[sdiy] [AH] Re: UA726 clone possible?

[Tom Bugs]

Well, yes, it is a learning project driven by curiousity rather than 
particular need & hopefully I can get some learnings along the way..
My understanding is that the CA/LM3046 has lower specs for the matched 
pair than the LS318 (or similar AD NPN pairs) - though of course it has 
been used in a heater arrangement in quite a few commercial synths over 
the years.
I would think that the heater element needs to produce a stable 
temperature (somewhere above ambient) and this temperature needs to be 
efficectively coupled to the matched pair - I would hope that as long as 
the device is potted (or using the heat transfer epoxy compound I have) 
this will be sufficient.
It could equally be possible to have all components on one side and have 
the PCB overhang the ua726 footprint (the current design already does a 
little).
Cheers, Tom


On 10/09/2015 22:54, Richie Burnett wrote:
>> Anyways, have managed to do a pretty compact test circuit - roughly 
>> 0.5" sqr(ish) with pins in a rough circle - matched pair on top and 
>> other components on the base.
>
> Isn't the whole idea of the uA726 that all of the components are 
> fabricated on the same silicon die so that they're always at the same 
> temperature? You're not going to get all the transistors at the same 
> temperature if they're plastic packaged devices soldered onto 
> different sides of an FR4 board.  Maybe I'm missing something here, 
> but it seems to me that you'd be better off starting with something 
> like a CA3046 transistor array where all devices are fabricated at the 
> same time on a common substrate.
>
> -Richie,
>

-- 
BugBrand LTD
UK company No. 07199808
VAT No. GB 988 2629 57
1 Ninetree Hill
BRISTOL BS1 3SB
United Kingdom
www.bugbrand.co.uk