Was all pumped for this build now spooked by the flames. Eek! I will have to shoot you a PM when I'm ready to order. I have the boards already.
Hey Stu, its for the TMC (transitional Miller compensation) feeding a small amount of the output back to the junction of C7/8 and into the VAS. Its optional (see the advanced options text) but gives the best distortion reduction.
I am still tweaking my Badgers but have only just got back into it after concentrating on work for most of the year. I am meant to be retired for goodness sake! Building some solid state relays to provide output and power supply protections and incorporate power on and off mute controls plus DC out detection. Had some boards made by Silver Circuits in Malaysia. Very satisfied.
Cheers
I am still tweaking my Badgers but have only just got back into it after concentrating on work for most of the year. I am meant to be retired for goodness sake! Building some solid state relays to provide output and power supply protections and incorporate power on and off mute controls plus DC out detection. Had some boards made by Silver Circuits in Malaysia. Very satisfied.
Cheers
Hi John,
Thank you so very much for that rapid response I really appreciate you jumping in and answering that so promptly. I will have to look into that a bit further as I must admit that I am not that familiar with the different compensation methods. I understand that they exist and I understand that they compensate for the gain reduction in the VAS transistor Q10 as frequency increases but I have only done LTspice simulations as per Bob Cordell's book where he usually uses only one capacitor in that location and calls it miller compensation. I know in chapter 4 of his book it goes into more detail on other compensation methods. so I will have to look into that again.
I hope that you are enjoying getting back into your badger amp. There is always something to do. I to have had a bit of a break, as I now have a 10 week old daughter so the past few months have been full on.
But I was helping a friend the other day and that part of the circuit was not clear to me.
Thanks again for your input. I will need to look in to TMC a bit further as I'm not sure how it is calculated. I noticed that on the first version of the HB's schematic R24 was 820 ohms and was not sure if it was related to the value of the voltage divider used for the negative feedback R5. But I see that in the latest version of the board its shows it as 1.5K I will need to check my board and see what I actually installed.
Thanks again
Stuart
Thank you so very much for that rapid response I really appreciate you jumping in and answering that so promptly. I will have to look into that a bit further as I must admit that I am not that familiar with the different compensation methods. I understand that they exist and I understand that they compensate for the gain reduction in the VAS transistor Q10 as frequency increases but I have only done LTspice simulations as per Bob Cordell's book where he usually uses only one capacitor in that location and calls it miller compensation. I know in chapter 4 of his book it goes into more detail on other compensation methods. so I will have to look into that again.
I hope that you are enjoying getting back into your badger amp. There is always something to do. I to have had a bit of a break, as I now have a 10 week old daughter so the past few months have been full on.
But I was helping a friend the other day and that part of the circuit was not clear to me.
Thanks again for your input. I will need to look in to TMC a bit further as I'm not sure how it is calculated. I noticed that on the first version of the HB's schematic R24 was 820 ohms and was not sure if it was related to the value of the voltage divider used for the negative feedback R5. But I see that in the latest version of the board its shows it as 1.5K I will need to check my board and see what I actually installed.
Thanks again
Stuart
Congratulations for the new arrival proud Dad - hope Mum is doing well.
The single capacitor version is called CMC, conventional Miller compensation or simply Miller compensation. It provides a frequency dependent local feedback loop around the VAS. The diode, D10 I think, softens the impact of clipping by preventing the collector of Q10 from going too far more positive than the base of Q9 - this prevents oscillation artifacts and latching.
One way to visualise the effect of all this is to run a frequency response out to 500Mhz in LTS to see where the positive feedback phase change happens with changes to the compensation capacitances and arrangements. Need the 180 degree point to be way out where the gain is minimal.
The single capacitor version is called CMC, conventional Miller compensation or simply Miller compensation. It provides a frequency dependent local feedback loop around the VAS. The diode, D10 I think, softens the impact of clipping by preventing the collector of Q10 from going too far more positive than the base of Q9 - this prevents oscillation artifacts and latching.
One way to visualise the effect of all this is to run a frequency response out to 500Mhz in LTS to see where the positive feedback phase change happens with changes to the compensation capacitances and arrangements. Need the 180 degree point to be way out where the gain is minimal.
Thanks John. That really helps point me in the right direction.
Going from what you said.
"TMC (transitional Miller compensation) feeding a small amount of the output back to the junction of C7/8 and into the VAS."
So without looking into it to much I am assuming that if the negative feedback trace is feeding a small amount of current back to C7/8. Then I am assuming that at that point C7/8 the voltage is nearly 180 degrees out of phase to the negative feedback voltage. Which is what I think you were say in your second post and will become clearer once I have time to simulate it.
Which actually makes sence otherwise if they were in phase the junction of C7/8 would be at a higher potential than the negative feedback trace. Which would be really problematic.
I guess this is were the slew rate speed comes into play too.
Stuart
Going from what you said.
"TMC (transitional Miller compensation) feeding a small amount of the output back to the junction of C7/8 and into the VAS."
So without looking into it to much I am assuming that if the negative feedback trace is feeding a small amount of current back to C7/8. Then I am assuming that at that point C7/8 the voltage is nearly 180 degrees out of phase to the negative feedback voltage. Which is what I think you were say in your second post and will become clearer once I have time to simulate it.
Which actually makes sence otherwise if they were in phase the junction of C7/8 would be at a higher potential than the negative feedback trace. Which would be really problematic.
I guess this is were the slew rate speed comes into play too.
Stuart
Not quite, the overall negative feedback will become positive when the phase delay at the output reaches 180 degrees at very high frequencies.
So the compensation increases the roll-off , delays the phase change as well as decreases the rate of change of the phase with increasing frequency. Somewhat of a balancing act.
It annoys me that I no longer have the maths and theory at my finger tips, probably never had full understanding.
So the compensation increases the roll-off , delays the phase change as well as decreases the rate of change of the phase with increasing frequency. Somewhat of a balancing act.
It annoys me that I no longer have the maths and theory at my finger tips, probably never had full understanding.
Hi Harry, I have had 2 PCB's made - first is a supply current monitor for the Badger, second manages the power up and down mute and DC out monitor as well as speaker output current. They are arranged so that an over-current or DC event will trip all the SSRs and latch - requiring a power cycle to reset. Perhaps overkill.??
Have spare boards if you are interested.
The mosfet driver and opto triac models for LTSpice if you wish to play, are included but you will need to create your own local symbols. Yell if you need help
Cheers
This stuff and more are in the LTSpice help files but I suppose it took me a while to get the hang of it first time....
Import the 3 files and save them in your LT directory
Open LT Spice
From LTS open 'JohnsModels.txt' It should now display in readable form
Scroll thru the file and find the line ".SUBCKT irfp4668 1 2 3"
Place the cursor on this line and Right Click your mouse
An LTS window will open and two thirds the way down is the option "Create Symbol" - click on that option
It will ask "Do you really really want to?" Yes you do. That symbol is then created and saved in the area "AutoGenerated" from where you can then add it to the schematic.
Do this again for VOM1271 and MOC3021 and 4N26 and MAT02
Now when you open the LTS sim files the symbols are used and the sim runs.
My model file is based on the file from Bob Cordell plus all the others I have had to download from manufacturers and others like the LT Spice user group, so in particular - many thanks Bob.
Import the 3 files and save them in your LT directory
Open LT Spice
From LTS open 'JohnsModels.txt' It should now display in readable form
Scroll thru the file and find the line ".SUBCKT irfp4668 1 2 3"
Place the cursor on this line and Right Click your mouse
An LTS window will open and two thirds the way down is the option "Create Symbol" - click on that option
It will ask "Do you really really want to?" Yes you do. That symbol is then created and saved in the area "AutoGenerated" from where you can then add it to the schematic.
Do this again for VOM1271 and MOC3021 and 4N26 and MAT02
Now when you open the LTS sim files the symbols are used and the sim runs.
My model file is based on the file from Bob Cordell plus all the others I have had to download from manufacturers and others like the LT Spice user group, so in particular - many thanks Bob.
Hi John,
I haven't done that before either so thank you so much for sharing that with us all.
I managed to find and add
.SUBCKT irfp4668 1 2 3
and
.subckt VOM1271 A K P N
but could not find MOC3021, 4N26 and MAT02 in the file JohnsModels.txt
I also used a text editor to search and those models or sub circuits were not found.
If you don't mind and you offer some more assistance.
Thanks
Stuart