The windings aren't tied together, so no there isn't any cancelation. They act like single ended coils and the tiny core will be saturated with anything over a minute dc current.
Rush
These transformers have quadfilar windings, meaning that the wires in the windings lay next to one another. This results better flux coupling and also more capacitive coupling. Here is an excellent book chapter on audio transformers: http://www.jensen-transformers.com/an/Audio Transformers Chapter.pdf. Equal but opposite DC currents in the windings should result in flux calcallation.
It is becoming obvious that there are many possible F6 biasing schemes. It suggest that DIYers design their breadboard or PC board layout in a modular way consisting 5 sections:
I have some ideas for an F6 PC layout conforming to the "Universal Mounting Spec". Alternatively, it is possible to cut traces on an the Burning Amp BA-CO board for the F6 output FETs and their source and gate resistors. A separate small board would then hold the input driver, transformer and bias circuits.
- Input JFET driver and transformer
- Upper and lower bias circuits
- Upper and lower output FETs and their gate and source resistors
I have some ideas for an F6 PC layout conforming to the "Universal Mounting Spec". Alternatively, it is possible to cut traces on an the Burning Amp BA-CO board for the F6 output FETs and their source and gate resistors. A separate small board would then hold the input driver, transformer and bias circuits.
winding doesn't know from where DC current is going and doesn't care for polarity
core knows what's happening ........ and is not giving a F from where coils are fed ..... but is pretty much giving a F for polarity
so small number of parts is really not asking for butchering any pcb
either PtP or veroboard
important thing is to save small signal xformer of stray magnetic fileds , everything else is piece of cake
Rush, and others: I keep going back to post #38 and its attached schematic OTLAmp1 of a successful and well-established power amp from the last century. I recall working with it in 1972. Some current is passing through the secondaries of the interstage transformer, and no capacitors were used to block it. I continue to struggle with this question. Why can't the output stage of simplified F6 schematic be exactly like that shown in the corresponding OTLAmp1; with the deliberate [and simple] scaling up of the bias voltages to accomodate the JFETs? Afterall, a bjt is also transconductance device like any other "FET". Thus, Mr. Pass will have little else up his sleeve as buzzforb hinted at in an earlier post.
Last edited:
The ones I played with were small from junked transistor radios. In schematic? One of the links in post #35 may have an answer for a reputable power amp.
If the proposed Jensen transformer is small; then can it be made bigger to tolerate a small dc current flowing through it?
yes , everything can be made
but this one certainly isn't made for that
so - what's the fuss - I know that I can toss at least 5mA of unbalanced current through my repeaters , without loosing anything ...... but through Jensen ....... I'm sure it's not the case
conclusion - use bigger xformer , and world is your oyster , regarding myriad biasing possibilities
if one is going to use delicate Jensen , DC through coil is no-no .......
canceling is impossible , because you need twins Semisouths and such thing doesn't exist
but this one certainly isn't made for that
so - what's the fuss - I know that I can toss at least 5mA of unbalanced current through my repeaters , without loosing anything ...... but through Jensen ....... I'm sure it's not the case
conclusion - use bigger xformer , and world is your oyster , regarding myriad biasing possibilities
if one is going to use delicate Jensen , DC through coil is no-no .......
canceling is impossible , because you need twins Semisouths and such thing doesn't exist
Your right, I wasn't thinking.
If the currents are equal and opposing should balance. If there is a small offset, should be OK, depends on core size.
Someone said, just try it.
Probably a good idea.
Rush
Mr. Pass is already using a Jensen transformer, and is assessing the subjective and objective performance of his F6. This by itself is useful information. A bigger transformer is not required. The bias Vgs for JFet is + 1.2 V [past post]; twice that of bjt. How bad can it really get in a circuit of a schematic like that of OTLAmp1? Gate current is in microAmps; and by contrast it is at a milliAmp level for bjt [base current].
Last edited:
example - SS as per pic
I=2A
40C : Ugs=1V576
70C : Ugs=1V581
I=2A
40C : Ugs=1V576
70C : Ugs=1V581
An externally hosted image should be here but it was not working when we last tested it.
ZM : you wrote:
I=2A
40C : Ugs=1V576
70C : Ugs=1V581
Ugs is almost a contant at 1.576-1.581 V in the temperature range. It implies great thermal stability. What is the attendant Gate current under these conditions ? It will be the talked about current flowing in the secondary of the transformer when I look at the schematic of OTLAmp1.
I=2A
40C : Ugs=1V576
70C : Ugs=1V581
Ugs is almost a contant at 1.576-1.581 V in the temperature range. It implies great thermal stability. What is the attendant Gate current under these conditions ? It will be the talked about current flowing in the secondary of the transformer when I look at the schematic of OTLAmp1.
Last edited:
apples and oranges
Jensen simply isn't made for that
from datasheet :
I know from experience what's happening when xformer - not made for it- is having any DC through coils ; just google for "signal xformer de-magnetising procedure"
yes - I measured that some time ago (while I was making disgraceful Babelfish J2)
current - dunno - don't remember ........ because Papa gave that info few times , and I just didn't found worth remembering it - irrelevant for common usage
I'm simply refusing any further talk about possible direct connection of coils in biasing scheme ; my fingers are sore from myriad of posts in this thread , contemplating about these issues
you didn't mention AC
not really ;
re-read thread few times
Last edited: