EMRFD Message Archive 4095

Message Date From Subject
4095 2010-01-29 14:31:44 steve white CA3055 ??
I ran across several RCA CA3055's in a one of the bins that I have and I
have searched google but to not avail. I can not find any info on this
device. Does anyone know what it is or even if it is worth keeping?



Steve NU0P



[Non-text portions of this message have been removed]
4096 2010-01-29 14:42:47 Chris Trask Re: CA3055 ??
>
> I ran across several RCA CA3055's in a one of the bins that I have and I
> have searched google but to not avail. I can not find any info on this
> device. Does anyone know what it is or even if it is worth keeping?
>

It's a somewhat noisy voltage regulator. It shows up in the 1970 RCA
Linear IC databook but not in the 1975 edition. Obviously not a popular
part, probably due to the noise (0.45-0.7mV) and poor internal voltage
reference (1.4-1.8V).

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4097 2010-01-29 14:43:35 Sam Morgan Re: CA3055 ??
steve white wrote:
> I ran across several RCA CA3055's in a one of the bins that I have and I
> have searched google but to not avail. I can not find any info on this
> device. Does anyone know what it is or even if it is worth keeping?
>
this one?

http://pdf1.alldatasheet.com/datasheet-pdf/view/134072/GESS/RCA3055.html



--
GB & 73
K5OAI
Sam Morgan
4098 2010-01-29 15:04:27 Lasse Re: CA3055 ??
Nope, thats the transistor 2N3055...
CA3055 is another animal ...

Sam Morgan skrev:
> steve white wrote:
>
>> I ran across several RCA CA3055's in a one of the bins that I have and I
>> have searched google but to not avail. I can not find any info on this
>> device. Does anyone know what it is or even if it is worth keeping?
>>
>>
> this one?
>
> http://pdf1.alldatasheet.com/datasheet-pdf/view/134072/GESS/RCA3055.html
>
>
>
>
4099 2010-01-29 16:34:21 steve white Re: CA3055 ??
Thanks Chris



I know what to do with them now. In the dumpster.



steve



_____

4100 2010-01-29 17:09:03 Chris Trask Re: CA3055 ??
>
> Thanks Chris
>
> I know what to do with them now. In the dumpster.
>
> steve
>

Eh!! Maybe put them in a bird feeder or use them for target practice.

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4101 2010-01-29 21:34:25 dave Re: CA3055 ??
The Ca3055 is supposed to be a variable positive voltage regulator.
Dave

4102 2010-01-30 04:40:46 Chris Trask Re: CA3055 ??
>
> The Ca3055 is supposed to be a variable positive voltage regulator.
> Dave
>

Nice to see that you read my earlier posting about it.

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4104 2010-01-30 22:58:49 dave Re: CA3055 ??
Hmmm, what dumpster are you throwing them in. lol
Dave

4106 2010-01-31 05:52:59 Chris Trask Re: CA3055 ??
Probably the same one with the banjo and the accordian. :{b

Chris

-----Original Message-----
>
4111 2010-01-31 06:55:44 Andy Re: CA3055 ??
> http://pdf1.alldatasheet.com/datasheet-pdf/view/134072/GESS/RCA3055.html

That webpage did give me a good chuckle, though:

"Your require pages is cannot open by blow"
"Reason : Connect this pages through directly deep link"

Gee, I wonder what was their native language is?

I used to collect spam emails with horrible English translations.
Sometimes even spam can be funny.

It's also slightly amusing that the "RCA3055" was a G.E. device,
according to Alldatasheet.com ... even though the CA3055 is an RCA
linear part number.

Is there any "collection" value in old parts like this? I wouldn't
think this one would be worth much ... but on the other hand when I
think about some of the parts I've seen trashed over the years, and
now wish I could have them back ....

Andy
4112 2010-01-31 07:07:43 Chris Trask Re: CA3055 ??
>

>
> It's also slightly amusing that the "RCA3055" was a G.E. device,
> according to Alldatasheet.com ... even though the CA3055 is an RCA
> linear part number.
>
> Is there any "collection" value in old parts like this? I wouldn't
> think this one would be worth much ... but on the other hand when I
> think about some of the parts I've seen trashed over the years, and
> now wish I could have them back ....
>

I wouldn't bother with that one, however I highly encourage people who indulge in discrete design to acquire any and all of the CA3xxx series transistor arrays, especially the CA3028, CA3046, CA3146, CA3054, CA3096, and CA3127. Once you've experienced the luxury of having multiple matched devices on a single die, you'll wonder how you managed to get along without them.

My first feedback mixers used the CA3054. The CA3046 can be used to make a complete superhet receiver (a fun project), and the CA3096 can make a very useful double-balanced phase detector with rail-to-rail output (a serious project).


Chris Trask
N7ZWY
WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4113 2010-01-31 07:10:40 David Craggs Re: CA3055 ??
And yours is ......?
"Gee, I wonder what was their native language is?"



[Non-text portions of this message have been removed]
4114 2010-01-31 07:29:29 Thomas S. Knutsen Re: CA3055 ??
Sigthly re-arranged for
2010/1/31 Andy <ai.egrps@gmail.com>
>
>
> It's also slightly amusing that the "RCA3055" was a G.E. device,
> according to Alldatasheet.com ... even though the CA3055 is an RCA
> linear part number.
>

RCA was once owned or an part of G.E. If you read Harold Beverage's
biography there are more about RCA early history.


>
> Is there any "collection" value in old parts like this? I wouldn't
> think this one would be worth much ... but on the other hand when I
> think about some of the parts I've seen trashed over the years, and
> now wish I could have them back ....
>

As Mr. Trask has said there are some transistor array that I would save.
Those are nice when trying to make an gilbert cell.


>
> >
> http://pdf1.alldatasheet.com/datasheet-pdf/view/134072/GESS/RCA3055.html
>
> That webpage did give me a good chuckle, though:
>
> "Your require pages is cannot open by blow"
> "Reason : Connect this pages through directly deep link"
>
> Gee, I wonder what was their native language is?
>
> Jeg snakker Norsk, hva med deg?


73's de Thomas LA3PNA.
--

Please avoid sending me Word or PowerPoint attachments.
See <http://www.gnu.org/philosophy/no-word-attachments.html>
Mike Ditka <http://www.brainyquote.com/quotes/authors/m/mike_ditka.html> -
"If God had wanted man to play soccer, he wouldn't have given us arms."


[Non-text portions of this message have been removed]
4115 2010-01-31 08:07:23 ehydra Re: CA3055 ??
Yes. And matched JFET, or current mirrors, e.g. BAV61 62.


- Henry


--
ehydra.dyndns.info



Chris Trask schrieb:
> >
>
>> It's also slightly amusing that the "RCA3055" was a G.E. device,
>> according to Alldatasheet.com ... even though the CA3055 is an RCA
>> linear part number.
>>
>> Is there any "collection" value in old parts like this? I wouldn't
>> think this one would be worth much ... but on the other hand when I
>> think about some of the parts I've seen trashed over the years, and
>> now wish I could have them back ....
>>
>
> I wouldn't bother with that one, however I highly encourage people who indulge in discrete design to acquire any and all of the CA3xxx series transistor arrays, especially the CA3028, CA3046, CA3146, CA3054, CA3096, and CA3127. Once you've experienced the luxury of having multiple matched devices on a single die, you'll wonder how you managed to get along without them.
>
> My first feedback mixers used the CA3054. The CA3046 can be used to make a complete superhet receiver (a fun project), and the CA3096 can make a very useful double-balanced phase detector with rail-to-rail output (a serious project).
>
4116 2010-01-31 08:26:46 Chris Trask Re: CA3055 ??
>
>Yes. And matched JFET, or current mirrors, e.g. BAV61 62.
>

Are you certain about that number? Google comes back with that being a mixer diode. Monolithic matched JFETs would be nice to have around.


Chris Trask
N7ZWY
WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4117 2010-01-31 08:29:51 Andy Re: CA3055 ??
> And yours is ......?
> "Gee, I wonder what was their native language is?"

:-)

Yes, the brain is not working yet today!

Thanks for that, I deserved it. ;-) ;-)

Andy
4118 2010-01-31 08:52:56 ehydra Re: CA3055 ??
Annoying the man who won't talk to me.

It was wrongly memorized. My harddisk now gave back BCV61 . Sorry for my
mistake.

- Henry


--
ehydra.dyndns.info


Chris Trask schrieb:
>> Yes. And matched JFET, or current mirrors, e.g. BAV61 62.
>>
>
> Are you certain about that number? Google comes back with that being a mixer diode. Monolithic matched JFETs would be nice to have around.
>
4119 2010-01-31 08:58:56 Chris Trask Re: CA3055 ??
>
> >
> > Yes. And matched JFET, or current mirrors, e.g. BAV61 62.
> >
>
> Are you certain about that number? Google comes back with
> that being a mixer diode. Monolithic matched JFETs would be
> nice to have around.
>

Digi-Key carries the PMBFJ620, which is a dual N-channel JFET suitable for VHF applications. Not the most linear, but it appears to be very much equivalent to a dual J310. I believe I will add some of these to my bench stock.


Chris Trask
N7ZWY
WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4120 2010-01-31 09:11:48 Chris Trask Re: CA3055 ??
>
> >
> > > Yes. And matched JFET, or current mirrors, e.g. BAV61 62.
> > >
> >
> > Are you certain about that number? Google comes back
> > with that being a mixer diode. Monolithic matched JFETs
> > would be nice to have around.
> >
>
> Annoying the man who won't talk to me.
>
> It was wrongly memorized. My harddisk now gave back BCV61 . Sorry
> for my mistake.
>

It's no wonder that I haven't seen this before. NXP/Philips uses the term "double transistor" rather than "dual transistor", and it appears that they have quite a few offerings of both dual (double) and complementary pair monolithic arrays, such as the BC846/7 series. Toshiba, though, has a wider selection available.


Chris Trask
N7ZWY
WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4122 2010-01-31 10:38:48 ehydra Re: CA3055 ??
I'm not sure, but the chips seem to be outdated at the distris??

- Henry


--
ehydra.dyndns.info


Chris Trask schrieb:
> I wouldn't bother with that one, however I highly encourage people who indulge in discrete design to acquire any and all of the CA3xxx series transistor arrays, especially the CA3028, CA3046, CA3146, CA3054, CA3096, and CA3127. Once you've experienced the luxury of having multiple matched devices on a single die, you'll wonder how you managed to get along without them.
>
> My first feedback mixers used the CA3054. The CA3046 can be used to make a complete superhet receiver (a fun project), and the CA3096 can make a very useful double-balanced phase detector with rail-to-rail output (a serious project).
>
4126 2010-02-01 06:16:22 Tim Re: CA3055 ??
4127 2010-02-01 06:27:38 Tim Re: CA3055 ??
CA3046 I think officially went out of production a couple years ago.
Many of the others were probably discontinued before that.

Last time I checked Jameco still had old-stock of CA3046 in DIP. I bought a bunch just a few years back and they had mid-70's date codes.

LM3046 is surface-mount and seems to be current production by National.

Intersil has some higher-frequency matched surface-mount transistor arrays with similar part numbers but way better specs e.g. HFA3046. These sorts of parts often have fT's in the GHz and I use them in some instrumentati
4128 2010-02-01 06:46:47 Chris Trask Re: CA3055 ??
>
> > I wouldn't bother with that one, however I highly encourage people
> > who indulge in discrete design to acquire any and all of the CA3xxx
series
> > transistor arrays, especially the CA3028, CA3046, CA3146, CA3054,
CA3096,
> > and CA3127. Once you've experienced the luxury of having multiple
matched
> > devices on a single die, you'll wonder how you managed to get along
without
> > them.
> >
> > My first feedback mixers used the CA3054. The CA3046 can be used
to
> > make a complete superhet receiver (a fun project), and the CA3096 can
make
> > a very useful double-balanced phase detector with rail-to-rail output (a
> > serious project).
>
> I agree, the CA3046 is a cool and exquisitely utilitarian part.
>

I have yet to run out of circumstances where it or the CA3096 turns out
to be the ideal part. Wavetek used the CA3096 in all of their earlier
sweepers for the temperature-compensated log sweep generator.

>
> While some distributors still have some DIP CA3046 in stock, I think
officially
> it has been out of production for a couple of years, with the exception of
the
> National Semi LM3046, which is still made in surface mount packages.
>

Lots of the DIP packaged parts ar available on eBay, though some sellers
are asking outrageous prices.

>
> Intersil has some higher-frequency matched transistor arrays that are
current
> production. e.g. HFA3046. In my use the transistors in a HFA3046 are
better
> matched than I ever saw in any CA3046. Of course it costs a lot more too!
This
> kinda prevents it from achieving the same jellybean status that the CA3046
had.
>

The HFA series are actually pretty nice. Slightly better linearity and
NF, and the fT is 1GHz, which is even better than the very rare CA3246
device. But, the Vceo is much lower than the CA series.

>
> The moral (but not matchingwise) equivalent in MOS transistors is the
CD4007.
> Which despite it having a CMOS logic-series part number, should usually
not be
> thought of as a digital part!
>

Gosh, I haven't heard of that part in years. You could make a nice
trio of push-pull series-shunt amplifiers with that. It's too bad the gates
of the complementary pairs are connected together, otherwise it would make
an interesting pair of current mirrors that I would have an immediate use
for. Instead, I'll use the CA3096 for what I need presently.

Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4129 2010-02-01 07:04:38 Chris Trask Re: CA3055 ??
>

> CA3046 I think officially went out of production a couple years ago.
> Many of the others were probably discontinued before that.
>
> Last time I checked Jameco still had old-stock of CA3046 in DIP. I bought
> a bunch just a few years back and they had mid-70's date codes.
>

I just looked and they still have them. At 89¢ each in small
quantities, they're a steal!

>
> LM3046 is surface-mount and seems to be current production by National.
>
> Intersil has some higher-frequency matched surface-mount transistor arrays
> with similar part numbers but way better specs e.g. HFA3046. These sorts
of
> parts often have fT's in the GHz and I use them in some instrumentation
front
> ends. They aren't nearly as cheap as the CA3046 was.
>

Digi-Key carries that entire series, and yes, they are pricey. But, the
HFA3134 (dual NPN and HFA3135 (dual PNP) would make really nice
single-balanced mixers at VHF and UHF, maybe into L-Band.

Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4130 2010-02-01 07:13:14 ehydra Re: CA3055 ??
Don't forget the HFA3102

- Henry


--
ehydra.dyndns.info



Chris Trask schrieb:

>> Intersil has some higher-frequency matched surface-mount transistor arrays
>> with similar part numbers but way better specs e.g. HFA3046. These sorts
> of
>> parts often have fT's in the GHz and I use them in some instrumentation
> front
>> ends. They aren't nearly as cheap as the CA3046 was.
>>
>
> Digi-Key carries that entire series, and yes, they are pricey. But, the
> HFA3134 (dual NPN and HFA3135 (dual PNP) would make really nice
> single-balanced mixers at VHF and UHF, maybe into L-Band.
4131 2010-02-01 07:22:40 k5nwa Re: CA3055 ??
Some CA3046 at a decent price cosidering it's obsolete.

<
http://www.goldmine-elec-products.com/prodinfo.asp?number=A20452&variation= >

At 08:15 AM 2/1/2010, you wrote:
>
>
>
4132 2010-02-01 07:44:47 k5nwa Re: CA3055 ??
They also have CA3083 for half the price, I
bought 25 of each to put in my stock.

Lately I've been thinking of getting some
transistor arrays, the thread motivated me to get some.

I assume they also make some with complementary matched transistors?


At 09:20 AM 2/1/2010, you wrote:
>
>
>Some CA3046 at a decent price cosidering it's obsolete.
>
><
><http://www.goldmine-elec-products.com/prodinfo.asp?number=A20452&variation=>http://www.goldmine-elec-products.com/prodinfo.asp?number=A20452&variation=
> >
>
>At 08:15 AM 2/1/2010, you wrote:
> >
> >
> >
4133 2010-02-01 07:59:49 Chris Trask Re: CA3055 ??
>
> Some CA3046 at a decent price cosidering it's obsolete.
>
> <
>
http://www.goldmine-elec-products.com/prodinfo.asp?number=A20452&variation=
>
>

Jameco is cheaper at 89¢ each. But, Goldmine has some really
interesting parts in their miscellaneous IC category.

Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4134 2010-02-01 08:16:35 Paul Re: CA3055 ??
I ran across this e-bay sale of a tube of 25 CA3054 for $39.95 (7 tubes available).

Ebay item number: 350278139554
Name: CA3054 ICs IC NEW RCA DUEL AMPLIFIERS 3054
Ebay store: Old School Surplus

Have to love the spelling of dual...

These are DIP, the only others I've found were SMT.

Thinking of buying a tube, but I'll never use all 25.

I've never dealt with this seller, so buyer beware.

The more easy to find arrays seem to be 5 transistors with one diff-connected pair.

Chris -- Isn't this the dual differential amp chip you used in your feedback mixer paper?



4139 2010-02-01 14:05:30 Chris Trask Re: CA3055 ??
Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4151 2010-02-02 10:05:27 Paul CA3054
I ended up making the seller an offer at $25 and he accepted.

So
4206 2010-02-05 19:25:23 Chris Trask Re: CA3054
>
> I ended up making the seller an offer at $25 and he accepted.
>
> Soon I'll have 25 of these, which is probably 20 more than I'll need.
>
> If anybody wants a few, I'll let them go for a buck apiece plus actual
postage
> costs to get it to you (your choice of USPS or UPS).
>
> I'll eat the $13 it cost to ship them me.
>
> Post a quick reply if interested and I'll work out the details one they
arrive
> (and I test a sample).
>

Whoa!! I was cruising eBay and ran across a listing for the CA3146 NOS
plastic DIP package being sold for just $0.25 each. He was selling them at
4/$1.00 and he had 36 pieces total. Smoke poured from the keyboard as I
cleaned him out and the bill came to $12.80 with postage.

The CA3146 is the 40V version of the CA3046, and the other specs remain
pretty much the same. The CA3246 was the lower voltage 3GHz version.

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4209 2010-02-06 08:47:19 Paul Re: CA3054
Saw these -- guy from Ohio if I recall. Same guy had a table at a Missouri hamfest in October and I bought a single 4-pack of the 3146 for $1.00. I didn't have a data book and was working from memory.
Also bought a bunch of op-amps. He seemed to specialize in $1.00 bags of IC's (either 1 2 or 4 per bag, depending on the part).

I was wondering if the matched nature of these transistors would allow me to parallel them directly for use in a 1W QRP cw final (maybe with a DIP heatsink
4211 2010-02-06 09:16:02 k5nwa Re: CA3054
At 10:46 AM 2/6/2010, you wrote:
>
>
>Saw these -- guy from Ohio if I recall. Same guy had a table at a
>Missouri hamfest in October and I bought a single 4-pack of the 3146
>for $1.00. I didn't have a data book and was working from memory.
>Also bought a bunch of op-amps. He seemed to specialize in $1.00
>bags of IC's (either 1 2 or 4 per bag, depending on the part).
>
>I was wondering if the matched nature of these transistors would
>allow me to parallel them directly for use in a 1W QRP cw final
>(maybe with a DIP heatsink
4212 2010-02-06 09:46:26 Chris Trask Re: CA3054
>
> > Whoa!! I was cruising eBay and ran across a listing for the CA3146
NOS
> > plastic DIP package being sold for just $0.25 each. He was selling them
at
> > 4/$1.00 and he had 36 pieces total. Smoke poured from the keyboard as I
> > cleaned him out and the bill came to $12.80 with postage.
> >
> > The CA3146 is the 40V version of the CA3046, and the other specs
remain
> > pretty much the same. The CA3246 was the lower voltage 3GHz version.
>
>
> Saw these -- guy from Ohio if I recall. Same guy had a table at a
Missouri
> hamfest in October and I bought a single 4-pack of the 3146 for $1.00. I
> didn't have a data book and was working from memory.
> Also bought a bunch of op-amps. He seemed to specialize in $1.00 bags of
> IC's (either 1 2 or 4 per bag, depending on the part).
>

It was a smokin' deal, and I've added him to my short list of preferred
sellers.

>
> I was wondering if the matched nature of these transistors would allow me
to
> parallel them directly for use in a 1W QRP cw final (maybe with a DIP
heatsink
> on top?).
>

Yes, I would use one device (Q5) as a biasing reference diode and then
run the remaining four as a pair of push-pull Class AB (linear) or Class B
(CW), and you should be able to get 1W peak output power with a 12V supply,
possibly 1W average if you use the CA3146 with a 15V supply. That's
something I would really like to see, as you could use a second CA3046 as an
exciter with a buffered VFO or XO (3 transistors), a keyed driver stage (1
transistor), and use the last device to temperature stabilize the VFO/XO or
as a keyer for the driver stage.

Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4213 2010-02-06 10:23:16 Chris Trask Re: CA3054
Now here's a thought: I was looking at the Super Simple Superhet, and
realized that my earlier CA3046 Super Flex Het would do the same thing with
just the five CA3046 transistors and a diode or two tossed in for a product
detector. How about a late winter project of making a transceiver with two
or three CA3046 devices? I'm really tied up with my DDS/PLL synthesizer for
now, but I can see that we have at least a couple of people who could pull
this off.

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4221 2010-02-08 12:41:29 Paul Re: CA3054
Regarding using the CA3146 as a QRP final amp.

I ran across the following page:

http://www.qrp4u.de/docs/en/rf_ampl/index.htm

The secti
4222 2010-02-08 12:51:45 Chris Trask Re: CA3054
>
> Regarding using the CA3146 as a QRP final amp.
>
> I ran across the following page:
>
> http://www.qrp4u.de/docs/en/rf_ampl/index.htm
>
> The section at the very bottom seems to match your description.
>
> I was thinking of using that as a starting point (but with 3146 instead of
3083).
>

Yes, that's very close. I cringe at putting transistors in parallel,
especially for power amplifier designs. I was thinking of two push-pull
amplifiers in parallel to play it safe.


Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4223 2010-02-08 13:38:29 Paul Re: CA3054
Thanks Chris,

> I cringe at putting transistors in parallel, especially for power amplifier designs.
> I was thinking of two push-pull amplifiers in parallel to play it safe

OK here is where I'm going to fully expose my ignorance. I'm not an EE, but I'm soaking up theory as fast as I can (primarly with EMRFD and Horowitz and Hill).

I understand the core issue about putting power transistors directly in parallel giving mismatched currents, giving dead transistors.
I was niavely hoping that having matched transistors was going to make this OK.

I'm a little stuck at how to parallel the two push pull amplifiers without ending up with the transistors in parallel.

The "push-side" emitter is of course connected directly to the "pull-side" emitter, as this is dictated by the internal Q1+Q2 connection in the CA3146.

I'm also pretty confident that the biasing would remain as drawn: assuming we are talking about a single 5 transistor array, with one transistor as a biasing diode. The "push-side" bases are connected together, as are the "pull-side" bases -- there doesn't seem to be another way.

Were you suggesting using a small (few ohms) resistor between the connected emitters of the each of the two separate push-pull amps and ground?

Or you do you mean somehow combing the collector outputs indirectly -- this is the possibility that I can't quite get my head around.

Or am I completely missing another alternative?

Thanks in advance for you patience.

Paul
4224 2010-02-08 14:30:00 Chris Trask Re: CA3054
>
> > I cringe at putting transistors in parallel, especially for power
amplifier designs.
> > I was thinking of two push-pull amplifiers in parallel to play it safe
>
> OK here is where I'm going to fully expose my ignorance. I'm not an EE,
but I'm
> soaking up theory as fast as I can (primarly with EMRFD and Horowitz and
Hill).
>
> I understand the core issue about putting power transistors directly in
parallel
> giving mismatched currents, giving dead transistors.
> I was niavely hoping that having matched transistors was going to make
this OK.
>
> I'm a little stuck at how to parallel the two push pull amplifiers without
ending
> up with the transistors in parallel.
>
> The "push-side" emitter is of course connected directly to the "pull-side"
emitter,
> as this is dictated by the internal Q1+Q2 connection in the CA3146.
>
> I'm also pretty confident that the biasing would remain as drawn:
assuming we are
> talking about a single 5 transistor array, with one transistor as a
biasing diode.
> The "push-side" bases are connected together, as are the "pull-side"
bases -- there
> doesn't seem to be another way.
>
> Were you suggesting using a small (few ohms) resistor between the
connected
> emitters of the each of the two separate push-pull amps and ground?
>
> Or you do you mean somehow combing the collector outputs indirectly --
this is the
> possibility that I can't quite get my head around.
>
> Or am I completely missing another alternative?
>
> Thanks in advance for you patience.
>

Basically, take his circuit and remove one transistor from each side, so
that you end up with an inpout transformer, two transistors, and an output
transformer. Now, duplicate that with the remaining two transistors that
you took out of his circuit. Then, connect the live end of the transformer
inputs and outputs together. This avoids having to make a power splitter
for the input and a power combiner for the output, which is overkill for
this.

The output tuning capacitor can now be a single variable and fixed
capacitor as he has, just connect them from the output side of the
transformer to ground so as to avoid having two variables and two fixed
capacitors.

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4225 2010-02-08 14:32:34 Nick Kennedy Re: CA3054
FWIW, my version of the "autodyne" WWV / 30 meter receiver was done with a
CA3046 array.

http://sites.google.com/site/kennnick/autodynereceiver

72-

Nick, WA5BDU


[Non-text portions of this message have been removed]
4226 2010-02-08 16:02:36 Paul Re: CA3054
So close and yet so far. I apologize for being obtuse.

> Basically, take his circuit and remove one transistor from each side, so that you end up with an input transformer, two transistors, and an output transformer.

Check.

> Now, duplicate that with the remaining two transistors that
you took out of his circuit.

OK -- but now I've drawn a 6th transistor (biasing diode) transistor that a single CA3046 does not have. I'll have to get rid of this when we re-combine...

> Then, connect the live end of the transformer inputs and outputs together. This avoids having to make a power splitter for the input and a power combiner for the output, which is overkill for this.

Your use of "live end" in this context is probably obvious to everybody else, and no doubt should be to me. But lets see if I can figure it out:

There are 4 raw (if not sane) possiblities:

BTW: Just to make sure I don't add any confusion: I'm always using primary and secondary in the sense of the signal flow (primary=left and secondary=right as drawn).

1) live end always means primary -- doesn't seem right as it leaves me with an unused output secondary.

2) live end always means secondary -- doesn't seem right as it leaves me with an unused input primary.

3) live end means the "inside" (i.e. the input secondary and output primary). Can't be right, because then I'm back to having transistors in parallel.

4) live end means the "outside" (i.e. the input primary and the output secondary). Winner by elimination? This looks promising, as it is like putting the two complete amps in parallel (your original comment). It also makes the most sense with your final comment regarding the single variable capacitor.

My only hang-up with 4) is that to get back to 5 transistors (assuming the whole time that we've been talking about a single CA3046), I must also connect the input secondary center taps together to apply bias, and the output primary center taps to the +DC rail via a choke. Once I've done that, aren't the transistors are in parallel again (at least from a DC point of view)? Is this OK?


If you've stayed with me this long I salute you!


Paul
4227 2010-02-08 16:12:28 Chris Trask Re: CA3054
>
> > Then, connect the live end of the transformer inputs and outputs
together.
> > This avoids having to make a power splitter for the input and a power
> > combiner for the output, which is overkill for this.
>
> Your use of "live end" in this context is probably obvious to everybody
else,
> and no doubt should be to me. But lets see if I can figure it out:
>
> There are 4 raw (if not sane) possiblities:
>

Let's try using "ungrounded end".

Chris

,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_
4228 2010-02-08 17:46:48 Paul Re: CA3054
The sound you may have just heard was my hand slapping my forehead. The term "live end" makes perfect sense in hindsight. Was locked into thinking horizontally and not vertically.

Thanks for the patience.

I'll be sure to actually build the thing now that you've been so kind to educate me.

Paul



4229 2010-02-08 17:56:34 Paul Re: CA3054
Hi Nick,

I saw your autodyne after I came back from a hamfest last fall with a bag of CA3146 and was looking for something to do with them.

Was torn between building that or the regen w/ frequency counter link that Chuck Adams K7QO has been writing about recently
4233 2010-02-09 19:05:38 Paul Re: CA3054
The 25 CA3054 arrived tonight. They appear to never have been removed from the original RCA tube as the end pegs were still intact. Nice easy to handle 14 pin DIPs.

I breadboarded a simple circuit to try one out: just a diff amp with current sink feed by a 0.3v 1Khz generator and putting out a nice looking square wave at over 5v. Not a comprehensive test, but at least I'm convinced the parts are not DOA.

I'm going to keep 5
Another 5 are on their way to KD4SGN

15 remain -- $1.00 each plus actual postage cost.

Contact me if you want a few.

Paul - K0EET

4234 2010-02-09 19:43:20 Phil Sittner Re: CA3054
Paul-

I'll take five as well if they're still available. Let me know how you'd like payment and I'll forward the address.

regards,
Phil kd6rm
----- Original Message -----
4235 2010-02-09 20:19:06 k5nwa Re: CA3054
At 09:04 PM 2/9/2010, you wrote:
>
>
>The 25 CA3054 arrived tonight. They appear to never have been
>removed from the original RCA tube as the end pegs were still
>intact. Nice easy to handle 14 pin DIPs.
>

I would like 5 also.

Thanks


Cecil
k5nwa
< www.softrockradio.org > < www.qrpradio.com >
< http://parts.softrockradio.org/ >

Never take life seriously. Nobody gets out alive anyway.
4236 2010-02-09 20:51:47 Paul Re: CA3054
OK -- all now accounted for.

5 for me
5 for KD4SGN
5 for WA5BDU
5 for KD6RM
5 for K5NWA

Phil, you have an off-list reply to snitters@... , come back here if you don't receive it.

Cecil, you also have an off list replay to k5nwa@...

4237 2010-02-09 21:56:40 k5nwa Re: CA3054
At 10:28 PM 2/9/2010, you wrote:
>
>
>OK -- all now accounted for.
>
>5 for me
>5 for KD4SGN
>5 for WA5BDU
>5 for KD6RM
>5 for K5NWA
>
>Phil, you have an off-list reply to snitters@... , come back here if
>you don't receive it.

Thanks

I been nosing around and many of these devices are still manufactured
by Intersil in SOIC packaging, some in the $.60 cent range others as
much as $12. I been seriously thinking lately that I might as well
start switching to SMT IC's for experimenting, the parts are a lot
easier to find.

Next week when I recoup some spending cash I will put in an order for
various SOIC devices, if people are interested we can make it a group
buy, I'll sell them to the group at cost + shipping expenses.


Cecil
k5nwa
< www.softrockradio.org > < www.qrpradio.com >
< http://parts.softrockradio.org/ >

Never take life seriously. Nobody gets out alive anyway.
4238 2010-02-10 05:43:06 Chris Trask Re: CA3054
>
>OK -- all now accounted for.
>
>5 for me
>5 for KD4SGN
>5 for WA5BDU
>5 for KD6RM
>5 for K5NWA
>

Boy, those didn't last long. You can make a pretty nice double-balanced mixer ir variable-gain IF strip with those.

Speaking of which, a CA3046 or CA3127 can make a decent LNA and single-balanced mixer in one package.


Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/
4239 2010-02-10 05:47:48 Phil Sittner Re: CA3054
Paul-
No I didn't receive it. The email address is sittners@sbcglobal.net. You have the "n" misplaced in this correspondence and perhaps in the email address as well. Thanks again.

Phil kd6rm
----- Original Message -----
4241 2010-02-10 06:59:28 Paul Re: CA3054
Sorry for the confusion Phil -- I was referring to the mail that you must have received (since you replied to it) before the post below made it out of moderation. I added to the confusi
4245 2010-02-11 10:46:24 Paul Re: CA3054
All four packages shipped late this morning. Each recipient should get an off-list message.

Thanks,
Paul - K0EET

4269 2010-02-16 20:37:24 Paul Re: CA3054
CA3146 as QRP final.

I built this last night and tried it out tonight. I didn't have enough cores around to try "parallel amps" trick, so I tried it as shown at
http://www.qrp4u.de/docs/en/rf_ampl/index.htm
but using CA3146.

I used the substrate-connected transistor as the bias diode, the pair with the pre-connected emitters as a push-pull, and the remaining pair also as a 2nd push-pull.

I first connected the emitters of each push-pull pair to ground via a separate 1-ohm resistor (as an attempt at balancing current). I then tried with all emitters grounded.

I drove it with the simple RF source from EMRFD ch 7. It will drive a 50-ohm dummy load to 2v p-p (10 mW if I calculate correctly). The voltage is measured by oscilloscope connected via 10x probe across the dummy resistor.

With 12v supply I was able to get 9v p-p out into a 50-ohm dummy load (200 mW, again if I haven't borked the calculation, same measurement technique as above). I tweaked the signal frequency rather than the capacitor to get the peak (just over 8Mhz). I got the same when using the 1-ohm emitter resistors and with them shorted.

The output waveform visually clean, once I solve some other issues I'll check for harmonic output with the SA.

The bottom of the chip was bonded to a few square inches of copper clad board with thermal epoxy and the top was somewhat uncomfortable to hold my finger on the top with continuous peak output. I have a few DIP heatsinks
4270 2010-02-17 05:22:26 Chris Trask Re: CA3054
>
> CA3146 as QRP final.
>
> I built this last night and tried it out tonight. I didn't have
> enough cores around to try "parallel amps" trick, so I tried it
> as shown at
>
> http://www.qrp4u.de/docs/en/rf_ampl/index.htm
>
> but using CA3146.
>
> I used the substrate-connected transistor as the bias diode, the
> pair with the pre-connected emitters as a push-pull, and the
> remaining pair also as a 2nd push-pull.
>

Yes, that's the best configuration.

<>

>
> During peak output I was drawing roughly 125ma from the supply
> at 12v.
>

How much were you drawing with no signal? You'll get your best power output if you're running class B or AB, so your idle current should be 1% of your peak current for Class B or 5-10% for Class AB.

With T2 being 1:1:1 (2CT:1), each side sees 100-ohms, which means that you can draw as much as 120mA peak current (85mA average). You should therefore be able to deliver just under 1W to a 50-ohm load. With the device dissipation benefits of Class B/AB, you'll be dissipating about 0.16W peak per device, and less than 0.075W average, which is well within the device limitation of 300mW.


Chris Trask
N7ZWY / WDX3HLB
Senior Member IEEE
http://www.home.earthlink.net/~christrask/