Topics

uBitx v6 board failure--suggestions? #v6 #ubitxv6


John Terrell
 

My uBitx V6 worked well for a few hours, then the power output stopped. (I smelled some smoke for a few seconds, and shut down, but it was too late.) I then could still hear a weak signal on my other rig indicating that the unit was working ahead of the finals. I obtained new IRF510 MOSFETS and switched the both out. Now I can't hear even the weak signals from the drivers, and furthermore there is a cycling white-out of the screen--that's new--that repeats again and again when I key the CW or SSB microphone. Please see the short video. Any suggestions? Do I need to get a new board?


Jerry Gaffke
 

It is possible that those IRF510's are now drawing too much current for your power supply,
and this is causing the "cycling white-out of the screen".

Did you figure out for sure that the old IRF510's were damaged before replacing them?
If they did get damaged, it was probably heat.  They were likely too hot to touch when they failed.

Did you figure out why the IRF510's smoked?  That should not have happened.
Perhaps you were running for long periods of time at 10W in digital modes, and didn't have an adequate heat sink?
Perhaps the antenna system presented more than a 2:1 SWR?   (Due to heat, IRF510 damage would not be instantaneous) 
Perhaps you fed your rig well over 12v to get a little bit more power?  (Damage due to IRF510 heat)

The gate bias required varies with each IRF510, they are not manufactured to be exactly the same.
That's why we have RV2 and RV3.
Did you adjust RV2 and RV3 for the new IRF510's, 100ma quiescent drain current on each?
A good way to determine if IRF510's are damaged is to attempt setting quiescent drain current, 
if you succeed then they are OK.

A good way to blow out your new IRF510's is to just turn the rig on and try transmitting
without first setting RV2 and RV3 properly.

My suggestion on how to proceed:
Make sure your power supply is 12v, and has a 3A fuse.
Adjust RV2 and RV3 for minimum gate bias (minimum is fully clockwise, this is backwards of most such knobs in this world)
Power up, see if you are back to the display working properly, and hearing the driver when attempting to transmit.

Then follow the instructions on hfsignals.com to set the gate bias on each IRF510, specifically:
  Monitor power supply current, I'm guessing on the v6 it might be around 300ma on receive, 500ma on transmit with zero gate bias.  (Could somebody confirm?)
  With PTT pressed and no audio into the mike, adjust RV2 counterclockwise very slowly until supply current increases by 100ma.
  Then adjust RV3 counterclockwise till the power supply increases by another 100ma.

You must turn RV2 and RV3 slowly, nothing will happen till about 4/5'ths of the way around when drain current will suddenly shoot up.
You can measure the gate voltage with a DVM, watch it rise from 0v to 5v as you turn the pot, current suddenly shoots up at about 4v on the gate.
to transmit.

If rotating RV2 and RV3 fully clockwise does not bring back the display and the signal from the driver, then remove the 12v supply on the IRF510 drains.
On older rev uBitx's there were two 12v supply wires into the main board, one for the IRF510's and one for everything else.
Unfortunately v6 only has one wire, you might try unsoldering one end of L8 and L9 instead.
I hope that if there is a v7, it will again have a separate 12v supply wire for the IRF510 drains going into the board.

Once 12v to the IRF510 drains is removed, you can turn the rig on and try transmitting.
The IRF510's won't be getting hot, or drawing so much current from your power supply that it fails to give 12v.
But you should be able to see the gate bias of each IRF510 rise from 0v to 5v as you turn RV2 or RV3 counterclockwise.

Tell us what you learn when you try the above.

Good Luck,
Jerry, KE7ER


On Tue, Jan 19, 2021 at 07:38 AM, <N6LN@...> wrote:
My uBitx V6 worked well for a few hours, then the power output stopped. (I smelled some smoke for a few seconds, and shut down, but it was too late.) I then could still hear a weak signal on my other rig indicating that the unit was working ahead of the finals. I obtained new IRF510 MOSFETS and switched the both out. Now I can't hear even the weak signals from the drivers, and furthermore there is a cycling white-out of the screen--that's new--that repeats again and again when I key the CW or SSB microphone. Please see the short video. Any suggestions? Do I need to get a new board?


Jerry Gaffke
 

Did you figure out where the smoke was coming from?
The IRF510's are a good first guess, but could also have been a board trace, or L8,L9.

I find that going out of the room can bring my nose back to where it can smell a faint wiff of burning epoxy.
Then perhaps a soda straw from board to nose to help pinpoint where the smell is coming from.

Or, when I first shut down, immediately press the board with my hand to figure out what got hot.

Jerry



On Tue, Jan 19, 2021 at 08:13 AM, Jerry Gaffke wrote:
I smelled some smoke for a few seconds, and shut down,


John Terrell
 

Thanks—I will look into all of that. I tested the 510s with a method found on the web alternating connections with a digital ohmmeter. I also tested the new ones before installing them. I also have not determined where the stove came from—none of the parts appeared charred.


John Terrell
 


Thanks Jerry. your suggestions have been very helpful. I've made progress but still haven't solved all the problems. First, it seems that the finals can be blown if RV2 and RV3 are not set correctly. They may not have been set properly at the factory. Here's why I think so: the slots on the pots came perfectly horizontal. I noticed on the antenna tuner (my ATU-100) that it was putting out 14.7 Watts just before I smelled smoke and it failed. The removed MOSFETs were indeed failed as tested with a multimeter, the new ones were good, and I followed your procedure. I rigged a current meter and adjusted carefully RV2 and RV3 to run 100 mA and then another 100 mA above idle on microphone keydown. (Baseline receive was 370 mA, and initial full clockwise RV2 and RV3 on mic keydown was 580 mA.) And listening on another receiver showed the signal coming up at that point. then running my CW keyer with continuous dashes showed each final warming up just a little, but not much. Unfortunately there is essentially ZERO output to the antenna. So something else must have happened...please see my video (attached)


On 1/19/2021 8:13 AM, Jerry Gaffke via groups.io wrote:
> It is possible that those IRF510's are now drawing too much current > for your power supply, and this is causing the "cycling white-out of > the screen". > > Did you figure out for sure that the old IRF510's were damaged > before replacing them? If they did get damaged, it was probably heat. > They were likely too hot to touch when they failed. > > Did you figure out why the IRF510's smoked? That should not have > happened. Perhaps you were running for long periods of time at 10W > in digital modes, and didn't have an adequate heat sink? Perhaps the > antenna system presented more than a 2:1 SWR? (Due to heat, > IRF510 damage would not be instantaneous) Perhaps you fed your rig > well over 12v to get a little bit more power? (Damage due to IRF510 > heat) > > The gate bias required varies with each IRF510, they are not > manufactured to be exactly the same. That's why we have RV2 and RV3. > Did you adjust RV2 and RV3 for the new IRF510's, 100ma quiescent > drain current on each? A good way to determine if IRF510's are > damaged is to attempt setting quiescent drain current, if you > succeed then they are OK. > > A good way to blow out your new IRF510's is to just turn the rig on > and try transmitting without first setting RV2 and RV3 properly. > > My suggestion on how to proceed: Make sure your power supply is 12v, > and has a 3A fuse. Adjust RV2 and RV3 for minimum gate bias (minimum > is fully clockwise, this is backwards of most such knobs in this > world) Power up, see if you are back to the display working > properly, and hearing the driver when attempting to transmit. > > Then follow the instructions on hfsignals.com to set the gate bias > on each IRF510, specifically: Monitor power supply current, I'm > guessing on the v6 it might be around 300ma on receive, 500ma on > transmit with zero gate bias. (Could somebody confirm?) With PTT > pressed and no audio into the mike, adjust RV2 counterclockwise very > slowly until supply current increases by 100ma. Then adjust RV3 > counterclockwise till the power supply increases by another 100ma. > > You must turn RV2 and RV3 slowly, nothing will happen till about > 4/5'ths of the way around when drain current will suddenly shoot up. > You can measure the gate voltage with a DVM, watch it rise from 0v > to 5v as you turn the pot, current suddenly shoots up at about 4v on > the gate. to transmit. > > If rotating RV2 and RV3 fully clockwise does not bring back the > display and the signal from the driver, then remove the 12v supply > on the IRF510 drains. On older rev uBitx's there were two 12v supply > wires into the main board, one for the IRF510's and one for > everything else. Unfortunately v6 only has one wire, you might try > unsoldering one end of L8 and L9 instead. I hope that if there is a > v7, it will again have a separate 12v supply wire for the IRF510 > drains going into the board. > > Once 12v to the IRF510 drains is removed, you can turn the rig on > and try transmitting. The IRF510's won't be getting hot, or drawing > so much current from your power supply that it fails to give 12v. > But you should be able to see the gate bias of each IRF510 rise from > 0v to 5v as you turn RV2 or RV3 counterclockwise. > > Tell us what you learn when you try the above. > > Good Luck, Jerry, KE7ER > > > On Tue, Jan 19, 2021 at 07:38 AM, <N6LN@...>
*
*


Jerry Gaffke
 

John,

Yes, running CW mode is a good way of checking out the final.

Are you confident your dummy load and wattmeter are cabled up and working properly?
Perhaps verify you have it right by plugging them into some other transmitter if you (or a friend) have one.

Perhaps a cable has opened up due to flexing.
Perhaps a bad antenna connector (some of the BNC connectors are machined poorly, no contact on the inner pin).
Perhaps one of the relay contacts is burned out.
I assume you have checked all of this visually to look for burned traces and broken parts.

If you have RF power from the drivers, and RV2,RV3 have been set for 100ma quiescent drain current,
and T11 does not have a broken or shorted wire, then key down should put RF power out at least as far as TP7.

If you have a scope or diode RF probe, have you looked at TP7 to see if you have RF power there?
You could use your dummy load and wattmeter as an RF probe, putting them between TP7 and ground,
leaving the normal antenna port unconnected.  Do you see RF power at TP7?

If you don't see RF power at TP7, perhaps there is a short to ground somewhere.
Perhaps cut the trace into KA1 pin 13, see if you then get RF power at TP7.

The movie is not very informative. 
Would be better if you can describe what you are doing and seeing.  In words.
With this poor internet connection, I'm not keen to grab megabytes of video even if it is informative.

Jerry, KE7ER

 

Thanks Jerry. your suggestions have been very helpful. I've made progress but still haven't solved all the problems. First, it seems that the finals can be blown if RV2 and RV3 are not set correctly. They may not have been set properly at the factory. Here's why I think so: the slots on the pots came perfectly horizontal. I noticed on the antenna tuner (my ATU-100) that it was putting out 14.7 Watts just before I smelled smoke and it failed. The removed MOSFETs were indeed failed as tested with a multimeter, the new ones were good, and I followed your procedure. I rigged a current meter and adjusted carefully RV2 and RV3 to run 100 mA and then another 100 mA above idle on microphone keydown. (Baseline receive was 370 mA, and initial full clockwise RV2 and RV3 on mic keydown was 580 mA.) And listening on another receiver showed the signal coming up at that point. then running my CW keyer with continuous dashes showed each final warming up just a little, but not much. Unfortunately there is essentially ZERO output to the antenna. So something else must have happened...please see my video (attached)


John Terrell
 

Will do. Thanks again. I'll let you know if there's any output.


John Terrell
 

For what it's worth, before I removed the blown 510s I put an inline current meter in, and on keydown the power supply current drain was over 4.5 amps--with a power supply rated for 3 amps. That may be why the screen kept rebooting. D6 was getting very hot very quickly! After removing both 510s the screen was fine and stable, and the receive current draw was about 50 mA. The MOSFETS appear to be very delicate...


John Terrell
 

I'll be installing a 3A fuse in the transceiver--not the power supply. I use different power supplies in different locations. Maybe that should be a recommended mod.


Jerry Gaffke
 

John,

Yes, that is indeed why the screen was misbehaving.

The IRF510 is surprisingly robust, it's hard to kill on the uBitx in normal use.
If running full power for long periods of time you need a bigger heatsink, but then
it's pretty much bulletproof assuming you keep the power supply voltage at 12v or so.

Any FET in that position will blow if RV2,RV3 are set for too much current.

Receive current was probably more than 50ma.

It would be unusual if RV2 and RV3 were not properly set before shipping from hfsignals.

Jerry, KE7ER



On Thu, Jan 21, 2021 at 01:15 PM, John Terrell wrote:

For what it's worth, before I removed the blown 510s I put an inline current meter in, and on keydown the power supply current drain was over 4.5 amps--with a power supply rated for 3 amps. That may be why the screen kept rebooting. D6 was getting very hot very quickly! After removing both 510s the screen was fine and stable, and the receive current draw was about 50 mA. The MOSFETS appear to be very delicate...


John Terrell
 

Thanks again Jerry. FWIW, I used to be an electronics repair tech at a factory that made large, 200 amp switchers about 40 years ago. But I'm not an RF engineer. 
I've done most of what you suggested. I looked for shorts. Couldn't find a burned component or trace. I removed, tested and re-installed T11. It's fine. I even cut the trace from T7 to pin 13 on relay KA1. That made no difference. 
 
There is output, enough to hear on my IC-7600 on another antenna (a separate magnetic receiving loop). Connecting or disconnecting the T7 - KA1 link (I cut the trace and repaired it with a short length of hookup wire) makes no difference. It's enough to pin the needle on direct wire connection to a field strength meter but not enough to get the needle to budge on two separate wattmeters. I can adjust output to the drivers using RV1, but that's it. Basically it has about as much output as my MFJ antenna analyzer. And this is with RV2 and RV3 adjusted as you instructed, raising idle current 100 mA for each MOSFET.
 
I think this unit is a great way to get people on the air with a HF rig without having to spend a lot of money. And I've learned quite a bit myself. But I'm done with this board for now.  I don't know what's wrong with it. I'd rather switch it out. How can I go about doing that?
 
John Terrell
N6LN


Evan Hand
 

John,
There is a static voltage chart in the Files section of this site.  Look for an Excel file from VU2ZAP.  It is for a v4, but the RF amp sections are pretty much the same, so the voltages should be good.  The static voltages around Q90 are a good indication if the transistor is blown.

73
Evan
AC9TU


Jerry Gaffke
 

John,

If it's getting your goat, perhaps set it aside for a week or two.
Maybe ask for some help at a local amateur radio club.
But I think you are close to solving this.

Sounds like you have the IRF510's set to 100ma of idle drain current each, so the IRF510's are ok.
And that when you put an RF wattmeter between TP7 and ground when transmitting in CW mode
you see nothing on that meter, even when the trace from TP7 to the first relay is cut.
So it seems you are not getting RF power into the IRF510 gates from the driver stage.

Evan is right, next step is to check static voltages on all those driver transistors with a DVM when
attempting to transmit in SSB mode with no audio.  Should see the emitter somewhat above ground
(indicating quiescient emitter current) and the base about 0.6 volts above the emitter (indicating
the transistor is biased into the active region).
Here is the file from VU2ZAP:  https://groups.io/g/BITX20/files/VU2ZAP

If the static checks look ok, I suggest you next look for an RF signal that gets progressively stronger
at TP2,3,4,5,6 and the IRF510 gates while attempting to transmit in CW mode.
You could build a diode RF probe (a 1n5711 will be easier to find than a 1n34a and sensitive enough)  
    http://www.n5ese.com/rfprobe2.htm
or you could clip 3 feet of wire to each test point as an antenna and look for it with your other receiver
(but make sure the wire is insulated so it doesn't short to something like ground or 12vdc).

Jerry, KE7ER


On Sat, Jan 23, 2021 at 05:07 PM, Evan Hand wrote:
There is a static voltage chart in the Files section of this site.  Look for an Excel file from VU2ZAP.  It is for a v4, but the RF amp sections are pretty much the same, so the voltages should be good.  The static voltages around Q90 are a good indication if the transistor is blown.


John Terrell
 

Thanks for the spreadsheet of voltages. I've put my measured voltages in a spreadsheet and I'm uploading it. Most of the voltages in the driver section seem to fit, along with the finals too. I know the spreadsheet was originally for a bitx v4, but a few things look odd compared with my board. First, some of the voltages seem reversed regarding transmit vs receive. Also, Q70's SMD pinout different regarding EBC from the 2n3904's. And it seems the v6 has no Q71-73. 

Unless I'm mistaken it seems that the drivers and the finals are working okay but they're not connected. And I still don't see any evidence of a heat damaged part.

I'll have the suggested 1N34 probe ready in a couple of days.

Thanks,

John Terrell
N6LN


John Terrell
 

Forgot the IRF510 voltages. Here is the updated spreadsheet


 

Q70 voltages of Base and emitter is interchanged I will correct the master in the files section.

Thanks!

Raj

At 25/01/2021, you wrote:
Thanks for the spreadsheet of voltages. I've put my measured voltages in a spreadsheet and I'm uploading it. Most of the voltages in the driver section seem to fit, along with the finals too. I know the spreadsheet was originally for a bitx v4, but a few things look odd compared with my board. First, some of the voltages seem reversed regarding transmit vs receive. Also, Q70's SMD pinout different regarding EBC from the 2n3904's. And it seems the v6 has no Q71-73.

Unless I'm mistaken it seems that the drivers and the finals are working okay but they're not connected. And I still don't see any evidence of a heat damaged part.

I'll have the suggested 1N34 probe ready in a couple of days.

Thanks,

John Terrell
N6LN


 

Your power supply is a bit low. if you are using an adjustable one then
increase to about 13V..

V6 has a diode in PS line and that will drop more than half a volt.

Good to see open office in use.

Raj

At 25/01/2021, you wrote:
Forgot the IRF510 voltages. Here is the updated spreadsheet


 

Is there continuity from antenna jack to ground during transmit - to make sure one of the PTHs have not burnt.

Raj


 

If IRFs were blown then D6 would have failed. Please replace with a robust diode or bypass for a test!

Raj

At 25/01/2021, you wrote:
Is there continuity from antenna jack to ground during transmit - to make sure one of the PTHs have not burnt.

Raj


_Dave_ AD0B
 

If you still have not fixed it, will throw in a data point, with a description of how I repaired one.

I was  given a ubitx to look at as it was not transmitting. Turns out the resistors r87 and r88 on the emitters of q911 and q912 had both fractured.

To diagnosis it I used an oscilloscope and measured rf out of every stage of the  transmitter key down. When I found the 911 and 912 stage not showing gain I replaced them to find that the it didn't change anything. Took a 10 diopter (2.5x) loupe to see the cracks in the resistors. I   reinstalled the original transistors and put in conventional resistors.

Testing under load and measuring output of large stages can quickly pin down which area needs more attention. My oscilloscope is a 20mhz design nothing fancy. I don't trust the calibration but the relative numbers showed the problem.
 
I would not have guessed that there could be enough current in that circuit to blow the resistors yet not damage the 2n3904s 

I have uploaded a file in a thread on the ubitx20 forum that shows the power flow of a v6 transmitter and receiver.
--
73
Dave
ADOB
Raduino bracket and Ham_Made_Keys