|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
157.52.10.96
I am using a 10/vt25 as the input/driver tube for another 10/vt25. From the RCA 10 data sheet average plate characteristics indicate that -5 volt bias and 150 volts plate the current consumption should be 10 ma. So I use a large 4 ohm resistor in series with the filament. Filament current is 1.25 amps, so my bias voltage s/b (4 0hms)(1.25 amps)=5v.
I have a B+ of 380 volts and I need a voltage drop of 230 volts @10 ma so I use 23Kohm resistor. R=230volts/.01 amps = 23kohms.
When I hook everything up I get 176 volts on the plate and current is 8.9 ma. Obviously not 150 volts, 10 ma @-5v. Any thoughts on why such a discrepancy between actual results and those predicted from the plate curves? TIA for any responses I'm really at a loss over this one.
Follow Ups:
Ale's curves suggest that with 150v plate, -5v and 145v plate-cathode you have 13mA.With HT of 380v you drop 230v across the pate resistor, and at 13mA that gives you a 17.7K resistor at 7W. My calculations!
I'd replace that resistor with something better like a plate choke or Ale's gyrator. Ale has all the values for 10y with his gyrator on his site.
At these values you could use a Hammond 126C interstage just as a plate choke - it's rated 15mA. I use this with a 2P29L at 15mA and it sounds fine. I prefer it as a plate choke with a 0.2uF Russian teflon coupling cap made up of 2 x 0.1 uF FT-2. It's a simple inexpensive solution if you don't want a gyrator. If you want to push the current higher than 15mA you can use the Hammond 126B rated 30mA.
I've attached my 2P29L circuit with the 126C to give you an idea. I really like it - Ale liked the 2P29L as well. Cheap as chips. You could keep your precious 10Ys for the outputs. I have a stash which are just accumulating value in my drawer.
Edits: 08/06/17 08/06/17 08/06/17
Thanks for all that. Very cool circuit and 2P29l's are very cheap. But I like to see my tubes glowing so unless there is a glass version its unlikely I will ever do anything with them. Perhaps not logical but if they don't glow whats the difference between it and a transistor?
Regards
Here's my calculations, based on Ale's curves.
Thanks! The circuit I am working on is a 10/10y/vt25 IT coupled to another 10/10Y/vt25. I am using a Hammond 126C. I have a 20volt DC @1.25 A supply for the filament. The heaters are connected in series (15V @ 1.25A) and bias for the input tube provided by a 4 ohm resistor in front of the filament supply. Bias for the second stage is provided by a negative voltage supply.
This is very similar to a circuit by Greenvalve Audio (Luddite 2 10).
The biggest problem I have encountered so far is trying to reconcile the values in the Greenvalve schematic with the RCA plate curves. Delighted to have found Ale's plate curves for this valve.
Regards
Don't you want a separate filament supply for each 10Y? That's usual.
As I was saying, try 126C as a plate choke - just the primaries - with a FT-2 or FT-3 coupling cap. I prefer the sound - tried it both ways. As an IT you have to connect the secondaries correctly otherwise you get treble rolloff. Primaries are red HT, blue plate. Secondaries are black grid, green ground.
What would be your objection to using one filament supply? Both tubes are in the same channel. I could understand objections in terms of channel to channel intermodulation of some sort, but that isn't the case here.
I have had a lot of trouble getting this to work as configured and feel like I am close to getting it to work like I want it to. Perhaps I will try your suggestion regarding plate choke once everything is dialed in and I am happy with the sound.
As regards connecting primary and secondary goes the primary leads are actually brown and blue. The connections as I have it are reversed from what you describe. I don't believe there is any absolute phase so as long as each winding is in phase relative to the other it probably won't make any difference. At least I have not detected any treble rolloff as I have it hooked up.
Thanks for all your help and suggestions.
Don't you want the second tube to have higher bias voltage?
Well to be strictly correct wouldn't that be lower voltage?. All kidding aside the second stage is biased with a separate negative voltage supply. I believe the term used is "fixed bias" but I have never understood this since most such supplies have provisions to allow adjustment of the voltage.
The difference in voltage from grid to cathode will be greater in magnitude for the second stage than the first stage.
Fixed bias is called fixed bias because the bias voltage doesn't wander during operation (transitioning from class A to B for example).
If your filament supply has considerable capacitance from each end of the filament to ground, then you may be able to bias up the first stage to several volts, then put fixed bias on the second stage and still run it on the same filament supply without the two stages interfering with each other too much.
Something like this?
So that brings even more complications, lol. The output tube has 3.75V of bias on the filament and some additional negative grid bias while the driver stage is biased at 11.25V with 0V of additional grid bias.
In this schematic the 22K resistor loading the 0D3 could be a 10W part and work just fine (5W is marginal). I don't know why it's drawn as an adjustable, that's strange. You do need one 0D3, loading resistor, and bypass cap per channel though.
The schematic is also missing the bias supply and the 20V filament supply. Again there's a weird 6 Ohm 25W adjustable resistor, not sure where you're going to buy that.
Not everything is immediately evident in this schematic. Filament power and the negative bias supply come from switch mode supplies. Also note the use of mercury vapor rectifier in the power supply which has a large switchable dropping resistor, not sure how this can prevent arcing before the tube has warmed up.
Well I have built my own version and to my ears it sounds excellent.
Filament supply uses an LED supply rated 20 volts, 1.25 amps with a fixed 4 ohm resistor. I have two 25 watt 6 ohm pots if you are interested. The input tube biases up to 170 volts @10 ma. I omitted the VR tube. The output stage uses my own conventional negative supply and with -25 volts I get 20 ma current @ 355volts. Its dead quiet, no discernable noise, and sounds great. Of course the fact that it makes any kind of noise at the end of the day is a constant source of amazement and pleasure for me :)
Regards
I visited Bartola Valves where I found plate curves for the vt25. At -5volts bias and 10 ma you actually do get 170 plate volts. Seems that the old data sheet plate curves are not too accurate.
Hi!
The 'old' data sheets are accurate. The curves usually refer to the midpoint of the filament and are taken with AC filament. If you apply DC filament voltage and raise it with a filament bias resistor to 5V, your actual bias is more like - 8.75V.
Anyways there are tolerances between tubes so not every tube sample will bias down to the same exact mA value at the same op point
BR
Thomas
In this particular case I think you are wrong. The plate curves on the RCA data sheet for the type 10 clearly state values derived with DC filament supply. But thanks for your input.
Thomas is correct. One end of your filament is at 5V of bias, and the other end is at 12.5V of bias. The average bias of the filament is 8.75V of bias.
Your voltage measurements agree with this.
I guess I am too thick to understand this. All I know is that what I measure agrees with Ale Moglia's plate curves and not with the RCA data sheet plate curves. And what the heck is "average filament bias"? Sorry to have wasted everyone's time.
Imagine that you set the tube up with a hum pot and put the cathode resistor between the center lug of the hum pot and ground. Let's use your case as the example, then you'd have 5V at the center lug of your hum pot. Assuming you're using a DC filament supply, you'll have 5V + (7.5V/2) on one end of the filament and 5V - (7.5V/2) on the other end, so the average bias is 5V.
This is part of the design process of filament bias, since there's no hum pot and bias isn't applied to a virtual center tap.
Not to be dismissive but -5v filament bias from a 4 ohm resistor and 1.25A of current gets me 10 ma current and 170 volts plate just as Ale Moglia's plate curves allow me to predict. For me it works and all this discussion about average bias I find difficult to understand. But thanks for taking the time to try and explain. Perhaps I should just buy a nice transistor amp, which I understand even less ;)
Ale probably referred the bias to the negative end of the DC filaments. That's why they exactly fit your application. But in any case, never expect every tube to behave exactly the same as the plate curves in the data sheet are only typical and tubes can vary considerably from sample to sample even when new.
"Filament current is 1.25 amps, so my bias voltage s/b (4 0hms)(1.25 amps)=5v."
Wouldn't that be 1.35 amps? 1.25 amps for the filament and .01 amps for the plate current?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Uh, not really, 1.25A plus 10 ma(.01A) is 1.26A. Plate to cathode current usually has a negligible effect on calculations.
Sorry about that. Brain fart.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: