|
Yup, I agree Allison.? ?I think if more people?make the recommended mod to add a third 470 in parallel, we might see more failures of the diode biasing chokes -- and the symptoms?will be sudden loss of almost all power.? ?The replacement I found was basically what you suggested; it came from a bitx 40 since I had no suitable cores on hand.? ?I need to order some!
I have no proof what did in the diodes.? ?I agree that almost anything could; I think it would be a good idea to current limit those diodes connected to a strong supply such as the 10A supply for the radio -- hence I added the 470 ohms.? ? ?The radio could get a very bad reputation if the incidence of failure becomes significant over time.? ?By itself, I don't think anything would fail there, but adding fallible HUMANS to the picture who are making various mods...and I think it is is a perilous trap.?
Thank you for your review!
I hope to get mine buttoned up soon, but with tree work, house work, ARES(R) group work, there is so little time left....and I'm getting older.? ? Can''t even get round to checking investments, and we depend on those for income......
Thanks! Gordon KX4Z
On Mon, Nov 13, 2023 at 5:48?PM ajparent1/kb1gmx < kb1gmx@...> wrote: Gordon,
3 parallel resistor in parallel, that 157 ohms.... In series with 12V to ground through a diode (.7V drop).
Than means 71 ma is flowing through the choke....? Now I assumed 12V DC power if higher its worse.?
Likely cause, it got cooked.
FYI: more than 30ma will not turn on the diodes and real amount more!.
Also if you increasing the current both ends of the LPF needs to be done
that's R201/202 and R203/4.? Two 470 ohm at 12V is about 48ma!
Quick and dirty 47uh choke is a FT37-43 toroid with 12-13 turns should do.
FyI the 1n4007 has a unique property is its charge recombination time is about 100nS
which makes it a terrible rectifier (meaning it barely rectifies) at more than 10mhz.? It was
designed? for 60hz circuits as a rectifier with a controlled avalanche characteristic
to allow for the 1000V capability.? Real PIN diodes are also similar in that they have
similar recombination times, for a larger price.
The 1n4148s? will fry if subjected to excess voltage, they are 70 or 100V PIV (vendors vary on this).
that is why 2 in series are used but damaging one (shorted) will lead to failure of the other.? Excessive
current? (under 100ma) will cook them as well.? The HV bias is normally very low current (under 1mA).
Slips when performing? fixes/mods or even solder splashes are dangerous!
Failures of any of the fast diodes (1n4148s) can add to a lot of extraneous harmonics and other
off behaviours.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
Gordon,
3 parallel resistor in parallel, that 157 ohms.... In series with 12V to ground through a diode (.7V drop).
Than means 71 ma is flowing through the choke....? Now I assumed 12V DC power if higher its worse.?
Likely cause, it got cooked.
FYI: more than 30ma will not turn on the diodes and real amount more!.
Also if you increasing the current both ends of the LPF needs to be done
that's R201/202 and R203/4.? Two 470 ohm at 12V is about 48ma!
Quick and dirty 47uh choke is a FT37-43 toroid with 12-13 turns should do.
FyI the 1n4007 has a unique property is its charge recombination time is about 100nS
which makes it a terrible rectifier (meaning it barely rectifies) at more than 10mhz.? It was
designed? for 60hz circuits as a rectifier with a controlled avalanche characteristic
to allow for the 1000V capability.? Real PIN diodes are also similar in that they have
similar recombination times, for a larger price.
The 1n4148s? will fry if subjected to excess voltage, they are 70 or 100V PIV (vendors vary on this).
that is why 2 in series are used but damaging one (shorted) will lead to failure of the other.? Excessive
current? (under 100ma) will cook them as well.? The HV bias is normally very low current (under 1mA).
Slips when performing? fixes/mods or even solder splashes are dangerous!
Failures of any of the fast diodes (1n4148s) can add to a lot of extraneous harmonics and other
off behaviours.
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
|
Attached is my effort at a "troubleshooting" guide for the T/R LPF switching.
Gordon Kx4Z
|
RATS!! Going carefully back through the record preserved in this thread, and also through my handwritten notes, it turns out that I had *not* made the 470 ohm bias resistor changes on my initial measurements that showed big problems on several bands.Although my first set of notes in my handwritten copy is not dated, I have a report on this thread dated Oct 6, and I have notes that I raised the idle bias on October 6th. -- and that October 6th? measurements (dated) showed still big problems on 20 meters on both 2nd and 3rd harmonics. My handwritten notes and the report here?? /g/BITX20/message/105460? indicate that I didn't make the 470 ohm change until 10/7, and that it didn't fix the 20 meter problem.? ?? So......I no longer have an explanation for how I could have damaged the D20/D21 diodes prior to the poor spectra recorded.? ? ? Although Evan's simulation seems both well done, and? useful, the spec of 2A impulse current is only for a duration of 1uSec.? If the duration is longer, it is probably the time integral of power delivered (=energy) over a short period that determines the temperature the diode dies reach.? ? ?The series dynamic resistance of the diode at higher currents appears to be in single digits.? ? To achieve a time constant of < 1? uSec, any capacitance being charged by the diodes needs to be 0.1 uF or lower....? ? The 100 pf in the circuit certainly meets this value, but SOMEHOW those diodes got shorted.? ?Higher voltage?? (ESD)? ? I presume the diodes were GOOD when inserted (all the others were) and the ONE that I could get a visual on, appeared to be inserted properly.? ?Had they been inserted backwards an entirely different sequence would have happened.? ?So I think they were good when inserted and inserted properly -- and subsequently destroyed BEFORE my very first measurements.?? There may be some sequence of operations or ESD discharge into the unit, that caused that failure.? ? Worse, this is a sneaky failure, because the unit will still produce good RF output -- it just will have lots of harmonics etc.? ?Finding the INCIDENCE would require testing a sample set of delivered units, and the incidence may INCREASE across time as the sequence of operations or ESD events will accumulate across time.? ? It would be difficult to have owners modify their? units to add series resistance there (like me, their chances of damaging OTHER items is significant!) but it might be quite wise to add series resistance in future delivered units, by hand modification of the circuitry.? ?This could be accomplished by tee-peeing a diode and a resistor or by cutting a trace and bridging with a resistor. Any user who makes a measurement using? normal VOM probes in that area might accidentally cause a current spike.? ?Fumble fingers.?? Manufacturing engineers might have more insight into these relative risks and user-error risks than I do.? ? ? Obviously I don't know the incidence of this, but for Ashhar's sake and the manufacturing design's sake, if it is non-trivial, I'm glad that it happened to me! Better to reduce the incidence now. My thoughts, take note that I am NOT a manufacturing engineer!? Ashhar has far more experience than I.? ?This is a great radio, and I'm looking forward to writing an enthusiastic article about? this version.? ? Gordon KX4Z
|
I was RELIEVED to find such a simple cause of the spurious problems I had encountered.? ?I suspect strongly that some false move of mine shorted out those diodes in the process of adding the 470 ohm resistors suggested to improve IMD response on 80 meters.? ??
I have about 5 pages of a "troubleshooting" document now written on the Transmit/Receive circuitry and how it functions and symptoms of failures.? ?I want to check out a thing or two with a simple VOM before it is "ready."? ?
The radio? is in some respects, "sparse" in order to keep costs down.? ?But for hobbyists like myself, it might be good to harden some portions and add some diagnostics.? ?Unless there is a downside that I'm not seeing, adding some resistance in series with D20/D21 (I used 470 ohms) is probably a good thing to protect against goofs such as I experienced.? ?Wow, that would potentially have avoided a ton of effort....
I'm not certain why L201 got fried.? I suspect a series resonance, but it could have been simple mechanical issues since I've been in and around it a few times.? Thank goodness for a choke from an older radio for a replacement!? ?One simple addition to this radio might be some SMT LED diodes (with series resistors) so the "state" of various parts of the circuit were obvious by looking at the board, rather than having to drag out the voltmeter.? ?An LED diode system across R201/R202 would have made it more obvious when I couldn't figure out what had happened to my output power, that the +12V drive to the INPUT diodes of the LPF switching system had disappeared.? ?Likewise, something similar on the T/R switching down around L27 might help? people understand what is happening there.? ? These are certainly not "required" but might make it a more "hobbyist friendly" radio.??
I hope to pull some spectral shots of a few bands in preparation for writing up an article for the NFL Section Newsletter, and then moving on to changing out the MOSFET insulators and adding the zener diodes and this hopefully begin to learn something about the CODE of the radio.? ?I'm surprised that no one has solved the FT8 switching timing issues by now!? ? That's not my "thing" but still surprised.
Can't wait to have this really on the air.? ?I still like the original gtk interface.? ? I just don't need the large left hand side of? the display.? ? Hoping someone comes u with far better CW decoder than FLDGI.? ?I don't need it (copy in my head) but FLDGI is just so sad at that.? ?Some bright person should be able to find the Goetzel algorighm or something and offer a big improvement.? ? And I want a real amplifier relay control output.? ?Experiences with the DE version demonstrated that analog +12 is very sensitive and should not be extended any at all for fear of RF coupling; instead digital signals should be optically isolated and brought out.? ? The WINKEYER that I have has optical output coupling.
Speaking of the WINKEYER, the PIC chip in it? apparently had (past tense) some ESD protection that got clobbered during Field Day or some of my many travels, and I had to get a replacement chip because it was chewing up batteries.? ?Steve, the designer, has put out a few versions in his time and I accidentally ended up with the wrong version of the PIC chip replacement, which caused some consternation and we finally figured it out and now it is working again.? ?That is a sideline for him and I think it is SO difficult to keep everything out there like a big huge company can; I can't find a schematic for the exact version that apparently I have which contributed to our difficulties -- delighted to have THAT working again thanks to his help!? ? It points to how difficult it can be with these various products to keep schematics etc.? ?The sBitx has done pretty well with that!? ? I added some additional diode/zener protection which Steve things is unnecessary -- but having ONE chip bite the dust, I'm thinking belt and suspenders both, is better than just whatever failed inside the PIC chip.? ? The same issues sort of with the sBitx.? ?It works!? Amazing design.? Incredible? step forward for hobbyists.? ?Even better if it can be made even more robust/resilient against us users!? ?The Al-N insulators are a possible example, gate protection another.? ?Fumble fingers like me can BREAK more stuff than we fix when we get inside such radios!
73
?Gordon KX4Z
|
Gary,
The theoretical response you propose does not consider the capabilities of the 1n4148 diode or a realistic rise time on the supply voltage.? I simulated a 12-volt supply turned on with a rise time of .1usec, and the maximum current spike was 40 ma.? The 1n4146 spec is 2 amps.? The value is way below the specification for the device.? I think the design is OK.
Gordon tried several connections that could have created the condition you described, including changing the values of capacitors.? I would bet that the diodes were faulty when he received the sbitx or were damaged in the tests.
Even with the above data, I would agree that adding the resistor is not a bad idea.? I will most likely do the mod if I ever get a V2 or V3 sbitx (I have a DE).
The transient response from RX to TX starts with the capacitor charged to the supply voltage.? The value will lead the transmitter output each cycle, so the diodes remain reversed-biased.? Hans Summers (the person Ashhar credits with the design) has verified this in multiple devices that use this design.
Please verify the above before using it as fact.? Anyone can make a mistake or bad assumption.
73
Evan
AC9TU
|
On Fri, Nov 10, 2023 at 12:45 PM, Gordon Gibby wrote:
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
?
They didn’t blow themselves; ?something caused i
Gordon,
Always amazed and inspired on how much you get done in a day.
On power up, C201 is at 0 volts.? The current is only limited by the power supply and the interconnect resistance, so the transient current can be very high for a short period of time.? ?Your choice of 470 Ohm series resistor is a conservative, wise one.? It limits the possible current to below the maximum forward continuous current rating of the diode with the resistor power dissipation in the resistor below 1/2 watt max.? Affirmation of a design weakness.?
I was having a hard time wrapping my head around the transient response from RX to TX of this filter's diode switching design implementation. You would need RF energy long enough to raise the voltage on the node "HF" for the diode switching to work properly. Until then, effectively all the filter paths would be at least partially in play, just like they were in play with the shorted diodes.? So I argue that the steady state problem you were chasing down still exists in the transient.? Am I incorrect?
I don't have a SBitx, so no skin in the game.? I do want to see it become the best it can be, and not miss feedback when moving forward in revisions.
Regards,
Gary
??
|
Gordon,
Adding the resistors could have been the cause, except I believe you reported the harmonic issues before you added the resistors.
73
Evan
AC9TU
|
Two comments: 1. ? It is VERY impressive how LITTLE distortion is coming from the 1N4007 "PIN" diodes on 15 meters at least.? ?If you compare the spectrum I measured going directly through the filter (see:?? /g/BITX20/message/105981)? to the spectrum I measured when I finally got it all back together again and working properly (see:?? /g/BITX20/message/105986)? they are really almost the same.? ? That is very impressive to me.? ?I would not have expected the diodes to have that little impact.? (I have not checked them on lower frequencies, where theory says there should be more issues) 2.? I have now had enough "gotchas" that I think I can put together a small "troubleshooting checklist" that covers some of the catastrophes that I somehow experienced through fiddling with this unit, and how to recognize them and fix them.? ?Voltages, waveforms, that sort of thing.? ?Might be helpful.? ?I'll work on it.? ? 3.? What killed L201?? ? ? Here is a data sheet for one particular variety of 47uH small thru-hole (leaded) chokes:? ?? ?I downloaded the catalog.? ?i was VERY surprised to see that the series resonant frequency for that model could be as low as 12 MHZ!? ?We're dealing with as much as 40 volts or so RMS of HF power!? ? If there were a series resonance or a low-impedance capacitive impedance from this choke, going to 470x3 in parallel to 12volts generously bypassed to ground --- you could have some VERY SIGNIFICANT CURRENT FLOWING through the (resonant) choke!!!? ?40Volts /? (470/3) =? 250 mA!!!? ?RF current going thru tht choke if it is resonant!? ?Guess what, that choke is only rated for 70 mA!!? ?That could explain why my L201 gave up the ghost!? ? I have not tested the little choke I pilfered from the Bitx40 but with its 10 or 12 turns I bet the series resonant frequency is much much higher....(won't know for sure until measured).? ?(L8 from Bitx40??)? ? Ashhar may know the exact type of inductor here and where their series resonance is, and that would be important -- would want to avoid operation there.? In the "old days" people sometimes burned up PLATE CHOKES in vacuum tube amps when they hit upon a series resonance, and so such chokes were a bit pricey when well made.? ?We might be using this rig (under different licensure) on many other different frequencies, so worth considering.....? This is the best explanation that I can come up with for L201 mysteriously opening. HF radios are far more intricate than meets the eye.? ?This is a huge advance Ashhar has made.?? 73 Gordon KX4Z
|
Gordon, ? I'm really enjoy reading about your debug adventure. ?Nice work!
toggle quoted message
Show quoted text
So I resoldered ?everything back together again as carefully as I could. ?I had damaged the original 47 pF and I tried a couple of ways of replacing it, got confused when another problem popped up and ended up putting in a 47 pfceramic 3 kV
What confused me was a sudden loss of output power and the strangest voltages in the low pass filters I’ve ever seen, extreme high negative voltages like -150. ?
This took quite a while for me to wrap my head around and figure out that the reason the output power had basically disappeared and such incredible voltages were showing up was that L201 47uH which was not even touched, (and showed absolutely no visible signs of damage ) had picked this moment to open up. ?Measured the resistance across it and sure enough it’s an open circuit, how the heck did that happen? ?
I don’t carry things like that but I found a 60 micro Henry on a real core in an old bitx40 I had gotten from somebody— so I cannibalized the unit and got my radio working again!
Now I had what looked like good output, ?and I was able to run 2 spectrums on the problem bands, 15 m and 20 m, and they look real good. ??
<image0.jpeg><image1.jpeg>
How long I’ve been working on this thing! It looks like finally it is going to be really on the air!!
I do hope to change out the insulators on the finals but not sure when I’ll get to that.? 73? Gordon Kx4z
? Correction. I do remember that Asher had asked for another 470 ohm resistor to be added, in each of two places. ( You can see them in the photographs. ) Those are right in the area of the D 20 and D 21. Perhaps somehow in the process the diode’s got destroyed? ?Current from the iron? ? That isn’t too hard to believe.
Gordon
? Thanks Evan. I went all the way back up to the beginning of my posts and found that I had documented the problem on the very first set of measurements before I had done anything to the radio
<image0.png>
At that point the only thing I had done was open it up to insert the raspberry pi. ? I can’t say what the previous owner might have done but his assertion was that he had never transmitted with it?
He had received with it. ? He seem to think that it worked and I did also; ?when I received it , it appeared to be in fine condition. ??
I looked around for what causes diets to fail shorted and generally it’s too much current. It can be too much in either direction but if you have way too much current then it blows it completely open. The way it gets shorted is to just fuse the silicon with excessive heat. ?
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
They didn’t blow themselves; ?something caused it. ?
Once upon a time I wrote assembly language code for an Altera 8080 for Minnesota power and light. ?The system would fail after a couple of weeks. I was convinced it was hardware. My boss told me to keep looking. Eventually I found it, a software goof on my part . It took me a long time to figure it out. ?There’s usually a cause when something rare happens. ? I just may not be smart enough to figure it out. ?
But “whatever” would be hard-pressed to destroy those diodes again— with a 470 ohm resistor right there in series with them!!!
Now I have to start the work of putting this whole thing back together again so I can redo all the measurements
Gordon Kx4z?
? You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
Congratulations Gordon!
I know this has been a tough journey, but boy the experiences you've been through. I followed each move and wished I had an sBitx? to give you some feedback. It truly is a journey!
Enjoy your radio!
Allen
KI4QCK
On November 10, 2023, at 6:55 PM, Gordon Gibby <docvacuumtubes@...> wrote: So I resoldered ?everything back together again as carefully as I could. ?I had damaged the original 47 pF and I tried a couple of ways of replacing it, got confused when another problem popped up and ended up putting in a 47 pfceramic 3 kV
What confused me was a sudden loss of output power and the strangest voltages in the low pass filters I’ve ever seen, extreme high negative voltages like -150. ?
This took quite a while for me to wrap my head around and figure out that the reason the output power had basically disappeared and such incredible voltages were showing up was that L201 47uH which was not even touched, (and showed absolutely no visible signs of damage ) had picked this moment to open up. ?Measured the resistance across it and sure enough it’s an open circuit, how the heck did that happen? ?
I don’t carry things like that but I found a 60 micro Henry on a real core in an old bitx40 I had gotten from somebody— so I cannibalized the unit and got my radio working again!
Now I had what looked like good output, ?and I was able to run 2 spectrums on the problem bands, 15 m and 20 m, and they look real good. ??
How long I’ve been working on this thing! It looks like finally it is going to be really on the air!!
I do hope to change out the insulators on the finals but not sure when I’ll get to that.? 73? Gordon Kx4z
On Nov 10, 2023, at 16:18, Gordon Gibby via <docvacuumtubes=[email protected]> wrote:
? Correction. I do remember that Asher had asked for another 470 ohm resistor to be added, in each of two places. ( You can see them in the photographs. ) Those are right in the area of the D 20 and D 21. Perhaps somehow in the process the diode’s got destroyed? ?Current from the iron? ? That isn’t too hard to believe.
Gordon
On Nov 10, 2023, at 15:45, Gordon Gibby via <docvacuumtubes=[email protected]> wrote:
? Thanks Evan. I went all the way back up to the beginning of my posts and found that I had documented the problem on the very first set of measurements before I had done anything to the radio
At that point the only thing I had done was open it up to insert the raspberry pi. ? I can’t say what the previous owner might have done but his assertion was that he had never transmitted with it?
He had received with it. ? He seem to think that it worked and I did also; ?when I received it , it appeared to be in fine condition. ??
I looked around for what causes diets to fail shorted and generally it’s too much current. It can be too much in either direction but if you have way too much current then it blows it completely open. The way it gets shorted is to just fuse the silicon with excessive heat. ?
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
They didn’t blow themselves; ?something caused it. ?
Once upon a time I wrote assembly language code for an Altera 8080 for Minnesota power and light. ?The system would fail after a couple of weeks. I was convinced it was hardware. My boss told me to keep looking. Eventually I found it, a software goof on my part . It took me a long time to figure it out. ?There’s usually a cause when something rare happens. ? I just may not be smart enough to figure it out. ?
But “whatever” would be hard-pressed to destroy those diodes again— with a 470 ohm resistor right there in series with them!!!
Now I have to start the work of putting this whole thing back together again so I can redo all the measurements
Gordon Kx4z?
? You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
So I resoldered ?everything back together again as carefully as I could. ?I had damaged the original 47 pF and I tried a couple of ways of replacing it, got confused when another problem popped up and ended up putting in a 47 pfceramic 3 kV
What confused me was a sudden loss of output power and the strangest voltages in the low pass filters I’ve ever seen, extreme high negative voltages like -150. ?
This took quite a while for me to wrap my head around and figure out that the reason the output power had basically disappeared and such incredible voltages were showing up was that L201 47uH which was not even touched, (and showed absolutely no visible signs of damage ) had picked this moment to open up. ?Measured the resistance across it and sure enough it’s an open circuit, how the heck did that happen? ?
I don’t carry things like that but I found a 60 micro Henry on a real core in an old bitx40 I had gotten from somebody— so I cannibalized the unit and got my radio working again!
Now I had what looked like good output, ?and I was able to run 2 spectrums on the problem bands, 15 m and 20 m, and they look real good. ??
How long I’ve been working on this thing! It looks like finally it is going to be really on the air!!
I do hope to change out the insulators on the finals but not sure when I’ll get to that.? 73?
toggle quoted message
Show quoted text
On Nov 10, 2023, at 16:18, Gordon Gibby via groups.io <docvacuumtubes@...> wrote:
? Correction. I do remember that Asher had asked for another 470 ohm resistor to be added, in each of two places. ( You can see them in the photographs. ) Those are right in the area of the D 20 and D 21. Perhaps somehow in the process the diode’s got destroyed? ?Current from the iron? ? That isn’t too hard to believe.
Gordon
On Nov 10, 2023, at 15:45, Gordon Gibby via groups.io <docvacuumtubes@...> wrote:
? Thanks Evan. I went all the way back up to the beginning of my posts and found that I had documented the problem on the very first set of measurements before I had done anything to the radio
At that point the only thing I had done was open it up to insert the raspberry pi. ? I can’t say what the previous owner might have done but his assertion was that he had never transmitted with it?
He had received with it. ? He seem to think that it worked and I did also; ?when I received it , it appeared to be in fine condition. ??
I looked around for what causes diets to fail shorted and generally it’s too much current. It can be too much in either direction but if you have way too much current then it blows it completely open. The way it gets shorted is to just fuse the silicon with excessive heat. ?
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
They didn’t blow themselves; ?something caused it. ?
Once upon a time I wrote assembly language code for an Altera 8080 for Minnesota power and light. ?The system would fail after a couple of weeks. I was convinced it was hardware. My boss told me to keep looking. Eventually I found it, a software goof on my part . It took me a long time to figure it out. ?There’s usually a cause when something rare happens. ? I just may not be smart enough to figure it out. ?
But “whatever” would be hard-pressed to destroy those diodes again— with a 470 ohm resistor right there in series with them!!!
Now I have to start the work of putting this whole thing back together again so I can redo all the measurements
Gordon Kx4z?
On Nov 10, 2023, at 15:33, Evan Hand <elhandjr@...> wrote:
? You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
Correction. I do remember that Asher had asked for another 470 ohm resistor to be added, in each of two places. ( You can see them in the photographs. ) Those are right in the area of the D 20 and D 21. Perhaps somehow in the process the diode’s got destroyed? ?Current from the iron? ? That isn’t too hard to believe.
Gordon
toggle quoted message
Show quoted text
On Nov 10, 2023, at 15:45, Gordon Gibby via groups.io <docvacuumtubes@...> wrote:
? Thanks Evan. I went all the way back up to the beginning of my posts and found that I had documented the problem on the very first set of measurements before I had done anything to the radio
At that point the only thing I had done was open it up to insert the raspberry pi. ? I can’t say what the previous owner might have done but his assertion was that he had never transmitted with it?
He had received with it. ? He seem to think that it worked and I did also; ?when I received it , it appeared to be in fine condition. ??
I looked around for what causes diets to fail shorted and generally it’s too much current. It can be too much in either direction but if you have way too much current then it blows it completely open. The way it gets shorted is to just fuse the silicon with excessive heat. ?
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
They didn’t blow themselves; ?something caused it. ?
Once upon a time I wrote assembly language code for an Altera 8080 for Minnesota power and light. ?The system would fail after a couple of weeks. I was convinced it was hardware. My boss told me to keep looking. Eventually I found it, a software goof on my part . It took me a long time to figure it out. ?There’s usually a cause when something rare happens. ? I just may not be smart enough to figure it out. ?
But “whatever” would be hard-pressed to destroy those diodes again— with a 470 ohm resistor right there in series with them!!!
Now I have to start the work of putting this whole thing back together again so I can redo all the measurements
Gordon Kx4z?
On Nov 10, 2023, at 15:33, Evan Hand <elhandjr@...> wrote:
? You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
Thanks Evan. I went all the way back up to the beginning of my posts and found that I had documented the problem on the very first set of measurements before I had done anything to the radio
At that point the only thing I had done was open it up to insert the raspberry pi. ? I can’t say what the previous owner might have done but his assertion was that he had never transmitted with it?
He had received with it. ? He seem to think that it worked and I did also; ?when I received it , it appeared to be in fine condition. ??
I looked around for what causes diets to fail shorted and generally it’s too much current. It can be too much in either direction but if you have way too much current then it blows it completely open. The way it gets shorted is to just fuse the silicon with excessive heat. ?
These diode’s are connected to a high capacity 12 V source, but on the other side they really only have a 100 pF capacitor, and several 470 K resistors. ? ?(unless I’m missing something).?
They didn’t blow themselves; ?something caused it. ?
Once upon a time I wrote assembly language code for an Altera 8080 for Minnesota power and light. ?The system would fail after a couple of weeks. I was convinced it was hardware. My boss told me to keep looking. Eventually I found it, a software goof on my part . It took me a long time to figure it out. ?There’s usually a cause when something rare happens. ? I just may not be smart enough to figure it out. ?
But “whatever” would be hard-pressed to destroy those diodes again— with a 470 ohm resistor right there in series with them!!!
Now I have to start the work of putting this whole thing back together again so I can redo all the measurements
Gordon Kx4z?
toggle quoted message
Show quoted text
On Nov 10, 2023, at 15:33, Evan Hand <elhandjr@...> wrote:
? You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
You need the 12volts when receiving and to start transmitting. ?There is no supply without the transmitter turned on.?
73
Evan
AC9TU
|
Gordon,
No, I did not see the post. ?
At this point we do not have a design issue until we can work out what caused the diodes to short.?
73
Evan
AC9TU
|
Hi Evan -- did you read this post?? ? /g/BITX20/message/105976D20 and D21 were shorted.? ?All of the diodes in the HV power system from RF were normal in correct directions. I have now put 470 ohms in series with the (replaced) D20/D21.? ?That seems to work fine also!? 60-61 volts at present. Here is a snapshot of the 15 meter spectral response (remember that the LPF is directly connected, not through diodes, so this could get different after I insert it BACK into the diode switching network).? ?All of the other LPFs are within the diode switching system -- and are now solidly OFF.  You can see that the 3rd harmonic (in tthis configuration) is now a whopping -47dBc!!!? ?This is great!! I hope it stays that way after I put it back into the switching network!! Gordon KX4Z
|
Gordon,
Thinking about your data, I came up with possible reasons for the failure and what the failure is.
Based on the scope measurement of the voltage on C201, I believe one or both D24 and D25 are open.? I base this on your measurement of around 20 volts on C201.? The values would be consistent with a rectified value of the peak voltage.
The failure would be an overvoltage on one, a short on one, then an overvoltage on the other.
73
Evan
AC9TU
|
I'm thinking it might be smart to put some resistance in SERIES with D20/D21.? ?I'm thinking they are there just to keep the LPF switching diodes from being heavily forward biased??? ? Maybe the RF-derived voltage doubler can drive current throughthem?? ?Do diodes fail SHORTED more from over-reverse voltage (avalanche)or from excessive forward current??? ? Maybe 470 ohms in series with them to limit 14volt-based current to ground to 30mA??? ?They are rated to as much as 150mA.? ??
Comments?? ?What causes diodes to? short?
Gordon KX4Z
On Fri, Nov 10, 2023 at 1:38?PM Gordon Gibby via <docvacuumtubes= [email protected]> wrote: (Note, I also went back to the 47pf capacitor instead of the big 330pf I had put in temporarily).? ??
Wow, this has taken a lot of time to figure out!!!
Gordon KX4Z
|
(Note, I also went back to the 47pf capacitor instead of the big 330pf I had put in temporarily).? ??
Wow, this has taken a lot of time to figure out!!!
Gordon KX4Z
|
