The input differential pair transistors in the simulation schematic are already perfectly matched, as PSpice assumes devices using the same part number are identical. For a real implementation of this amp, I'll probably try to match within 5-10 mV for the input pair. Having said that though, I haven't even started trying to lay out the board yet. I plan to house the amp in a 3U Dissipante case, which of course, I have yet to order. Unfortunately, that case is not available yet domestically from the Diyaudio store, so it has to come all the way from Italy. Expensive, too.
Last edited:
I may try selecting a population of devices with similar HFE using multimeter test, then select from that population for Vbe. That's what I'm doing for the auto bias circuit of my 6LR8 amp-in-progress. I select a bunch of MPSA92 and 2N6517 for high HFE, then select from that lot for matching Vbe.
The are a couple of reasons for the approach mentioned above, both of which will help to reduce the output offset voltage. First, by matching HFEs of the dif stage, you are going for equal bias current for each side of the diff amp, which will help to keep it balanced. That may be more important than closely matching the Vbe of the two transistors in the diff stage, though that also plays a factor in balancing the diff stage and reducing output offset. .
I use this method for matching transistors for both of my Leach amps:
https://leachlegacy.ece.gatech.edu/lowtim/part2.html
DC offset below 10mV for both...
https://leachlegacy.ece.gatech.edu/lowtim/part2.html
DC offset below 10mV for both...
I've been playing around with the value of the compensation cap in simulation. Response is a little slow with the 68pF original value. Reducing the cap to 39 pF perks things up quite a bit, and the amp still looks stable (nice, overdamped response) with 0.2V square wave excitation. Whether I can actually get away with something like that in a real implementation of this amp remains to be seen. I think a session with the gain-phase analyzer will be revealing.
Last edited:
I was actually able to find some 15000uF, 50V caps at Digi-Key that are cheaper than the ones from Jameco I was looking at, and are considerably shorter than the Apex Jr. caps. I may order up a bunch, as the quantity discount is considerably better than what Jameco offers.
wrenchone:
I bought a Dissipante 3U/400 mm chassis for a Wolverine integrated amp project last year and HiFi2000 drilled and tapped the UMS hole pattern onto its heat sinks for me. The price was quite reasonable and they made my life a lot easier. If your pcbs are adhering to the UMS specification, you might want to add it to your order.
Regards (and good luck with your project),
Scott
I bought a Dissipante 3U/400 mm chassis for a Wolverine integrated amp project last year and HiFi2000 drilled and tapped the UMS hole pattern onto its heat sinks for me. The price was quite reasonable and they made my life a lot easier. If your pcbs are adhering to the UMS specification, you might want to add it to your order.
Regards (and good luck with your project),
Scott