Just confirms the statement - this is definitely built for somebody who likes pushing the limits 😛 Looking great by the way
I want to use it to drive the subwoofers though. So mostly it will end up clipping or to be on safer side.
The SOA at 70V is about 2A for 2sc5949 so thats my requirement.
Looks a bit dangerous, running full cage driving without helmet and 4-6 point harness have killed people before because they hit their head in the pipes when they have an accident.
In a street car I would prefer some scaffolding in the backseat only to maintain some headroom while rolling while keeping the space around the driver seat free.
My approach when I had a street toy was to keep the center of gravity low to prevent rolling in the first place. 😉
Attachments
you built this?🙂, what other expertise do you have apart from building superb amps cases and cars?😉
I've seen lots of lethal cages and know the dangers. This one's a bush buggie/ street machine. There's lots of headroom and the cage is fully padded and upolstered. As an added safety bonus, the driver isn't very tall.
SG-1 input stage
I have it biased with 4 red led's , runs the Slew OPS nice at 10-20ppm THD20k.
Below is the .ASC (+ keen-cordell.txt) and my "version" of this nice circuit.
The final VAS current seems independent of differential current and rail voltage.
R21 and 22 in my example seem to be the only way to trim final current.
This IPS is quite different than all my others - in every way. 😱
PS- I'll explain it next post (once I understand it fully myself)
OS
I have it biased with 4 red led's , runs the Slew OPS nice at 10-20ppm THD20k.
Below is the .ASC (+ keen-cordell.txt) and my "version" of this nice circuit.
The final VAS current seems independent of differential current and rail voltage.
R21 and 22 in my example seem to be the only way to trim final current.
This IPS is quite different than all my others - in every way. 😱
PS- I'll explain it next post (once I understand it fully myself)
OS
Attachments
Last edited:
SG-1 input stage (2)
Tweaked it to 8ppm THD-20k / 3ppm 1K !!
(below 1) the Bode is more like a CFA , I could get good stability with even a
22pf @ 3.8mhz UGLF.
At 47p (C5) UGLF is still 2mhz with 80db phase margin (ultra stable).
(below 2) is PSRR - good as the best of the VFA's. -140db at 1-2K is exceptional.
This sort of predicts the slew rate (unity gain) frequency tends to determine
speed. This should be quite a fast IPS .....(next tests - clip and slew).
OS
Tweaked it to 8ppm THD-20k / 3ppm 1K !!
(below 1) the Bode is more like a CFA , I could get good stability with even a
22pf @ 3.8mhz UGLF.
At 47p (C5) UGLF is still 2mhz with 80db phase margin (ultra stable).
(below 2) is PSRR - good as the best of the VFA's. -140db at 1-2K is exceptional.
This sort of predicts the slew rate (unity gain) frequency tends to determine
speed. This should be quite a fast IPS .....(next tests - clip and slew).
OS
Attachments
I "wupped" this one quick.
Add the clamps - Q29-30 - R5/6 (like Groner did) .... (below 1).
Previous post schema had horrendous saturation - drew 100ma on overload
through Q12 and 28 !!! 😡
clamps can mitigate this . I do the same in the wolverine VAS (options).
Works just as well symmetrically - in fact , it works better - NO CCS to worry
about.
(below 2) is as good as the K-ND or C ... clip wise.
Slew was 150V/us - OK by me.
This IPS should be a easy "hit". I'm impressed by Mr. Groner's creation.
OS
Add the clamps - Q29-30 - R5/6 (like Groner did) .... (below 1).
Previous post schema had horrendous saturation - drew 100ma on overload
through Q12 and 28 !!! 😡
clamps can mitigate this . I do the same in the wolverine VAS (options).
Works just as well symmetrically - in fact , it works better - NO CCS to worry
about.
(below 2) is as good as the K-ND or C ... clip wise.
Slew was 150V/us - OK by me.
This IPS should be a easy "hit". I'm impressed by Mr. Groner's creation.
OS
Attachments
Last edited:
All done ....
"listening impression"
I have not built it yet ??
It should be as excellent as all the others.
It has the BEST PSRR of them all. THD is ultra low , it is fast ....
(below ) it is finalized.
Hope Mr. Groner won't mind if we actually listen to it (in slewmaster form).😀
.ASC (final) is posted. Can't make it fail on the simulator - unless I go to
nearly no Q1/2 degeneration and <22pF for C5.
OS
"listening impression"
I have not built it yet ??
It should be as excellent as all the others.
It has the BEST PSRR of them all. THD is ultra low , it is fast ....
(below ) it is finalized.
Hope Mr. Groner won't mind if we actually listen to it (in slewmaster form).😀
.ASC (final) is posted. Can't make it fail on the simulator - unless I go to
nearly no Q1/2 degeneration and <22pF for C5.
OS
Attachments
Vzaichenko tested and built the single current sourced version.
http://www.diyaudio.com/forums/solid...ml#post4586493
My middlebrook probe says plenty of phase margin ,
can't see why it shouldn't work ?
Edit - SG-1 is quite close to Groner's original ... http://www.eetimes.com/document.asp?doc_id=1280059&page_number=1
(figure 12) ....
My simulation is a "tamer" version , but exhibited the same overload and PSRR characteristics.
Just about every other observation in the article showed up in my simulation.
OS
http://www.diyaudio.com/forums/solid...ml#post4586493
My middlebrook probe says plenty of phase margin ,
can't see why it shouldn't work ?
Edit - SG-1 is quite close to Groner's original ... http://www.eetimes.com/document.asp?doc_id=1280059&page_number=1
(figure 12) ....
My simulation is a "tamer" version , but exhibited the same overload and PSRR characteristics.
Just about every other observation in the article showed up in my simulation.
OS
Last edited:
What I am asking about - maybe in some time there will be greated a PCB in same size as the Kypton ND / Symasui so on?
Thanks
For the Tian probe, please DC couple the circuit, the 220uf cap destroys the results'
You can also add a 1000Gheny coil to the FB loop, that is very close to the Tian, and much easier to do.
You can also add a 1000Gheny coil to the FB loop, that is very close to the Tian, and much easier to do.
Of course it was DC coupled! C3 was short circuited. Input capacitor was bypassed too. Have reduced the 1000µF cap across bias generator C14 to 1µF. For me using the Tianprobe isn't difficult (only with parameter stepping using the table command) and it gives very good predictable stability results.
Have found the reason for the disturbed "TianProbe()" plot: the capacitance multiplier to better the PSRR should be disabled/removed during simulation.
Has someone checked the 42V/µs slew?
BR, Toni
http://www.diyaudio.com/forums/soli...e-old-ideas-1970s-ips-ops-10.html#post4586493
Sorry , should check my links ...
OS
Sorry , should check my links ...
OS
Nice to see you back Pete, sorry to hear about the Parkinson's that is a real drag. Not sure how you can do some of the soldering and smt may get to be out of the question. I've been dealing with the worst Sciatica the past couple of weeks, nothing kills the pain and I surely can't sleep but it will pass.
I'm wondering if you still plan on doing any single board designs with smt devices? I've been getting pushed hard to go the class-d route for my speakers and since you were gone for so long haven't had a decent argument against it without your input. I'm looking at a class-d amp on the cone driver, TPA3251D in a bptl single ended configuration and a class-ab for the tweeter.No way will the specs or distortion be in the same league as your designs, nothing better than 1% distortion at 100 watts output. I do believe that I am going to be forced to use a smps power supply just for size considerations but I don't see any reason that a Slewmaster wouldn't work just fine without a linear power supply.
Anyway glad to see your return and best wishes.
I'm wondering if you still plan on doing any single board designs with smt devices? I've been getting pushed hard to go the class-d route for my speakers and since you were gone for so long haven't had a decent argument against it without your input. I'm looking at a class-d amp on the cone driver, TPA3251D in a bptl single ended configuration and a class-ab for the tweeter.No way will the specs or distortion be in the same league as your designs, nothing better than 1% distortion at 100 watts output. I do believe that I am going to be forced to use a smps power supply just for size considerations but I don't see any reason that a Slewmaster wouldn't work just fine without a linear power supply.
Anyway glad to see your return and best wishes.
When I solder , I shake a little ..... I concentrate harder - I can still make a
professional junction. It just takes longer.
I am going to choose at least 2 of the best IPS's and make a integrated
OPS/IPS/protection PCB. Hopefully JW will help a bit (with the SMD).
OS
professional junction. It just takes longer.
I am going to choose at least 2 of the best IPS's and make a integrated
OPS/IPS/protection PCB. Hopefully JW will help a bit (with the SMD).
OS