Diamond Differential Amplifier

Jump to Latest
#22
borys said:
Mile
I just was wondering if there is anythink special in diamond topology, diamond buffers are all over the net, so I just added 2 current sources and tef stage into it and it worked just fine and is simple. The results are there SR around 60u/us and thd nearly not mesurable:

http://www.diyaudio.com/forums/solid-state/240848-bg3-diamond-amp-simple.html

I was thinking how to improve the input stage but I have lack of knowlage and ideas.

Your amp is very interesting, buffored front stage is really good idea (I will defenetley build it).
There is one think that I do not know how to explain, I did listening tests of 2 nearly same circuits (one has toshibas TO92-mod second has 2n5xx1 tranies), there was audiable diffrence heard strigt away in bass area, the one with 2n5xx1 was bit deeper and more soft.
Another think - I just had changed 2 resistors in second diamond stage (reducing biasing by 50%) and harmonics character has changed totally - I do not know if this is specyfic case of my circuit or kind of diamond topology characteristics. It looks like that there is a way to fine tune the amp with listening test.

In your amp prapobly there will be a way too to fine tune the amp with listening sessions. I would put j-fets at input buffer anyway.
Click to expand...

Simple schematic work good but need more tuning than complex circuit, I think my schematic is also simple.
I suggest to revert resistors value on your diamond amp, R12/R13 and R11/R14.
Regards
 
#24
Rieds said:
I tried the schematic on ltspice.
I replaced R10K with 1K8 & added some capacitors for amp stability.
Click to expand...

If you put big compensation caps the amp will be slower.
Just replace in first stage amplifiers 5.6k resistors with 2.2k, than 1.8k with 3.3k in second stage and amp will remain stable with only one pair 33pF caps, and will be much faster.
FFt looks nice.

Mile
Thanks for advice, You are right, the second stage is bit warm (50deg) so I will swap resistors values.
 

Attachments

  • hd50.png
    hd50.png
    78.3 KB · Views: 1,077
Last edited:
#27
borys said:
If you put big compensation caps the amp will be slower.
Just replace in first stage amplifiers 5.6k resistors with 2.2k, than 1.8k with 3.3k in second stage and amp will remain stable with only one pair 33pF caps, and will be much faster.
FFt looks nice.

Mile
Thanks for advice, You are right, the second stage is bit warm (50deg) so I will swap resistors values.
Click to expand...
I tried your idea, PA is stable, the vas current decrease to 8mA .
thanks for the advice

Regards
 
Last edited:
#28
Rieds said:
I think You don't need 'Triple Darlington' or triple emitter folower, if you need high power you can replace the driver with the bigger transistor like MJE15034/35 or 2sc5200&2SA1943.

Regards.
Click to expand...
No, I don't think so... I thought TD or TEF just like you now(for more power),
but
more gain there at output stage mean less stress to VAS stage &
lower the distortion 🙂

I have build one, it is from apex too (AX-17 or 20)
MJE340+350 ---> 2SA1837+C4793 ---> 2SA1493+2SC5200
They are quite stable this arrangements, I've also changed to
MJE340+350--->MJE15032+15033--->2SA1216+2SC2922
and I prefer this one, more rugged (the heat is not so hot when biased 100mA)
The above is more warmer in the sound & hotter the heat sink with same bias.
I thing it might be the same here, try to build one & you know what I mean 🙂

ps: I can't simulate like you do, but I think should try sometime 🙄
do you mind to try to simulate it now? TEF or TD...?

Regards
 
Last edited:
#30
borys said:
If you put big compensation caps the amp will be slower.
Just replace in first stage amplifiers 5.6k resistors with 2.2k, than 1.8k with 3.3k in second stage and amp will remain stable with only one pair 33pF caps, and will be much faster.
FFt looks nice.

Mile
Thanks for advice, You are right, the second stage is bit warm (50deg) so I will swap resistors values.
Click to expand...

or replace 10R resistors with 33R in second stage...
 
#31
I think prototype should be made anyway. In symulation amp looks stable but square with high amplitude doesnt look nice, there is large overshoot and capactive load is making the amp to oscillate badly.

Mile
I have one question:
Can I make rail resistors in frot diff amps not symetric? For eg. 1k and 2.2k (if I will replace R1 and R7 with 1k)
 

Attachments

  • hd50v5.png
    hd50v5.png
    65.4 KB · Views: 947
  • hd50v51.png
    hd50v51.png
    32.1 KB · Views: 894
  • hd50v52.png
    hd50v52.png
    78.7 KB · Views: 981
#33
JOHN BALI said:
No, I don't think so... I thought TD or TEF just like you now(for more power),
but
more gain there at output stage mean less stress to VAS stage &
lower the distortion 🙂

I have build one, it is from apex too (AX-17 or 20)
MJE340+350 ---> 2SA1837+C4793 ---> 2SA1493+2SC5200
They are quite stable this arrangements, I've also changed to
MJE340+350--->MJE15032+15033--->2SA1216+2SC2922
and I prefer this one, more rugged (the heat is not so hot when biased 100mA)
The above is more warmer in the sound & hotter the heat sink with same bias.
I thing it might be the same here, try to build one & you know what I mean 🙂

ps: I can't simulate like you do, but I think should try sometime 🙄
do you mind to try to simulate it now? TEF or TD...?

Regards
Click to expand...
in simulation we can use tef from ax-17 or ax-20 without any stability problem. but most important for this time is the stable prototype 😉

Regards.
 

Attachments

  • tef.jpg
    tef.jpg
    60.7 KB · Views: 882
#35
Interesting harmonic distribution. Ensure that as the phase closes 180 degrees that your gain has fallen to 0 and below
borys said:
If you put big compensation caps the amp will be slower.
Just replace in first stage amplifiers 5.6k resistors with 2.2k, than 1.8k with 3.3k in second stage and amp will remain stable with only one pair 33pF caps, and will be much faster.
FFt looks nice.

Mile
Thanks for advice, You are right, the second stage is bit warm (50deg) so I will swap resistors values.
Click to expand...
 
#36
Great job you have going there Borys. Your out on a truth hunt. Well would you like to compare your current amplifiers with any amplifier here 😉 HAL: Amplifiers . You never know what you might find 😀. Let us know what you think.

kind regards,

Harrison.

borys said:
Mile
I just was wondering if there is anythink special in diamond topology, diamond buffers are all over the net, so I just added 2 current sources and tef stage into it and it worked just fine and is simple. The results are there SR around 60u/us and thd nearly not mesurable:
Click to expand...
 
#39
Rieds said:
in simulation we can use tef from ax-17 or ax-20 without any stability problem. but most important for this time is the stable prototype 😉

Regards.
Click to expand...

I want to test new topology with simple basic circuit, and add CCS, TEF, cascodes... in high power versions. With TEF output stage VAS can work with 6mA instead 20mA and DD with 3mA instead 9mA... prototype must be made.
Regards
 
You must log in or register to reply here.
Top Bottom