"The Wire AMP" Class A/AB Power Amplifier based on the LME49830 with Lateral Mosfets

[wirewiggler]

Owen,
No point in sweating about what is out of your control, it changes nothing other than making yourself stressed and there are enough things in your control for that. In any case since my order is grouped with the headphone amp and the other boards I would like to get started on them sooner, but don't bust a nut over it when it is convenient for you. When you get the laterals in let me know how much for freight and I will through in a few extra bucks so you can buy yourself a beer. Many thanks for your hard work.

Bill

[HYPERTUNE]

Owen,

The only concern I have regarding the delay is that it might put you off doing more projects like this in future.

Please don't let that be the case!

I'd like to have the rest of my parts sent if possible as it will help with the heatsink/chassis designs. I'll have some parts to send you by the time the fets turn up and we can trade shipping costs if it suits you.

[npdot]

For anyone interested in using the 2SK1058/2SJ162 lateral power MOSFET devices as an alternative to the double die output devices currently on back order, I offer the following information from my personal experience.

About a year ago I purchased a bunch of 2SK1058 AND 2SJ162 on eBay for about $3.25 each. (See http://www.ebay.com/itm/1-pcs-2SJ162-1-pcs-2SK1058-P-Channel-N-Channel-MOSFET-160V-7A-Free-Shipping-/400268172940?pt=LH_DefaultDomain_0&hash=item5d31d7 9e8c.) Out of concern that they might be fakes, I also bought several from a big local distributor in the US, and tested all of them against the output characteristics in the datasheet. I was happy to find that all were consistent with each other and with the datasheet. In all, 12 of each N and P channel were tested over the entire SOA in 1Vgs steps from 1-10 Vgs (up to 50Vds for 1-3Vgs). Although they were very rugged they were not indestructible.

The project was to build a 4 channel amp using a modified version of Rod Elliot’s P-101 circuit. I purchased some boards from him and then made a new layout to permit a different physical chassis requirement. The testing permitted using matching pairs of outputs. With a few tweaks, the P-101 design was quite respectable (see attached performance chart).

As Owen points out, the Renesas devices can serve as an exact replacement for the ACD101NDD and ACD103PDD double die lateral MOSFETs up to about 125W. Beyond that you will need to use two pairs. For two pairs you will need to add a source resistor (to force current charing) for each output and maybe do a little tweaking on the gate resistor values and compensation. Owen? The challenge with using two pairs of outputs will include how to add the additional components and connect them to Owen’s PCB.

Furthermore, since Owen has already paid for the DD devices, changing our orders now and going with the Renesas devices will cause Owen a lot of additional work.

I’m sticking with option 2.

All the best,
Neel

[qusp]

I was under the impression that laterals did a pretty decent job of current sharing on their own? not speaking from experience as i havent tried it, just some reading i did the other night seemed to suggest that with 2 parallel devices on the same heatsink, with one devices biased directly, the other would mirror the response. it was late so maybe i misunderstood; seemed a curious effect.

[npdot]

qusp,

I've read the same thing about current sharing with parallel laterals, but never tried it. Nobody seems to be recommending it and I can't recall a single design without source resistors in a double pair OPS. What do the double die laterals look like under the skin? This is an interesting subject to research. Maybe Bob Cordell or one of the others at his level could enlighten us.

Cheers,
Neel

[qusp]

yeah i thought i had, well if you think about it, the double dies are just 2 parallel devices in the one package afaik, no room for a second resistor there is there? this would seem to be in line with that. Andrew T probably knows, being the fountain of knowledge he usually is

[holminator]

I'm good with option 2

[leejh]

#2 is fine for me.
Thanks.

[arthur]

I prefer to opt #2

[exa065]

I'm good with option 2