Crucial is the output impedance of your DAC or phono stage.with high output impedance you can not drive AVT and input transformers properly
The DAC output impedance = 540 ohms.
Z = (A1*RL/A2) - RL = (0.53 * 1250 / 0.37) - 1250 = 540 ohms
Where:
Z = Output impedance
A1 = Open circuit voltage amplitude
A2 = Loaded voltage amplitude
RL = Load resistance
Z = (A1*RL/A2) - RL = (0.53 * 1250 / 0.37) - 1250 = 540 ohms
Where:
Z = Output impedance
A1 = Open circuit voltage amplitude
A2 = Loaded voltage amplitude
RL = Load resistance
Regarding the implementation of the 1LH4 tube, the initial operating point I am targeting is what is published as "typical operating conditions" for class A. 90V, 0.15mA and 0V bias. See attached.
To get the 0V bias, is any resistor and/or cap needed for cathode bias?
Attachments
OK, Cagomat. Please help me out, as I am still learning. Looking at the curves, where would you suggest the operating point be?
In the datasheet is the bias voltage refered to the negative filament and not to ground, so you need bias via resistor. From the data this will be a horrible driver tube
Cagomat, this is an experiment with a DHT driver. It cost $15 to buy 3 tubes, so I said "go for it" and play around. Curious to learn about a mu=65 and DHT driving a DHT. Will this be my "forever" driver? doubtful. But I want to learn.
Cagomat - Also, The DAC output impedance = 540 ohms. Thoughts on this going into the Elmaformer --> LL7903 transformer --> DHT --> 45B?
Why is a dht a better driver than idht?
1.) Headroom = low distortion
when you need 2V input for full power and the tube are biased with 1V you get a lot of distortion
For example I biased the 10Y with 18V and EML 20AM with 12V
2.) Possible voltage swing
3.) Linearity
I see with your equipment 2 possibilities for best sound and overload protection
EML30a with a different interstage transformer or high inductance choke load
C3m/C3o in pentode mode biased with 5V
1.) Headroom = low distortion
when you need 2V input for full power and the tube are biased with 1V you get a lot of distortion
For example I biased the 10Y with 18V and EML 20AM with 12V
2.) Possible voltage swing
3.) Linearity
I see with your equipment 2 possibilities for best sound and overload protection
EML30a with a different interstage transformer or high inductance choke load
C3m/C3o in pentode mode biased with 5V
It will work, but not good, I use for example a DHT linestage with output impedance of 200ohm. 540 is to high to drive an autoformer with input transformer. For 10Y you need 1:4 or 1:8 ratio
I'm with Cagomat here - don't be so fixed on the idea of a DHT. Start with a tube that offers decent drive, a reasonably low plate impedance and a realistic operating point. I would suggest this means around 7mA current minimum, at least -1.5v bias and a plate impedance of not much over 12K. You can go better than this, but that's fairly basic.
I have 89db sensitive speakers and don't listen loud, so I'm OK with driver with a gain of x30. A gain of x40 would be better for most. If the 10Y has a mu of 8 then the step-up required looks like x4 as Cagomat is pointing out. With low sensitivity speakers it might be more. A step-up of x4 does indeed require a low impedance input.
I spent quite a lot of last year trying to use a DHT driver for a 300b amp, and gave up in the end. It's not that it can't be done, it's just that you can get good sounding results with indirectly heated tubes. As a DHT addict that was a big step for me. I didn't believe indirectly heated tubes could be that close to a DHT, but in fact they can. Not perfectly comparable, but pretty close and no input transformer needed.
I have 89db sensitive speakers and don't listen loud, so I'm OK with driver with a gain of x30. A gain of x40 would be better for most. If the 10Y has a mu of 8 then the step-up required looks like x4 as Cagomat is pointing out. With low sensitivity speakers it might be more. A step-up of x4 does indeed require a low impedance input.
I spent quite a lot of last year trying to use a DHT driver for a 300b amp, and gave up in the end. It's not that it can't be done, it's just that you can get good sounding results with indirectly heated tubes. As a DHT addict that was a big step for me. I didn't believe indirectly heated tubes could be that close to a DHT, but in fact they can. Not perfectly comparable, but pretty close and no input transformer needed.
Andy, totally agree, a amplifier is always a composition and all parts must work together
A friend use this driver tube for a PX4
Its the russian 6E5P, a lot of gain very low distortion, a perfect linearity and a good overload margin and a very low anode resistance perfect for your IT
look at these curves
6Э5П / 6E5P [triode] | VT52.com
These tubes are dirt cheap on Ebay
A friend use this driver tube for a PX4
Its the russian 6E5P, a lot of gain very low distortion, a perfect linearity and a good overload margin and a very low anode resistance perfect for your IT
look at these curves
6Э5П / 6E5P [triode] | VT52.com
These tubes are dirt cheap on Ebay
Banpuku this is a nice explaination of loadline and headroom
VinylSavor: The search for that 'magical' operating point
VinylSavor: The search for that 'magical' operating point
I love Thomas' website and articles. Very easy to understand. Thank you, Cagomat. Thomas is an inspiration, along with Ale. First class.
First of all, I have the DHT 1LH4 coming soon and will give it a go, knowing in advance that the IT is not optimized for its use. This is an experiment for $15, so nothing gain/lost. I should be able to report back on Sunday/Monday.
Putting aside the 1LH4 experiment, as I re-read this entire stream of posts, there is quite a bit of back-n-forth with ideas / questions. So, let's re-examine my goals/desires and pick a solid path forward.
Goals:
1. continue learning about tubes/design. This should be fun! Actually, I regularly check my emails hoping that you guys have responded.
2. Create "lifelike" sonics for chamber music using this amplifier
3. Keep using the 45B, 5K OPT and power supply design.
4. Minimize / eliminate caps in the signal path
Considerations for change:
1. DHT vs. IDHT Driver tube
3. filament DC supply
4. filament chokes
5. DAC source
DHT DRIVER TUBE: If a DHT driver is to be used, then we have a couple of option:
a. EML 30B (mu=30). Downside is that a new IT would be needed. I do not want to consider RC coupling.
b. 10Y driver with a high quality input transformer. Downside, again, is a new IT is needed. Also, would the high quality input transformer degrade the sound?
IDHT DRIVER TUBE: If I continue with IDHT, then I would continue with the C3g in triode mode. I would also consider the C3o in pentode. If the C3g is going to continue to be used, then we examine other methods to tweak the amplifier to get the most out of the C3g and 45B.
DC FILAMENT SUPPLY: I have all the necessary Coleman Raw DC and Filament regulators, including the power transformers. BUT, I don't have room in my chassis for the 10Vac power transformer. So, as of now, this is a non-starter.
FILAMENT CHOKES: Waste of time / money??? I have no idea.
DAC: Maybe I should get a different DAC with lower impedance to better align with the passive volume control / power amp input (FWIW, my DAC is nothing special, so I don't ming spending $s to upgrade and get a lower impedance DAC).
First of all, I have the DHT 1LH4 coming soon and will give it a go, knowing in advance that the IT is not optimized for its use. This is an experiment for $15, so nothing gain/lost. I should be able to report back on Sunday/Monday.
Putting aside the 1LH4 experiment, as I re-read this entire stream of posts, there is quite a bit of back-n-forth with ideas / questions. So, let's re-examine my goals/desires and pick a solid path forward.
Goals:
1. continue learning about tubes/design. This should be fun! Actually, I regularly check my emails hoping that you guys have responded.
2. Create "lifelike" sonics for chamber music using this amplifier
3. Keep using the 45B, 5K OPT and power supply design.
4. Minimize / eliminate caps in the signal path
Considerations for change:
1. DHT vs. IDHT Driver tube
3. filament DC supply
4. filament chokes
5. DAC source
DHT DRIVER TUBE: If a DHT driver is to be used, then we have a couple of option:
a. EML 30B (mu=30). Downside is that a new IT would be needed. I do not want to consider RC coupling.
b. 10Y driver with a high quality input transformer. Downside, again, is a new IT is needed. Also, would the high quality input transformer degrade the sound?
IDHT DRIVER TUBE: If I continue with IDHT, then I would continue with the C3g in triode mode. I would also consider the C3o in pentode. If the C3g is going to continue to be used, then we examine other methods to tweak the amplifier to get the most out of the C3g and 45B.
DC FILAMENT SUPPLY: I have all the necessary Coleman Raw DC and Filament regulators, including the power transformers. BUT, I don't have room in my chassis for the 10Vac power transformer. So, as of now, this is a non-starter.
FILAMENT CHOKES: Waste of time / money??? I have no idea.
DAC: Maybe I should get a different DAC with lower impedance to better align with the passive volume control / power amp input (FWIW, my DAC is nothing special, so I don't ming spending $s to upgrade and get a lower impedance DAC).
Price performance wise C3o pentode or the russian one in triode. You win nothing with the dht in your given equipment
You're welcome to read my own 2020 experiments in trying to use a DHT driver with a 300b output tube. It's in this fairly long thread. It goes over all the issues you describe with plenty of input from wiser guys than me:
300b SE with a DHT driver
This covers input transformers, op-amp stages, DHTs versus IDHTs, and plenty more. Towards the end I get fed up and start using an alternative "inverse DHT" amp with a 10Y in the input and a EL38 in the output. This actually worked very well. But in the end I went back to the 300b with an indirectly heated triode input stage with a 150H amorphous plate choke and Russian FT-2 teflon coupling caps. No cathode bypass cap.
I tried most of the common topologies, including an active load and Cree diodes in the cathode, both of which I rejected in the end. I had a go at a pentode stage, but it was far from optimal and I didn't like what I heard. If I'd have done it better no doubt the sound would have surely improved. I'm now very happy with the results. Still experimenting with input tubes, but very happy nevertheless.
I feed the amp directly from a generic AK4490 DAC, which is connected by USB to my Mac Pro which holds all my digital media. All very simple in the end. No wacky circuits and all very cheap to build apart from the big NP Acoustics amorphous plate chokes, the OPTs and the 300bs. In fact the Lundahl LL1682/50mA OPTs I'm using are £99 each in the UK and the EH 300bs are around £150 a pair. So it's all standard parts and economical to put together. I took a fair bit of care over the filament supplies to the 300b and used choke input to Rod's regs. I also used a large and heavy PSU choke and a pair of AZ1 mesh rectifiers which are lovely, and they are for me a worthwhile investment. I haven't heard better.
.
300b SE with a DHT driver
This covers input transformers, op-amp stages, DHTs versus IDHTs, and plenty more. Towards the end I get fed up and start using an alternative "inverse DHT" amp with a 10Y in the input and a EL38 in the output. This actually worked very well. But in the end I went back to the 300b with an indirectly heated triode input stage with a 150H amorphous plate choke and Russian FT-2 teflon coupling caps. No cathode bypass cap.
I tried most of the common topologies, including an active load and Cree diodes in the cathode, both of which I rejected in the end. I had a go at a pentode stage, but it was far from optimal and I didn't like what I heard. If I'd have done it better no doubt the sound would have surely improved. I'm now very happy with the results. Still experimenting with input tubes, but very happy nevertheless.
I feed the amp directly from a generic AK4490 DAC, which is connected by USB to my Mac Pro which holds all my digital media. All very simple in the end. No wacky circuits and all very cheap to build apart from the big NP Acoustics amorphous plate chokes, the OPTs and the 300bs. In fact the Lundahl LL1682/50mA OPTs I'm using are £99 each in the UK and the EH 300bs are around £150 a pair. So it's all standard parts and economical to put together. I took a fair bit of care over the filament supplies to the 300b and used choke input to Rod's regs. I also used a large and heavy PSU choke and a pair of AZ1 mesh rectifiers which are lovely, and they are for me a worthwhile investment. I haven't heard better.
.
Last edited: