Dear Diy friends , i have joined a 10M45S chip here just as shown at this schematic:
I like the sonic result but i am questioning myself , what is happenning with current values now . I have set a 300R resistor for 10 ma at K and 1K resistor between G and K . I haven' t measured the miliamps output until now , but for sure something must be done to restore current levels feeding a 6C45pi (WE437A) equiv.
Please give me advice .
tt
An externally hosted image should be here but it was not working when we last tested it.
I like the sonic result but i am questioning myself , what is happenning with current values now . I have set a 300R resistor for 10 ma at K and 1K resistor between G and K . I haven' t measured the miliamps output until now , but for sure something must be done to restore current levels feeding a 6C45pi (WE437A) equiv.
Please give me advice .
tt
If I understand your question right . . . .
By increasing the current through the 437a you are changing the voltages at the divider upstream. You either need to take a look at the entire input/driver section and supply to see if you need to re-spec' or redesign it, or return the 437a to it's designed for Ip of 6.5mA.
For the design plate current of .0065A change the set resistor to a 470 (465.38) Ohm on the IXYS chip and get rid of the 1K as well as the original 10K plate load resistor. (You don't need to take it out to try. You can just bypass it with wire) You can put a resistor on the output (the 1K) for a buffer at rf if you need it but to start it's best to keep it simple.
By increasing the current through the 437a you are changing the voltages at the divider upstream. You either need to take a look at the entire input/driver section and supply to see if you need to re-spec' or redesign it, or return the 437a to it's designed for Ip of 6.5mA.
For the design plate current of .0065A change the set resistor to a 470 (465.38) Ohm on the IXYS chip and get rid of the 1K as well as the original 10K plate load resistor. (You don't need to take it out to try. You can just bypass it with wire) You can put a resistor on the output (the 1K) for a buffer at rf if you need it but to start it's best to keep it simple.
Thanks Hearinspace . I will bypass the 10k resistor and change 300r for 470r . As for the 1k between G and K i have tried it without and the bass had become not so good as it were . Tubelab says it prevent oscilation .
Do you think the reg will present 10ma at output K regardless it has 6.5ma at input A, just because i set 300R at K for 10ma regulation ?
Do you think the reg will present 10ma at output K regardless it has 6.5ma at input A, just because i set 300R at K for 10ma regulation ?
Your high voltage (B+) connects to the centre pin A. The Bias resistor (470 Ohm) is across pins G & K (1 & 3). The output is taken from G (pin 1). The bias resistor is what controls the current that the 10M45S can pass. The ~470 Ohm value sets the current at ~6.5mA. As it is being used as a two terminal device (ie. only two interface connections with the rest of the circuit) there is only one possibility - what goes in must come out (and vice versa) so if 6.5mA is coming out , 6.5mA is going in.
As always . . . Safety First!!! You do know the rules for safe testing etc. right?
As always . . . Safety First!!! You do know the rules for safe testing etc. right?
Yeah it's OK . Don't worry. My description sounds different when you're thinking of it like K is the output with the current set resistor in series with the load. I do a lot of temporary test builds and it's easier for me to put the resistor across the pins directly and bolt the chip to chassis or plate. It all amounts to the same thing
Take a look at this page on Pete Millet's site. There's a diagram showing the hook-up. Just put in your 470 Ohm instead of the 600 Ohm one shown.
Take a look at this page on Pete Millet's site. There's a diagram showing the hook-up. Just put in your 470 Ohm instead of the 600 Ohm one shown.
Tap an output at the plate, and its just a triode driving a load.
The constant current provides only bias current for the plate,
and a high CCS impedance does not much otherwise interact
with the either the plate or the load.
-----------
Tap output at CCS-K, now its a source follower of extremely
high Gm. Following the Triode's pure Mu. The Triode's plate
sees a nearly constant current, barely any of the real load.
Could be good or bad, depending wether you don't mind
losing the normal triode interaction between the plate and
the load. As the real load will appear driven only by sand.
-----------
Tap an output halfway up the current sense resistor, and you
have a Triode driving half the load (or twice the apparent load
impedance) as seen in series with the lower half of the current
sense resistance.
The CCS is tricked into driving the other half of the load exactly
in opposition to the Triode, in series with the other half of the
current sense resistance. A near perfect "through the listening
glass" anti-clone of the reference Triode!
Net result: like Twin triodes in Parallel SE. Driving the real load
in series with 1/4 of the total current sense resistance. Most
of the triode flavor and single-endedness is preserved. But you
get some useful boost out of the sand, in addition to constant
bias current.
This recurring concept has been re-invented over and over:
Anti-Triode, Aleph, SEPP, Paraphase. Exactly the same thing!
Only different in name and specifics of implementation.
This is a real viable option to consider with the 10M45S.
The constant current provides only bias current for the plate,
and a high CCS impedance does not much otherwise interact
with the either the plate or the load.
-----------
Tap output at CCS-K, now its a source follower of extremely
high Gm. Following the Triode's pure Mu. The Triode's plate
sees a nearly constant current, barely any of the real load.
Could be good or bad, depending wether you don't mind
losing the normal triode interaction between the plate and
the load. As the real load will appear driven only by sand.
-----------
Tap an output halfway up the current sense resistor, and you
have a Triode driving half the load (or twice the apparent load
impedance) as seen in series with the lower half of the current
sense resistance.
The CCS is tricked into driving the other half of the load exactly
in opposition to the Triode, in series with the other half of the
current sense resistance. A near perfect "through the listening
glass" anti-clone of the reference Triode!
Net result: like Twin triodes in Parallel SE. Driving the real load
in series with 1/4 of the total current sense resistance. Most
of the triode flavor and single-endedness is preserved. But you
get some useful boost out of the sand, in addition to constant
bias current.
This recurring concept has been re-invented over and over:
Anti-Triode, Aleph, SEPP, Paraphase. Exactly the same thing!
Only different in name and specifics of implementation.
This is a real viable option to consider with the 10M45S.
Expertise? I'm only the village idiot. My ideas are waaay
offa the deep end. On the sandy path to madness...
I've used the 10M45S current source before only twice.
And only as a plate load for a 12AT7. In our application,
it did seem to be working well. Does that "expertise"
apply in any way to your situation? Take your chances.
But that don't stop me from rambling on about junk I
know nothing about, nosirree Bob! What idiots do.
-----------------------------------------
One caveat I neglected to mention: Both the Plate tap
and the middle tap have symmetrical pull up/down slew
capabilities. But the middle tap has resistance in series
you must not forget.
The CCS "K" tap has sandy near infinite pull-up! But only
one triode plate's worth of pull down. Don't be fooled by
the ultra low source follower "impedance" of the "K".
Because of the asymmetry, its not any lower impedance
than plate or middle taps. Unless you are directly driving
an equally assymetrical load. Like a screen or grid in A2
that wants extra current in mostly the pull-up direction.
So, what I'm saying: The highest slew is at the plate. The
lowest "impedance" is at the K. The middle tap, is a nice
compromise. But it all depends what you are driving next,
there is no one best answer.
offa the deep end. On the sandy path to madness...
I've used the 10M45S current source before only twice.
And only as a plate load for a 12AT7. In our application,
it did seem to be working well. Does that "expertise"
apply in any way to your situation? Take your chances.
But that don't stop me from rambling on about junk I
know nothing about, nosirree Bob! What idiots do.
-----------------------------------------
One caveat I neglected to mention: Both the Plate tap
and the middle tap have symmetrical pull up/down slew
capabilities. But the middle tap has resistance in series
you must not forget.
The CCS "K" tap has sandy near infinite pull-up! But only
one triode plate's worth of pull down. Don't be fooled by
the ultra low source follower "impedance" of the "K".
Because of the asymmetry, its not any lower impedance
than plate or middle taps. Unless you are directly driving
an equally assymetrical load. Like a screen or grid in A2
that wants extra current in mostly the pull-up direction.
So, what I'm saying: The highest slew is at the plate. The
lowest "impedance" is at the K. The middle tap, is a nice
compromise. But it all depends what you are driving next,
there is no one best answer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.