It appears the supply of LME49710NA ICs is dwindling. Newark/Farnell is now the only distributor that has them in stock.
Thankfully, using the TO-99 can version (LME49710HA) on a DIP-8 footprint is pretty easy. Pin one goes to pin one, two to two, etc. Just bend the leads to make the chip fit. The LME49710HA is still in production.
If the whole lead bending exercise isn't your cup of tea, you can use the LMP8671MA with an SOIC -> DIP adapter (punch in DIP-ADAPTER-EVM in the search bar on Mouser and you shall find a pretty nice one).
Tom
Thankfully, using the TO-99 can version (LME49710HA) on a DIP-8 footprint is pretty easy. Pin one goes to pin one, two to two, etc. Just bend the leads to make the chip fit. The LME49710HA is still in production.
If the whole lead bending exercise isn't your cup of tea, you can use the LMP8671MA with an SOIC -> DIP adapter (punch in DIP-ADAPTER-EVM in the search bar on Mouser and you shall find a pretty nice one).
Tom
They make a plastic adaptor that holds a To99 set of leads and bends them to the Dip8 pattern as you push the leads into and through the adaptor.
The leads can still be soldered to the PCB, or inserted into a dip socket.
the white one on the right
https://www.google.co.uk/search?q=t...w=1420&bih=1103&dpr=0.8#imgrc=gmIk-DU3ttckeM:
The leads can still be soldered to the PCB, or inserted into a dip socket.
the white one on the right
https://www.google.co.uk/search?q=t...w=1420&bih=1103&dpr=0.8#imgrc=gmIk-DU3ttckeM:
Thanks for the heads up. I'm building in stages, and haven't ordered all of my amp boards yet. I did order some LME49710NA from Newark last night. They have 500 something left, and I believe they have more coming in a week or two.
I picked up a tap/dye set last night; and I hope to make time to build chassis, and solder lm3886 to mod86 boards this weekend. I've got lxmini parts on hand too, so I can get started on that. I just need to get one more mod86 board, (then I'll have 4), and I can delete my old a/v amp entirely from my system. This has been a long time coming,..exciting for me.
I picked up a tap/dye set last night; and I hope to make time to build chassis, and solder lm3886 to mod86 boards this weekend. I've got lxmini parts on hand too, so I can get started on that. I just need to get one more mod86 board, (then I'll have 4), and I can delete my old a/v amp entirely from my system. This has been a long time coming,..exciting for me.
Oh, that's right. I forgot about those. Thanks for mentioning them. One could use a small piece of perf board as well.
Newark has been pretty good at keeping product in stock. They were the last to run out of LM4780s as well.
If you need more boards and want everything to be identical, I suggest ordering soon. I have six PAR86 boards, eight LM4780s, and 18 MOD86 R2.0 boards left.
I will do everything in my power to keep the Modulus-86 going. I'm currently working on figuring out which path to take. For an all leaded build, the best option is to use the LME49710HA and just change the footprint to a TO-99. That'd be a Rev. 2.1 and would be the same price as the Rev. 2.0. It may make more sense to use one of the newer op-amps in SMD packages, though.
Whether the solution, the Modulus-86 will still offer world class performance and be the best performing, best documented, best supported DIY amp out there.
The Parallel-86 will sunset with the LM4780 once the remaining board stock runs out.
Tom
Sob. Farnell still has a small stock of 4780s I was going to get 4 in just in case I needed more powa in 12 months time. Ah well c'est la vie
Yeah... C'est la vie. I'm sure I can come up with a moar powah solution for you if it's really needed. Opamp + VAS + output stage would get you there, though, it's certainly more involved to build than the typical chip amp plug-n-play solution.
Tom
I appreciate the emphasis on "should". With composite amps, you really have to be careful with opamp substitutions. The OPA1611 with some component tweaks is my current candidate for a replacement.
You're absolutely right. SOIC-8 is not bad at all, assuming you have the right tools. I wouldn't mind an SMD build, but have received feedback along the lines of "SMD is a showstopper for me" form some builders.
In a noise optimized design, one would need to use 0.25 W resistors due to the low impedances involved, which puts you in 0805 or 1206 packages. Even an SMD build would use sizeable components. I have soldered 0402 by hand and that gets to be a pain. For 0603 and below, I prefer to have a microscope on hand - especially if the end product goes to a customer. 0805 and above, I can solder by naked eye or with the help of reading glasses.
Meanwhile, I've changed the Mouser project to point to an LME49710HA. The show must go on!
Tom
Last edited:
Yeah, I certainly didn't want to be presumptuous, as there's a lot that can go sideways with composites that doesn't show up in the spec sheets and preliminary examinations of their loop gain/phase curves.
I could see a full-on SMD board respin being off-putting to many (even with 0805/1206 parts), which is why I was wondering if any of the soic's you're considering could be brought in on a dip-8->soic-8 daughter/adapter board without too much negative recourse (local ps bypassing on the adapter to keep it super close to the pins?). Might help mitigate some buyers reluctance if you/someone can provide an already-soldered solution (at an understandable surcharge).
Either way, I'm thinking aloud and you obviously know your beans.
I could see a full-on SMD board respin being off-putting to many (even with 0805/1206 parts), which is why I was wondering if any of the soic's you're considering could be brought in on a dip-8->soic-8 daughter/adapter board without too much negative recourse (local ps bypassing on the adapter to keep it super close to the pins?). Might help mitigate some buyers reluctance if you/someone can provide an already-soldered solution (at an understandable surcharge).
Either way, I'm thinking aloud and you obviously know your beans.
With the more exciting parts being SMD, I see two paths:
One would be a hybrid design (for lack of a better term). So mostly leaded with a couple of SMD bits. If it's a matter of 1-2 ICs, I'd just offer to solder them in exchange for a cup of coffee (or its monetary equivalent).
The other path would be an all-out SMD design. For that, I'd need to get the assembly done elsewhere to get the cost down. It would be one of those where you could buy the board with all the SMDs on it and assemble the rest, or I could deliver assembled and tested modules.
Getting "DIP" ICs made from SOIC + adapter is probably going to be as expensive as getting the boards assembled.
For the near term, using the TO-99 LME49710HA makes the most sense.
Either way, I'm thinking aloud and you obviously know your beans.
No worries. I like to think I know my stuff anyway...
Tom
Tom FYI I just ordered parts for my build (2 Power 86 + 2 Parallel 86 + 2 Modulus 86) and Mouser had the LME47910NA in stock for immediate shipping with an available quantity on hand of 1,101. The only not available part I encountered was the LM4780 which I had previously sourced from Newark. The only item out of stock and more than a couple weeks lead-time I ran into were the Power 86 bridge rectifiers projected for late-April.
I really appreciate the heads up on limited qty. of Mod86 rev2.0 boards. I just ordered the other one I'll need. I'm building LXminis, so it would kinda suck to not have the 4 channels matching.
The next pair of amps I'll need will be for subwoofer duty, but that's over a year away. I'm thinking right now of HFrames (or UFrame) for subs but I may change my mind. I'll do a practice run on hframe idea with some 15" woofers I have on hand, and temporarily they can run on my first build, a six year old lm3875 which still works well. Someday though, moa powa for the subs will probably be desired.
The next pair of amps I'll need will be for subwoofer duty, but that's over a year away. I'm thinking right now of HFrames (or UFrame) for subs but I may change my mind. I'll do a practice run on hframe idea with some 15" woofers I have on hand, and temporarily they can run on my first build, a six year old lm3875 which still works well. Someday though, moa powa for the subs will probably be desired.
No argument here. I've yet to build a reflow oven. I still hand-assemble my SMD builds.
If you just ordered (as in yesterday or today), you'll get the LME49710HA (TO-99), which is still in production. The NA is available at Newark and Mouser has 260 of the MA (SOIC) in stock. Mouser is out of the NA (DIP).
Yeah. That's the thing with Mouser. They run out of stuff. Digikey does a better job at keeping stock on hand but most DIYers prefer Mouser. Digikey doesn't carry the THAT parts. Tradeoffs, tradeoffs...
The GBU2504 (Mouser: 750-GBU2504-G) is a good substitute.
I hear ya. Thanks for your order. Your board will be underway by tonight.
Tom
Tom,
Ah, so you updated the BOM...looks like I'll be doing some careful and triple checked lead bending then!
Yeah... See the last sentence of Post 1408.
Rather than fielding a bunch of support email letting me know that Mouser is out of the NA, I decided to change the BOM. That way people get to make progress on their projects and I get to make progress on mine.
I suspect there'll be a few questions about how to fit the round peg in the square hole, but I'll field those when they occur.
I'll add a more prominent note on my website so the expectations are communicated a bit clearer up front.
Tom
Finally finished wiring everything up, a quick test with the light bulb and everything is fine, so onto voltage measurements across J3. I get the following measurements: -15.1V/14.8V on one board, and -15.0V/14.6 V on the second board.
According to the build instructions, measurements should be within a few percent of +/- 15V - I guess mine are ok to carry on?
According to the build instructions, measurements should be within a few percent of +/- 15V - I guess mine are ok to carry on?
I got impatient and figured that my voltages were ok, and progressed to the next step. I only have a Voltmeter so I used the 200mV setting and measured 0.00 mV on one channel, and 0.00 - 0.01 mV on the other (it was mainly at 0.01 mV, but occasionally went to zero.) Both of these were after about 10 seconds, and I watched the initial 20-30 mV come rapidly down as the DC servo kicked in. Anyway, both of those measurements would seem to check out.
I don't have any test speakers so I will definately wait unti someone says my measurements check out before connecting to my main speakers.
I don't have any test speakers so I will definately wait unti someone says my measurements check out before connecting to my main speakers.
I'm guessing those voltages are measured on U4 rather than J3, right? J3 (power connector) should have ±28 V. On U4, pin 4 you should see -15 V (within a few percent) and on U4, pin 8, you should see +15 V (within a few percent).
Assuming my guess on your supply measurements is correct, it sounds like everything is working properly.
Tom