Topics

Pre driver and driver transistors

ohwenzelph
 

Is replacing those 2N3904s with 2n2222As a bad idea?

Gordon Gibby
 

As I recall, discussions a year or two ago, it was a good idea but you also need to change the biased point to reach maximum gain at the frequencies of interest. That requires a bit of an understanding of common amateur amplifier biased, but Alyson had given lots of information on how to do it, you might search for her very very detailed posts from back. As I recall, discussions a year or two ago, it was a good idea but you also needed to change the biased point to reach maximum gain at the frequencies of interest. That requires a bit of an understanding of common amateur amplifier bias, but Alyson had given lots of information on how to do it, you might search for her very very detailed posts from back then


On May 30, 2020, at 12:19, ohwenzelph via groups.io <Ohwenzelph@...> wrote:

Is replacing those 2N3904s with 2n2222As a bad idea?

ajparent1/kb1gmx
 

And all of it is in the wiki... and the world longest thread as well.

Allison
---------------------------------
No direct email, it goes to bit bucket due address harvesting in groups.IO

 

Link to Wiki page on the topic:

uBITX RF Power and Gain Issues

Maurice Bersan <Vk6hly@...>
 

Follow Allisons instruction for the complete mod....you'll be glad you did.
Maurice.   Vk6hly




On Sun, May 31, 2020 at 11:31 AM, Mark - N7EKU
<n7eku@...> wrote:
Link to Wiki page on the topic:

uBITX RF Power and Gain Issues

roebi@...
 

I'm thinking about these mods as well.

In the wiki for Allison's solution, there is the easy part with just replacing 2n2222's plus Q90 and C81, and then there is the more complicated mod. And then there is the recommended solution from ubitx.net: http://ubitx.net/fix-fix-uneven-tx-output-across-bands/ with the easy part of Allisions solution and a 2nd part with relais get the power to equal levels across the bands. This is probably no the nicest solution, but seems a lot easier than the rest of Allisons mods.

So now the questions:

What output power across the different bands can be expected with just the easy part of Allison mod?

Has anyone here done the easy part plus relais solution?

Are these solutions capable of digital modes without overheating, or is it necessary to somehow reduce power for FT8/WISPR?
As 10W is probably too much for digital modes anyway, the relais solution could easily be extended with an additionals switch and potentiometer for manual power reduction.

Thanks,

Robert

ohwenzelph
 

fwiw
Actually, I am not so concerned about the power out being equal across the bands. The concern is to have reasonably enough power out on the 10 meter end. And if 10 and 12 and 15 meters are open you don’t need very much power. At the 80/40 end I don’t mind if there is a lot more power, granted not QRP, but if summer thunderstorms are making a.a racket and you want to be heard, then it becomes a question of how do I cool those heat sinks. Why do they need to be equal? With increased power harmonics and spurs became more of a concern but my impression is that Farhan has fixed those. One can always try to  optimise efficiency but that is less of a concern if not running on batteries or portable. I am still trying to understand the concept of IMD and the square root of negative one...
jer aa1of

Gérard
 

Hello,

I have not yet asked myself the question to replace some 3904. For me the 2n2222 is an "old Go anywhere " transistor.It's work fine, but it is at least over 50 years old, I was already changing some in the years  70/80. We really should study the case.
Take a more "modern"? Ex :Bf 959 Bandwitch 1G against 300mo to a 2n2222.
 But the change does not stop there. It may be necessary to adjust resistances emitters, base. See if anyone has already put transistors other than 2n2222. Take the µbitx version6 schematic  and find an equivalent to the BFR93w.


Le mer. 3 juin 2020 à 17:32, ohwenzelph via groups.io <Ohwenzelph=yahoo.com@groups.io> a écrit :
fwiw
Actually, I am not so concerned about the power out being equal across the bands. The concern is to have reasonably enough power out on the 10 meter end. And if 10 and 12 and 15 meters are open you don’t need very much power. At the 80/40 end I don’t mind if there is a lot more power, granted not QRP, but if summer thunderstorms are making a.a racket and you want to be heard, then it becomes a question of how do I cool those heat sinks. Why do they need to be equal? With increased power harmonics and spurs became more of a concern but my impression is that Farhan has fixed those. One can always try to  optimise efficiency but that is less of a concern if not running on batteries or portable. I am still trying to understand the concept of IMD and the square root of negative one...
jer aa1of

MadRadioModder
 

You may be taking this “equal across the bands” statement too literal.  If you go back to the actual post conversations, that was not necessarily the original goal.  It started out as the statement “why is the power so low on 10 meters”…

 

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of ohwenzelph via groups.io
Sent: Wednesday, June 3, 2020 10:33 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Pre driver and driver transistors

 

fwiw
Actually, I am not so concerned about the power out being equal across the bands. The concern is to have reasonably enough power out on the 10 meter end. And if 10 and 12 and 15 meters are open you don’t need very much power. At the 80/40 end I don’t mind if there is a lot more power, granted not QRP, but if summer thunderstorms are making a.a racket and you want to be heard, then it becomes a question of how do I cool those heat sinks. Why do they need to be equal? With increased power harmonics and spurs became more of a concern but my impression is that Farhan has fixed those. One can always try to  optimise efficiency but that is less of a concern if not running on batteries or portable. I am still trying to understand the concept of IMD and the square root of negative one...
jer aa1of


Virus-free. www.avg.com

--

…_. _._

ajparent1/kb1gmx
 

Over the last three years (see the old and very large thread) the following were tried...

MPSA10,
2n2222a
2n3866
2n5109

The BF959 is a low power VHF transistor and would be a poor performer
except as replacement for Q90  maybe, the BFR106 was better.   The BF959
gain falls off fast above 20ma, undesired for high peak current applications.

Desired characteristics are:
Pdiss greater than 800mw (most metal cans do better than plastic).
Vcc max not less than 30V
Max collector current not less than 200ma.
Ft more than 800mhz , and does not decrease at high current.

I tried a few UHF parts I had for the 1W region with Ft of 5ghz and
they did not like the board layout (undesired oscillation). One issue 
is get enough gain at 30mhz and it is unstable due to layout.

Allison
---------------------------------
No direct email, it goes to bit bucket due address harvesting in groups.IO

MadRadioModder
 

Again, you-all might do a little research by reading all of the posts from about a year and a half ago.  No sense reinventing the wheel.  There were several people working this issue… redesigning output transformers, changing the feedback components on the finals (oh yes, those parts are frequency dependent), adding another drive state, changing bypassing). As I recall, the group consisted of a DL, a VK, a K9, and another euro station… some of these people have since left the group.  You might reread all of that dialogue and weigh that in with  Allison’s ever-present posts.

 

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of Gérard
Sent: Wednesday, June 3, 2020 11:12 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Pre driver and driver transistors

 

Hello,

 

I have not yet asked myself the question to replace some 3904. For me the 2n2222 is an "old Go anywhere " transistor.It's work fine, but it is at least over 50 years old, I was already changing some in the years  70/80. We really should study the case.
Take a more "modern"? Ex :Bf 959 Bandwitch 1G against 300mo to a 2n2222.

 But the change does not stop there. It may be necessary to adjust resistances emitters, base. See if anyone has already put transistors other than 2n2222. Take the µbitx version6 schematic  and find an equivalent to the BFR93w.

 

 

Le mer. 3 juin 2020 à 17:32, ohwenzelph via groups.io <Ohwenzelph=yahoo.com@groups.io> a écrit :

fwiw
Actually, I am not so concerned about the power out being equal across the bands. The concern is to have reasonably enough power out on the 10 meter end. And if 10 and 12 and 15 meters are open you don’t need very much power. At the 80/40 end I don’t mind if there is a lot more power, granted not QRP, but if summer thunderstorms are making a.a racket and you want to be heard, then it becomes a question of how do I cool those heat sinks. Why do they need to be equal? With increased power harmonics and spurs became more of a concern but my impression is that Farhan has fixed those. One can always try to  optimise efficiency but that is less of a concern if not running on batteries or portable. I am still trying to understand the concept of IMD and the square root of negative one...
jer aa1of


Virus-free. www.avg.com

--

…_. _._