Octal based "universal tone test-bench" (power) amp

[roberto]

Hello everyone,

I am wondering to build (or at least design) a "test-bench" amp (or just power amp) to be able to test different pair of tubes in push-pull with different configurations.
The idea is to have a rotary switcher (any suggestion?) for voltages, while just different speaker output jacks for impedances.

VVR for the screens (extreme case, distributed load could be through a tertiary winding), or at least different caps and dropping resistors to have a better view of their compression effect.

Voltages will be switched by having multiple secondaries on the PT to be able to have:
300 - 350 - 400 - 450 - 500 Vdc as B+ (covering from 6V6 to 6L6GC to EL34 to 6550 to all KTs)

While for the output transformer there will be a fixed primary with 20 and 40% taps, then multiple secondary taps in order have a reflected impedance (with a 16 Ohm cab) ranging from 2 to 10 kOhm (still to be determined in detail) considering that some values will be obtained by connecting a 8 Ohm cab instead of a 16 Ohm one on the same output.

Bias will be set in a cold voltage range before switching on the standby, then set.

I would like to dedicate a VVR to the PI as well, to make it independent from the output tubes supply, but at the same time the interaction is part of the sound IMHO.

I do some programming on Arduino and it could be made automatic, but I don't want to.
I'm even not considering to have autobias on the second tube, just to keep it super manual.

I'm considering a power amp in order to connect different preamp on it, or to just connect a very simple cleanish preamp.

Any suggestion?
Has anyone ever built something similar?

Thanks

[TUBEDUDE]

How and where are you going to spec the transformers for that? I'd like to see the prices on the quotes. That power xfmr would be nice to have as a bench supply.

[Colossal]

How and where are you going to spec the transformers for that? I'd like to see the prices on the quotes. That power xfmr would be nice to have as a bench supply.

Check this out:

https://edcorusa.com/tbpwr-1

[roberto]

Thanks Colossal,

exactly like those ones that costs less than 100 USD.

As for the output transformer, the price will not be so much higher than a hammond.
I don't expect to pay more than 250 euros for PT plus OPT.

[TUBEDUDE]

Oh hell yeah, I have a use for this! Thanx again!

0005821_tbpwr-hv1.png

[Andy Le Blanc]

I used a generic "bassman" PT and OPT as a standard bench circuit, a reference point. Its a good approach and saves a lot of time.

If you are going to noodle with the high voltage mains with a switch, make sure you use a "break before make" switch, no sense shorting windings on your PT.

Have fun with it.

[roberto]

Thanks Andy,

so around 430 Vdc and 4k Raa?
I have alot of amps with similar conditions, but what I would like to do is have the choice to try different voltages and loads.
Before doing so I need to calculate the needed primary inductance, in order not to cut too much lows when using higher loads.

[Andy Le Blanc]

Do you have a single generator with enough juice to figure the turns ratio of the OPT?
You can get creative to a certain extent with loading without swapping the iron out, but...
you need the turns ratio to provide the figure.

[roberto]

Hi Andy,

everything will be custom, that's why I wrote:

Voltages will be switched by having multiple secondaries on the PT to be able to have:
300 - 350 - 400 - 450 - 500 Vdc as B+ (covering from 6V6 to 6L6GC to EL34 to 6550 to all KTs)

and

While for the output transformer there will be a fixed primary with 20 and 40% taps, then multiple secondary taps in order have a reflected impedance (with a 16 Ohm cab) ranging from 2 to 10 kOhm (still to be determined in detail) considering that some values will be obtained by connecting a 8 Ohm cab instead of a 16 Ohm one on the same output.

So, considering worst case for the OPT, it will be needed to have let say -1 dB at 80 Hz with highest impedance, so 10k Raa primary impedance for EL84s (yes, I will plan adaptors) in pentode, so 10k||76k = 8k8 then not less than 35 H (Hammond 1650E has 53 H).

So the primary has a fixed number of turns with 20%v and 40%v DL taps, and the primary impedance is varied by connecting the 8/16 Ohm speaker(s) to different secondary taps.

So it can have:
- 3.4 kOhm connecting the 8 Ohm speaker into a 8 Ohm tap;
- 4.4 kOhm connecting the 8 Ohm speaker into a 6.2 Ohm tap;
- 5.2 kOhm connecting the 8 Ohm speaker into a 5.2 Ohm tap;
- 6.8 kOhm connecting the 16 Ohm speaker into a 8 Ohm tap;
- 8.8 kOhm connecting the 16 Ohm speaker into a 6.2 Ohm tap;
- 10.4 kOhm connecting the 16 Ohm speaker into a 5.2 Ohm tap;

Then, with a 64 Ohm tap and grounding the 8 ohm tap, I can apply cathode feedback to the output tubes too.

As for the power transformer, having in mind from 250 to 500Vdc with full bridge rectification, it will be: something like:
0-180Vac rectified and filtered by its own plus a second winding on top with: 0-40-75-110-145-180 Vac
This way I can also test tubes with lower screen voltages. 600 mA seems a safe value.
For the heaters 3.15-0-3.15V 6A to be safe.
0-100V 100mA for the bias.
100-0-70V 50mA in case I'd try AB2 with 6L6GCs.

[Andy Le Blanc]

Holy mosses, your all over it... Double check the the ma. rating of your OPT primary for a max design limit.
If you explore tube multiples p-p, or,.. the higher KT-- 's, the max dissipation can cook a meal and heat the shop thru winter.

[roberto]

I’ve considered 360 V x 600 mA = 216 W.
It seems reasonable to me, even considering a pair of KT88 or even GU50. I don’t plan to test KT150 due to all negative feedback they got about lack of reliability.

[roberto]

Just to show some examples, in attachment:

[TUBEDUDE]

200+ watfs out of a pair of KT-88's is asking a lot isn't it? Are there any reliable examples out there?

[roberto]

It is not possible to get 200 Wrms from a pair of KT88 in push pull.
With maximum 500V B+ you can get around 80 Wrms.

[Andy Le Blanc]

Plate dissipation,... not necessarily Wrms at output,.. even a pair of 6550 can dissipate 80w at design max, kt88 can go 90w.
The max rating for the kt88 cathode is over 200ma. per tube. The current rating of an OPT primary becomes a limiting factor.

I've made this very assumption,.. no fun when the potting wax starts to run.

Your defining your own design center for the project, and a corresponding system design max.

say kt88 run at design max,... 450v.... 45w... thats .1 A... or 100 ma. per tube. A quad is close to .4A 180w total plate dissipation,
that the opt primary has to handle the 400ma. (200ma. per side) with enough extra capacity to not result in significant heating or core saturation.

I like your PS, but I might suggest more the one OPT, flexible loading is quite different from power handling.
The current goes up when you drop the plate voltage, to maintain the dissipation figures. 45w at 300v is about 150ma. (x4 is 600ma.)