Topics

2N2222A getting super hot

Joe Puma
 

Hi,

I performed Alison's mod and replaced the 3904’s with 2N2222A T08. I also paralleled the 22 ohm resistors to get the 11 ohms. Not to mention I changed q90 and replaced the resister with 2.55kohms. The 2n2222a’s that have their resistors lowered to 11 ohms get really hot. the other set of drivers that don’t have their resistors lowered ( im not sure why) dont get that hot but they get warm.

when testing amp draw from brown wire in line with amp meter and rv2 and 3 fully clockwise I get 0amp on key up. I turn up one 100ma, then the other. now im at 200ma. If I talk in the mic It swings to 400ma+. With CW key down it jumps to under 1amp.

I am on a dummy load

I dont have a power meter.

Does this make sense after doing that mod. is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?


Joe,
KD2NFC

Gordon Gibby <ggibby@...>
 

Maybe check with an ohmmeter to make sure you ended up with 11 ohms. Then measure the voltage across the resistors while turned on, use ohms law to figure out the current through the transistors

Allyson told people what current she preferred.

Measure the voltage from the emitter to the collector, multiply by the current through the transistor, to figure out the resting dissipation; compared to the specified for the device.

Somewhere in there you’ll probably find your problem. There might be an incorrect biasing set up.

Gordon

On Mar 29, 2019, at 02:44, Joe Puma <@kd2nfc> wrote:

Hi,

I performed Alison's mod and replaced the 3904’s with 2N2222A T08. I also paralleled the 22 ohm resistors to get the 11 ohms. Not to mention I changed q90 and replaced the resister with 2.55kohms. The 2n2222a’s that have their resistors lowered to 11 ohms get really hot. the other set of drivers that don’t have their resistors lowered ( im not sure why) dont get that hot but they get warm.

when testing amp draw from brown wire in line with amp meter and rv2 and 3 fully clockwise I get 0amp on key up. I turn up one 100ma, then the other. now im at 200ma. If I talk in the mic It swings to 400ma+. With CW key down it jumps to under 1amp.

I am on a dummy load

I dont have a power meter.

Does this make sense after doing that mod. is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?


Joe,
KD2NFC





Ted
 

Joe,

RV1 & 2 "can" be brought up to more than 100ma each, at the cost (very possibly) of toasting your IRF510 final output devices :)   I don't recommend it.

Like yourself, I note the 2N2222 replacements in the predriver and driver getting warm. My "warm" may be your "hot" due to perspective, maybe not. Can you keep a finger on it for 10 seconds without experiencing real pain?   If you get a genuine burn injury,  it's probably running hot; my suspicion is that, as this is a terribly common upgrade,  a bit warm or very warm is just fine.

As an explanation, follow the circuitry.  Those transistors are basically shorting 12 volts to earth/ground at an RF rate, save for the 11-ohm resistors that govern the process.  The result at each point is to take a sample of the stage before it and make that energy "wiggle" the next transistor, whose job it is to almost short out the next stages' 12-volt feed at an HF frequency rate. Virtually draining B+ current through the resistor creates heat (how an oven works, etc etc). The whole reason for swapping the 22-ohm resistors for 11-ohm chips is that the 2n2222 can handle the greater workload - the more direct conduct of B+ to ground due to reduced resistance  - and delivers greater output as a result.  That more heat is produced, should be expected within reason.

Hope this helps.


Ted
K3RTA

 

Joe,

Which version is this ? In my V4 boards even the 3904 run hot.

Raj

At 29-03-19, you wrote:
Hi,

I performed Alison's mod and replaced the 3904’s with 2N2222A T08. I also paralleled the 22 ohm resistors to get the 11 ohms. Not to mention I changed q90 and replaced the resister with 2.55kohms. The 2n2222a’s that have their resistors lowered to 11 ohms get really hot. the other set of drivers that don’t have their resistors lowered ( im not sure why) dont get that hot but they get warm.

when testing amp draw from brown wire in line with amp meter and rv2 and 3 fully clockwise I get 0amp on key up. I turn up one 100ma, then the other. now im at 200ma. If I talk in the mic It swings to 400ma+. With CW key down it jumps to under 1amp.

I am on a dummy load

I dont have a power meter.

Does this make sense after doing that mod. is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?


Joe,
KD2NFC

Jim Strohm
 

This is a little off-thread --

With the advent of SMDs everywhere -- does anybody know of a CHEAP thermal / IR camera for PCB inspection?

I know that you can get a thermal /IR videocam for several hundred dollars, but we basically just need the imaging module and enough of a lens that we can focus on a 6x6 inch PCB, or section thereof.  

Thoughts?  Some of the older Sony camcorders used to see into the infrared, and with an infrared filter over the lens gave a decent IR image, although in B&W. 

A compact, cheap thermal / IR cam would be a very useful tool for troubleshooting.

73
Jim N6OTQ

Tom, wb6b
 

On Fri, Mar 29, 2019 at 05:39 AM, Jim Strohm wrote:
does anybody know of a CHEAP thermal /
If you are into DYI here are a couple of possibilities. 

https://learn.adafruit.com/adafruit-amg8833-8x8-thermal-camera-sensor/arduino-thermal-camera

https://shop.pimoroni.com/products/mlx90640-thermal-camera-breakout

I have one of the Melexis MLX90640 devices I bought from Digikey. The example software I found (that someone hacked to run on the RPi) wrote directly to the frame buffer of my RPi, so had issues with it being overwritten by the desktop. I bought a tiny color OLED display and when that day comes when I have some spare time I'll port the example code to display on the external OLED display rather than the main HDMI monitor. 

The example code uses too much RAM to work on a basic Arduino, but might work on a "Blue Pill" microprocessor board. For the fun of it I tried running the SD card from my "regular" Raspberry Pi on a Raspberry Pi Zero. Seems the Desktop UI of a full RPi linux install has grown to the point the RPi zero could not reliably run it. (Although I need to check with a different brand/size SD card to be sure SD card compatibility was not the issue.) But, a cut down linux install probably would work on the Zero; making for a small portable thermal camera.

Tom, wb6b

Allen Hill
 

Jim, do a web search for USB powered microscopes. Some are < $50 and amazingly useful. I have one that connects to PC has built in LED lighting, allows you to snap pics, etc. Some also have numerous accessories (stand options, etc). Not infared, but mine has saved me many times.

KI4QCK
73



On March 29, 2019, at 8:40 AM, Jim Strohm <jim.strohm@...> wrote:


This is a little off-thread --

With the advent of SMDs everywhere -- does anybody know of a CHEAP thermal / IR camera for PCB inspection?

I know that you can get a thermal /IR videocam for several hundred dollars, but we basically just need the imaging module and enough of a lens that we can focus on a 6x6 inch PCB, or section thereof.  

Thoughts?  Some of the older Sony camcorders used to see into the infrared, and with an infrared filter over the lens gave a decent IR image, although in B&W. 

A compact, cheap thermal / IR cam would be a very useful tool for troubleshooting.

73
Jim N6OTQ

Joe Puma
 

V3 board.

I ran a WSPR on 40m and measured the 2N2222’s with a heat gun, I was reading close to 60c, didn’t go over. It got a bit hotter on 80m. I checked spec and the transitor max temp is 146c, so I figured I’m still safe.

I guess I never realized that the pre and driver get extra warm. Even hotter then the finals. I’m gonna continue to test with the suggestions of everyone here. Thank you guys. I also did read that some others who did this mod noticed the heat as well and went back to 22ohms.

Joe

On Mar 29, 2019, at 8:01 AM, Raj vu2zap <@Raj> wrote:

Joe,

Which version is this ? In my V4 boards even the 3904 run hot.

Raj

At 29-03-19, you wrote:
Hi,

I performed Alison's mod and replaced the 3904’s with 2N2222A T08. I also paralleled the 22 ohm resistors to get the 11 ohms. Not to mention I changed q90 and replaced the resister with 2.55kohms. The 2n2222a’s that have their resistors lowered to 11 ohms get really hot. the other set of drivers that don’t have their resistors lowered ( im not sure why) dont get that hot but they get warm.

when testing amp draw from brown wire in line with amp meter and rv2 and 3 fully clockwise I get 0amp on key up. I turn up one 100ma, then the other. now im at 200ma. If I talk in the mic It swings to 400ma+. With CW key down it jumps to under 1amp.

I am on a dummy load

I dont have a power meter.

Does this make sense after doing that mod. is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?


Joe,
KD2NFC


ajparent1/kb1gmx
 

is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?

Because they have higher gain and are drawing more current.  The correction is changing the bias resistors
to get the idling current down to about 20ma per transistor.

its in the wiki.  The vale needed will have to be determined experimentally.  I'd start at 2.2k.

Allison

Clark Martin
 


Clark Martin
KK6ISP

On Mar 29, 2019, at 9:17 AM, Joe Puma <kd2nfc@...> wrote:

V3 board. 

I ran a WSPR on 40m and measured the 2N2222’s with a heat gun, I was reading close to 60c, didn’t go over. It got a bit hotter on 80m. I checked spec and the transitor max temp is 146c, so I figured I’m still safe. 

146°C would be the junction temperature max.  According to the datasheet I’m looking at, the thermal resistance from junction to case is 83.3°C/W.  Which means if it’s dissipating 1W that 60°C case temp means 143°C junction temperature.  So you need to figure out the power dissipation to determine if the 60°C was temperature is acceptable.


I guess I never realized that the pre and driver get extra warm. Even hotter then the finals. I’m gonna continue to test with the suggestions of everyone here. Thank you guys.  I also did read that some others who did this mod noticed the heat as well and went back to 22ohms. 

Joe Puma
 

Hi Allison,

I did start at 2.2k I now have 2.55kohm, does the current go up as you raise resistance?

I’m not sure how to test the current on the transistors so I don’t know what they are running at, but I think Gorgon gave me some suggestions on how to do that. I will do some testing. 

Btw I didn’t change the resistors you suggested to 1.5kohms. 


Joe


On Mar 29, 2019, at 12:45 PM, ajparent1/KB1GMX <kb1gmx@...> wrote:

is there a reason why those 2n2222a are getting so hot, can i bring RV1 and 2 up so its drawing more current?

Because they have higher gain and are drawing more current.  The correction is changing the bias resistors
to get the idling current down to about 20ma per transistor.

its in the wiki.  The vale needed will have to be determined experimentally.  I'd start at 2.2k.

Allison

Joe Puma
 

Thanks Clark,

The heat was gradual too. It took toward the end of the WSPR transmission before I saw 50-56c. 

Joe



On Mar 29, 2019, at 12:46 PM, Clark Martin <kk6isp@...> wrote:


Clark Martin
KK6ISP

On Mar 29, 2019, at 9:17 AM, Joe Puma <kd2nfc@...> wrote:

V3 board. 

I ran a WSPR on 40m and measured the 2N2222’s with a heat gun, I was reading close to 60c, didn’t go over. It got a bit hotter on 80m. I checked spec and the transitor max temp is 146c, so I figured I’m still safe. 

146°C would be the junction temperature max.  According to the datasheet I’m looking at, the thermal resistance from junction to case is 83.3°C/W.  Which means if it’s dissipating 1W that 60°C case temp means 143°C junction temperature.  So you need to figure out the power dissipation to determine if the 60°C was temperature is acceptable.


I guess I never realized that the pre and driver get extra warm. Even hotter then the finals. I’m gonna continue to test with the suggestions of everyone here. Thank you guys.  I also did read that some others who did this mod noticed the heat as well and went back to 22ohms. 

iz oos
 

My first power meter was the RF-Probe by Hendricks. http://www.qrpkits.com/files/RF_Probe_How_To.pdf. It proved to be useful even to check the output power of oscillators. You can make one with just a few components.


Il 29/mar/2019 18:07, "Joe Puma" <kd2nfc@...> ha scritto:
Thanks Clark,

The heat was gradual too. It took toward the end of the WSPR transmission before I saw 50-56c. 

Joe



On Mar 29, 2019, at 12:46 PM, Clark Martin <kk6isp@...> wrote:


Clark Martin
KK6ISP

On Mar 29, 2019, at 9:17 AM, Joe Puma <kd2nfc@...> wrote:

V3 board. 

I ran a WSPR on 40m and measured the 2N2222’s with a heat gun, I was reading close to 60c, didn’t go over. It got a bit hotter on 80m. I checked spec and the transitor max temp is 146c, so I figured I’m still safe. 

146°C would be the junction temperature max.  According to the datasheet I’m looking at, the thermal resistance from junction to case is 83.3°C/W.  Which means if it’s dissipating 1W that 60°C case temp means 143°C junction temperature.  So you need to figure out the power dissipation to determine if the 60°C was temperature is acceptable.


I guess I never realized that the pre and driver get extra warm. Even hotter then the finals. I’m gonna continue to test with the suggestions of everyone here. Thank you guys.  I also did read that some others who did this mod noticed the heat as well and went back to 22ohms. 

 

Hi Joe,

I thought Allison gave a pretty good explanation for how to do it in the Wiki:

==================== first amp q90 ======================== 

Replace Q90 with BFR106.  (Mouser about 18cents each) . Reset bias of Q90 with BFR106 , Set bias for less than 10ma (r86 2.7K for mine).

Make C81 470pf, and R83 8.2ohms. (lower power at 80m improves 10m)

 

To set bias and measure simple measure across the devices emitter resistor voltage/resistance=current  so 100ohms with 1Volt is .01A or 10ma.

Change C82 to 470pf (peaking for 10M and less power at 80m, optional if your not playing at more than 20mhz.)

 

This method can be used for the other amplifiers, just the resistor value used in the calculation changes.

The bias resistor changed is part of a resistive potentiometer divider.  This is a classic way to bias the base of a transistor.  Look up how to calculate the voltage out of a divider and then do the calculation for changing that resistor.  You can then see how the voltage provided to the base will change.  Do you know what changing the voltage on the base of an NPN transistor will do to the current running through the base to emitter and thus the collector to emitter?

73,


Mark.

MVS Sarma
 

Thanks for reminding.


On Sat, 30 Mar 2019, 7:15 pm Mark - N7EKU <n7eku@... wrote:
Hi Joe,

I thought Allison gave a pretty good explanation for how to do it in the Wiki:

==================== first amp q90 ======================== 

Replace Q90 with BFR106.  (Mouser about 18cents each) . Reset bias of Q90 with BFR106 , Set bias for less than 10ma (r86 2.7K for mine).

Make C81 470pf, and R83 8.2ohms. (lower power at 80m improves 10m)

 

To set bias and measure simple measure across the devices emitter resistor voltage/resistance=current  so 100ohms with 1Volt is .01A or 10ma.

Change C82 to 470pf (peaking for 10M and less power at 80m, optional if your not playing at more than 20mhz.)

 

This method can be used for the other amplifiers, just the resistor value used in the calculation changes.

The bias resistor changed is part of a resistive potentiometer divider.  This is a classic way to bias the base of a transistor.  Look up how to calculate the voltage out of a divider and then do the calculation for changing that resistor.  You can then see how the voltage provided to the base will change.  Do you know what changing the voltage on the base of an NPN transistor will do to the current running through the base to emitter and thus the collector to emitter?

73,


Mark.

ajparent1/kb1gmx
 

Joe,

Read the wiki, all of it.  I put all that there so people would have it and and could
find it, while not having to repeat it every three months.

The easy way to measure the transistor current...  Transmit, no audio.

Measure the voltge across the emitter resistor.  That voltage divided by
the resistor value will give you the current.

Its less than 20ma (10-12 is better) for Q90.  For the predrivers and
drivers its not more than 20ma (plus or minus a few).

the finals need to be at about 100ma EACH.

Allison


Joe Puma
 

Thank you Allison and everyone one else. I was able to follow the part changing with no problem, I just might need a little hand holding with knowing how to check the amperages.  

To the gentleman that asked me if I know what’s happening in the NPN transistor as it pertains to the drivers in the radio. I would say not a hell of a lot lol. Well maybe I do a little, but I just don’t have a good visual picture of it in my head.  I know if you apply voltage to the base, current can flow from the collector to emitter? 

I’m am going to make some test readings today and do some simple ohms law math. I’ll let everyone know what I find. 

I’ve mentioned having a handy service many available for us tinkerers but I understand that if you have the right skill set, all this stuff is obvious for the seasoned engineer. But some of us need the fisher price version of how to maybe do these simple testing tasks. 

I solder really well if that helps :)

Joe
KD2NFC



On Mar 30, 2019, at 11:30 AM, ajparent1/KB1GMX <kb1gmx@...> wrote:

Joe,

Read the wiki, all of it.  I put all that there so people would have it and and could
find it, while not having to repeat it every three months.

The easy way to measure the transistor current...  Transmit, no audio.

Measure the voltge across the emitter resistor.  That voltage divided by
the resistor value will give you the current.

Its less than 20ma (10-12 is better) for Q90.  For the predrivers and
drivers its not more than 20ma (plus or minus a few).

the finals need to be at about 100ma EACH.

Allison


 

Hi Joe,

I was just asking because, like you said, people are at different levels of their learning.  There are many sites on the web for this:

https://learn.sparkfun.com/tutorials/transistors/all

That one is gives a good overall view.

Of course there is always the Handbook:

https://archive.org/details/arrl_1968_handbook/page/n80

Often it just takes a half hour or so of study, or a few pages reading here and there, to start learning the basics!

But restricting oneself to only reading forum posts or questions and answers doesn't really work that well for learning.  But for sure working through problems and posting your answers for confirmation can be very helpful so go for it!

73,


Mark.

ajparent1/kb1gmx
 

For those people that only use groups.IO as a mail-list and have no idea that the site
can be accessed complete with history, wiki and files sections....

This is where the wiki can be found... using a wed browser...

https://groups.io/g/BITX20/wiki/home

and your Emails end up here....

https://groups.io/g/BITX20/topics

Allison
Please no direct/private emails.

Joe Puma
 

Hey Mark,

Thanks for the links. I learned some basic EE in a tech school I went to after HS in the late 80’s. I was taught enough to know what I should be doing here, but some things just don’t click at first like if I wanted to know current draw on Q90, to measure the voltage on that resistor that bias’s the transistor and do ohms law, Duh.  As I refresh myself things become more obvious. And of course with the help from this group. 

The uBitx has been slowly bringing me back to something I started about 30 years ago but never finished properly, which is a intimate understanding in EE. That was my desire, but computers took off at that time and I jumped on the bandwagon building clone PC’s and networks. 

Joe
KD2NFC 





On Mar 30, 2019, at 2:08 PM, Mark - N7EKU <n7eku@...> wrote:

Hi Joe,

I was just asking because, like you said, people are at different levels of their learning.  There are many sites on the web for this:

https://learn.sparkfun.com/tutorials/transistors/all

That one is gives a good overall view.

Of course there is always the Handbook:

https://archive.org/details/arrl_1968_handbook/page/n80

Often it just takes a half hour or so of study, or a few pages reading here and there, to start learning the basics!

But restricting oneself to only reading forum posts or questions and answers doesn't really work that well for learning.  But for sure working through problems and posting your answers for confirmation can be very helpful so go for it!

73,


Mark.