|
I've had and enjoyed a somewhat modified v4 uBitx for years. Last week we had some area thunderstorms...nothing nearby, but I suspect static buildup on the antenna has zapped my rig. It still receives and appears to operate normally, but has no power out when transmitting on CW or SSB. Even though I added the back-to-back diodes some time back, I immediately checked Q90 and it was bad, so I replaced it, but voltages around that transistor are screwy (Base at 8.6 and Collector and emitter over 7), so something else is wrong and it still won't transmit. When I key it with a nearby receiver on frequency, I do hear it coming on and it modulates, so something in the final transmitter stages I suspect. I plan to check the finals, but with grand kids this weekend I have no time to be on the bench, so I thought I'd post to get any suggestions for when I get back into it. Any thoughts?
=Vic=
|
Vic,
R80 is blown. Replace it!
Raj
toggle quoted message
Show quoted text
At 04/06/2022, you wrote: I've had and enjoyed a somewhat modified v4 uBitx for years. Last week we had some area thunderstorms...nothing nearby, but I suspect static buildup on the antenna has zapped my rig. It still receives and appears to operate normally, but has no power out when transmitting on CW or SSB. Even though I added the back-to-back diodes some time back, I immediately checked Q90 and it was bad, so I replaced it, but voltages around that transistor are screwy (Base at 8.6 and Collector and emitter over 7), so something else is wrong and it still won't transmit. When I key it with a nearby receiver on frequency, I do hear it coming on and it modulates, so something in the final transmitter stages I suspect. I plan to check the finals, but with grand kids this weekend I have no time to be on the bench, so I thought I'd post to get any suggestions for when I get back into it. Any thoughts?
=Vic=
|
If R80 is blown and it still does not transmit then C80 needs to be replaced.
Raj
toggle quoted message
Show quoted text
At 04/06/2022, you wrote: Vic,
R80 is blown. Replace it!
Raj
At 04/06/2022, you wrote:
I've had and enjoyed a somewhat modified v4 uBitx for years. Last week we had some area thunderstorms...nothing nearby, but I suspect static buildup on the antenna has zapped my rig. It still receives and appears to operate normally, but has no power out when transmitting on CW or SSB. Even though I added the back-to-back diodes some time back, I immediately checked Q90 and it was bad, so I replaced it, but voltages around that transistor are screwy (Base at 8.6 and Collector and emitter over 7), so something else is wrong and it still won't transmit. When I key it with a nearby receiver on frequency, I do hear it coming on and it modulates, so something in the final transmitter stages I suspect. I plan to check the finals, but with grand kids this weekend I have no time to be on the bench, so I thought I'd post to get any suggestions for when I get back into it. Any thoughts?
|
Well I found what was causing the voltage problems...after checking all around the system I recalled that I had replaced a leaded 2N3904 with a 2N2222A for Q90 since they are equivalent, right? Well, yes, but the pin outs are reversed and even further from the original pin layout of the original SMD 3904. Correcting that mistake now gives correct voltages on all transmit transistors from Q90 on through the finals.
However, still no power out. It is oscillating on frequency and sounds fine...just no measurable power. Raj, I do get blocking action on C80, so can I just bridge it with a leaded cap of the appropriate value? I haven't touched RV1, but maybe I just swing that through its range as well. After spending literally hours on it this afternoon, though, I am going to take a break before moving forward now.
=Vic=
|
Vic,
Remember to check R83, no need to desolder it but check in situ with a multimeter.
Raj
toggle quoted message
Show quoted text
At 07/06/2022, you wrote: Well I found what was causing the voltage problems...after checking all around the system I recalled that I had replaced a leaded 2N3904 with a 2N2222A for Q90 since they are equivalent, right? Well, yes, but the pin outs are reversed and even further from the original pin layout of the original SMD 3904. Correcting that mistake now gives correct voltages on all transmit transistors from Q90 on through the finals.
However, still no power out. It is oscillating on frequency and sounds fine...just no measurable power. Raj, I do get blocking action on C80, so can I just bridge it with a leaded cap of the appropriate value? I haven't touched RV1, but maybe I just swing that through its range as well. After spending literally hours on it this afternoon, though, I am going to take a break before moving forward now.
=Vic=
|
Well, I am puzzled. All transistors from Q90 through the finals now have correct bias and indicated voltages (given a slightly reduced power supply, so typically 0.2v lower everywhere) and I can hear a CW or modulated signal in a nearby receiver. As I adjust RV1 the signal in the nearby receiver goes up and down. It was and is again set for maximum, so I think everything through Q90 is probably OK. I've checked R83 as suggested, put another cap in parallel with C80, and nothing makes any difference. I confirmed the power meter is still working with another QRP rig. Other than the area I have been working around Q90, there are no "distressed" looking traces or components.
So what now? If the finals were "dead" would they still bias correctly?
=Vic=
|
The proper way would be to build a diode RF probe for use with your DVM.
Should be roughly 10dB of power gain (so a voltage gain of around 3)
going from TP3 to TP4, and from TP4 to TP5, and TP5 to TP6, TP6 to TP7.
The improper way would be to take a 12 inch piece of wire for use as an antenna
and touch it to each test point in succession, see if that receiver hears an
increasingly strong signal.
If the IRF510 finals were dead, they would NOT bias correctly.
I'm betting one of the transformer wires has opened up or is shorted.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Tue, Jun 7, 2022 at 01:01 PM, Vic WA4THR wrote:
Well, I am puzzled. All transistors from Q90 through the finals now have correct bias and indicated voltages (given a slightly reduced power supply, so typically 0.2v lower everywhere) and I can hear a CW or modulated signal in a nearby receiver. As I adjust RV1 the signal in the nearby receiver goes up and down. It was and is again set for maximum, so I think everything through Q90 is probably OK. I've checked R83 as suggested, put another cap in parallel with C80, and nothing makes any difference. I confirmed the power meter is still working with another QRP rig. Other than the area I have been working around Q90, there are no "distressed" looking traces or components.
So what now? If the finals were "dead" would they still bias correctly?
=Vic=
|
Continuity between C80 input and L1 ?
Resolder L1/2/3/4
Assuming static killed some parts check D1/2 - if a diode has failed you will get lower Rx sensitivity.
Raj
|
Hi Vic
Did you say that there was no power out on the key and on ssb with q90 in correctly??
Much of the static bias son the finals is set by 12v, ground and a 7805 regulator.
Have you looked at the output connector to the radio?
--
73
Dave
|
Possible open trace to antenna connection?
73 David ac9xh
On Wednesday, June 8, 2022, 08:00:46 AM EDT, _Dave_ AD0B <davesters@...> wrote:
Hi Vic
Did you say that there was no power out on the key and on ssb with q90 in correctly??
Much of the static bias son the finals is set by 12v, ground and a 7805 regulator.
Have you looked at the output connector to the radio?
--
73
Dave
|
Thanks, Jerry. I think that suggestion has helped a lot! I put together a simple RF probe and poked around to find that the signal is all but lost between TP3 and TP4, while all other stages show expected gain.
TP2 100
TP3 322
TP4 7
TP5 100
etc.
I am getting TX voltage and bias at Q911/912, so I assume the primary of T9 is OK, but the pin numbering has me a bit confused. I get continuity between some pads, not others, but I am not sure what the pattern there is. I am seeing some discoloration on one side of T10, but otherwise nothing appears out of the ordinary. Voltages around Q94-5 Q96-7, and Q911-12 all look fine.
Raj...D1-2? Is that the BAT54SL between T1 & T2? Where is that on the board?
|
T10 discolored?? As in burnt insulation?
The TX node is 12vdc during transmit, so a short to ground at TP5 or TP6 would indeed cause some discoloration.
What is the current draw from the battery when transmitting?
I assume you looked at Q92,93.? Those voltages look good?
As I recall, the Kicad decal used for the transformers is that of a 300 mil 6 pin DIP package.
Pin 1 is square, go down to pins 2 and 3, across to pin 4 and up through pins 5 and 6.
Pin 6 is opposite pin 1.
Odd it went from 7 to 100 going from TP4 to TP5, way more gain than we expect..
Perhaps half of that push-pull circuit is shorted out.
Is there a signal at C91 but not at C90?? They should give the same RF probe readings.
How do TP5 and TP6 compare, these should also have equivalent RF probe readings.
Most of those transformers are center tapped, so the actual signal strength is double the voltage you measure against ground.
Of course, for a given signal power level, the voltage you see will depend strongly on the load.
The load presented by the IRF510 in particular varies wildly, especially with frequency, but also with the bias point and power level out.
You can get pretty far with just an ohmmeter and a hard look at the schematic,
For example, the resistance seen between the antenna wire and TP7 should be near zero with rig power off.
With power off, only the 10/12 meter filter is engaged (L11,12,13), so you might look for transmit power on 10m
if you thought you were having trouble back there.? (Vic is not.)
Also, we have in the distant past had reports of poorly machined BNC socket connectors, where the center pin of the?
antenna coax does not make a good connection with the BNC socket.
Good to hear that my remark might have helped.
I try to post something on topic occasionally so the moderators don't kick me off the forum.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Thu, Jun 9, 2022 at 12:10 PM, Vic WA4THR wrote:
Thanks, Jerry. I think that suggestion has helped a lot! I put together a simple RF probe and poked around to find that the signal is all but lost between TP3 and TP4, while all other stages show expected gain.
|
If you had enough current through a primary winding of T10 to burn it,
then R961 is likely cooked as well.?
Not sure why you have so much voltage gain between TP4 and TP5 if T10 is damaged.
Perhaps probe a bit more around there and see if things are balanced.
That RF probe should the show the same voltages on pins of Q92 and Q96 when transmitting.
Should see the same DC voltages too.
toggle quoted message
Show quoted text
On Thu, Jun 9, 2022 at 04:58 PM, Jerry Gaffke wrote:
T10 discolored?? As in burnt insulation?
The TX node is 12vdc during transmit, so a short to ground at TP5 or TP6 would indeed cause some discoloration.
|
More probing...voltage checks on transistors Q92-3, Q95-7, Q911-912 all read C:10.0, B:0.9, E:0.2 today
RF probe checks:
TP2: 103
TP3: 231
TP4: 5? <== so this is the oddball
TP5:152
TP6:142
TP7:445
TP8:1120
TP10: 1035
TP11:1033
C90: 5 <== certainly agrees with TP4
C91: 2 <== even lower!
So does this suggest that T9 itself is the problem? The base on it is not discolored and wiring looks OK. I am guessing this is a 3-layer board as the traces are hidden between what I see on top and the mostly ground plane on the bottom? T10 looks a little funky, but the fact that the bias is OK on the transistors suggests continuity there.
=Vic=
|
Vic,
I am 99% sure it is only two layers for the board.?
I will look at my v4 again to see what I get for voltages using my oscilloscope ?
73
Evan
AC9TU
?
|
I've got a v3, but I believe all versions are just 2 layer boards.
Adding internal layers would be a considerable jump in price.
Could well be that the issue is around T9, and the discoloration you reported on T10 doesn't matter.
That seems more consistent with your RF probe readings of TP4 having a very small RF signal.
Let's assume that is the case.
I assume your DC voltage readings at pins of Q911,912, 92,93, 96,97 are all still
in agreement with what has been reported by others when transmitting.
For transistors in parallel, you only need to check one of the two transistors.
Note that the TX node supplies 12v to those transistor collectors only when transmitting.
Look directly at the base pin of Q911 using the RF probe, should see roughly the same
signal level as you see at TP3.? If not, then perhaps C83 is fractured open.
Look at pins 4,3,2 of T9 using the RF probe.? Pin 4 should have a very healthy signal,
much bigger than TP3.? Pin 3 should be about half that.? Pin 2 should have no signal?
(but does have a 12v DC supply present during transmit).? If that is true then the primary
of T9 is ok and Q911,912 are doing their job.?
If not then remove C90 and see if that improves things, since a shorted secondary?
can cause the primary of T9 to not have appropriately large RF signals.
If you still don't see the big signals at T9 pin 4, then I would probably cut the trace from T9 pin 4
that goes to the collectors of Q911,912.? From the collectors of 911 and 912 add a 200 ohm 1/4 watt resistor
to 12v.? Do you now see a good RF signal signal on those collectors?? If so, then T9 is bad.?
If T9 primary checks out as outlined above, double check that there is still no RF signal
at C90 and C91.? At this point I would remove C90 and short T9 pin 1 to ground.
If T9 is working properly, you should now see an RF signal at TP4 that is about half of
what is on T9 pin 4.
If T9 is not working properly, then pull it off the board by clipping the wires, leaving nubbins long
enough to grab with a pair of needle nose pliers, and pull them out of the board with help from
a soldering iron.? Blow the hot solder out of the 6 holes for T9.? Rewind T9 with new wire.
and re-install.? Verify you now have a good RF signal at TP4, then replace C90 and remove that?
ground to T9 pin 1.
Or, if that's all too complicated, you could just shotgun it:? Replace or rewind T9.
Jerry, KE7ER
.
toggle quoted message
Show quoted text
On Fri, Jun 10, 2022 at 01:47 PM, Vic WA4THR wrote:
More probing...voltage checks on transistors Q92-3, Q95-7, Q911-912 all read C:10.0, B:0.9, E:0.2 today
RF probe checks:
TP2: 103
TP3: 231
TP4: 5? <== so this is the oddball
TP5:152
TP6:142
TP7:445
TP8:1120
TP10: 1035
TP11:1033
C90: 5 <== certainly agrees with TP4
C91: 2 <== even lower!
So does this suggest that T9 itself is the problem? The base on it is not discolored and wiring looks OK. I am guessing this is a 3-layer board as the traces are hidden between what I see on top and the mostly ground plane on the bottom? T10 looks a little funky, but the fact that the bias is OK on the tran
|
Vic,
As promised, here are the scope captures of the test points TP2 through 7.
Hope this helps.
73
Evan
AC9TU
|
Vic,
Your power supply is also low,please check voltage. You should
get more voltage on the collector.
On 11/06/2022 2:17 AM, Vic WA4THR via
groups.io wrote:
toggle quoted message
Show quoted text
More probing...voltage checks on transistors Q92-3, Q95-7,
Q911-912 all read C:10.0, B:0.9, E:0.2 today
RF probe checks:
TP2: 103
TP3: 231
TP4: 5? <== so this is the oddball
TP5:152
TP6:142
TP7:445
TP8:1120
TP10: 1035
TP11:1033
C90: 5 <== certainly agrees with TP4
C91: 2 <== even lower!
So does this suggest that T9 itself is the problem? The base on it
is not discolored and wiring looks OK. I am guessing this is a
3-layer board as the traces are hidden between what I see on top
and the mostly ground plane on the bottom? T10 looks a little
funky, but the fact that the bias is OK on the transistors
suggests continuity there.
=Vic=
|
Evan,
?
You¡¯ve done a fantastic job with these oscilloscope screen copies.
But would it be even more interesting to actually have a scope screen of all the test points?? (20 i think)
It¡¯s a little more work, but it¡¯s worth it for troubleshooting and homebrewing.
Why not also integrate some screenviews of the hot spots (Ex Q90, mixer,...)
cdt
|
Evan,
I forgot, specify the bandwidth and model of your oscilloscope, the measurements can be different depending of our oscillo.
Thank's
?
|
Vic WA4THR
Raj vu2zap
_Dave_ AD0B
Gerard