will be conducting before the output tubes are warmed up.

If the first cap value stays above about 10uf or so in a CLC PS, it won't change your final B+. Reducing the value of the first cap to less than 8-10uf or so will start to reduce your B+ voltage as the first cap value gets smaller. So you can always try some 1uf - 5uf or so caps for C1 to bring down your B+ if req'd.

Model your PS in PSUDII and play with the choke DCR and the value of the first cap for a quick and easy demo/sim.