reviving my sbitx v3
|
Some sources the IRF shipped is some cheap transistor reprinted and such may blow up on power up.
toggle quoted message
Show quoted text
Raj On 23/03/2025 2:04 AM, Gordon Gibby KX4Z via groups.io wrote:
It is unfortunately very difficult to make conclusions; Evan is doing his best to help.? ?I'm still thinking the actual measurements are needing careful repeat and clarification; for example the sources of the final MOSFETS are solidly SOLDERED TO THE GROUND PLANE, so yes, they should always show a zero ohms to ground.? ? The DRAINS go to capacitors and the power supply so will show a transient response. |
|
Hi Mike,
?
I am sorry to read of your troubles.? Here are some thoughts interspersed in our post:
?
Evan? ? So I have had everything functional for a few days. I managed to get the bias set, spent a good deal of time calibrating the output, which I also have a question about that I will mention further down. No sooner than I was about to put it together and in service, I blew a final. What were you doing when the final blew? I replaced the 510, and as I was considering what caused the last 510 to die, all went dark. As there is variability between batches and manufacturers, it is best always to replace the IRF510s as a pair from the same order. Sitting in steady state receive, I managed to get a short to ground. Do you have any idea where the short may have occurred?? What were you doing when this happened? I have pulled and checked Q9, Q101, and both IRF510s. All checked out, even though one of the 510s began to smoke. If one of the IRF510s smoked, I would replace both after figuring out the reason for the overload.? Maybe the bias is set too high. With these still removed, I have checked Q22 and Q29, both drains show continuity to ground as well as Q22 having continuity to source. I've been checking paths for shorts/opens, etc.? ? ? It would be difficult to measure the resistance of Q22 and Q29 in the circuit as D32 and D33 are across the Q22 source to drain.? To verify, you can take out the diodes or remove the MOSFETs for testing. ? You can verify that the resistance across the MOSFET source to drain is very high with no power. ? ?I found that checking the pads for the 510s source, that there is brief continuity to ground, milliseconds, before opening. Is this possibly the string of caps , C104,106,107,110 between the T7 primary and the +12v supply? Do you mean the drains of the IRF510?? The source is connected to ground. ?Next question, this last attempt to set the bias resulted in no change when adjusting the pot. Current is ~.9A in receive and bounces around 1-1.3A ptt engaged, no drive, 40m, usb. Removed pot, checked okay. Did you start with the PABIAS1 Pot fully counterclockwise, note the current, and then SLOWLY adjust the pot until the current starts to increase?? A .3A current variation seems high.? Without audio in USB mode, the current should not vary more than .1A. ? Now, the calibration question. This supposed to be a scale adjustment as I understand, but relative to what? Which signal is being adjusted? Going from the default values, direction and amount of adjustment is very erratic from band to band. Some have had to be be adjusted by a factor of 10. The scale value is used to set the gain of the signal inside the software.? The higher the value, the more audio IF is sent to the modulator.? This is the AUDIO-MOD signal. So I am going to clean the board up the best I can and look at it again?tomorrow. Since I have quite a few of the components removed, I will leave them until I hear back from you, in case I need to pull anymore.? ?Don't feel to pressed to think about this right away, I know it's mouthful! I'm definitely learning a good deal with this project though, as frustrating as it may be. Thanks again.? What software are you running?? The 64-bit upgrade from JJ (W9JES) has a toolbox app that helps adjust the scaling factor.? It is still a tedious task.? I start from the default settings in the hw_settings_v3.ini file and adjust by no more than 10% each time.? The power will not increase at some point, and going higher can decrease the output.??Note that the higher frequency bands cannot get to 25 watts.? The output starts to drop at 20 meters for some of my rigs. ? Others in the group may have suggestions as well. ? I hope the above helps. 73 Evan AC9TU |
|
Sorry to interrupt the conversation.
I'm only interested in what power supply you use during operation and what you use to adjust the Bias.
Just so you know, I've had the sBitx v2 since April 2023 and only started using it in mid-November 2023. Until then, I had a lot of external work and busyness. At first, I had no output power because the D27 in the power measurement circuit was faulty and did not give good data for operation.
Then I managed to make several improvements in heat dissipation.
Since the device was initially powered by a battery, one of the output stage FETs smoked during a measurement.
I measured the FETs in advance for opening voltage and installed almost identical ones. The Bias setting was 12.5V at 2.00 A I did it with current limitation.
I am currently using the v2 with a voltage of 12.5V and a current limit of 6.50A, and the FETs I replaced are working.
I am writing this as an encouragement that the device works well with proper use.
I am currently close to 10,000 QSOs and I use it on 3 bands (7-14-21MHz) with a vertical loop antenna mostly in digi modes.
I know that proper operation is in the details and I paid attention to where and what needs to be changed.
--
Gyula HA3HZ |
开云体育Evan?
? So I have had everything functional for a few days. I managed to get the bias set, spent a good deal of time calibrating the output, which I also have a question about that I will mention further down. No sooner than I was about to put it
together and in service, I blew a final. I replaced the 510, and as I was considering what caused the last 510 to die, all went dark. Sitting in steady state receive, I managed to get a short to ground. I have pulled and checked Q9, Q101, and both IRF510s.
All checked out, even though one of the 510s began to smoke. With these still removed, I have checked Q22 and Q29, both drains show continuity to ground as well as Q22 having continuity to source. I've been checking paths for shorts/opens, etc.?
?I found that checking the pads for the 510s source, that there is brief continuity to ground, milliseconds, before opening. Is this possibly the string of caps , C104,106,107,110 between the T7 primary and the +12v supply?
?Next question, this last attempt to set the bias resulted in no change when adjusting the pot. Current is ~.9A in receive and bounces around 1-1.3A ptt engaged, no drive, 40m, usb. Removed pot, checked okay. Now, the calibration question. This
supposed to be a scale adjustment as I understand, but relative to what? Which signal is being adjusted? Going from the default values, direction and amount of adjustment is very erratic from band to band. Some have had to be be adjusted by a factor of 10.
So I am going to clean the board up the best I can and look at it again tomorrow. Since I have quite a few of the components removed, I will leave them until I hear back from you, in case I need to pull anymore.?
?Don't feel to pressed to think about this right away, I know it's mouthful! I'm definitely learning a good deal with this project though, as frustrating as it may be. Thanks again.?
-------- Original message --------
From: "Evan Hand via groups.io" <elhandjr@...>
Date: 3/10/25 10:10 PM (GMT-05:00)
Subject: Re: [BITX20] reviving my sbitx v3
Michael,
?
The first step is to verify that the Q10 Drain goes to zero when TX goes to 3 volts (or higher).? If that happens, Q10 is OK.? If not, it is bad.
?
Once Q10 works, verify that the Q9 tab (Drain) goes to 12 volts when the TX signal drives the Q10 drain low.? If that does not happen, remove Q9 and check the operation of Q101.? Do this with Q9 removed to prevent damaging both devices again.
?
Q101 drain goes to zero when TX and Q10 are off.? When TX is low, the Q101 gate should go above 5 volts to make Q101 turn on.? When Q10 turns on with the TX signal, Q101 should turn off.? If Q9 is removed, you can verify that the Drain-to-source resistance
is very high with an Ohm meter.? Q101 Drain to Source resistance should be close to zero when TX and Q10 are on.
?
An alternative approach is to remove Q101 so there is no short on the 12V line during receive, Verify and replace bad parts, and then install a verified working Q101.? I built a test jig on a small copper-clad board with wires to insert into a component
tester.? I hold the device with tweezers to connect each lead to a separate pad.
?
?
73
Evan
AC9TU
-- Mike - K8MCB? |
开云体育Evan, or anyone else who may know. Looking at the sBitx v3 schematic, Share github, named sBitx v3 production, I am unable to locate D8 or D9. I lost D9 in the process of replacing yet another pair of 510s. It is in such close proximity that
it got heated and fell off the board.?
?It appears that their purpose is gate protection of 510s. Anyway, does anyone know what diode I need to put here?
-------- Original message --------
From: "Evan Hand via groups.io" <elhandjr@...>
Date: 3/10/25 10:10 PM (GMT-05:00)
Subject: Re: [BITX20] reviving my sbitx v3
Michael,
?
The first step is to verify that the Q10 Drain goes to zero when TX goes to 3 volts (or higher).? If that happens, Q10 is OK.? If not, it is bad.
?
Once Q10 works, verify that the Q9 tab (Drain) goes to 12 volts when the TX signal drives the Q10 drain low.? If that does not happen, remove Q9 and check the operation of Q101.? Do this with Q9 removed to prevent damaging both devices again.
?
Q101 drain goes to zero when TX and Q10 are off.? When TX is low, the Q101 gate should go above 5 volts to make Q101 turn on.? When Q10 turns on with the TX signal, Q101 should turn off.? If Q9 is removed, you can verify that the Drain-to-source resistance
is very high with an Ohm meter.? Q101 Drain to Source resistance should be close to zero when TX and Q10 are on.
?
An alternative approach is to remove Q101 so there is no short on the 12V line during receive, Verify and replace bad parts, and then install a verified working Q101.? I built a test jig on a small copper-clad board with wires to insert into a component
tester.? I hold the device with tweezers to connect each lead to a separate pad.
?
?
73
Evan
AC9TU
-- Mike - K8MCB? |
开云体育I figured it had something to do with the reception, as when I first turned it back on and the receiver appeared to be "deaf", but yes, right around the halfway point it picked up again. I appreciate the help getting the Tx switching fixed too!
Thx and 73
-------- Original message --------
From: "Evan Hand via groups.io" <elhandjr@...>
Date: 3/13/25 8:50 PM (GMT-05:00)
Subject: Re: [BITX20] reviving my sbitx v3
Glad you fixed it!? The other pot is used to adjust the "bias" for the mixer:
?
?
?
The adjustment does not seem critical.? I would set it to the midpoint and verify that the receiver is sensitive and that the transmitted CW signal does not have spurs.
?
73
Evan
AC9TU
-- Mike - K8MCB? |
开云体育Evan, I finally got it going! My problem was having more than one problem at a time. Once I replaced everything in the circuit, fixed dry or bridged joints, and put a new set of finals in, it now switches and has output. However, before I even
touch the bias pot, does the other pot have anything to do with this circuit? It doesn't appear to, but I mistakenly put it to full ccw with the bias pot too. What is the pot for and how should I set it?
-------- Original message --------
From: "Evan Hand via groups.io" <elhandjr@...>
Date: 3/10/25 10:10 PM (GMT-05:00)
Subject: Re: [BITX20] reviving my sbitx v3
Michael,
?
The first step is to verify that the Q10 Drain goes to zero when TX goes to 3 volts (or higher).? If that happens, Q10 is OK.? If not, it is bad.
?
Once Q10 works, verify that the Q9 tab (Drain) goes to 12 volts when the TX signal drives the Q10 drain low.? If that does not happen, remove Q9 and check the operation of Q101.? Do this with Q9 removed to prevent damaging both devices again.
?
Q101 drain goes to zero when TX and Q10 are off.? When TX is low, the Q101 gate should go above 5 volts to make Q101 turn on.? When Q10 turns on with the TX signal, Q101 should turn off.? If Q9 is removed, you can verify that the Drain-to-source resistance
is very high with an Ohm meter.? Q101 Drain to Source resistance should be close to zero when TX and Q10 are on.
?
An alternative approach is to remove Q101 so there is no short on the 12V line during receive, Verify and replace bad parts, and then install a verified working Q101.? I built a test jig on a small copper-clad board with wires to insert into a component
tester.? I hold the device with tweezers to connect each lead to a separate pad.
?
?
73
Evan
AC9TU
-- Mike - K8MCB? |
Raj vu2zap
HA3HZ