Balanced F5 question

Status
Not open for further replies.
Jump to Latest
#82
bobodioulasso said:
Matching is compulsary for a balanced amp.
Or it is not balanced.
Rather to build a regular F5 with them.
Click to expand...

An N channel and a P channel typically deal with the different halves of the sine wave and I believe these can never really be matched correctly. In a balanced amp with unmatched parts, will it be any different? My understanding is that matching is more for current sharing and dc ofset, but they are not sharing current in this config and Im not sure how much offset I would need to deal with.
 
#84
Matching resistors wheatstone Bridge


Well busy now got same dudas to match
 

Attachments

  • 300px-Wheatstonebridge_svg.png
    300px-Wheatstonebridge_svg.png
    7.7 KB · Views: 736
  • DSCF0799.jpg
    DSCF0799.jpg
    591 KB · Views: 735
#88
Gentlemen,

I proposed a resistor for you which IMHO is good value for money and very good sonically. Just look at what they use in high end Japanese amps. If you want to use more fancy resistors, then you can try Isabellehuette PBH, Caddock MP930, Texas Components T220Z, ....... Plenty of choices. The PCB will have 9mm & 5mm pitch, for MPC74 (faithful to my original design) and any TO220 boutique resistors.

If you don't want to GB MPC74, it is just fine with me. I am just thinking ahead for those who want to build it exactly as I did mine.

As to resistor matching, this is how I do it myself (see attachment). But feel free to do it your way.

I'll still do the following as promised :

1) get a Gerber for my PCB layout done and send it to Variac, with standard DIYA mounting hole patterns,
2) give guidance and help to AR2 to build his prototype according to the way that I built mine.

He has the generosity to then share his building experience with you, and help those out who is new to the game.

What I will not do is to answer question like "what if I chance your circuit by this and that for xyz special application", ......, etc. I have not tried any other configurations than those I am going to publish, so I have no idea whether your proposed changes will work or not. I can only encourage to experiment and report back your experience. It is not so bad to burn a few transistors -- all part of the education.

😉

So now I need to get the layout updated for Variac to check and agree, then Bksabath will get the ACAD files to convert to Gerber, according to my instructions.


That'll be it for tonight,
Patrick


PS My source says MPC74 +/-5%. I have no problems getting quad sets to much less than 0.5% for a batch of 40.

.
 

Attachments

Last edited:
#91
Incorrect. The balanced input signal is still being amplified by a factor of 6.

You do get higher output voltage level at your speakers, but this is not because the gain is higher.
It is because the input signal (= Vin+ - Vin-) is now 2x that of a single ended signal (e.g. Vin+ alone).


Patrick
 
Last edited:
#97
I thought I would bring this up because some people might be frightened off to build this.

I agree the N devices need to be matched to each other and the P devices need to be matched to each other. However I don't think it is absolutely critical to get perfect matching between N and P. If that was the case Nelson would not use IRFP devices which are not perfect complementary devices to start with.
Even the Toshibas which are better are still not perfect complementary devices.

So could either Papa/Patrick/Juma/Zenstar, comment about how critical all of this is.
This project should be about encouraging people to participate in it, rather than scaring people away.
It would be nice to put things into perspective for people, rather than just saying you need perfect matching.
 
#98
current matching between all input jfets is a must ;
EUVL's combined source resistor/different Idss - as tricky bonus , for fun and builder's pride

current matching of same polarity of outputs is a must
current matching between polarity of outputs is splitting a hair , but can be fun if you like it

using THD meter and Papa's trick of additional source resistor in appropriate place in output is a cherry on top of cream pie ( and poke in the eye of our usual overengineering .... and above hair splitting :rofl: )

but - some people don't care for cherry , especially not 2000 .

:clown:

btw. - did I told you that I don't care which mosfets are in output ?

I like it lush
 
Last edited:
Status
Not open for further replies.
Top Bottom