|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
37.152.212.125
In Reply to: RE: Can someone explain DHTs to me? posted by coin on April 15, 2016 at 00:03:13
There are 2 voltages here:
1. Voltage across the filament. Think of this as "floating".
2. Voltage from one side (or centre point) of filament to ground
Voltage across the filament in a 300b is 5v
Voltage to ground could be 70v
In filament bias, the filament supply (which is floating) goes first through the filament as above, but then continues its journey through the cathode resistor to ground. So for a 26, You would have one amp of filament current going through the filament and then down the cathode resistor to ground. So the cathode resistor in this case is about 10 ohms, so you don't need to bypass it.
Follow Ups:
"You would have one amp of filament current going through the filament and then down the cathode resistor to ground."
Andy, the filament is "referenced" to ground through the cathode resistor but no filament current flows through the cathode resistor.
Only the plate current (cathode current) flows through the cathode resistor.
The current loop for the filament current (heating current) is just the filament itself (acting as the heater).
The current loop for the audio signal (the plate current) is ground - cathode resistor - filament (acting as the cathode) - plate - plate load - power supply - ground.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I'm talking about a 26 in filament bias. As I said, in filament bias the DC current reg - in this case Rod Coleman's reg which is floating - is connected at the plus side to one side of the filament, and the other side to ground. The other side of the filament goes to ground through the cathode resistor. So the current reg goes through the filament and then through the cathode resistor.
The audio signal also goes through the cathode resistor. So the total current going through the cathode resistor is 1 amp for the filament plus e.g. 6mA for the audio signal.
That's how filament bias works.
"I'm talking about a 26 in filament bias."Of course you were. I should have known that.
I don't know how I missed that.
Sorry Andy.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 04/16/16
There's a lot of users now that would consider filament bias the new standard for tubes like 26, 4P1L, 01A and probably 10Y as well. No cathode bypass and a resistor nor much larger than 10 ohms is a nice design option. It's a great choice for output tubes as well, if you use 4P1L.But it's a DHT choice, and once the bias gets too high combined with an amp or two of filament current, it's just not possible or practical. So no good for 45, 2a3, 300b unfortunately.
But what it does, it does really well. Thomas Mayer popularised it and Rod Coleman created filament regs that made a clean DC supply cheap and easy. A lot of us jumped on the bandwagon when we saw the advantages. I've been using it since around 2009 now in all my amplification. Rod's designs started to be published around 2010. Seems a while ago now.
Edits: 04/17/16
So if I understand correctly the I can put whatever voltage bias I want on the filament as long as the other end has a 5v differential?Does this mean I can use a DHT as a cathode follower if I use a DC filament supply and just keep the filament at a 5vDC differential?
How do I maintain this offset?Also, which end counts as the differential to the plate voltage?
If I have 50v on one end of the filament and 45v on the other, which end counts as the cathode to plate voltage? How does it know the difference?How bout a white cathode follower? Clearly the tube cannot output enough current to run the other filament, but how much filament current is transfered to the plate? Can I just a source follower on the "cathode" of the top triode in a WCF to power the filament and then use the bottom triode to operate the push pull?
In an IDHT the current goes straight from the cathode to the plate, but since the filament has its own current how does it work?
Edits: 04/15/16 04/15/16 04/15/16
"How bout a white cathode follower?"
IMHO designing WCF with DHTs is too complicated, solvable, but not practical.
Lower output impedance is an advantage, but with other solutions (other tubes, DHT/IDHT mix etc.) give better results.
As you can see the upper tube has DC filament (current or voltage is equal in this respect), the lower use filament bias. The anode current flow trough filament bias resistor, but it is negligible beside the filament current.
Hmmmmm. I thoroughly enjoy my current WCF output stage and I figured since DHTs are so much more linear than the 6as7/6080s I'm using that it simply must sound better. Am I right in assuming this?
In any case I would go for an OTL solution. The filaments complicate it a bit but it won't be too hard to null the offset of the filaments and still get the 1 amp current.
Although I'm still a bit confused on what the cathode to plate voltage is when there are two voltages on either end of the filament.
I'm curious what you meant when you talked about DHT/IDHT mixing and other tubes.
"I'm curious what you meant when you talked about DHT/IDHT mixing and other tubes."
I regurarly use (sand or tube based) CCS loaded DHT tubes, sometimes CCS built with IDHT tube (for example E180F/E280F/D3a, 6N6P etc.).
Where did you get that DHT model for LTspice?
"If I have 50v on one end of the filament and 45v on the other, which end counts as the cathode to plate voltage? How does it know the difference?"
As you can see at the VT25 CF schematic, this simple practice centered the filament.
The CF use negative supply (practicaly direct drive to output tube) and "LND150 CCS bias" to adjust via CF tube the output tube bias.
Distortion a bit high, but if you use appropriate CCS insted of R4, the distortion will decrease one order of magnitude.
Bah, I'm still confused. Are you saying the "cathode" voltage of the cathode to plate differential is the center voltage between both filament terminals?
Also why are you feeding the grid with a current source? Grids don't consume current unless they are positive right?
"Also why are you feeding the grid with a current source?"
This simple bias circuit use CCS (its current largely independent of changing of the PSU voltage) to pruduce bias voltage on the R103 resistor.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: