Mosfet driver

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have a 5E3 clone that I built and use it for band practice and some gigging depending on the music we are playing. I usually have the amp's volume maxed and I use my guitar volume as a distortion control. The tinkerer in me got the idea to drive the output tubes with some mosfets seeing how I drive them so hard most of the time. I am hoping somebody has tried this (most likely tubelab) and is there a noticeable difference in distortion characteristics (audibly)? I am hoping it tightens up the distortion a bit. I have reduced the size of the coupling caps which helps with the flabby low end. I have a schematic worked up but I don't have much room in the chassis so I would like to make a small PCB to do this. I am new to making PCB's so I will post one when I am done along with the schematic so people can double check my work for me (if they are so inclined of course).

Another question, don't mosfets have a large input capacitance and will this be a problem because the 5E3 PI is only drawing 800uA?
 
Yes MOSFETs have large input capacitance but that is irrelevant since adding MOSFET Source followers to a cathodyne phase splitter is a waste of time. The cathodyne already has incredibly low output ompedance (Zout approx. = 1/gm)
If you want to improve the overdrive characteristics of a cathodyne phase splitter amp, like the 5E3, then read this:
The Valve Wizard
The "trick" is to run large gridstop resistors on the output tubes, try 47K.
Cheers,
Ian
 
Juan,
OK lets have a go at that OLD arguement - here is something "I prepared earlier". I used "u" for mu in the algebra below..

The cathodyne has equal output impedances at the anode and cathode SO LONG as the loads on the anode and cathode are equal. When overdriving the amp then the loads will become un-equal when the output tubes start to draw grid current (positive peaks of the grid drive).

Maths (skip to the summary below if maths "gets on your wick".
For equal loads on Anode and cathode

Zout = RL.ra/RL(u+2)+ra

The ra term on the bottom line is insignificant compared to RL(u+2) term so drop it. Then the RL terms top and bottom lines cancel leaving

Zout approx = ra/u+2

At typical values of u (>=20) u+2 approx = u, so simplify again

Zout approx = ra/u = 1/gm

If driving output stage directly The equal loads on Anode and Cathode will NOT be guaranteed if:
Output Stage strays into grid current.


If the Anode load drops significantly then:

Zout cathode = RL+ra/(u+2) x ra/RL, the ra/RL term is insignigicant so

Zout cathode approx = RL+ra/u+2, at usual values of u

Zout cathode approx = RL/u + ra/u = RL/u + 1/gm

That is it increases by RL/u

If the cathode load drops significantly then:

Zout anode = RLxRL(u+1)+RL.ra / RL(u+2)+ra

RL squared (u+1) is much larger than RL.ra and RL(u+2) is much larger than ra so

Zout anode approx = RLxRL(u+1)/RL(u+2)

and at reasonable values of u

Zout anode approx = RL

Summary:
As the loads on Anode and cathode become unbalanced then

Zout anode increases from 1/gm toward RL
Zout cathode increase from 1/gm by maximum factor of RL/u

Thats why the HIFI guys suggest that low u is better when driving other than Class A Output Stage.

AND

That is why folk like the Wiz recommend (for guitar amps) increasing the grid stops on the output tubes to a value like 47K. This minimises the load change when output tubes swing into grid current. You can "tune" the amount of distortion in the power amp in overdrive by adjusting those grid stop resistors. I've seen up to 100K used. I use 47K routinely on 6V6 and 6DQ6A amps with cathodyne splitters.

Cheers,
Ian
 
Last edited:
Dear gingertube :)
The cathodyne has equal output impedances at the anode and cathode SO LONG as the loads on the anode and cathode are equal.
I never said that but quite the contrary :)
And as far as Math is concerned, it's me the one who usually tries to introduce it in DIY threads ;)

So don't try to convince me of what I'm already convinced :)

By the way, my field is Guitar Amps (well over 10000 built by 2004/2005 where I lost count) so overdriven amps are my bread and butter.

And I particularly dislike cathodynes for their unsymmetry when heavily overdriven :(

Take care :)
 
Dear gingertube :)

By the way, my field is Guitar Amps (well over 10000 built by 2004/2005 where I lost count) so overdriven amps are my bread and butter.

And I particularly dislike cathodynes for their unsymmetry when heavily overdriven :(

Take care :)

Which is why you are valued so highly here (by me and others) and why I always take particular note of what you have to say.

I agree that the cathodyne is not the splitter of choice for heavily overdriven amps (the differential splitter is a much better option for them) and why I sometimes wonder why so many folk want to build 5E3 clones.

If you do have an existing amp with a cathodyne splitter (which was the case for the 6V6 and 6DQ6A Amps I mentioned), or feel that you really do want to build something like the 5E3 then the Wiz's recommended mods will go a long way toward adressing the cathodynes weakness in overdriven amps.

The algebra I posted just shows why large grid stops on the output tubes does work to improve to situation.

To be clear to other readers - if you are building an amp that you want to give you the best overdriven sounds then Juan and I are both saying DO NOT use the cathodyne phase splitter.

If you already have an amp with a cathodyne splitter and you want to improve its overdrive performance than do the Wiz's mods.

The original analysis comes from Priesman
http://www.aikenamps.com/images/Documents/cathodyne.pdf

More recently this analysis has been documented by Morgan Jones (Valve Amplifiers) and simulation results from SY (see the Red Light District Amp on the HiFi Forum here) have confirmed that the Anode and Cathode source impedances are equal when the loads are equal and also confirmed that using high value grid stops on the output tubes to minimize the amount that the loads become UN equal (when in overdrive) helps.

Best Regards,
Ian
 
I am going to sit down and run some simulations and also do the maths of course but I am still unsure of what you guys are saying. I agree that the the cathodyne distorts badly and why I wish to fix it. I do have a large grid resistor feeding the PI as the wiz suggests, I have not tried 100k on the 6V6 grids yet. The problem in my opinion is blocking distortion which even thought the output impedance of the cathodyne is low it doesn't fix the problem of the coupling capacitor.

Like I said I haven't done the maths yet but I don't think I am going to get into class AB2 with just a cathodyne PI, or even fix the overload issues with just basic resistors at the grids of the 6V6's. Another part of the issue of imbalanced loading of the PI is that I plan on maybe trying class A2 by using a LM317 to auto bias the power tubes, from what I am thinking this will make matter worse for the cathodyne PI because the the tube with the distortion on grid current onset will not be in cutoff like in AB operation, this also is leading me to try to drive the 6V6's with mosfets.

I highly appreciate both of your inputs so thank you very much. I am also a hobbiest and am still learning.

I would really like to hear what you have to say JMFahey because like you said over driven amps are your "bread and butter" :) I will take any advice you give me in my quest for low power overdriven guitar nirvana. I do not wish to build a whole different amp for I have been out of work so I would like to simply modify my existing build. If I had the choice I would have one amp of each that does just one thing well but I can't. So I am thinking of changing my 5E3 into a class A2 power house, or maybe just class AB2 but I really don't need too much power. The 5E3 is bias so hot towards class A I don't think changing to CCS for bias will be any worse in eating away at my stash of grey glass RCA VT-107-A's.

Thanks
-bird
 
Last edited:
Hi Bird,
I am going to reverse what I said originally. MOSFET Source Followers on the Cathodyne Phase Splitter will help because it puts a Class A operating load onto the Cathodyne so that it is never subjected to the very low load impedance of the an output tube with grid current and therefore the Anode/Cathode loads remain balanced.

Here is a 200Watt AM Transmitter Modulator circuit for reference - note that the Cathodyne phase splitter has cathode followers on it. Note the output tube grids are direct coupled from the followers and the bias is applied at the follower grids. You could do this with MOSFET Source followers. Ignore the 6AU6 and 6AQ5 which are just a screen supply regulator.

Regards,
Ian
 

Attachments

  • Vike_Val_Mod.gif
    Vike_Val_Mod.gif
    29.9 KB · Views: 226
Last edited:
Hi Ian,
Thanks for the reference, it confirms the way I was going to implement the source follower. I will post a schematic once I figure this out some more, for I can't seem to get it to work correctly in a simulation:confused:

I omitted the cathode bias scheme for the 6V6's and tied the cathodes to ground through 1R resistors. For mosfets I used IRF630's and tied the drain to 200v source, the source to a -200v source through 10k resistors, and the gate was tied to a -100 through a voltage divider for the bias. No matter what I change the -bias to I get 200mA through each 6V6:headbash: The IRF630's are AC coupled to the cathodyne PI. What am I missing?


-bird
 

Attachments

  • sourcefollower.gif
    sourcefollower.gif
    6.8 KB · Views: 195
bird,
I don't see anything wrong with the schematic sketch so would suggest that there is something wrong with your modelling schematic capture or similar.
I do exactly this in my HiFi Amp except "spiffed up" with current source loads (instead of resistors) on the source follower. Here is that for your possible intertest. The 1.7mA source follower idle current was sunbsequently increased to 3mA but 1.7mA would be plenty for a guitar amp.
Regards,
Ian
 

Attachments

  • BH_Fixed_bias_B.jpg
    BH_Fixed_bias_B.jpg
    92.9 KB · Views: 220
Here is what I have so far. I am thinking of using a DPDT switch, one side to put the 6V6 cathodes to the LM317 or straight to ground. The other side will switch between a positive bias supply to use with the LM317 CCS and a negative supply when the cathodes are at ground. This should give me a lower powered class A2 mode and a higher powered class AB2 mode. The mosfets are running at 2mA and the LM317 is setup up for 80mA putting the cathodes of the 6V6's at 20v.
 

Attachments

  • 5e3.png
    5e3.png
    143.3 KB · Views: 191
Last edited:
If you want to go for AB2 then don't forget a voltage/current limitation function at the control grid of the power pentodes.
Otherwise you risk to blow up the control grid in case of overdrive because the Mosfets can deliver high current.

Do you happen to have any schematics with this "voltage/current limitation" implementation? And would I have to worry about the grids melting if I just go with A2 operation?

Of the various resources online that I have seen that uses cathode followers or mosfets to drive power tubes I haven't seen such circuitry. I do plan on using Pete Millets board for this project which has protection diodes to prevent the mosfets from failing due to reverse breakdown voltage.


-bird
 
Aa an example you will find the power stage of my 400W bassamp. The limitation is done with R9a/ZD3 ans R10a/ZD4. The max positive voltage at the source of the IRF 840 not more than ~25V. The grid stoppers R19/27/35/20/22/36 are preventing oscillations and were used to measure the grid current at the GU50 during development. This tube can handle 1W dissipation at the control grid according to specification.

I don't know that value for a 6V6, my personal guess is less than 0,3 W. The value of the Zenerdiode needs to be adjusted as well as the gridstopper value.
To start from the safe side I would first go with a grid stopper of 1K and measure the gridcurrent with a differential probe with increasing drive to calculate the grid dissipation.

Hans- Georg
 

Attachments

  • Bass 400 Power Amp (AB2 mode with control grid current limitation) .png
    Bass 400 Power Amp (AB2 mode with control grid current limitation) .png
    37.3 KB · Views: 200
RG Keene has done an article in his ''fun with mosfets'' series. It describes that circuit , some theory and what to look out for in implementation. As you mention the point is to avoid grid conduction, bias shifting, which results in harsh distortion in the output tubes operation such as class B crossover distortion and grid current conduction distortion. I am inclined to agree that you will greatly improve the overdrive characteristics by getting plate saturation, while avoiding the other two situations that occur when driving that hard plate saturation.
Good luck and and have fun
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.