IRF matching, very simple, but it will take you some time hehe
Do not try to find or make a selection of IRF610 and IRF9610 according to Vgs voltage. That is not important. Usually they differentiate for 0,5 V but that is OK as offset setting eliminates this.
Transconductance curves of each type have to be parallel as much as possible. That you will get from three sets of measurements for each type of mosfet. Mark each mosfet with number. Start with 1 by IRF610 and the same with IRF9610.
IRF610 step 1
- take first IRF610 and set potentiometer that you read 100 ohm, read out Vgs and record the value as Vgs-initial-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at initial Vgs in a table chart
IRF610 step 2
- take first IRF610 and set potentiometer that you read 50 ohm, read out Vgs and record the value as Vgs-medium-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at medium Vgs in a table chart
IRF610 step 3
- take first IRF610 and set potentiometer that you read 10 ohm, read out Vgs and record the value as Vgs-maximum-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at maximum Vgs in a table chart
IRF9610 step 1
- take first IRF9610 and set potentiometer that you read 100 ohm, read out Vgs and record the value as Vgs-initial-9610
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at initial Vgs in a table chart
IRF9610 step 2 (attention)
- take first IRF9610 and set potentiometer to Vgs-medium-9610= Vgs-initial-9610 + (Vgs-medium-610 - Vgs-initial-610), read out and record resitance value in a table chart
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at medium Vgs in a table chart
IRF9610 step 3 (attention)
- take first IRF9610 and set potentiometer to Vgs-maximum-9610= Vgs-medium-9610 + (Vgs-maximum-610 - Vgs-medium-610), read out and record resitance value in a table chart
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at maximum Vgs in a table chart
From IRF610 and IRF9610 table charts you can select the first pair from the ones that have most similar resistance values at all three Vgs voltages. In this way transconductance curves of the first pair are closely parallel-matched. Second pair is the second most matching combination from all three Vgs measurements, etc.
Do not try to find or make a selection of IRF610 and IRF9610 according to Vgs voltage. That is not important. Usually they differentiate for 0,5 V but that is OK as offset setting eliminates this.
Transconductance curves of each type have to be parallel as much as possible. That you will get from three sets of measurements for each type of mosfet. Mark each mosfet with number. Start with 1 by IRF610 and the same with IRF9610.
IRF610 step 1
- take first IRF610 and set potentiometer that you read 100 ohm, read out Vgs and record the value as Vgs-initial-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at initial Vgs in a table chart
IRF610 step 2
- take first IRF610 and set potentiometer that you read 50 ohm, read out Vgs and record the value as Vgs-medium-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at medium Vgs in a table chart
IRF610 step 3
- take first IRF610 and set potentiometer that you read 10 ohm, read out Vgs and record the value as Vgs-maximum-610
- leave potentiometer as it is and measure all other IRF610
- for each numbered IRF610 put a measured resistance value at maximum Vgs in a table chart
IRF9610 step 1
- take first IRF9610 and set potentiometer that you read 100 ohm, read out Vgs and record the value as Vgs-initial-9610
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at initial Vgs in a table chart
IRF9610 step 2 (attention)
- take first IRF9610 and set potentiometer to Vgs-medium-9610= Vgs-initial-9610 + (Vgs-medium-610 - Vgs-initial-610), read out and record resitance value in a table chart
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at medium Vgs in a table chart
IRF9610 step 3 (attention)
- take first IRF9610 and set potentiometer to Vgs-maximum-9610= Vgs-medium-9610 + (Vgs-maximum-610 - Vgs-medium-610), read out and record resitance value in a table chart
- leave potentiometer as it is and measure all other IRF6910
- for each numbered IRF9610 put a measured resistance value at maximum Vgs in a table chart
From IRF610 and IRF9610 table charts you can select the first pair from the ones that have most similar resistance values at all three Vgs voltages. In this way transconductance curves of the first pair are closely parallel-matched. Second pair is the second most matching combination from all three Vgs measurements, etc.
Attachments
Last edited:
I lost track as of why IRF matching is discussed in this thread. Anyway, I have been waiting for anyone to mention about TSSA-HEXFET. TSSA-HEXFET is very different with other versions of TSSA/SSA. It is very cooperative (and very useful) and the performance is more like the F5 topology than the other SSA versions.
I'm just helping to select mosfet driver for BIGBT which is useful for anyone who will try this kind of output.
Hey Jay which specific hexfet are you interested in?
I didn't have time to work more on a TSSA basic, since I was ocupied with SSA upgrade but I promise I will. If you read the thread you can see that Sonnya and Gerhard already made it.
Hey Jay which specific hexfet are you interested in?
I didn't have time to work more on a TSSA basic, since I was ocupied with SSA upgrade but I promise I will. If you read the thread you can see that Sonnya and Gerhard already made it.
Yes it is really not the best starting position, but first we have to know few things about (driven) cascodes here. KSA/KSC cascode BJT's hFE in TSSA can hardly influence only to initial DC offset (different voltage drop on R8, R9 as a result of their base current difference), which is in any case calibrated to zero with trimmers.
hFE of KSA/KSC doesn't influence the open loop gain value nor the linearity of TSSA amplifier. Their main job as cascodes is to maintain Vce of VAS gain transistor constant, that results in a reduced power dissipation (stable hFE), constant and smaller Cob (bandwidth extension) of VAS gain BJT. Cascode BJT role is mostly supporting one, main (gain) job here is still on VAS gain transistors (BC550/560C), which have to be matched to be complementary equal.
Last edited:
So the main point lay in BC560C/550C matching. We have more room on KSA/KSC as when can calibrate initial DC offset with trimmer, the goal is to not have KSA and KSC at oposite hfe range (60 and 200) but to try to maintain they closed possible to the area 100-120. Have you a cross reference for KSA/KSC maybe 2SC3423/2SA1360 could do the job?
Marc
Marc
Nice supporting circuits you have Idefixes. In a soft-start circuit you could replace R4-R7 with this one.
That's a good idee as it save a huge place. This could be used to build the speaker protection in the same board as softstart=> results in wiring simplification.
Marc
Last edited:
To resume IRF matching described in post #261 in short.
IRF matching is reduced to a form similar like finding the most equal "potentiometers", where mechanical scale is set to a certain position by accu & P, at which Ohm-meter readout is recorded for comparison.
Yes LC that was i have understanding from your process description. It's quite clear.
Marc
Huge PCB space reduction yes. Only one CL-90 soft starts my stereo SSA BIGBT HP. Attached is the graph chart of inrush AC mains current. Explanation here.
Attachments
Huge PCB space reduction yes. Only one CL-90 soft starts my stereo SSA BIGBT HP. Attached is the graph chart of inrush AC mains current. Explanation here.
Thanks for link.....I have quite similaire condition, lower rail (+/-36Vac) but higher Va rating (500Va) and 88.000µF per chanel...I will adopte for sure CL90.
Marc
Note : I love the SSA picture two post after...chirurgical clean and compact layout/architecture...
Last edited:
LC
I'm very impressed with this nice ongoing thread, especially Marc is very intuitive here and I wonder to
congrats both of You for a really very nice and synergistic follow up the simple SSA to become an art.
Andrej I wonder to ask You from Yours enormous experiences which front end would be the best
for the bal. V12 engine to drive: the TSSA or SSA ?
Is the topology for inv. bal pin2 the same as is for the SSA?
After my experience with Nemo I listen music very rarely and wonder to chose the very best front end
for my bal. V12 which is in my slowprogress obtaining all semis needed for the proto.
Which Tosh. VAS drivers would You suggest instead the IRF?
I attach the guessed bal. IP for the TSSA. Your sharp eye suggestion & comment would be much appreciated.
Thnx for the answ.
Hi Smiley
I didn't forget about full balanced input T/SSA configuration. I just didn't find time to evaluate all details since I upgraded SSA with CCS and supporting Idefixes to test TSSA ASAP. After that I'll have enough data to design balanced version. At first sight TSSA would be easier task since there's very little thermal issue present. I'll be in touch soon.
P.S. Toshiba mosfet TO-220 drivers suitable are 2SJ313 & 2SK2013 or Renesas 2SJ79 & 2SK216 also.
Last edited:
Ecellent
Thank You Andrej for Yours Great Share,
it'll be a nice music feedback to You . .
As I can see Your best would be TSSA?
Thank You for drvs. info.
Thorsten
My sch was intend question to LC as he knows the main TSSA sch in detail. . .
Here the pics for You to easy see my question to Andrej,
Cheers
LCHi Smiley
I didn't forget about full balanced input T/SSA configuration. I just didn't find time to evaluate all details since I upgraded SSA with CCS and supporting Idefixes to test TSSA ASAP. After that I'll have enough data to design balanced version. At first sight TSSA would be easier task since there's very little thermal issue present. I'll be in touch soon.
P.S. Toshiba mosfet TO-220 drivers suitable are 2SJ313 & 2SK2013 or Renesas 2SJ79 & 2SK216 also.
Thank You Andrej for Yours Great Share
, it'll be a nice music feedback to You . .
As I can see Your best would be TSSA?
Thank You for drvs. info.
Hello,
Thorsten
My sch was intend question to LC as he knows the main TSSA sch in detail. . .
Here the pics for You to easy see my question to Andrej,
Cheers
Attachments
Oh, how could I forget that the BIGBT used IRF610
I used IRF9630/IRF632. I have other ("better") hexfets but I could get only the subcircuit for LTSpice simulation (SUBCKT). All hexfet subcircuits that I found on the net have strange behavior as if the Drain and the Gate is mixed up.
I have tried to change the pin orientation in the first line of the subckt so that the simulation looked "normal". But I'm not sure if it is the only thing that I have to change and I don't know yet how the subckt subroutine works, so I kept those hexfet for later when I know more about the software.
I didn't have time to work more on a TSSA basic, since I was ocupied with SSA upgrade but I promise I will. If you read the thread you can see that Sonnya and Gerhard already made it.
Sonnya used latfet (Exicon) isn't he? And I'm not sure if there has been a TSSA-IRF. Please have a look at this, because it is very "different". I expect many will build a TSSA-HEXFET (I mean TSSA-IRF) because of its "usefulness".
I didn't mean presenting there new commercial amp you're working on, but rather you in person, we can go and drink some beer
Back to TSSA.
you have got it under good control this time Andrej. The removing of dc coupled gain has improved things a lot.
Yeah suddenly amp relaxed from being weighted by low resistance GND FB resistor, causing major differences to input bias current. DC voltage gain is fixed to 1 but who cares if the amp now has -3dB/3 Hz cutting point, maybe it is even better considering safety in DIY arena.
Looking to sch, calculating all the details and being faced with SSA experience, I indubitably assume that TSSA will be even better performer with top edge sound.
Last edited:
Oh, how could I forget that the BIGBT used IRF610
Guys thanks for good wishes, appreciate greatly.
Yes IRF-s, I know, cannot get rid of them, more or less obsolete according todays standards, but what can I do.
I would love to implement Renesas 2SJ79 and 2SK216 into BIGBT, but the problem is where to get them. I would take couple of hundreds if someone could provide me reliable source. Any help with this mosfets would be greatly appreciated. They even have gate protection zeners integrated, OMG
I didn't mean presenting there new commercial amp you're working on, but rather you in person, we can go and drink some beer
I would love too be there but i cannot find space in the calendar next week. I have two orders that needs to be shipped next week. And we are running as fast as we can. Where in sweden are you located?