Will do - Thx!
100mA per Vertical MOSFET is a good starting point and a good rule of thumb! Lateral MOSFET can be driven with lower bias and still achieve very good THD levels as I have found during test of the Exicons
I have come to rely on Borbely's recommendation to use >=500mA of total output stage bias irrespective of device numbers.
I am listening to your amp now, with 600mA output stage bias (previously reported a while back).
I am listening to your amp now, with 600mA output stage bias (previously reported a while back).
... this was one of the major reasons to develop the 3pair SA2015 and 3pair SA2016 pcb's. Minimum 450 - 500mA bias total for the first watt being class A. Another reasons was SOA, noise, lower distortion. higher linearity.
... a typical FFW-Amp - First Four Watts amplifier!
Sorry - have forgotten to add the "s" to "watt" 😉
Last edited:
Hi Toni, I need to order some more heatsinks for my SA2014 Mono build. It is hard to find the ones with pins except from Reichelt. What heatsinks are used on the SA2015 and SA2016? I may as well get a bunch given the shipping cost.
PS: I am also having difficulty finding the Kemet 2.2uF caps PHE426HD7220JR06L2 as Mouser received the incorrect parts in their last resupply order. At the moment I only need 2 but can purchase in lots of 10. Are they used in SA2015/16?
I've been away for a week and trying to catch up on things now
PS: I am also having difficulty finding the Kemet 2.2uF caps PHE426HD7220JR06L2 as Mouser received the incorrect parts in their last resupply order. At the moment I only need 2 but can purchase in lots of 10. Are they used in SA2015/16?
I've been away for a week and trying to catch up on things now
- SA2014 uses 2 pieces 2.2µF/250Vdc MKP PHE426HD7220JR06L2 ( Reichelt/WIMA alternative: MKP4-250 2,2µ)
- SA2015/SA2016 uses 1 piece 4.7µF/250Vdc MKP PHE426HD7470JR06L2
Just have seen: the PHE426 2.2µF are available after 19.04.2017 at mouser - I can send you 2 pieces if you need them earlier but www.distrelec.biz has them also on stock if you don't want to buy the alternatively useable WIMA MKP .
Heatsinks for SA2014 you need per channel:
- 1 piece SK75/37.5 for TO220 (Reichelt: V SK 75-37 STS)
- 2 pieces SK76/37.5 for TO220 (Reichelt: V SK 76-37 STS)
Heatsinks for SA2015/SA2016 you need per channel:
- 1 piece SK75/25 for TO220 (Reichelt: V SK 75-25 STS)
- 2 pieces SK76/25 for TO220 (Reichelt: V SK 76-25 STS)
SK75/SK76 are hard to get. But I have designed the PCB so that you may use an aluminium stripe formed to an L-shape.
BR, Toni
Thanks. I placed an order with Reichelt including the WIMA MKP.
Here is the new SMD Main Control Unit wired up. (The IPS/VAS boards sit on a shelf above the 2-channel star/silent ground board - removed for the pic.) I now have just one "aux" PSU powering the MCU, input boards, output relays and inrush relays. Previously there were three PSUs (1 aux psu per channel and the MCU itself). The high performance PSU board installed now is complete overkill for this application, but I had it to hand and perhaps it helps the BAL-SE input boards perform better.
I also added drain wires to chassis from the soft iron tubes through which output wires run.
(The aux PSU toroid is shielded and potted, and it sits on an iron base plate. Should be better than 3 unshielded PCB transformers. At one point I had made the shelf they all sat on, with 10mm spacers, in powder coated iron with a view that it would shield the twin main PSU toroids below. However, I could clip leads to this shelf and observe in ARTA the increase in radiated EMI as each of these PCB transformers powered up.)
EDIT: I added an earlier pic of the previous setup
Here is the new SMD Main Control Unit wired up. (The IPS/VAS boards sit on a shelf above the 2-channel star/silent ground board - removed for the pic.) I now have just one "aux" PSU powering the MCU, input boards, output relays and inrush relays. Previously there were three PSUs (1 aux psu per channel and the MCU itself). The high performance PSU board installed now is complete overkill for this application, but I had it to hand and perhaps it helps the BAL-SE input boards perform better.
I also added drain wires to chassis from the soft iron tubes through which output wires run.
(The aux PSU toroid is shielded and potted, and it sits on an iron base plate. Should be better than 3 unshielded PCB transformers. At one point I had made the shelf they all sat on, with 10mm spacers, in powder coated iron with a view that it would shield the twin main PSU toroids below. However, I could clip leads to this shelf and observe in ARTA the increase in radiated EMI as each of these PCB transformers powered up.)
EDIT: I added an earlier pic of the previous setup
Attachments
Last edited:
Awesome! Absolutely awesome!
I'm sure you will get the additional noise headroom as expected!
Best regards,
Toni
I'm sure you will get the additional noise headroom as expected!
Best regards,
Toni
SA2015: using vertical mosfets IXTH80N20L and IXTH48P20P
Test sample using 1 pair of IXTH80N20L and IXTH48P20P up und running!
Quick THD+N tests (80k bw) with SOA circuit in situ:
Rgate 15R. No oscillations detected. No signs of instability at +/-50V supply.
More tests and tuning data will follow as soon as I have repaired my test sample (a f...y short circuit during a f...y screw driver accident during tests 😱)
Dave Zan's simulation models seems to match pretty good with the devices in reality. Bias settings in reality as expected from spice design!
Bias temperature stability similar to SA2015 IRFP240/IRFP9240.
Have fun,
Toni
Test sample using 1 pair of IXTH80N20L and IXTH48P20P up und running!
Quick THD+N tests (80k bw) with SOA circuit in situ:
- THD+N 20k@100W@8R: 0.0035%
- THD+N 20k@200W@4R: 0.0042%
Rgate 15R. No oscillations detected. No signs of instability at +/-50V supply.
More tests and tuning data will follow as soon as I have repaired my test sample (a f...y short circuit during a f...y screw driver accident during tests 😱)
Dave Zan's simulation models seems to match pretty good with the devices in reality. Bias settings in reality as expected from spice design!
Bias temperature stability similar to SA2015 IRFP240/IRFP9240.
Have fun,
Toni
Attachments
The whole installation looks nice, especially the shielded toroid, where is that from?
That's also nice.
And that's nice too😉
Some of the parameters of a FET can vary quite a bit from nominal datasheet values, which makes it difficult to be accurate without some adjustments to match measured results.
So this is excellent news.
Best wishes
David
Thanks! I purchase my toroids from https://sklep.toroidy.pl/en_US/c/Toroidal-transformers-SUPREME-AUDIO-GRADE-V2/99
The service and pricing is excellent. I have had a number of custom ones made as well. They make each to order anyway.
There are two big guys underneath that little one...
Attachments
Last edited:
After repair - IXTH48P20P has past away with a short circuit between drain/source ~ 200 mOhm - I started some additional tests.
SOA circuit disabled, +/-40V, VAS-diode removed, Bias 150mA
THD+N 20k@60W@8R: 0.0018% (80k bw)
- Rgate 15R
- no oscillations detected
- no signs of instability at +/-40V supply.
- small signal square wave: nearly same as previous amplifiers (SA2014/SA2015 and SA2016 variants).
- slew rate 100V/µs
- 131 µV - no bw limit
- 56 µV - 80k bw
- 31 µV - 20k bw
- 23 µV - A-w
Bias circuit/stability: thermal runaway starts above 650mA bias. Above 650mA this may destroy your MOSFET very quickly! This behaviour seems to be exponentially.
BR, Toni
Last edited:
R.Cordell in his interviews' Threads gave us a procedure to check for thermal instability. It is repeated in his book.
=> case is from "Modu" and the model "Pesante Dissipante" includes this heatsinks.
BR, Toni
Last edited:
Dear AndrewT,
THX! A link, thread name and/or post number or a book page number would be helpful here ...🙂
BR, Toni
