Hello DIYers,
I am attempting to design an amplifier with a C3m wired as a pentode direct coupled to an F2a wired as a triode using the "Monkey" topology. Aware there are some non-standard resistor values in there. That aside, there are a few things I am struggling with:
I am having trouble figuring out how to nail down Vg1 for the C3m. Right now, I have a battery, but I am not sure it would work.
150v for Vg2 is something of a placeholder - plan to use an 0D3 in the power supply to get this. Any reason to wire the 0D3 into the amp instead? Perhaps use a voltage divider to also supply the 5V required for Vg1?
This is my first crack at this sort of thing so I would not exactly be surprised if there are errors or omissions, so anything you see that strikes you as odd, please let it be known.
Thanks!
I am attempting to design an amplifier with a C3m wired as a pentode direct coupled to an F2a wired as a triode using the "Monkey" topology. Aware there are some non-standard resistor values in there. That aside, there are a few things I am struggling with:
I am having trouble figuring out how to nail down Vg1 for the C3m. Right now, I have a battery, but I am not sure it would work.
150v for Vg2 is something of a placeholder - plan to use an 0D3 in the power supply to get this. Any reason to wire the 0D3 into the amp instead? Perhaps use a voltage divider to also supply the 5V required for Vg1?
This is my first crack at this sort of thing so I would not exactly be surprised if there are errors or omissions, so anything you see that strikes you as odd, please let it be known.
Thanks!
An externally hosted image should be here but it was not working when we last tested it.
Using a C3M in pentode connection with an inductive load may cause problems with LF response due to the rather high plate resistance of pentode connection. (You want low rp with choke loading) I've not seen monkey configuration used with anything but triodes I suspect for this reason.
You can use LEDs for biasing the C3M or conventional cathode bias with a bypass cap. (Again you don't want to do anything to unnecessarily raise the rp of the C3M.) You can also use negative bias at the grid, (input cap required) and if you make it adjustable you can then easily set the operating point of the output stage as well.
You can use LEDs for biasing the C3M or conventional cathode bias with a bypass cap. (Again you don't want to do anything to unnecessarily raise the rp of the C3M.) You can also use negative bias at the grid, (input cap required) and if you make it adjustable you can then easily set the operating point of the output stage as well.
Last edited:
G3 of the F2a is internally connected to the cathode; you can't connect it to the anode. Which is good, 'cause you'd fry the tube.
The datasheet here suggests 16mA through 1 250Ohm cathode resistor, which means a Vg of about -4V.
I'm not familiar with the 'Monkey topology'; does it have anything to do with drawing 16mA through the grid of the F2a? I have zero experience with this tube, so perhaps it's reasonable, but it strikes me as odd, especially since in your diagram, Vg1 of the F2a is effectively -21V (compared to the cathode of the same tube),and with those parameters, the tube shouldn't draw 71mA of anode current. So either I have no understanding of this circuit at all (which I don't, I admit), or there's something not completely in order with the values.
The datasheet here suggests 16mA through 1 250Ohm cathode resistor, which means a Vg of about -4V.
I'm not familiar with the 'Monkey topology'; does it have anything to do with drawing 16mA through the grid of the F2a? I have zero experience with this tube, so perhaps it's reasonable, but it strikes me as odd, especially since in your diagram, Vg1 of the F2a is effectively -21V (compared to the cathode of the same tube),and with those parameters, the tube shouldn't draw 71mA of anode current. So either I have no understanding of this circuit at all (which I don't, I admit), or there's something not completely in order with the values.
The current isn't drawn from the grid, but from the cathode through the dcr of the choke and an additional series resistor. The plate current of the C3M through the series resistance of the choke and resistor is what provides the actual negative bias for the output tube. Can't talk to the value of effective bias on the F2a but the topology does work quite well with other tubes. The driver does IMHO have to be triode connected in order to work properly with the choke reactance if any sort of LF response is desired.
Right, that makes sense, and I sort of get it. But in that case, the C3m should see a plate voltage of about 234V (241V of the F2a cathode minus the 16mA through the choke and series resistor, assuming the ). That means the grid of the F2a sees the same voltage, so the F2a is biased at -6.8V, and the tube should draw about 100mA. That in turn means that the cathode of the F2a won't sit at 241V, but at about 375V. This topples the whole bias again, exceeds the limiting values of the C3m... Can't really work this one out from the top of my head, guess I'm too stoopid for DC coupled circuits 
Apart from the above, what is the advantage of this concept? The C3m effectively acts as a choke loaded stage, but what is the advantage of feeding it from the cathode of the F2a? Sort of a push pull effect? Could anyone link me to a topic/site where this is covered?
Apart from the above, what is the advantage of this concept? The C3m effectively acts as a choke loaded stage, but what is the advantage of feeding it from the cathode of the F2a? Sort of a push pull effect? Could anyone link me to a topic/site where this is covered?
Look up DRDin this forum. I went on a journey in understanding this topology, named Loftin White, Monkey, Free Lunch, DRD, Ultrapath, etc. I think there is a resistor missing between the cathode of the F2a and the series resistor and choke feeding the C3m. The vendor of this design is that it is a safer way to do direct coupling. The way I understand it is that there are two resistors at the cathode of the F2a, the top one sets the bias for the F2a and the bottom sets the plate voltage of the C3m. Running this at 670V+ means more caps (in series) in the PSU and everything is go to be rated at 1000V plus I reckon.
On another note, I am thinking of building a C3g-F2a amp, but using a 2:1 IT instead of this because of the higher voltage. I do have all the parts ready to build a 45DRD, so can comment later when I build it.
On another note, I am thinking of building a C3g-F2a amp, but using a 2:1 IT instead of this because of the higher voltage. I do have all the parts ready to build a 45DRD, so can comment later when I build it.
I was (obviously) not aware inductive loading presented such a problem for pentodes - appreciate the heads up. Will look at subbing in a 6c45pi or a triode-strapped D3a in place of the C3m; new schematic will follow after I run the numbers.
eeyore - out of curiosity, are you planning to run the C3g as a triode or a pentode? Looking forward to seeing your design. I actually ran the loadlines based on F2a curves I entered into Steve Bench's spreadsheet (available here: http://greygum.net/sbench/sbench101/). I just picked a number based on what maximized power output, not what would give the best sonic performance (which is what I am ultimately interested in).
eeyore - out of curiosity, are you planning to run the C3g as a triode or a pentode? Looking forward to seeing your design. I actually ran the loadlines based on F2a curves I entered into Steve Bench's spreadsheet (available here: http://greygum.net/sbench/sbench101/). I just picked a number based on what maximized power output, not what would give the best sonic performance (which is what I am ultimately interested in).
Last edited:
Yeah, I have only seen Triodes in this topology. Will be using the C3g in triode mode only. D3a might have too much gain at a mu of around 70, but that could mean you can remove a previous stage. The F2a doesn't need heaps of voltage to drive, around +/- 20~30V would be enough. You can even try some of the industrial dual triode, 5687 family, 6n6p, 6h30, 5842, etc. Lots of options.
When I get around to it, I have been sitting on this project for 2 years due to kids... Anyway, from what I understand, the higher the load, the lower the distortion as it approaches an infinite load. Similar reason, I suspect to why using CCS loads and inductive loads are better, the loadline is in the more linear region of the curves. But the lower the load, the easier and better it is to build a transformer. So I guess there is a balance between the two. Best way is to download the F2a Siemens datasheet and draw your own loadline. From memory, I had mine at 400V, -25V on the grid and 5k load. I think that gave around 8W power before positive grid current. Be careful also not to exceed the maximum plate and screen voltage for the F2a, I believe it is around 425V.
Have a good read around the net, as there are heaps of information on this topology and design. Also be careful of the higher B+ voltages. Good luck!
When I get around to it, I have been sitting on this project for 2 years due to kids... Anyway, from what I understand, the higher the load, the lower the distortion as it approaches an infinite load. Similar reason, I suspect to why using CCS loads and inductive loads are better, the loadline is in the more linear region of the curves. But the lower the load, the easier and better it is to build a transformer. So I guess there is a balance between the two. Best way is to download the F2a Siemens datasheet and draw your own loadline. From memory, I had mine at 400V, -25V on the grid and 5k load. I think that gave around 8W power before positive grid current. Be careful also not to exceed the maximum plate and screen voltage for the F2a, I believe it is around 425V.
Have a good read around the net, as there are heaps of information on this topology and design. Also be careful of the higher B+ voltages. Good luck!
It's very easy to work out form the plate curves. Mark your operating point and draw your loadline, then calculate change in operating point voltage over change in Vg1. This will work out to approx 15 -17, depending on how you operate the tube..
In triode mode, Siemens recommend an operating point of 330V, -13.7V, 98mA (combined anode + g2 current), with a transformer impedance of 1K5 and cathode resistor value 140R. this gives 5.5W @ 10% THD.
Last edited:
I'm not sure that is necessarily the case with pentodes, as it is with triodes. For example, a 6L6 is loaded nicely with 2.5KΩ despite it's high plate resistance. The plate resistance/load ratio may be a problem with output transformers when considering damping factor, but suitable inductive loads for pentodes should be considered and tested individually rather than making a general statement like the one which is true in the case of triodes.
John
Here is the first revision of the amplifier. Sticking with the C3m, but triode strapped. Also added a cap across the resistors to improve noise rejection from the power supply, as per Thorsten's recommendation on direct-coupled circuits.
An externally hosted image should be here but it was not working when we last tested it.
Hi John,
Kevin is right with his general staement.
One can look upon a pentode like a AC currentsource.
A 6L6 or any pentode works properly with load but not without. You compare a 6L6 loaded with 2,5k(depending on secondary loading and winding ratio) to a C3m, as in this case, loaded with next to nothing.
Just like the 6L6, C3m will need a load of just a few kohm to work properly. Will not be doable here.
So as I said a few posts ago, a pentode without load does not work in any circuit.
Hi,
Some notes.
First, the C3m as Pentode is not really suited to inductive load, unless you parallel a low Value resistor with the C3m. Normal Anode load would be around 12K, which means you will need around 240H inductance, at around 15mA for an 8Hz -3dB Point.
I think you are better off adapting the Pentode Loftin White Circuit with resistive load from +B instead.
Secondly, the best way to derive the screen voltage is of course by tapping it of the Cathode Resistor of the F2a (again, classic Loftin White).
Thirdly, I REALLY, REALLY think you are better off deriving the cathode voltage for the C3m also from the resistor chain on the F2a, I would not put an Electrolytic capacitor into the cathode of this circuit, much less a battery, honestly. There are also other benefits for this, discussed in the part of my DC Coupled Amp series that covered the pentode driven LW. I'd probably also try the F2a as pentode and use some NFB to get a sensible output impedance, probably F2A anode to C3m Cathode...
Ciao T
Some notes.
First, the C3m as Pentode is not really suited to inductive load, unless you parallel a low Value resistor with the C3m. Normal Anode load would be around 12K, which means you will need around 240H inductance, at around 15mA for an 8Hz -3dB Point.
I think you are better off adapting the Pentode Loftin White Circuit with resistive load from +B instead.
Secondly, the best way to derive the screen voltage is of course by tapping it of the Cathode Resistor of the F2a (again, classic Loftin White).
Thirdly, I REALLY, REALLY think you are better off deriving the cathode voltage for the C3m also from the resistor chain on the F2a, I would not put an Electrolytic capacitor into the cathode of this circuit, much less a battery, honestly. There are also other benefits for this, discussed in the part of my DC Coupled Amp series that covered the pentode driven LW. I'd probably also try the F2a as pentode and use some NFB to get a sensible output impedance, probably F2A anode to C3m Cathode...
Ciao T
I have yanked the battery in the new schematic in favor of the resistor chain approach. The 90 uf cap is based on the fourth post in this thread - figure this can be a motor run cap, assuming it belongs where I think it does. Will revisit the thread to see what I am missing and check out the pentode-driven LW.
Thanks!
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Hey EvilM,
If you still want to go monkey, you should lower Ua and Ia of the C3m somewhat. In the ballpark of 150-175V/15mA would be better. Check tube data below:
http://www.scottbecker.net/tube/sheets/128/c/C3m.pdf
I this particular case the load line will be close to horizontal.
If you still want to go monkey, you should lower Ua and Ia of the C3m somewhat. In the ballpark of 150-175V/15mA would be better. Check tube data below:
http://www.scottbecker.net/tube/sheets/128/c/C3m.pdf
I this particular case the load line will be close to horizontal.
nic
my new project is 6aw8 +f2a in shindo circuit ..use nfb from f2a anode to 6aw8....i have expriance in se amp ...but this circuit is different .f2a in pentod is powerfull and musical.... highend sound .
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.