Hi Thomas,
With 3500 drive impedance it was rolling of just below 50Hz.
4.5:1 for DAC analog stage which has 130mVAC signal at full output. That stepdown is pretty severe for this, voltage gain.. I could get away with gain of 70 or so. A lot, I know. How did you find them to perform in 4:4.5 ?
Regards,
Shane
With 3500 drive impedance it was rolling of just below 50Hz.
4.5:1 for DAC analog stage which has 130mVAC signal at full output. That stepdown is pretty severe for this, voltage gain.. I could get away with gain of 70 or so. A lot, I know. How did you find them to perform in 4:4.5 ?
Regards,
Shane
Hi!
I thought you used a 71A which would have a plate resistance of less than 2kOhms. How did you get 3500Ohms drive impedance? Was the cathode resistor of the 71A properly bypassed or was it fixed bias? That transformer was definetily not a godd fit but maybe something else in the circuit was wrong too?
You could use a D3a with this transformer. It works as 4:4,5 but as mentiond give it in step down a try to explore it's full capability.
maybe try 41PL with it in a linestage?
Thoms
I thought you used a 71A which would have a plate resistance of less than 2kOhms. How did you get 3500Ohms drive impedance? Was the cathode resistor of the 71A properly bypassed or was it fixed bias? That transformer was definetily not a godd fit but maybe something else in the circuit was wrong too?
You could use a D3a with this transformer. It works as 4:4,5 but as mentiond give it in step down a try to explore it's full capability.
maybe try 41PL with it in a linestage?
Thoms
Hi Thomas,
71A, unbypassed 150R filament bias resistor. CCS fed shunt reg B supply using two VR tubes in series bypassed with RC to keep impedance of the VR string at 1k5 for passband. Does the impedance of the VR tubes contribute to the drive impedance?
I'll try the d3a or something similar, see how it goes. Otherwise I can try the 41PL linestage as you suggest.
Regards,
Shane
71A, unbypassed 150R filament bias resistor. CCS fed shunt reg B supply using two VR tubes in series bypassed with RC to keep impedance of the VR string at 1k5 for passband. Does the impedance of the VR tubes contribute to the drive impedance?
I'll try the d3a or something similar, see how it goes. Otherwise I can try the 41PL linestage as you suggest.
Regards,
Shane
Hi Shane,
I don't quite get this. Was there anything in series with the signal to artificially increase the drive impedance?
That's something you do not want to do as it can have all kinds of unwanted side effects. The drive impedance numbers given in the datasheet are no hard numbers which have to be kept. They are a compromise between bandwidth and ringing.
Best rgards
Thomas
I don't quite get this. Was there anything in series with the signal to artificially increase the drive impedance?
That's something you do not want to do as it can have all kinds of unwanted side effects. The drive impedance numbers given in the datasheet are no hard numbers which have to be kept. They are a compromise between bandwidth and ringing.
Best rgards
Thomas
Hi Thomas,
In series, well depends on how you view it.. ground - grid - plate - xfmr primary - VR tubes - ground would be the AC loop.
A friend of mine who knows more than I do (EE) seemed quite sure that the AC impedance of the VR tubes added to the drive impedance as seen by the transformer primary. I have no other basis for this, and it was based on this that an RC compesation branch was put in parallel with the VR string to keep the drive impedance reasonably flat at 3500 ohms, the figure that the datasheet suggested the 1660 wanted to see.
On the d3A thing.. your two stage DHT for phono has me thinking, type 40 DC to 112A 4.5:1.. total voltage gain of about 50, with an output impedance of 300 ohms. 250mA filaments.. hmm.
Regards,
Shane
In series, well depends on how you view it.. ground - grid - plate - xfmr primary - VR tubes - ground would be the AC loop.
A friend of mine who knows more than I do (EE) seemed quite sure that the AC impedance of the VR tubes added to the drive impedance as seen by the transformer primary. I have no other basis for this, and it was based on this that an RC compesation branch was put in parallel with the VR string to keep the drive impedance reasonably flat at 3500 ohms, the figure that the datasheet suggested the 1660 wanted to see.
On the d3A thing.. your two stage DHT for phono has me thinking, type 40 DC to 112A 4.5:1.. total voltage gain of about 50, with an output impedance of 300 ohms. 250mA filaments.. hmm.
Regards,
Shane
Last edited:
Hi Shane,
those VR tubes should provide a fairly low AC impedance to ground and that is what you want on the B+ side of the primary. I keep that side simple with a cap only. You do not want to add any impedance there.
That 3.5k is a guideline. Artificially rasing the drive impedance is not what you would want to do.
The 40 has a very high plate resistance which cannot really drive anything except maybe a cathode follower. Even DC coupled to a 112 you will not get a good bandwidth. I experimented with the 40 in the past. The high plate resistance is frustrating. The 841 is a much better alternative
Thomas
those VR tubes should provide a fairly low AC impedance to ground and that is what you want on the B+ side of the primary. I keep that side simple with a cap only. You do not want to add any impedance there.
That 3.5k is a guideline. Artificially rasing the drive impedance is not what you would want to do.
The 40 has a very high plate resistance which cannot really drive anything except maybe a cathode follower. Even DC coupled to a 112 you will not get a good bandwidth. I experimented with the 40 in the past. The high plate resistance is frustrating. The 841 is a much better alternative
Thomas
Hi Thomas,
To correct an oversight on how I presented the function of the RC bypass to the VR tube string. Of course, the impedance is in parallel with VR tubes and the overall impedance will of course always be lower than that of the VR tubes themselves.
The idea is that the RC bypass will limit the overall impedance of the drive stage to 3500 ohms at some high frequency and beyond, where the impedance of the VR tubes can be quite high. Of course if the impedance of the VR tubes doesnt infact get that high, the overall impedance will still be less than the impedance of the VR tubes - so there was no need to be measuring and selecting VR tubes to suit the purpose of the circuit, the max drive impedance would always be below the max. specified by the transformer datasheet.
At anypoint, below lets say 200Hz.. the drive impedance would be almost exclusively dominated by the driver tube, so 1800-2000 ohms would be a good approximate and the BW limitations of the transformer as previously stated still apply - as they should even with 3500 ohms drive at 25Hz - which brings me back to the question; At what frequency and drive level do they measure primary inductance on this model?.
All of which is to no real end as the transformers have been replaced, and as noted there are better applications for this particular model. I hope this clears up any issues for anyone following this thread, or for anyone is trying to glean something from it in the future.
841 would be ideal, but availability and pricing concerns are real to me. With type 40, I was thinking to load it with a CCS and driving the 12A from from the 'mu' output for some several hundred ohms output impedance. I've since noticed the tube sets up with Ip around 200uA!. I'm not sure if I can construct a CCS cct that is stable at such a low current. Still, I can see a nice pre 1930's pair NIB online for less than USD40.00 and it might be fun to try.
Regards,
Shane
To correct an oversight on how I presented the function of the RC bypass to the VR tube string. Of course, the impedance is in parallel with VR tubes and the overall impedance will of course always be lower than that of the VR tubes themselves.
The idea is that the RC bypass will limit the overall impedance of the drive stage to 3500 ohms at some high frequency and beyond, where the impedance of the VR tubes can be quite high. Of course if the impedance of the VR tubes doesnt infact get that high, the overall impedance will still be less than the impedance of the VR tubes - so there was no need to be measuring and selecting VR tubes to suit the purpose of the circuit, the max drive impedance would always be below the max. specified by the transformer datasheet.
At anypoint, below lets say 200Hz.. the drive impedance would be almost exclusively dominated by the driver tube, so 1800-2000 ohms would be a good approximate and the BW limitations of the transformer as previously stated still apply - as they should even with 3500 ohms drive at 25Hz - which brings me back to the question; At what frequency and drive level do they measure primary inductance on this model?.
All of which is to no real end as the transformers have been replaced, and as noted there are better applications for this particular model. I hope this clears up any issues for anyone following this thread, or for anyone is trying to glean something from it in the future.
841 would be ideal, but availability and pricing concerns are real to me. With type 40, I was thinking to load it with a CCS and driving the 12A from from the 'mu' output for some several hundred ohms output impedance. I've since noticed the tube sets up with Ip around 200uA!. I'm not sure if I can construct a CCS cct that is stable at such a low current. Still, I can see a nice pre 1930's pair NIB online for less than USD40.00 and it might be fun to try.
Regards,
Shane
Hi Shane,
this explanation with the VR tubes limiting the impedance does not really make sense. You also mentioned an unbypassed cathode resistor, which would also increase the drive impedance.
As stated before: the 3,5k recommendation is valid for the 10mA model. For yours this would be more like 2,5k maybe even less. Still should work reasonable with the 71A, that's why I assume something was not quite right with the circuit.
Anyways, you can easily test this: connect the primary of the transformer through for example a 2k resistor to the signal generator, then measure the frequency response. This will show the behaviour for 2k drive impedance
Best regards
Thomas
this explanation with the VR tubes limiting the impedance does not really make sense. You also mentioned an unbypassed cathode resistor, which would also increase the drive impedance.
As stated before: the 3,5k recommendation is valid for the 10mA model. For yours this would be more like 2,5k maybe even less. Still should work reasonable with the 71A, that's why I assume something was not quite right with the circuit.
Anyways, you can easily test this: connect the primary of the transformer through for example a 2k resistor to the signal generator, then measure the frequency response. This will show the behaviour for 2k drive impedance
Best regards
Thomas
Hi Thomas,
The RC bypass limits the impedance of the VR tubes at higher frequencies and its the RC bypass in parallel with the VR tube impedance which adds to the overall drive impedance. ie. Drive impedance should be Rp + (z. VR//z. RC) = ~1900 ohms.
In other words its the RC bypass which limits the impedance of the VR tubes - and its the VR tubes that add to the drive impedance.
The unbypassed cathode resistor of 150 ohms; [Rp + (mu-1)*Rk] = 1900 + 2(150) has the total drive impedance at LF at approx. ~2200 ohms, well within spec.
Regards,
Shane
The RC bypass limits the impedance of the VR tubes at higher frequencies and its the RC bypass in parallel with the VR tube impedance which adds to the overall drive impedance. ie. Drive impedance should be Rp + (z. VR//z. RC) = ~1900 ohms.
In other words its the RC bypass which limits the impedance of the VR tubes - and its the VR tubes that add to the drive impedance.
The unbypassed cathode resistor of 150 ohms; [Rp + (mu-1)*Rk] = 1900 + 2(150) has the total drive impedance at LF at approx. ~2200 ohms, well within spec.
Regards,
Shane
Hi!
That added impedance of your VR tubes and RC network should not have any apreciable effect in the audio band and even way beyond that.
Simplest thing would be to run a quick measurement as suggested above. This will tell you the inductance of the transformers. Note that the rp of tubes can vary considerably from tube to tube and as it ages. A controlled test set up as above is the only way to get any solid data
Thomas
That added impedance of your VR tubes and RC network should not have any apreciable effect in the audio band and even way beyond that.
Simplest thing would be to run a quick measurement as suggested above. This will tell you the inductance of the transformers. Note that the rp of tubes can vary considerably from tube to tube and as it ages. A controlled test set up as above is the only way to get any solid data
Thomas
- Status
- Not open for further replies.