Nobody has published Gerbers of an M2x input stage using the (solid state equivalent of a) White Cathode Follower. Even though Douglas Self praises it to the sky in his "Small Signal Audio Design" textbook, and even though it was the runaway winner of his Audio Precision Distortion Shootout.
Nor has anybody published Gerbers for an M2x daughter card, where every single transistor in the signal path is cascoded ... even the output devices.
Nor has anybody published Gerbers for an all-MOSFET M2x daughter card. In fact, I don't even think anybody has installed TI's all-MOSFET spectacular performing opamp (OPA1656) into one of the existing cards.
There is plenty of unmined ore still in the ground.
_
Nor has anybody published Gerbers for an M2x daughter card, where every single transistor in the signal path is cascoded ... even the output devices.
Nor has anybody published Gerbers for an all-MOSFET M2x daughter card. In fact, I don't even think anybody has installed TI's all-MOSFET spectacular performing opamp (OPA1656) into one of the existing cards.
There is plenty of unmined ore still in the ground.
_
Attachments
The Calvin Buffer I mentioned here :
The diyAudio First Watt M2x
is a sort of white follower with cascoded JFETs.
I chose not to use either of them, for the simple reasons that :
1) The BFB as is is already too high in bandwidth and easy to pick up RF noise.
Cascoding will not make this better.
2) The White follower will increase current driving capabilities, at the expense of poor PSRR.
The BFB already is more than sufficient to drive the 600R transformer.
It will be much better to use the Taylor follower rather than the White follower if one wants to use a modulated current source.
Taylor Source Follower
Cheers,
Patrick
The diyAudio First Watt M2x
is a sort of white follower with cascoded JFETs.
I chose not to use either of them, for the simple reasons that :
1) The BFB as is is already too high in bandwidth and easy to pick up RF noise.
Cascoding will not make this better.
2) The White follower will increase current driving capabilities, at the expense of poor PSRR.
The BFB already is more than sufficient to drive the 600R transformer.
It will be much better to use the Taylor follower rather than the White follower if one wants to use a modulated current source.
Taylor Source Follower
Cheers,
Patrick
If I turn the signalgenerator of the background remain the same.
Yes I know the I´m exceding the opamp maxvoltage with no zeners.
Yes I know the I´m exceding the opamp maxvoltage with no zeners.
Last edited:
Do all M2x input stage cards produce that same oscilloscope picture? If Austin and MountainView do the same thing as IPS7 during your testing, and if those two sound good in the amplifier, maybe this oscilloscope display is "normal". Have you checked?
Do both IPS7 input stage cards produce that same oscilloscope picture? Or is it only one of the two?
When you install a dual opamp like NE5532 or TL072 or OPA2134 or LM4562 into the IPS7, do you get the same oscilloscope picture that you get with a single opamp like OPA134 or TL071 or LT1122 or OPA604? On both of the two IPS7 cards?
If the "not normal" behavior happens to occur when using a single opamp in IPS7, does it also occur when you plug that exact same individual opamp chip into a Tucson input card?
etc.
Do both IPS7 input stage cards produce that same oscilloscope picture? Or is it only one of the two?
When you install a dual opamp like NE5532 or TL072 or OPA2134 or LM4562 into the IPS7, do you get the same oscilloscope picture that you get with a single opamp like OPA134 or TL071 or LT1122 or OPA604? On both of the two IPS7 cards?
If the "not normal" behavior happens to occur when using a single opamp in IPS7, does it also occur when you plug that exact same individual opamp chip into a Tucson input card?
etc.
Last edited:
I do not run devices anywhere close to their maximum rating.
For me bad practice.
The 2SK2145 is rated at 300mW for a tiny SOT23-5.
I normally do not want to use more than 1/5 of that.
Which means say 12V 2.5mA per JFET.
In the BFB, they are running at exactly 12V 2.5mA.
The BFB has a Class A bias of 30mA and a Zout of 3.5R.
To achieve the same Zout, you will need to have 18x 2SK2145 in parallel at 4mA each.
And then you will have > 18x the capacitance to drive than a single 2SK209 in the BFB.
So yes, normally I also don't like using CFPs.
But if you have a heavy load and want low bias, then they become quite handy.
Just need to watch out for oscillations.
Patrick
For me bad practice.
The 2SK2145 is rated at 300mW for a tiny SOT23-5.
I normally do not want to use more than 1/5 of that.
Which means say 12V 2.5mA per JFET.
In the BFB, they are running at exactly 12V 2.5mA.
The BFB has a Class A bias of 30mA and a Zout of 3.5R.
To achieve the same Zout, you will need to have 18x 2SK2145 in parallel at 4mA each.
And then you will have > 18x the capacitance to drive than a single 2SK209 in the BFB.
So yes, normally I also don't like using CFPs.
But if you have a heavy load and want low bias, then they become quite handy.
Just need to watch out for oscillations.
Patrick
Great BFB indeed!
I will use 1/2 2Sk2145 instead of each 2sk209 because I have 100 in stock
Can-I replace the toshiba BJT with BC560C ?
I know the dissipation will be a problem...
I will use 1/2 2Sk2145 instead of each 2sk209 because I have 100 in stock
Can-I replace the toshiba BJT with BC560C ?
I know the dissipation will be a problem...
> Great BFB indeed!
All credit to Calvin, hence the name BFB (where he lives).
> I will use 1/2 2Sk2145 instead of each 2sk209 because I have 100 in stock
Unless you have GR grade, you will need to change the circuit to suit the high Idss.
> Can-I replace the toshiba BJT with BC560C ?
The diyAudio First Watt M2x
All spice files are posted.
Please help youself.
Patrick
All credit to Calvin, hence the name BFB (where he lives).
> I will use 1/2 2Sk2145 instead of each 2sk209 because I have 100 in stock
Unless you have GR grade, you will need to change the circuit to suit the high Idss.
> Can-I replace the toshiba BJT with BC560C ?
The diyAudio First Watt M2x
All spice files are posted.
Please help youself.
Patrick
Build it on breadboard before making PCBs.
You will need to check stability on the real thing, like I did.
Won't always show up in Spice.
Patrick
You will need to check stability on the real thing, like I did.
Won't always show up in Spice.
Patrick
I was told that a few of you had asked for the BFB Alpha PCBs.
While I thank you for your interest, please take my words seriously in post #4659:
The diyAudio First Watt M2x
I will not be able to nurse you through this with remote debugging if you get into trouble.
You need to understand how the circuit works, and know your way to find bugs and solve them.
Do not hope to just solder and switch on, and all is fine.
Otherwise it is not known as Alpha testing.
Don't say you have not been warned.
Patrick
While I thank you for your interest, please take my words seriously in post #4659:
The diyAudio First Watt M2x
I will not be able to nurse you through this with remote debugging if you get into trouble.
You need to understand how the circuit works, and know your way to find bugs and solve them.
Do not hope to just solder and switch on, and all is fine.
Otherwise it is not known as Alpha testing.
Don't say you have not been warned.
Patrick
For those who seek an M2x IPS daughter card with very VERY low output impedance, I suggest taking a look at post #3448 of this thread. One of its attachments is copied below.
(BTW check this opamp part number in the listening evaluations of #4603)
_
(BTW check this opamp part number in the listening evaluations of #4603)
_
Last edited:
I am 2 zener short for the other IPS7 so I can only test with one (this).
Yes its the same with dual opamp.I only have the IPS6 and IPS8 to compare with,and they dont have that green background.
But I see that when I unplug the input cable the green dissapeares.
So It´s sommething with the input,any way listening to music through it sounds as it should.
Edit:Anyone else checked IPS7 with scope?
Last edited:
I see you are conservative with power dissipation, grat for DC ofset stability and reliability
Maybe 2 X BC560C could cope with 30mA bias? (emitter resistor needed? )
if not I will order the toshiba transistors...
If you have, use BD139/140, though they are less linear.
If you have to buy, then buy a bunch of the Toshiba's.
Plenty of use, e.g. also in the Vbe multiplier.
And be prepared to have to adjust values, especially capacitors.
Patrick
If you have to buy, then buy a bunch of the Toshiba's.
Plenty of use, e.g. also in the Vbe multiplier.
And be prepared to have to adjust values, especially capacitors.
Patrick
To mister BB in AZ ... please check your PMs
Experimental prototypes sent your way on Friday.
Experimental prototypes sent your way on Friday.
My IPS6 curse has been lifted.
Third time was the charm! After installing J113 backwards on the first 2 builds (silly me trusted the data sheets), I completed my third attempt at IPS6 this morning, and currently listening, and they sound good.
Thanks for all your help Mark Johnson!
At least now, I took the Cedarburgs out, and can clean them finally.
Next project:
I got a couple of other Analog Device op amps - AD744BQ and AD8475Q, and hope to compare between these, on the IPS7, and the AD797 on the Cedarburg (IPS8)
Also, I got a Meanwell 24v power supply, and hope to figure a way to test/set daughter cards for trimming, or biasing. I plan to insert some stand offs on a piece of wood (carefully marking the +/-/in/out) to be able to simply bolt in the card being tested, and flip on the PSU. This would allow me to do some of the functions without having to measure, and set inside the M2x case. I welcome suggestions and criticisms.
Third time was the charm! After installing J113 backwards on the first 2 builds (silly me trusted the data sheets), I completed my third attempt at IPS6 this morning, and currently listening, and they sound good.
Thanks for all your help Mark Johnson!
At least now, I took the Cedarburgs out, and can clean them finally.
Next project:
I got a couple of other Analog Device op amps - AD744BQ and AD8475Q, and hope to compare between these, on the IPS7, and the AD797 on the Cedarburg (IPS8)
Also, I got a Meanwell 24v power supply, and hope to figure a way to test/set daughter cards for trimming, or biasing. I plan to insert some stand offs on a piece of wood (carefully marking the +/-/in/out) to be able to simply bolt in the card being tested, and flip on the PSU. This would allow me to do some of the functions without having to measure, and set inside the M2x case. I welcome suggestions and criticisms.
So your BFB is already working ?
Don't forget the input resistor / capacitor network on the main board.
Patrick
Don't forget the input resistor / capacitor network on the main board.
Patrick