audio-talk

whilst buggering about on one of the tube datasheet sites i came across some datasheets for nuvistors. ive never come across these apart from the musical fidelity nu vista amp i heard many moons ago and havent heard the like of since (cant you tell i liked that one)

the nuvistors look like a viable proposition for faffing with as they dont seem to be any more expensive than common or garden driver valves of decent quality.

plus i was looking at the 7f8w phono stage that i keep coming back to and these little beasties seem ideal for a look in that use.

how does one look at them? same as any other valve?

purely theoretical you understand, as the plans for speakers have hit the skids

ant

Yep, just like any other valve, its all in the spec sheet.

having a play today with the datasheet and this is what i came up with. based it on a choke input supply using one of my 275 0 275 mains tx, and the class a1 data on the datasheet. when i worked out the anode and cathode resistors off the graph i got about the same values so i used my own values. hopefully theyre correct

i used the riaa correction off the 7f8w schematic so im not sure if thats correct in relation to the use of the 6ds4 instead, so i made the first cap bigger. just because nick commented that on the 7f8w one it looked abit low. any thoughts on this particular aspect are appreciated...................

the 6ds4 has a gain of 63 so should get a nice amount of gain out of the stage unless its wrong

any comments?

cheers ant

The big question you need to know is what is the anode resistance of the first valve, thats what controls at least the first RIAA resistor.

anode resistance according to the datasheet is 7000 ohms, what do i need to do now to work it out? i cant find the anode resistance for the 7f8w so im not sure what the first resistor figure is based on

Hi cressy,

As Nick says you need to know the load R you would like and the internal anode resistance of the triode, its usually called Ra, I think.

-- Andrew

Looking at the curves, I would guess about 11k for the 7f8w, so you could gestimate adding 4k to the first resistor value.

You could just fiddle with what is at the mo the 300k, 'til it sounds right


See you've got the 1M bias up the front

.

http://en.wikipedia.org/wiki/Nuvistor


"nuvistor", "New Vista"

.

being having a good read of this
http://www.platenspeler.com/background/ ... und_1.html

so far over my head i struggled to see it!
the bits i could understand were very good though, i couldnt understand any of the equations........................... as per usual, but i did find a mac version of spice so if it works i'll see if i can figure out how to use it and sim what i have

bugger couldnt get it to work............need to update to os10.4 apparently even though it says 10.3.9 will be fine

could anyone model it in spice?

i'll stick it on a usb memory stick thing and send it up to me old mans, his computer should run it

Its easy, a filter is like a see-saw, and for it to operate correctly it has to balance resistively, i.e. the resistances must be the same.

Just like two end of a see-saw will balance if they have the same mass on them.

The RIAA curve correction is just a filter.

Now, a filter can balance in one of two ways and there are three resistive parts to designing a filter, "its input"->"the filter"->"its output".

Either the input and the filter must be resistively equal and the output a high resistor or the input high and the filter an the output must be equal.

OK, so the balancing act in this case is that the input to the filter and the filter itself must have the same resistance. The input is our triode's output resistance.

We also know the nominal resistance of the RIAA, as we designed it, let's say its 300K, this is a good one for a valve.

So the valves output resistance must also equal 300K to get the balance. Now we know our triode is likely to be much less than this so we work out what its own contribution is to the resistance and add a series resistance to make it up to 300K.

The output resistance of a valve is its anode load in parallel with Ra, assuming its got a bypassed cathode.

I think that's got, been a long day.

-- Andrew

so for this 18k in paralell with 7k, which is 5.04k?

so if the resistance of the network is 300k then the first resistor would be 295k?

or have i got that arse about face (i always get alarm bells when someone says 'its easy!!!!!! )

Assuming the numbers are right then you maths looks fine.

I get 5.4k for 7K || with 18K and I think, if memory serves me it was a 300K RIAA, so 295K would do it.

-- Andrew

Here's what we did for your 7F8W design.

http://www.audio-talk.co.uk/phpBB2/view ... w&start=30

Basically, we also attempted to compensate for the Miller cap of V2 by changing the RIAA slightly, what the Miller cap of these are the gain you have?

-- Andrew