From: krw on
On Sun, 18 Jul 2010 16:24:03 -0700, John Larkin
<jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:

>On Sun, 18 Jul 2010 17:55:09 -0500, "krw(a)att.bizzzzzzzzzzzz"
><krw(a)att.bizzzzzzzzzzzz> wrote:
>
>>On Sun, 18 Jul 2010 15:51:54 -0700, John Larkin
>><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>On Sun, 18 Jul 2010 14:26:45 -0700, "Joel Koltner"
>>><zapwireDASHgroups(a)yahoo.com> wrote:
>>>
>>>>krw(a)att.bizzzzzzzzzzzz wrote:
>>>>> Sure. The issue is two-fold, really. First, the thing works. A
>>>>> good chunk of the possible savings is a newer DSP (savings of $6-7
>>>>> per), but they had so much trouble with TI getting this one working
>>>>> that the owner is afraid of touching another unknown.
>>>>
>>>>I've often heard that the Analog Devices SHARC DSPs are rather "friendlier"
>>>>and work better (in terms of the actual behavior relative to what the data
>>>>sheet claims)... have anyone there who has the time and inclination to do a
>>>>bit of a skunkworks project?
>>>>
>>>>We're sticking with TI DSPs as well for the moment, though -- not enough spare
>>>>people/time to be contemplating a change right now.
>>>>
>>>>---Joel
>>>
>>>So far, we've been happy with a 68K or an ARM for the slow stuff and
>>>an FPGA for the fast signal processing. That's a different way to
>>>slice peoblems.
>>
>>Different problems, different solutions. Certainly there is an overlap, but
>>it's not all that much.
>
>We just did a Spartan6 design that runs 16 channels of A/D conversion,
>each channel having an SPI interface to its A/D, two programmable
>8-pole lowpass filters, a data interpolator, and a 4K FIFO. The FPGA
>also does our VME interface and some trigger logic that you couldn't
>do in a DSP. Brutal parallelism is nice sometimes.

Right. Different problem, different solution. I'd love to do our whole
widget on an FPGA, but it would be a huge waste.

>Using a separate uP and FPGA partitions the problem nicely, at the
>cost of interconnect pins.

Sure, combining them on the same silicon gets rid of the interconnect problem
while maintaining that separation. Virtex is, or at least was, incredibly
expensive, but there are other alternatives now. ARM on FPGA, complete with
A/D is now available, cheap.
From: krw on
On Sun, 18 Jul 2010 16:29:41 -0700, "Joel Koltner"
<zapwireDASHgroups(a)yahoo.com> wrote:

><krw(a)att.bizzzzzzzzzzzz> wrote in message
>news:ug17465nsqejfu252efn3rm6ehjk8ci7o2(a)4ax.com...
>> On Sun, 18 Jul 2010 15:51:54 -0700, John Larkin
>> <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>>>So far, we've been happy with a 68K or an ARM for the slow stuff and
>>>an FPGA for the fast signal processing. That's a different way to
>>>slice peoblems.
>> Different problems, different solutions. Certainly there is an overlap, but
>> it's not all that much.
>
>...and I don't think any of John's stuff is battery powered, whereas Keith's
>and mine is.

Good point. Our serious number crunching is in the base, though. Add five
copies of the remote, plus echo cancellation, mixing, and all that and there
is a lot an FPGA could do. It wouldn't save anything, however. The DSP is
still needed.

>G.729 is also a pretty complex algorithm; it'd be a awful lot of work to put
>the bulk of it into an FPGA.

Since we bought the function...
From: Phil Hobbs on
On 7/18/2010 4:42 PM, Joerg wrote:
> krw(a)att.bizzzzzzzzzzzz wrote:
>> On Sun, 18 Jul 2010 09:52:00 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>
>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>> On Sat, 17 Jul 2010 17:21:23 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>>>
>>>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>>>> On Sat, 17 Jul 2010 15:56:12 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>>>>>
>>>>>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>>>>>> On Sat, 17 Jul 2010 09:38:50 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>>>>>>>
>>>>>>>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>>>>>>>> On Sat, 17 Jul 2010 09:11:49 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>>>>>>>>>
>>>>>>>>>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>>>>>>>>>> On Sat, 17 Jul 2010 06:33:23 -0700, Joerg<invalid(a)invalid.invalid> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>> Frank Buss wrote:
>>>>>>>>>>>>>> Joerg wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> I believe Newark carries them.
>>>>>>>>>>>>>> Digikey has it, too. Search for "keystone test points" and you'll get a
>>>>>>>>>>>>>> nice selection for diameter, color etc. Looks like it is popular, because
>>>>>>>>>>>>>> some types are available in container quantities :-)
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=5002K-ND
>>>>>>>>>>>>>>
>>>>>>>>>>>>> Thanks, I didn't know they carried these. Unfortunately such test points
>>>>>>>>>>>>> are comparatively large, for 0.040" holes and similar.
>>>>>>>>>>>> Come on Joerg. They're only three for ten bucks. You can afford them!
>>>>>>>>>>>>
>>>>>>>>>>>> <yikes!> BTW, we use something similar for less than $.10 each. ...and those
>>>>>>>>>>>> are usually deleted when production settles down.
>>>>>>>>>>> I didn't say I can't afford them :-)
>>>>>>>>>> ;-)
>>>>>>>>>>
>>>>>>>>>>> They are just a bit large for modern SMT layouts. Technologically a bit
>>>>>>>>>>> long in the tooth.
>>>>>>>>>> We use them on power supplies and a couple of communications lines (RS-422).
>>>>>>>>>> They're useful for scope grounds.
>>>>>>>>>>
>>>>>>>>>> There once were connectors that would accept a scope tip. Has anyone seen
>>>>>>>>>> those recently? They were *expensive* but for test jigs they would be useful.
>>>>>>>>> Not recently. I often use needles, the ones that come with higher end
>>>>>>>>> shirts. You know, where Murphy says that you'll always forget one when
>>>>>>>>> trying it on and ... OUCH. Many can be soldered to (not all) and then
>>>>>>>>> occasionally I use a rubber band to tug a bit on the cable or the needle
>>>>>>>>> so it won't fly out.
>>>>>>>> You mean "pins". Needles have a slot for the thread. ;-)
>>>>>>>>
>>>>>>> Sorry, yes, pins. As you can see it's my wife who does all the sewing
>>>>>>> around here :-)
>>>>>> You think I do? There is a reason SWMBO, MBO. ;-)
>>>>>>
>>>>> But tonight I'm doing the cookin', standing near the charcoal Weber at 100F.
>>>> She left you outside in that condition?
>>>>
>>> As long as there's food that's ok :-)
>>
>> Man does not live by food alone. ;-)
>>
>
> Well, ok, food plus beverages then :-)
>
> [...]
>
>
>>>>> Different here. We rarely run the A/C so I wear short and T-shirts. The
>>>>> office often hovers around 90F. When doing heavy-duty SPICE a lot more.
>>>> I wear them year around. AC or not. I don't own any others.
>>>>
>>> Wow. I'd be drenched in sweat and it would not be a pretty picture.
>>
>> Only when I'm working outside, or in the attic. The attic work likely won't
>> start again until at least September. I thought I might get a break today and
>> work out there, but nope (91F w/50% humidity and sunny). It's like working in
>> the rain. You can only get so wet. ;-)
>>
>
> That's what one guy thought. Until the whole deck gave way and turned
> itself into a raft. He lucked out and got off, somehow, but the deck was
> later found a few miles downstream.
>
>
>>>>>>> It does make sense to buy quality stuff. Cheap shirts can look bad after
>>>>>>> just a dozen washes while good ones last many years.
>>>>>> Yes, but you don't have to spend a lot of money on them.
>>>>>>
>>>>> I'd rather have stuff that lasts.
>>>> Same shirts, different day.
>>>>
>>>>>>> But hey, I knew a guy who had more money that Uncle Scrooge, was
>>>>>>> chauffeured around in a Bentley, yet he usually ate herring and
>>>>>>> potatoes. Now with food, we sometimes splurge. Same with beer, you won't
>>>>>>> find any cheap ones here.
>>>>>> ...but you do go cheap on electronics. ;-)
>>>>>>
>>>>> Yes, there's always a cost calculator running inside my brain. Even when
>>>>> working on hi-rel stuff where cost really doesn't matter, it simply
>>>>> won't turn off. Once a client asked me what the electronics would cost,
>>>>> more casually. "45 bucks" ... some jaws dropped ... "We can't even buy a
>>>>> decent wrench for that, heck, our coffee maker cost more".
>>>> We're on a cost kick now. The owner decided that it's his money.
>>>> Unfortunately, a lot of the decisions were made long ago and aren't easy or
>>>> cheap to change now. It's tough when the rules change after the game is in
>>>> play.
>>>>
>>> Mostly that will result in a complete redesign. That is when the NRE
>>> that entails must be factored in, to see how fast the effort truly
>>> amortizes. Anything north of four years is generally frowned upon.
>>
>> Yep. I'm doing a redesign now, but at least on this pass the ground rules
>> state that no firmware can be changed. That alone cuts the possibilities some
>> 80% (I've already done the design, including firmware changes). The real
>> bigies would hit the case, though. There is well into six figures in the
>> molds so there is no way that's going to pay.
>>
>
> That would put enclosure changes off-limits, unless you make a bazillion
> of these per year. I've often run into this "no SW change" or "no
> firmware change" requirement and never really understood it. There are
> nowadays so many trick you can play in firmware that save beaucoup bucks
> because whole chunks of hardware could simply vanish.
>

Depends whether the original firmware guy is still there,
though...often, getting somebody to change somebody else's working
firmware is a complete can of worms.

OTOH, since you can still make the previous version, at least there's a
revenue stream paying for all that mucking about.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net
From: krw on
On Sun, 18 Jul 2010 20:38:30 -0400, Phil Hobbs
<pcdhSpamMeSenseless(a)electrooptical.net> wrote:

>On 7/18/2010 4:42 PM, Joerg wrote:
>> krw(a)att.bizzzzzzzzzzzz wrote:
>>> On Sun, 18 Jul 2010 09:52:00 -0700, Joerg<invalid(a)invalid.invalid> wrote:
<snip>

>>>> Mostly that will result in a complete redesign. That is when the NRE
>>>> that entails must be factored in, to see how fast the effort truly
>>>> amortizes. Anything north of four years is generally frowned upon.
>>>
>>> Yep. I'm doing a redesign now, but at least on this pass the ground rules
>>> state that no firmware can be changed. That alone cuts the possibilities some
>>> 80% (I've already done the design, including firmware changes). The real
>>> bigies would hit the case, though. There is well into six figures in the
>>> molds so there is no way that's going to pay.
>>>
>>
>> That would put enclosure changes off-limits, unless you make a bazillion
>> of these per year. I've often run into this "no SW change" or "no
>> firmware change" requirement and never really understood it. There are
>> nowadays so many trick you can play in firmware that save beaucoup bucks
>> because whole chunks of hardware could simply vanish.
>>
>
>Depends whether the original firmware guy is still there,
>though...often, getting somebody to change somebody else's working
>firmware is a complete can of worms.

The lead programmer is still around. Actually, they all are with the
exception of the guy who went to the FBI. Chasing terrorists was more fun,
evidently.

>OTOH, since you can still make the previous version, at least there's a
>revenue stream paying for all that mucking about.

It's an opportunity cost issue. That's above my pay grade. I give them the
numbers and they tell me what to work on. Well, I gave them the numbers nine
months ago...
From: Phil Hobbs on
krw(a)att.bizzzzzzzzzzzz wrote:
> On Sun, 18 Jul 2010 20:38:30 -0400, Phil Hobbs
> <pcdhSpamMeSenseless(a)electrooptical.net> wrote:
>
>> On 7/18/2010 4:42 PM, Joerg wrote:
>>> krw(a)att.bizzzzzzzzzzzz wrote:
>>>> On Sun, 18 Jul 2010 09:52:00 -0700, Joerg<invalid(a)invalid.invalid> wrote:
> <snip>
>
>>>>> Mostly that will result in a complete redesign. That is when the NRE
>>>>> that entails must be factored in, to see how fast the effort truly
>>>>> amortizes. Anything north of four years is generally frowned upon.
>>>> Yep. I'm doing a redesign now, but at least on this pass the ground rules
>>>> state that no firmware can be changed. That alone cuts the possibilities some
>>>> 80% (I've already done the design, including firmware changes). The real
>>>> bigies would hit the case, though. There is well into six figures in the
>>>> molds so there is no way that's going to pay.
>>>>
>>> That would put enclosure changes off-limits, unless you make a bazillion
>>> of these per year. I've often run into this "no SW change" or "no
>>> firmware change" requirement and never really understood it. There are
>>> nowadays so many trick you can play in firmware that save beaucoup bucks
>>> because whole chunks of hardware could simply vanish.
>>>
>> Depends whether the original firmware guy is still there,
>> though...often, getting somebody to change somebody else's working
>> firmware is a complete can of worms.
>
> The lead programmer is still around. Actually, they all are with the
> exception of the guy who went to the FBI. Chasing terrorists was more fun,
> evidently.
>
>> OTOH, since you can still make the previous version, at least there's a
>> revenue stream paying for all that mucking about.
>
> It's an opportunity cost issue. That's above my pay grade. I give them the
> numbers and they tell me what to work on. Well, I gave them the numbers nine
> months ago...

Being on my own, and working mainly with startups, I can assure you that
opportunity cost is a much nicer thing to worry about than cash flow!

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net