DHT CF

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

5v333

Well-known member
Joined
Jun 30, 2013
Messages
631
Location
Gothenburg
Im thinking about making a circuit using a DHT Pentode, feeding the filament with a regulated voltage and also use a constant current source to bias the tube.

the voltage source gives filament voltage and the current source is set to source filament current + bias current.
is this the right way to think when doing a CF this way?
will the current source deal with the bias by its own or do i need an additional resistor?

here is a scetch of how i think it works like. the tube is not the actual one i have in mind...

 

Attachments

  • DHT.pdf
    10.9 KB · Views: 43
Another way you could do it with less parts is to use a LM317/LM337 as a current regulator. Apply it to the DHT instead of to the indirect heated cathode. Just a thought. Don't know if its what your going for.
 

Attachments

  • 12-28_LM337_2.gif
    12-28_LM337_2.gif
    7.4 KB · Views: 36
5v333 said:
Im thinking about making a circuit using a DHT Pentode, feeding the filament with a regulated voltage and also use a constant current source to bias the tube.

the voltage source gives filament voltage and the current source is set to source filament current + bias current.
You can't regulate both the voltage and the current in a variable load, such as a filament that changes in a 1:3 ratio between cold and hot. The way you have set your system, the total cathode+heater current is regulated. As the filament temp increases, the cath voltage increases too, reducing the cathode current, which means there's more current available for the filament, which temp thus increases, and so on, so you're in a vicious circle that probably ends with too much filament current and zero cathode current.
The datasheet is not much help there, since the app circuit shows the two tubes are used with different bias; in the 1st tube bias averages at 0.375V and the 2nd at 1.125V. It doesn't seem very consistent to me, but that's the way it worked. It suggests these tubes were quite tolerant.
 

Attachments

  • ck505 circuit.jpg
    ck505 circuit.jpg
    20.3 KB · Views: 38
> is this the right way

No. As Abbey says, if you nail one end to +1V, the other end of the low-resistance filament can't go anywhere.

That's (one reason) why there were NO cathode followers in days of filament cathodes.

Note that if you threw out the tube and took even one of the two transistors you have added, you would have a "better cathode follower" than any tube.

In concept you "could" do what you ask with two current sources. One feeding down from above and the other taking that down to a negative voltage. In practice it fails right away because two current sources in series will "fight" if not eXactly matched. Any difference must be absorbed by a resistor. (Actually the tube will mis-bias trying to absorb the difference. But a small difference in filament-feed current is a large difference to cathode current.)

You DO realize that all you have to do is throw a D-battery in there? Life may be many-many evenings or sessions. A rechargeable can be put on switch or relay so it re-fills while you sleep. 
 
> the 1st tube bias averages at 0.375V and the 2nd at 1.125V.

Both stages, the gridleak returns to the - end of the filament/cathode. Bias is a mix of half the 0.75V across the filament, and the grid current in the large 5Meg gridleak. The bias is nearly the same, except stage 2 only has 43.5VXXX 44.3V of the 45V B+ to work with.
 
PRR said:
> the 1st tube bias averages at 0.375V and the 2nd at 1.125V.

Both stages, the gridleak returns to the - end of the filament/cathode. Bias is a mix of half the 0.75V across the filament, and the grid current in the large 5Meg gridleak. The bias is nearly the same, except stage 2 only has 43.5VXXX 44.3V of the 45V B+ to work with.
Correct, I didn't see that. It was 5a.m. when I wrote that.
 
thanks!
a little bit hard to follow so far...
DHT´s are new to me.

PRR said:
> is this the right way

No. As Abbey says, if you nail one end to +1V, the other end of the low-resistance filament can't go anywhere.

what if we put a cap across B-E making the voltage source more like a gyrator?

 
5v333 said:
thanks!
a little bit hard to follow so far...
DHT´s are new to me.

what if we put a cap across B-E making the voltage source more like a gyrator?
Putting a capacitor between base and emitter doesn't make a gyrator. Anyway, what do you expect it to do? Your concern is above all a DC one.
 
Im missunderstanding what you guys are saying.

I need to find more texts on the subject i guess.
 
Hi again!

my last idea was an embarrassnent.
Cant find much about filament bias either so i left my CF for a while.
until today when i saw the old m49 circuit, some braincells started moving in my head. still not sure how bias works in this new idea.

attached is a DHT tube that gets 1.2VDC heater voltage on one side of filament and the other end is tied to ground.
at this point is it safe to say that the tube is also biased by this 1.2V heater voltage (Vgc -1.2V)

im actually aiming for that Vgc region.

underneith the tube is the dc load resistor. im hoping that they are the only components needed for the bias.

the grid resistor is hooked to the cathode so it will look like a bigger resistor for the source.
although i was thinking that if the 1.2v gets up to the grid it would de-bias the tube, so i connected it after the output cap.

have i once again embarrassed my myself?
 

Attachments

  • DHT CF.pdf
    6.6 KB · Views: 8
> embarrassed

I think so.  :-[  :-[  :-[

The grounded 1.2V heater supply shorts-out the signal on the cathode/filament.

Also there is no good from screen grid drop and filter. Wire triode.

The average half-1.2V grid-filament IS the proper bias (they planned it that way). BUT you have to get the heater power from a dedicated floating power source (such as a battery cell).

Output impedance is around 3k ohms. So you won't drive a long line without a transformer.

Grounded cathode gives greater power gain, and you can use a grounded heater supply shared across all tubes in your box. The high power gain can be transformed for higher voltage gain or lower output impedance.


 

Attachments

  • 5v333-CF-42.gif
    5v333-CF-42.gif
    11.2 KB · Views: 26
where did you find the battery symbol in eagle? :)

Ok. Ive still got stains of shame but this starts to get some where.

if its not possible to do a voltage supply out of two transistor voltage/current sources i guess it might be interesting enough to try it out with a battery and some kind of recharging plan.

the tube i have in mind only shows curves in pentode mode Vg2 45V. so there will only be  a question about dropping a few volts.
Triode mode would be interesting to try out, i guess lower impedance, maybe even a switch between them.
the load and bias i have in mind would give about 600R output Z in pentode mode.
final load is a stepdown transformer with about 50k on the primary.

So in your example, does the average 0.6V filament voltage put the tube in 0.2mA bias on the 10k  dc loadline?
 
> where did you find the battery symbol in eagle? :)

There's a whole website for Eagle Battery.

CK505 is a hearing aid tube. If switched off after every session, a D-cell will probably hold charge until the case leaks, a couple years. Don't be over-Green. Carbon/Zinc is milder to the Earth than Lithyates and little charge-packs.
 

Latest posts

Back
Top