I used higher hfe rated Cdil BD139 in JLH69 Class-A & it worked fine. The price is also very cheap. You can buy 10 unit by spending only $1.
Capacitor Decoupling the front end from the output stages means more than adding a few capacitors to earth at the Vas and Bootstrap connections to the supply rails.
If you would allow space for resistors to form RC low pass filter to prevent spuriae on the supply rails entering the Vas and bootstrap connections to get a more effective result. 47 Ohms and 47uF//100n should be enough.
I have seen circuits where RC decoupling has been used on only one rail, here that would be the negative one. The LTP tail is a CCS which has some rejection properties which the bootstrap lacks. The latter will ingest any negative supply rail.spuriae unless steps are taken to prevent such ingress.
Last edited:
The formula where turnover is - 3dB is 1/2pi. R.C. that works out at 72 Hz the filter. Frequencies above 72 Hz will be attenuated at a rate of 6 dB per Octave so the higher the spurious voltage is the greater this will be attenuated. There are on line calculators you can use. You can find one of these in a selection on digikey website. From one of these you should be able to see the attenuation at 100kHz.
just randomly throwing decoupling along the rails won't do anything ? 10-22R between driver/output with the caps behind (at the driver/vas)
decouples that nested FB path ,I do this on the wolverine/slewmaster EF3's. The EF3 also can exhibit instability depending on the decoupling
an the Ft of the pre/driver/ output. I've modeled the OP in isolation. EF3 peaks near Ft , CFP @ 10-12 Mhz.
Last edited:
Grounding is often the trickiest part of audio PCB design. Ideally, the positive and negative grounds should resolve locally before connecting to the rest of the ground network. If you look at the ESP's P3a PCB, you'll see this in practice. Addressing this on your PCB would require a bit of reworking.
The Wolverine PCBs are an excellent example of audio grounding done right, using a star on star approach with a ground lift. It's quite a bit more intricate that simpler grounding topologies.
The Wolverine PCBs are an excellent example of audio grounding done right, using a star on star approach with a ground lift. It's quite a bit more intricate that simpler grounding topologies.
Attachments
The "Good" version is what is generally recommended and how I see most grounding implemented on most successful PCB designs. I have not tried a PCB with the "BAD" Layout.
Douglas Self describes the reasoning for this in his book Audio Power Amplifier Design (chapter 25):
You can see this approach used in the Wolverine and ESP P3a layout.
Douglas Self describes the reasoning for this in his book Audio Power Amplifier Design (chapter 25):
You can see this approach used in the Wolverine and ESP P3a layout.
A power amplifier has a mix of high and low currents. You want to control the flow of these currents (as described in the quote from Douglas Self above). With a ground plane, you loose this control.
However, I have seen ground planes used is specific parts of the circuit. Like just the small signal sections. I think Self does this in PCBs he sells for his Blameless amp.
You'll generally see ground planes avoided and specific routing used to control the flow of currents and keep the large output currents away from the small input current.
However, I have seen ground planes used is specific parts of the circuit. Like just the small signal sections. I think Self does this in PCBs he sells for his Blameless amp.
You'll generally see ground planes avoided and specific routing used to control the flow of currents and keep the large output currents away from the small input current.
I outlined the first point in post 203 in connection to the image 121807 and provided the RC formula and where online calculators could be found. The circuit in post 207 image 1218448 showed no change.
Moving on, in another oversight the zobel network now shows R15 as 220R?? Even with a proper value this is needs help from additional components.
Thanks for the catch on R15.
With the RC networks on the supply rails I thought I was surely ready to go.
I've seen this kind of grounding layout used on many boards around here so I thought I was on a good path.
I've got a lot more work to do it seems.
Time to study my wolverine boards some more....
With the RC networks on the supply rails I thought I was surely ready to go.
I've seen this kind of grounding layout used on many boards around here so I thought I was on a good path.
I've got a lot more work to do it seems.
Time to study my wolverine boards some more....