• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

How to design the straight load line for choke loaded tube ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
We are building a 6C45P choke anodic load (40 Hy – 600 DCR) – direct coupled to a 300B.
We’d like to simulate the behavior of the 6C45P on the V-A characteristics curves…but we are not able to design the straight load line. How to design it ?

Which type of input signal we have to consider ? …the p-p or the RMS Volt value ?

Which is the output signal (Volt RMS) of the actual DVD- or CD-Players ?
 
Most CD/DVD players output 2Vrms.

It isn't hard to plot a straight loadline. You get the anode characteristics and draw a horizontal line across it, through the idle operating point. The inductive load will allow the anode voltage to swing above the voltage rail, so Va will be B+ minus the voltage drop across the DC resistance of the anode choke.

The horizontal loadline is a close enough approximation if the grid leak resistor of the next stage is large. At low frequencies and high frequencies it becomes elliptical, but hopefully shouldn't affect anything within the normal audio band.

You would aim to bias the valve such that it will not draw grid current under normal input voltages, which will also depend on the overall input sensitivity of the amplifier.
 
sdfsdf
 

Attachments

  • schematic-3.jpg
    schematic-3.jpg
    94.7 KB · Views: 668
Something is wrong in what I have understood.

If I trace an orizzontal line accross the operating point (-2.4V, 170V, 15 mA) ... and considering an input voltage of 2 VRMS, I have an anodic voltage variation of 200V (275-75) with a null anodic curent variation...
What is wrong ?
 

Attachments

  • curve.jpg
    curve.jpg
    92.8 KB · Views: 598
Jason,
...first of all ...Thank you for your replay.

I'd like to ask some specific question about that. In particular:

1) Does "2 VRMS input" mean that we have a variation of +2V on the left and -2V on the right of the operating point ? .. or do we have to consider the V p-p value to design this variation (... +/- 2.828 V) ?
In this last case, we could reach the positivity of the grid (+2.828-2.4=+0.428 V)!

2) How is it possible the 200 V variation without current variation ?

3) Have you had a look at the schematic of the amplifier ? ... what do you think about it ?
 
1) 2Vrms is equivalent to 2.828Vpk or 5.657Vp-p. With -2.4V bias, a 2Vrms input signal will cause input clipping by causing grid current. However, the 300B needs only 70Vpk or so to be driven to full power, and the gain proviced by the 6C45P is around 50, so we only need 1.4Vpk, or 0.99Vrms to drive the output stage into clipping. The output stage clips before the input, so all is well. This will mean that some attenuation will always be needed at the input such as a volume pot, but that is pretty common anyway.

2) There is some miniscule current variation, it's just that the impedance presented by the 300B's grid and the anode choke is so large that it is represented as infinite. At low frequencies, and at high frequencies, there will be much current swing, due to the reducing impedance of the choke at low frequencies, and Miller capacitance at high frequencies.

3) A proper Free Lunch/Monkey/DRD design may be simpler, yet still offering the advantages of direct coupling and a choke loaded driver while using far fewer parts (and only needs one power supply). Have a look at this offering from Thorsten:
An externally hosted image should be here but it was not working when we last tested it.
 
...Some points I do not like about that schematic:

1) Parafeed.
2) Capacitors (directely or indirectely) in the signal path.
3) Single power supply.

As to the straight load line ...

I have now two versions about the straight load lines for choke load.

- One is that you proposed.
- Another one is that proposed by SomeJoe ... that consider the straight load line variable with the frequency (not always parallel as you tell) (see post at Tube DIY Asylum http://www.audioasylum.com/forums/tubediy/bbs.html

Where is the true ?
 
Antonio Tucci said:
1) Parafeed.

No.

2) Capacitors (directely or indirectely) in the signal path.

*cough sputter*

Oh, and no.

3) Single power supply.

Yes very wasteful of power...

- Another one is that proposed by SomeJoe ... that consider the straight load line variable with the frequency (not always parallel as you tell) (see post at Tube DIY Asylum http://www.audioasylum.com/forums/tubediy/bbs.html

Where is the true ?

I don't know where your post is and frankly I don't care to dig through a *BBS* to find it, but if it is as you describe, it is false. If you must be completely exact about this, the line will be flat horizontal at resonance, eliptical off (due to varying amounts of inductance (at LF) and capacitance (at HF)) and how much depends on the reactance at that frequency.

To a basic approximation, you can ignore reactance if it is equal to the circuit's impedance (here defined by the plate resistance) at the lowest frequency of interest.

Tim
 
Antonio Tucci said:
1) Parafeed.

No. It is series feed.

Antonio Tucci said:
2) Capacitors (directely or indirectely) in the signal path.

You cannot eliminate capacitors in the signal path. The AC current loop between the cathode/filament and the anode has to be completed somehow, and unless you have a short between HT and the cathode...

Antonio Tucci said:
3) Single power supply.

True, this wastes some power, though the "Free Lunch/Monkey/DRD" design allows some of the power that would otherwise be wasted in the cathode resistor to be reused to power the driver valve. The reduction in the quantity of parts may very well allow the use of higher quality parts.

Yes, a single supply will waste power, but don't forget that power will also be wasted in the additional rectifier. The filament of the 5U4 alone takes 15W... If greater efficiency is the goal, just use cap coupling between the stages.

Antonio Tucci said:
As to the straight load line ...

I have now two versions about the straight load lines for choke load.

- One is that you proposed.
- Another one is that proposed by SomeJoe ... that consider the straight load line variable with the frequency (not always parallel as you tell) (see post at Tube DIY Asylum http://www.audioasylum.com/forums/tubediy/bbs.html

Where is the true ?

I never said it was always horizontal. I said that it was a good enough approximation within the audio band. We would hope that the anode choke has enough inductance so that the loadline goes significantly elliptical at LF below the audio band, and that the Miller capacitance of the next stage in parallel with the shunt capacitance of the anode choke causes the loadline to go significantly elliptical above the audio band.

I have not searched the entire AA for that post.
 
I report the post by SomeJoe in Tube DIY Asylum.

" ...In a choke loaded stage there are two aspects to consider: the DC operation conditions and the AC (signal) conditions.

The DC conditions is governed by the DCR of the choke. It will be a pretty steep loadline, because the choke won't drop many volts. The AC conditions are governed by the impedance of the choke at a given frequency.

Since a choke's impedance rises with frequency, there won't be one straight load line (as you would get with a constant-current source load), but instead you will draw at least two loadlines, one for the impedance at say 20Hz, and one for 20 KHz. The low freq. loadline will be more slanted than the high freq one.

That implies that the gain of the stage won't be the same for low freq than high freq (higher load Z => higher gain). The difference between the extremes can be lowered by using low Ra tubes, since gain = mu*(RL/RL+Ra). I think negative feedback, either local or global, can improve matters too, but i don't have the formula at hand and besides i don't use that (isssh!!!).

All the loadlines must cross the choosen op point.

The formula for calculating a choke's impedance at a given frequency: Z = 2*Pi*f*L, where f is the frequency considered, L is the choke's inductance. Use that value to compute one line at 20Hz, and another at 20Khz, all crossing the DC op point.

For this reason, I find it easier to pick an op point first, then arrange the B+ to fit. Consider the max input signal (peak), and choose the op point so to stay out of the grid current zone (-0.9V typical) on the positive alternances of the input signal.

No all line-level sources have equal output amplitude, and i'm not sure there is a standard. Morgan Jones in his book talk about 2V, but he doesn't mention if it's peak, so i assume it's an RMS value, meaning the peak value would be about 2.83V peak.

As an example, in my current amp when i added a -1V margin i got a theorical bias of -3.83V, but settled on 3.2V (i use LED bias, i am forced to multiples of 1.6V)

By the way, 40H seems too little a load for a load, but that's just my opinion. I use a 260H on a 5687; but then again YMMV and all that.

Hope that helps--SomeJoe
 
Okay, I searced the AudioAsylum (it actually wasn't as hard as I thought it would be ;) ), and have come up with the relavent parallel thread.

You have misquoted me in a post you made on AudioAsylum.
Another one is that proposed by Jason ... that consider the straight load line always to be parallel in the case of a choke load

audiousername said:
The horizontal loadline is a close enough approximation if the grid leak resistor of the next stage is large. At low frequencies and high frequencies it becomes elliptical, but hopefully shouldn't affect anything within the normal audio band.

At this point I probably should add that it is actually elliptical to some extent at all frequencies (if it wasn't, this would imply the valve is loaded with a pure resistance, which it is not).

40H is plenty for a 6C45P. The suggestion of using 260H on a 5687 is simply overkill. Downloading this Excel spreadsheet from Jim de Kort's website illustrates this graphically.
 
Jason,
Many many many thanks for the included .xls file.
It is very interesting and helpful.

Now, I think it is clear to me.

I thank you again ... and profiting of your competence I'd like to ask if
- we can use normal relays (250 Vac - 10A; coil:12Vcc) in the schematic reported in the post #4;
- we can utilize an appropirate R (ohm? W?) in spite of NTC in the anodic line of the 300B and 6C45P.

Regards
 
The relays aren't necessary. The valve rectifier doesn't heat up instantly anyway. Changing to a 5AR4 or perhaps even a quad of damper diodes will even a more gradual turn-on.

Where did you get this schematic from? It is needlessly complex.

If you must use relays, you probably can substitute a resistor in lieu of the NTC thermistors, but I'm not completely sure about using normal 250Vac mains relays though because I've never used them in this application. You can use whatever value you want, the larger it is, the less current it will let through. The wattage rating will depend on the size of the resistor.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.