Hi Luvdunhill,
if you are accepting order, I can pay in advance for 16 X 2SJ201 and 16 X 2SK1530.
you have a mail.
THX
luvdunhil, I agree that the device drifts WRT to temperature, and this does present some issue when matching.
But what temperature are you chosing to match at?
And, does it really matter, as long as the temp that the devices are matched at are all the same??
I have not found much of an issue when matching on my Tek curve tracer, as it is possible to switch quickly between two DUTs and compare.
But you might consider bringing the device up to temp, then having a means to switch to a different bias point and note the current, getting a two point "vector" on the curve? Just an idea.
Something that could be tried on devices you already have and see if there is any variance between "matched" devices or not. If there is insigificant variance amongst the existing matched devices you now have then there's nothing to talk about...
_-_-bear
But what temperature are you chosing to match at?
And, does it really matter, as long as the temp that the devices are matched at are all the same??
I have not found much of an issue when matching on my Tek curve tracer, as it is possible to switch quickly between two DUTs and compare.
But you might consider bringing the device up to temp, then having a means to switch to a different bias point and note the current, getting a two point "vector" on the curve? Just an idea.
Something that could be tried on devices you already have and see if there is any variance between "matched" devices or not. If there is insigificant variance amongst the existing matched devices you now have then there's nothing to talk about...
_-_-bear
bear, I think your answers might be in the first post
Everyone else:
I'm really close to finalizing things, and it's looking good Just have to bounce some details with Blues and we should be off to the races.
Plan is to have 50 pairs on hand soon, and 50 more pairs in 4-6 weeks. So, there should be plenty to go around.
I may ask a mod to move this thread to the official Group Buy section, if that's okay with everyone here.
Everyone else:
I'm really close to finalizing things, and it's looking good
Just have to bounce some details with Blues and we should be off to the races.Plan is to have 50 pairs on hand soon, and 50 more pairs in 4-6 weeks. So, there should be plenty to go around.
I may ask a mod to move this thread to the official Group Buy section, if that's okay with everyone here.
bear, I think your answers might be in the first post
Everyone else:
I'm really close to finalizing things, and it's looking good
Plan is to have 50 pairs on hand soon, and 50 more pairs in 4-6 weeks. So, there should be plenty to go around.
I may ask a mod to move this thread to the official Group Buy section, if that's okay with everyone here.
I'd still like to keep it the way it is now and not turn it into a Group Buy where diyers pre-pay.
1st come 1st serve -I'll go back thru the thread and start a list on who requested first and how many. Please confirm by then.
Thanks!
Andrew,
It is an issue with a pair or more per rail.
Also, wrt a PP configuration, one of the main reasons to go to Toshibas, more expensive devices is that allegedly their transfer curves match almost perfectly between N & P devices. I'm sure you know this already.
The fact that you can match static currents via a bias adjustment does not assure that you have matched or best matched to an optimal point along the linear part of the transfer curve.
Of course you may argue that is the job of negative feedback, and to some extent it is, but others may argue that it is less than perfect and this sort of thing alters the harmonic spectra, especially when the curves diverge between the N & P devices while being driven.
Certainly for most builders it isn't much of an issue, since the Fairchilds and IRF devices already are quite satisfactory.
For me, it is something worth considering and investigating, and testing.
_-_-
PS. do you post in every thread?
It is an issue with a pair or more per rail.
Also, wrt a PP configuration, one of the main reasons to go to Toshibas, more expensive devices is that allegedly their transfer curves match almost perfectly between N & P devices. I'm sure you know this already.
The fact that you can match static currents via a bias adjustment does not assure that you have matched or best matched to an optimal point along the linear part of the transfer curve.
Of course you may argue that is the job of negative feedback, and to some extent it is, but others may argue that it is less than perfect and this sort of thing alters the harmonic spectra, especially when the curves diverge between the N & P devices while being driven.
Certainly for most builders it isn't much of an issue, since the Fairchilds and IRF devices already are quite satisfactory.
For me, it is something worth considering and investigating, and testing.
_-_-
PS. do you post in every thread?
note I said a single pair output stage, this is what the standard F5 has.
You are talking about a pair or more per rail. That is a completely different scenario and is not an F5.
I read what you said, Andrew.
The idea of more than one output device has been, as you know, discussed a whole lot over on the F5 thread. It is still an F5, imo. Cviller is making PCBs that accomodate pairs of devices. So, for all those who are using his boards matching is an important consideration.
Even so, I maintain that matching is a benefit if your goal is to optimize the operation of the circuit so that it compares favorably to the simulated circuit.
The whole point of the Toshibas is that they "match" curves between N & P, so why would one not want to optimize the matching?
Yep, "not required"... but we're discussing matching... got any insights into matching Andrew?
_-_-bear
The idea of more than one output device has been, as you know, discussed a whole lot over on the F5 thread. It is still an F5, imo. Cviller is making PCBs that accomodate pairs of devices. So, for all those who are using his boards matching is an important consideration.
Even so, I maintain that matching is a benefit if your goal is to optimize the operation of the circuit so that it compares favorably to the simulated circuit.
The whole point of the Toshibas is that they "match" curves between N & P, so why would one not want to optimize the matching?
Yep, "not required"... but we're discussing matching... got any insights into matching Andrew?
_-_-bear
see what he is asking.
I answered that first question. I did not stray into modified F5 territory.
You have answered the N to P part of his enquiry. You do not need to make reference to mine unless you think what i said is wrong.
Whatever - a better match means a more ideal operating condition in the real world. If it is "required" or not depends entirely upon degree. Is there a reason to go to the Toshiba's in the first place, if N & P matching is not of interest and merely setting static bias current is sufficient?
Everything is relative and depends upon degree... I guess that is the answer?
_-_-bear
PS. did you get my email?
Everything is relative and depends upon degree... I guess that is the answer?
_-_-bear
PS. did you get my email?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.