PSUD designer simulation for TSEII build

[lykkedk]

Hello.

While finishing up my amp., i was thinking of upgrading the Triad choke to some DC-Toroidal with the following specs. ::

Toroidal choke 10H 250mA

Through the use of innovative toroidal cores production and winding technology, compared to EI-chokes, we have achieved exceptional electrical parameters. Major advantages are: minimal external magnetic leakage fields, noiseless operation and much lower DC resistance.
Technical data
Core type toroidal
Inductance L 10H
DC Resistance 45Ω
Nominal DC Current 250mA
Nominal voltage 400V
Dimensions 88 mm (OD) x 35 mm (h), weight: 1,3 kg

I have some quistions regarding the psu-part of my TSEII (ripple, B+ etc...)
I use "stock" components, aka' C4=47uF, C5=150uF, Choke=Triad C14X (6H / 150 Ohm)
Transformer 325 - 0 - 325 @ ~220mA (Toroidal custom wound)

I found an older thread here, where this was also discussed, but didn't answer my quistions through. https://www.diyaudio.com/forums/tubelab/163641-lowering-voltage-4.html

Playing with PSUD designer, i can simulate the two different chokes.
Looks like ripple goes from 0.06 to 0.03 swapping the Triad with the Toroidal one.
Same time voltage rises around 20VDC with approx. load of 165mA

What i cannot figure out is, will the higher voltage stay there with B+ around 20vdc more or will it lower when actually in amp. loaded with my 300b biased at ~67mA or will i "win" nearly 20vdc of B+ (guess not ?)

Ohh... one thing more, is the load set at 165mA on psudesigner right?

Any other things i have to consider? - Consequences perhaps?

Jesper.

 

[lykkedk]

Hi...

This is the toroid i am considering to buy :
TDS10/250 - Toroidal Choke 10H 250mA DC - Shop Toroidy.pl

I tried running a lot of simuations, and the clear view is that bigger Henry gives lower ripple, an 10H choke is about double as good as an 6H one.
Also the lover resistance, the higher B+.

When i "load" the simulated PSU with 230mA i have 355vdc@B+ with the C14X choke excatly as in real life here. When i replace the C14X with the toroid one, the ripple goes to nearly half and B+ is rised to 380vdc...

I'am about to order one... but like some opinion first cause i'am still newbee into the tubes

Rgds; Jesper.

 

[cogitech]

I wouldn't bother spending the money for such a small change in ripple. Yes it is "half" the ripple of the C14X, but the ripple is extremely small to begin with.

If you want the higher B+ then that seems like a valid reason to change the choke.

 

[itsikhefez]

I think 230mA is higher than what the circuit will actually be.
At 67mA per tube, I estimate around 175mA.

C4=47uF is the default value and is what I use on the 300B as well.
C5=150uF is without the additional PP or oil cap.

To my understanding, if you have around 50mV ripple on the primary side, that will result in 2mV ripple on the secondary of a 5K:8R transformer (25:1 ratio).
It depends on your speaker sensitivity, but if you dont want any hum, I think 1mV ripple or less (at the speaker) is a good target.
(I'm no expert here so best to get further confirmation)

With the 10H choke, you get pretty close with 35mV PS ripple (simulated, ofcourse).
With the voltage drop of the opt, the 300B B+ will be around 385V which is an operating point alot of people like.
If your opt's are compatible you could increase the bias to around 80mA (thats about where I have it)

If you didn't add the supplemental cap to C5, another 80-100uF help reduce the ripple a bit

 

[lykkedk]

Thanks guy's...

Well for a start i didn't calculate with the motorrun
cap. in the PSUD program (was not sure if it would count there)

I have C5@150uF and motorrun@100uF.

Would this change things a lot?
Regarding the B+ i think it would also be good enough, i just need to
be sure (well simulated) it would not be more than 385 or so!

Btw: I have none hum, but i still think ir would be a good upgrade.

Rgds; Jesper.

 

[itsikhefez]

Yes hum would depend on speaker sensitivity at those levels. I personally had that issue in my first TSE that I built as a headphone amp, as those were 103db and I used a 5K:32 OPT which has a smaller turn ratio.

You can change the second cap in the simulation to 250uF.. it will slightly bring down the ripple.

 

[lykkedk]

Morning ( well in Europe )

Preciate all reply's here...

I tried to simulate with load@175mA and chokes 6H@150Ohm // 10H@45Ohm now, with C5@150uF+100uF(motorrun)

I end up with simulated B+@375vdc with my current build (B+ is actually 355vdc in real life)
And if i choose 10H@45Ohm toroidal choke i end simulated with B+@392vdc.

Soo my quistion, why is there a difference at approx. 20vdc between my current build simulated and the build in real situ/real life ?

Rgds; Jesper.

 

[itsikhefez]

Unfortunately I dont know the exact answer to that.. but if I were in your shoes the takeaway for me would be that if I swap in the new choke, B+ will be in the range of 372V-392V.

One additional thing you can check is the transformer value. I see there that you have 325V, but that doesn't take into account the transformer regulation and its rating.
If you double-click on T1 325V, then on the ellipsis (...), plug in the actual values of your transformer and see if it changes anything.

 

[itsikhefez]

One more possible explanation is that the simulation only considers B+ and not the negative side of the circuit that is also connected to the HV tap.

Someone that actually understands how this works would need to validate if that makes sense.

 

[lykkedk]

Unfortunately I dont know the exact answer to that.. but if I were in your shoes the takeaway for me would be that if I swap in the new choke, B+ will be in the range of 372V-392V.

One additional thing you can check is the transformer value. I see there that you have 325V, but that doesn't take into account the transformer regulation and its rating.
If you double-click on T1 325V, then on the ellipsis (...), plug in the actual values of your transformer and see if it changes anything.

So it's getting fun this

With load at 175mA ::

I tried to make load regulation at 14% on mains trafo (Toroidy say's it's between 12-15%).
Setting the C5 to 250uF, this simulation now gives within 1vdc the real value with my 6H@150Ohm choke i have now. (~355vdc@B+)

When i simulate the Torid choke 10H@45Ohm i get ~375vdc@B+

So i think (think? ) the 20vdc i couldent figure out before, is duo to load regulation.

So i guess i have to buy this toroidal choke to see if it's calculated right.
Don't we agree

Rgds; Jesper.

 

[cogitech]

Sure. Buy it!

I found some interesting info about ripple, but I am not sure if I am doing the math right. Maybe you are interested in this too (?)

Anyway, this site Vacuum Tube Power Supply Design states the following:

In chapter 14 of his book "Theory and Applications of Electron Tubes", Herbert Reich gives some good rules of thumb for just how much ripple may be acceptable for various types of circuits. Based on Reich's recommendation I generally require a B+ voltage ripple level of at least -90dBv or about 0.0031% for any general purpose audio amplifier such as the Ghost amp. So this is where we'll start with our filter design.

Using the data you posted above - your ripple at approximately .06v with the Triad vs. .03v with the toroid) and B+ at (let's average it to) 360v:

.06 / 360 = 0.00016666667%
.03 / 360 = 0.00008333333%

So, if 0.0031% = -91dBv then either of the two results above must be "really effing low".

My math could be wrong, and perhaps you find it irrelevant anyway because the new choke certainly appears to give you more B+.

 

[lykkedk]

cogitech & crew

I think you calculated right.

I am sure i can't resist ordering this choke.

Will see if another wise would tell something more? - Certently got very good info today here
Thank's all.

Rgds; Jesper.

 

[lykkedk]

AutomaticMouseoverClickOnBuyStuff

So i did order this choke now, by a little luck i could have it by end of next week.

I'am gonna start all over with adjusting the amp., as soon the new choke is installed between the two OPT. -Starting it all up at first with my variac

excited to see/hear/meassure the result's...

Jesper.

 

[lykkedk]

Evening all here

So finally after 2 week's i recieved the 10H/250mA choke...
I calculated some B+ voltage with the new choke to be around 375vdc with bias at 68mA or so... With the new choke i endup with 365vdc but with a higher bias at 75mA.

When bias is 68mA the B+ is 368vdc so ~7vdc lower as i (PSUDesigner ) calculated...

I will try to simulate with the higher biasload and post results here one of the day's.

Back to test / assample the good tubelab amp.

BTW : Good news is also that choke don't get hot at all, the Triad was hot to the touch after ~1hour around 70C. (a bit more than i really like, also knowing that insulationclass on the cu windings proberly are rated F witch is not good at >125C., but okay a good margin through)

Jesper.