Topics

receive works great, but no trans power, #ubitx-help

Don - KM4UDX
 

My beloved and long suffering (like my wife) uBitx has no output power in transmit. 
When I key down the display changes from "USB A:" to "CW"
When I PPT it changes from "USB A:" to "TX".

So all that looks good. But. Neither produces a output at the antenna.

What should I check...fix...validate...replace..?

Background...receive works great!!!!!
CEC 1.9x
no mods beyond the firmware update, which seemed to work fine.
And I got no transmit power on the original firmware anyway. 

Thank you for any help!!!!

Don
km4udx

Lowell Haney NE4EB
 

Have you checked the finals for 12 VDC on the tabs?

Don - KM4UDX
 

Yes. 12.45, same as power source. Also, when i hit the PTT w/o audio, both finals have 3.8v/12.45/0 for g/d/s.  The gate goes from 0 to 3.8 when I hit the PPT.

Don - KM4UDX
 

I watched a bunch of youtubes, and tested the IRF510s with my DVMM.  They flow only one way between two pins, so they appear good. And the voltage on the pins is close to expected values. So I have to assume the problem is BEFORE the finals, or the signal that tells them to get going on TX is not there?  Oh wise ones, guide your lowly student to the light......

I'm going to follow the TX Quiescent Testpoings, Group3, then Group2 and finally group 1, based on the very nice "Diagnostic Guide" by Ufi Auttorri, May 16, 2018.

My fear is that the values will be fine, so the problem is something stupid that I've done. Which is typical. I'm just saying. 

Thoughts or shortcuts much appreciated...

Clark Martin
 

Build yourself an RF Probe such as at: http://www.qsl.net/g3pto/probe.html.  You can use it to trace the signal (or lack thereof) through the transmit path.


Clark Martin
KK6ISP

On Jul 15, 2018, at 8:54 PM, Don - KM4UDX <dontAy155@...> wrote:

I watched a bunch of youtubes, and tested the IRF510s with my DVMM.  They flow only one way between two pins, so they appear good. And the voltage on the pins is close to expected values. So I have to assume the problem is BEFORE the finals, or the signal that tells them to get going on TX is not there?  Oh wise ones, guide your lowly student to the light......

I'm going to follow the TX Quiescent Testpoings, Group3, then Group2 and finally group 1, based on the very nice "Diagnostic Guide" by Ufi Auttorri, May 16, 2018. 

My fear is that the values will be fine, so the problem is something stupid that I've done. Which is typical. I'm just saying. 

Thoughts or shortcuts much appreciated...
_._,_._,_

Jerry Gaffke
 

Yes, an RF probe is a good idea to have around when debugging a uBitx.

That looks like a reasonable circuit, but good luck finding a GEX66.
Seems to be a germanium point contact diode from 50 years ago, meant for UHF,
5v max reverse voltage.
I'd just use a Schottky diode such as the 1n5711 or a BAT54*.
And maybe adjust those cap values, 100pf at the input seems low for under 100mhz. 

See this post, and the thread around it.
    https://groups.io/g/BITX20/message/52563

Jerry



On Sun, Jul 15, 2018 at 09:47 PM, Clark Martin wrote:
Build yourself an RF Probe such as at: http://www.qsl.net/g3pto/probe.html.  You can use it to trace the signal (or lack thereof) through the transmit path.
 

Don - KM4UDX
 

Clark & Jerry -- thank you for the guidance! 

I think I'm getting in over my head here. I'll try some more basic debugging, but if I can't figure it out, I may have to put it in the dump and call it a "lesions learned" experience.

The ratio of working/communicating is getting way too high. I'm sadden that so many other uBitx folks are working the world, but I can't get out of my workbench. sigh. 

Don

Bo Barry <wn4ghv@...>
 

Please post the link to the guide again. Tnx Bo W4GHV 

Jerry Gaffke
 

I certainly wouldn't dump it.
Could sell it on the forum here.
Could state your location in the forum, see if others live close.
Could haul it into a local ham club meeting, see if anybody is 
interested in helping you debug it.
Could park it on the shelf for a week or two, then take another look.

Could be as simple as a bad connection at the antenna BNC connector.
The BNC socket provided with the kit is not machined to tight enough tolerances,
in many cases the middle pin does not make contact.
That was a problem with my uBitx, and others here have reported the same issue.
Try adding wires to go around that BNC connector directly to your dummy load.

Another common point of failure is the torroid transformers, inspect them closely.
Could be that the enamel wasn't scraped off the wire somewhere, and so the solder joint isn't working.
Or a broken wire, they are kind of heavy so rough handling while shipping can break them.
Following the schematic, check with an ohmmeter.

##############
We should have a standard procedure for checkout out the transmitter.
Something like this:

Build a specific version of the diode RF probe, driving A7 for voltage readings to the LCD.
Create special Raduino firmware to support the debug procedure.

Set up the rig to transmit CW at 7mhz, disconnect PA-PWR to the IRF510's so they don't burn up.
Verify DC and AC levels using DVM and the RF probe at all points through T2, Q90, Q911, Q912, Q92,93,96,97.
Compare your readings with a chart for a properly functioning rig.

Apply PA-PWR, add a dummy load, measure power at IRF510 drains and at the dummy load.

Drive mike input from sidetone, go to special code that puts rig in SSB transmit mode with sidetone operating.
Verify DC and AC levels through Q6, D6, Q40,41,42, Q20,21,22
(The remainder of the transmitter was already checked out in CW mode.)

Something similar could be done for the receiver.
Perhaps generate 7mhz on CLK0 (normally the BFO) and apply that through an attenuator to the antenna port.
Trace this using the RF probe all the way through the 45mhz and 12mhz crystal filters to D5.
Now restore the BFO to 12mhz, disconnect it from the antenna, and unbalance the first mixer using CW-KEY.
Inject 45mhz at CLK2, and verify the product detector and audio chain.
Again, we need to compile a list of expected DC and AC voltages at each stage.

Parts for a good 12W dummy load (parallel four of Mouser part number 283-200-RC) 
plus a diode RF probe (1n5711 or similar, a couple caps and resistors) plus appropriate attenuators (a few resistors)
would cost about $1.  Could be included with every kit.

This capacitance meter using a Nano might also be of interest:
    https://groups.io/g/BITX20/message/52471

Am curious if anybody has a good way to measure inductance using the Nano.
I'd probably measure a 100pf cap using the above capacitance meter, then sweep a parallel combination
of the inductor being tested plus the 100pf cap using an output from the si5351 through a series resistor.
Look for the lowest freq dip using the RF probe.  Other caps such as 10pf, 1nf, ... might also be of use
depending on the size of the inductor.
The si5351 can go from about 4khz to 200mhz, up to 290mhz if pressed.

Jerry, KE7ER


On Mon, Jul 16, 2018 at 07:59 AM, Don - KM4UDX wrote:
Clark & Jerry -- thank you for the guidance! 

I think I'm getting in over my head here. I'll try some more basic debugging, but if I can't figure it out, I may have to put it in the dump and call it a "lesions learned" experience.

The ratio of working/communicating is getting way too high. I'm sadden that so many other uBitx folks are working the world, but I can't get out of my workbench. sigh. 

Gordon Gibby <ggibby@...>
 

If the ham radio operator has another receiver,  they can use it to listen for signal at earlier stages.  

CW would allow them to know exactly what frequency to listen for on earlier stages.  

The way you eat an elephant is one stage at a time.   




On Jul 16, 2018, at 12:50, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I certainly wouldn't dump it.
Could sell it on the forum here.
Could state your location in the forum, see if others live close.
Could haul it into a local ham club meeting, see if anybody is 
interested in helping you debug it.
Could park it on the shelf for a week or two, then take another look.

Could be as simple as a bad connection at the antenna BNC connector.
The BNC socket provided with the kit is not machined to tight enough tolerances,
in many cases the middle pin does not make contact.
That was a problem with my uBitx, and others here have reported the same issue.
Try adding wires to go around that BNC connector directly to your dummy load.

Another common point of failure is the torroid transformers, inspect them closely.
Could be that the enamel wasn't scraped off the wire somewhere, and so the solder joint isn't working.
Or a broken wire, they are kind of heavy so rough handling while shipping can break them.
Following the schematic, check with an ohmmeter.

##############
We should have a standard procedure for checkout out the transmitter.
Something like this:

Build a specific version of the diode RF probe, driving A7 for voltage readings to the LCD.
Create special Raduino firmware to support the debug procedure.

Set up the rig to transmit CW at 7mhz, disconnect PA-PWR to the IRF510's so they don't burn up.
Verify DC and AC levels using DVM and the RF probe at all points through T2, Q90, Q911, Q912, Q92,93,96,97.
Compare your readings with a chart for a properly functioning rig.

Apply PA-PWR, add a dummy load, measure power at IRF510 drains and at the dummy load.

Drive mike input from sidetone, go to special code that puts rig in SSB transmit mode with sidetone operating.
Verify DC and AC levels through Q6, D6, Q40,41,42, Q20,21,22
(The remainder of the transmitter was already checked out in CW mode.)

Something similar could be done for the receiver.
Perhaps generate 7mhz on CLK0 (normally the BFO) and apply that through an attenuator to the antenna port.
Trace this using the RF probe all the way through the 45mhz and 12mhz crystal filters to D5.
Now restore the BFO to 12mhz, disconnect it from the antenna, and unbalance the first mixer using CW-KEY.
Inject 45mhz at CLK2, and verify the product detector and audio chain.
Again, we need to compile a list of expected DC and AC voltages at each stage.

Parts for a good 12W dummy load (parallel four of Mouser part number 283-200-RC) 
plus a diode RF probe (1n5711 or similar, a couple caps and resistors) plus appropriate attenuators (a few resistors)
would cost about $1.  Could be included with every kit.

This capacitance meter using a Nano might also be of interest:
    https://groups.io/g/BITX20/message/52471

Am curious if anybody has a good way to measure inductance using the Nano.
I'd probably measure a 100pf cap using the above capacitance meter, then sweep a parallel combination
of the inductor being tested plus the 100pf cap using an output from the si5351 through a series resistor.
Look for the lowest freq dip using the RF probe.  Other caps such as 10pf, 1nf, ... might also be of use
depending on the size of the inductor.
The si5351 can go from about 4khz to 200mhz, up to 290mhz if pressed.

Jerry, KE7ER


On Mon, Jul 16, 2018 at 07:59 AM, Don - KM4UDX wrote:
Clark & Jerry -- thank you for the guidance! 

I think I'm getting in over my head here. I'll try some more basic debugging, but if I can't figure it out, I may have to put it in the dump and call it a "lesions learned" experience.

The ratio of working/communicating is getting way too high. I'm sadden that so many other uBitx folks are working the world, but I can't get out of my workbench. sigh. 

Don - KM4UDX
 

here is the link to the troubleshooting guide....it is a great document...and thanks to Ufi Auttorri, wherever you are!

https://groups.io/g/BITX20/attachment/52578/0/uBitxTroubleshooting%20manuala.pdf

Peter Parker
 

This fix may be helpful to someone. 

* Symptom: Low then almost no transmit power.  Power meter might move up to about 1w for a short time then drop down to very low.  Tx signal audible in local receiver but weaker than it should be.  PA transistor bias settings seem to be OK, indicating the IRF510s are OK. 

* Probing around with an RF probe indicated the burst from higher to low power was noticed on the PA transistor drain and back through the driver stage. 

* DC voltage check found all were OK except the emitter of Q90. Instead of 3 volts there was only 0.9v. 

* Q90 was removed and replaced with leaded transistor (I used a BC548 which is similar to a 2N3904).  Used adjacent parts to solder leads to for better spacing.  

Q90 turned out to be the problem.  With the new transistor the emitter voltage reads normal and the rig now transmits. 

73, Peter VK3YE

Gordon Gibby <ggibby@...>
 

Q 90 is exactly the one to be destroyed by a strong signal coming in from Antenna.  


*If they are being destroyed by strong signals, that’s just the diode mixer is not protecting them.


On Jul 17, 2018, at 15:50, Peter Parker <parkerp@...> wrote:

This fix may be helpful to someone. 

* Symptom: Low then almost no transmit power.  Power meter might move up to about 1w for a short time then drop down to very low.  Tx signal audible in local receiver but weaker than it should be.  PA transistor bias settings seem to be OK, indicating the IRF510s are OK. 

* Probing around with an RF probe indicated the burst from higher to low power was noticed on the PA transistor drain and back through the driver stage. 

* DC voltage check found all were OK except the emitter of Q90. Instead of 3 volts there was only 0.9v. 

* Q90 was removed and replaced with leaded transistor (I used a BC548 which is similar to a 2N3904).  Used adjacent parts to solder leads to for better spacing.  

Q90 turned out to be the problem.  With the new transistor the emitter voltage reads normal and the rig now transmits. 

73, Peter VK3YE

Jerry Gaffke
 

Curious.
Q90 is the one that would fry if the antenna was hooked up and a nearby QRO station was operating.
Or a nearby electrical storm was operating.

Q13 of the Bitx40 often blow in this way.
The back-to-back diode fix of the Bitx40 is not such a good idea on the uBitx because
the uBitx is wideband up to the 45mhz filter, and the diodes can distort all that incoming RF.

We don't hear much about Q90 blowing on the uBitx, 
likely because the diodes in the first mixer conduct before there's real trouble.
Here's an old thread on this:  https://groups.io/g/BITX20/topic/7715656

Jerry,   KE7ER


On Tue, Jul 17, 2018 at 12:50 PM, Peter Parker wrote:

This fix may be helpful to someone. 

* Symptom: Low then almost no transmit power.  Power meter might move up to about 1w for a short time then drop down to very low.  Tx signal audible in local receiver but weaker than it should be.  PA transistor bias settings seem to be OK, indicating the IRF510s are OK. 

* Probing around with an RF probe indicated the burst from higher to low power was noticed on the PA transistor drain and back through the driver stage. 

* DC voltage check found all were OK except the emitter of Q90. Instead of 3 volts there was only 0.9v. 

* Q90 was removed and replaced with leaded transistor (I used a BC548 which is similar to a 2N3904).  Used adjacent parts to solder leads to for better spacing.  

Q90 turned out to be the problem.  With the new transistor the emitter voltage reads normal and the rig now transmits. 

73, Peter VK3YE

Gordon Gibby <ggibby@...>
 

Stuff that doesn’t get through the filter can fry Q 90 and never get clipped by the diode mixer.....

For example, 2 meters 
TV 

I still don’t buy the distortion argument, otherwise Q90  would be creating exactly the same distortion.   We haven’t heard anyone complaining of that have we?

Gordon KX4Z 



On Jul 17, 2018, at 16:19, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

Curious.
Q90 is the one that would fry if the antenna was hooked up and a nearby QRO station was operating.
Or a nearby electrical storm was operating.

Q13 of the Bitx40 often blow in this way.
The back-to-back diode fix of the Bitx40 is not such a good idea on the uBitx because
the uBitx is wideband up to the 45mhz filter, and the diodes can distort all that incoming RF.

We don't hear much about Q90 blowing on the uBitx, 
likely because the diodes in the first mixer conduct before there's real trouble.
Here's an old thread on this:  https://groups.io/g/BITX20/topic/7715656

Jerry,   KE7ER

On Tue, Jul 17, 2018 at 12:50 PM, Peter Parker wrote:

This fix may be helpful to someone. 

* Symptom: Low then almost no transmit power.  Power meter might move up to about 1w for a short time then drop down to very low.  Tx signal audible in local receiver but weaker than it should be.  PA transistor bias settings seem to be OK, indicating the IRF510s are OK. 

* Probing around with an RF probe indicated the burst from higher to low power was noticed on the PA transistor drain and back through the driver stage. 

* DC voltage check found all were OK except the emitter of Q90. Instead of 3 volts there was only 0.9v. 

* Q90 was removed and replaced with leaded transistor (I used a BC548 which is similar to a 2N3904).  Used adjacent parts to solder leads to for better spacing.  

Q90 turned out to be the problem.  With the new transistor the emitter voltage reads normal and the rig now transmits. 

73, Peter VK3YE

ajparent1/KB1GMX <kb1gmx@...>
 

It the signals are strong enough to fry Q90 the mixer will soon be gone too.
The BE junction is as robust as the mixer diodes.  so whatever you are seeing
is quite strong and could even be ESD.

For ubitx (V3 or 4) a pair of back to back 1n194/4148 diodes from K3 Pin14 to ground
is the best bet as then its only there for RX and has no impact on TX.

Bitx 20 and 40 the solution is different.

Its easily done on the back of the board.

Allison

Jerry Gaffke
 

We have had a couple reports of Q90 frying, but not the mixer.
Perhaps as Gordon notes, these were due to a strong VHF signal.

I believe Gordon has suggested in the past the the failure mode might be 
due to reverse voltage at the BE junction, not forward current.

I have no idea if the back-to-back diodes would significantly contribute to the spurs.
Doubt it would cause trouble where I'm at, as there are no strong local broadcasters.
I should try it, include a strap to remove the diodes if curious about some new spur.
We've got plenty of electrical storms.

Jerry

 

On Tue, Jul 17, 2018 at 01:28 PM, ajparent1/KB1GMX wrote:
It the signals are strong enough to fry Q90 the mixer will soon be gone too.
The BE junction is as robust as the mixer diodes.  so whatever you are seeing
is quite strong and could even be ESD.

For ubitx (V3 or 4) a pair of back to back 1n194/4148 diodes from K3 Pin14 to ground
is the best bet as then its only there for RX and has no impact on TX.

Bitx 20 and 40 the solution is different.

Its easily done on the back of the board.

Allison

Gordon Gibby <ggibby@...>
 

That would make for a very useful experiment!   If you try it, let us know what you find out!

It’s possible that different transistors have different reverse (base-emitter) break down voltages, which would dramatically change the amount of energy deposited on the base-emitter junction by a strong signals.... I just don’t know enough about the devices to say for sure.


On Jul 17, 2018, at 23:30, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

We have had a couple reports of Q90 frying, but not the mixer.
Perhaps as Gordon notes, these were due to a strong VHF signal.

I believe Gordon has suggested in the past the the failure mode might be 
due to reverse voltage at the BE junction, not forward current.

I have no idea if the back-to-back diodes would significantly contribute to the spurs.
Doubt it would cause trouble where I'm at, as there are no strong local broadcasters.
I should try it, include a strap to remove the diodes if curious about some new spur.
We've got plenty of electrical storms.

Jerry

 
On Tue, Jul 17, 2018 at 01:28 PM, ajparent1/KB1GMX wrote:
It the signals are strong enough to fry Q90 the mixer will soon be gone too.
The BE junction is as robust as the mixer diodes.  so whatever you are seeing
is quite strong and could even be ESD.

For ubitx (V3 or 4) a pair of back to back 1n194/4148 diodes from K3 Pin14 to ground
is the best bet as then its only there for RX and has no impact on TX.

Bitx 20 and 40 the solution is different.

Its easily done on the back of the board.

Allison

Jerry Gaffke
 

That strap could cut Q90-base out of the circuit too, in case that causes RX spurs.
Probably OK to leave the back-to-back diodes in there near Q90-base (to the left of C80)
during transmit, as I doubt we have much more than 0.1vrms there.

Unfortunately, I've got a very long list of stuff to do while the weather is good.
Fiddling with the uBitx isn't on it.

Jerry, KE7ER



On Wed, Jul 18, 2018 at 03:53 AM, Gordon Gibby wrote:
That would make for a very useful experiment!   If you try it, let us know what you find out!
 
It’s possible that different transistors have different reverse (base-emitter) break down voltages, which would dramatically change the amount of energy deposited on the base-emitter junction by a strong signals.... I just don’t know enough about the devices to say for sure.

 
 

RCBoatGuy
 

Allison wrote:

"For ubitx (V3 or 4) a pair of back to back 1n194/4148 diodes from K3 Pin14 to ground
is the best bet as then its only there for RX and has no impact on TX." 

If there is a possibility that these diodes could cause spurs during TX (which I doubt), then it seems the safer approach is to connect the back-to-'back diodes from K1 pin 12 to ground.  K1 pin 12 is grounded and isolated from the RX/TX circuitry connected to K1 pin 14  during TX, which guarantees that the diodes can't cause spurs during TX.  Placing the the diodes from K1 pin 12 to ground should still provide the same protection for Q90 during receive as connecting them from K1 pin 14 to ground. 

73,

Carl,  K0MWC