Topic: Blitzortung RED Amp debugging

[W7DRM (old call was W3DRM)]

Don, here's some voltage measurements of the Red Amp devices under normal operation... .
Maybe you can use them.
Mike

Mike,

Finally got the time go compare my readings against what you sent me. I have found some significant differences. I am wondering if we have different versions of our Amplifier boards. Mine is labeled PCB Ver. 3 * 7/2013. The schematic on the current (8/23) Blitzortung manual, matches what I have as far pinout for the various components. I see there has been some discussion relative to a different IC being used for IC's 1, 3, 5, & 7. I received the MCP6S91 chips with my RED system. Others have gotten the MCP6S21 chips. They are both identical except for temperature ratings.

I have attached both an XLSX file showing our differences and a PDF of the same. NOTE: Can't upload a XLS file so have renamed it with a .TXT extension. Just rename it to .XLSX (actually, just remove the .TXT) and it should be fine.

Good catch...! I reversed 11 and 12 on the controller when I created the document  and missed it during proof. 17,18,19 will gradually increase on the DC scale because of signal... I just put in a dash... depending on meter... You are also correct about pins 6 and 7 on IC 1 and IC 3 ... they should measure the same as controller pins 17 and 19... sheesh, no wonder I don't do electronics for a living anymore... musta been a really bad day for me.
Mike

No problem, it made me dig out the schematic and trace the wiring to see if I could find a difference in the board traces. It's been a long time since I have gone through a schematic myself... Thanks for the voltages though. At least I do compare pretty closely (DC-wise, that is) to what you see so I'm not sure what my problem may be. Still don't have a scope yet. All solder connections look good. Channel-B is still dead. I even moved the cable input on the Controller board so it is using the secondary channel and no change at all in the output. Am thinking it probably is something bad on the amp board. Just haven't found it yet.

[saratogaWX]

Don
I've discovered that my chan-B is dead too... Just starting the trace for why that is happening.

Best regards,
Ken

[W7DRM (old call was W3DRM)]

Don
I've discovered that my chan-B is dead too... Just starting the trace for why that is happening.

Best regards,
Ken

Sorry to hear that Ken. I was wondering how you were doing with your troubleshooting.

So far, I haven't been able to find anything wrong. The solder joints all look good and the DC voltage levels compare to another users readings. All socketed IC's have been removed and reseated (and swapped between channels) with no changes observed. I may have to break-down and buy an o'scope to find the problem. My old Bell & Howell Scope (ala Heathkit origin, I think) is dead. The CRT filaments appear to be open.

One question about your circuit boards and the rev level of them. My boards have a rev stamp on them, as follows:

• Amplifier PCB: "PCB 12 Ver. 3 * 7/2013"
• Controller PCB: "PCB 10 Ver. 3b * 8/2013"

Are yours marked similarly?

Best of luck - hope you can find the source of your problems quickly...

[gwwilk]

I finally succeeded in getting the B channel on my amplifier to come to life a couple of days ago.  This was after over a week of frustration, going over the board multiple times with a soldering iron, and even trying to obtain a new amplifier from Egon.  He won't be available until next week, so I persisted.  Even after going over the board multiple times I still found a couple of pins on the first DIP channel B socket that didn't look quite right.  I touched up those two along with a couple of other possibly cold solder joints and Voila!  My channel B came alive.  It doesn't see the same number of strokes as channel A when configured the same, but at least it's giving me some data now.

We had a huge thunderstorm train overhead yesterday evening with tornadoes a few miles SE of here, and my Blitzortung kept going into interference mode just because of the number of strikes it was seeing.  My Boltek got up to over 500/min. but with the Blitzortung's interference mode kicking in at 30/sec I spent the evening 'tuning' my Blitzortung to keep it from going offline due to 'interference'.  At the peak of the storm I had all amps set to gain of 4 with a threshold of 220 mV each.  Overnight while I slept my detection became much less as the storms moved a little further away.  I'm currently set at gain on A 4 & 5 with a 240 mV threshold and gain on B 5 & 5 with a 190 mV threshold.

My overall impression is that babysitting a lightning detector during a thunderstorm isn't much fun.  I certainly hope that Egon, Tobi, et al get the 'Automatic' mode working real soon now.  Interference mode could also use a little fine tuning for those of us who see severe thunderstorms on occasion.

I installed MyBlitzortung and it works fine standalone, but I can't make it play nice at all in a template.  Better coding skills than mine are needed for that job, but right now I'm out of ideas!

In summary, don't give up.  Even after you have all of the components installed correctly there may still be cold joints hidden here and there.  Once you reheat and/or resolder them your board should come alive.  I thought I had a broken trace for the longest time due to backwards electrolytic capacitor removals, but not so.  I didn't ruin my board like I feared through going over and over it with a soldering iron.  BTW, I went back to my small-wedge-tipped iron from 25 yrs. ago after I gave up trying to get good joints with my new off-brand digital temperature-controlled iron.  Just like 25 yrs. ago, the old iron didn't destroy anything.  It's higher temperature allowed me to solder much much more quickly and effectively.  Two to fours hours a day of bending over a circuit board seems about all my old back and eyes can tolerate any more.

[Jumpin Joe]

Congratulations Jerry! 

I had said in some of my earlier posts that this is what I had experienced as well. Good to know that there are others that have older eyes too.

Joe

[saratogaWX]

Don
I've discovered that my chan-B is dead too... Just starting the trace for why that is happening.

Best regards,
Ken

Sorry to hear that Ken. I was wondering how you were doing with your troubleshooting.

So far, I haven't been able to find anything wrong. The solder joints all look good and the DC voltage levels compare to another users readings. All socketed IC's have been removed and reseated (and swapped between channels) with no changes observed. I may have to break-down and buy an o'scope to find the problem. My old Bell & Howell Scope (ala Heathkit origin, I think) is dead. The CRT filaments appear to be open.

One question about your circuit boards and the rev level of them. My boards have a rev stamp on them, as follows:

• Amplifier PCB: "PCB 12 Ver. 3 * 7/2013"
• Controller PCB: "PCB 10 Ver. 3b * 8/2013"

Are yours marked similarly?

Best of luck - hope you can find the source of your problems quickly...

Don,
My boards are:

Amplifier:   PCB 12 Ver. 3b 8/2013
Controller: PCB 10 Ver. 3b 8/2013

I'm going to use the o'scope to follow the gain through both A and B amps to try and spot where it's going dead.

Congrats Jerry!  I'm thinking I may have a similar problem with the B channel that you did

Best regards,
Ken

[gwwilk]

My rather primitive method of following the signal back to the antenna was to find with a jeweler's screwdriver a pin on the A channel that produced an interference tone on the buzzer, and then see if the same pin produced the same results on the B channel.  Crude, but effective.  I'm still alive and the board's working.  An 'interference' tone produced by touching the copper shielding I installed on the ferrite antennas was the last indication that the A and B channels were working.

[W7DRM (old call was W3DRM)]

Hi Ken,

So, the Blitzortung folks are making updates to their boards as they fine-tune the RED system. This is fine but, I wish they had someplace we could go to and read the revision information. That might make things a bit easier for those of us trying to troubleshoot problems. I've looked for a list of revisions on the blitzortung.org website but haven't found anything so far.

[W7DRM (old call was W3DRM)]

My rather primitive method of following the signal back to the antenna was to find with a jeweler's screwdriver a pin on the A channel that produced an interference tone on the buzzer, and then see if the same pin produced the same results on the B channel.  Crude, but effective.  I'm still alive and the board's working.  An 'interference' tone produced by touching the copper shielding I installed on the ferrite antennas was the last indication that the A and B channels were working.

Good idea Jerry!

I noticed when I was taking the DC voltage readings last night that, as I touched some pins with the DVM probe, I would hear the 'interference' tone on the controller board. I didn't think about comparing the matching pins to the other channel. I'll give that a shot and see what happens.

[Jumpin Joe]

Don
I've discovered that my chan-B is dead too... Just starting the trace for why that is happening.

Best regards,
Ken

Sorry to hear that Ken. I was wondering how you were doing with your troubleshooting.

So far, I haven't been able to find anything wrong. The solder joints all look good and the DC voltage levels compare to another users readings. All socketed IC's have been removed and reseated (and swapped between channels) with no changes observed. I may have to break-down and buy an o'scope to find the problem. My old Bell & Howell Scope (ala Heathkit origin, I think) is dead. The CRT filaments appear to be open.

One question about your circuit boards and the rev level of them. My boards have a rev stamp on them, as follows:

• Amplifier PCB: "PCB 12 Ver. 3 * 7/2013"
• Controller PCB: "PCB 10 Ver. 3b * 8/2013"

Are yours marked similarly?

Best of luck - hope you can find the source of your problems quickly...

Don,
My boards are:

Amplifier:   PCB 12 Ver. 3b 8/2013
Controller: PCB 10 Ver. 3b 8/2013

I'm going to use the o'scope to follow the gain through both A and B amps to try and spot where it's going dead.

Congrats Jerry!  I'm thinking I may have a similar problem with the B channel that you did

Best regards,
Ken

I have the same boards as Ken.

[Cutty Sark Sailor]

A thought: A lot of those components use "thru connections", and the components should connect to paths on both sides. It might be possible to have a 'loose' through that's missing the 'component' side connection, but good on the 'bottom' side. Might examine them on both sides.

I say this because this "thru" connection design drove us nuts at times, prior to my retirement in 97... we didn't have the 'metal' thrus in all cases, just bare pads on both sides... and we fixed many a 'bad thru' especially for intermittents. I made sure that my solder wicked up thru the board to secure the top trace connection to the 'thru'...

I also made it a point to insure all the diodes on the amp input on the controller were installed in proper polarity!

[gwwilk]

A thought: A lot of those components use "thru connections", and the components should connect to paths on both sides. It might be possible to have a 'loose' through that's missing the 'component' side connection, but good on the 'bottom' side. Might examine them on both sides.

I say this because this "thru" connection design drove us nuts at times, prior to my retirement in 97... we didn't have the 'metal' thrus in all cases, just bare pads on both sides... and we fixed many a 'bad thru' especially for intermittents. I made sure that my solder wicked up thru the board to secure the top trace connection to the 'thru'...

I also made it a point to insure all the diodes on the amp input on the controller were installed in proper polarity!

Some of those components sucked up lots and lots of solder before the solder mounded up to indicate a solid joint.  Until I mounded up ALL of the connections on the bottom of the board the B channel was dead, so you may just have hit on the reason why.

[miraculon]

Interesting. We had some bad PTH (Plated Through Holes) problems at work a few years ago. They failed during a very tough thermal cycling test though, not room temperature. I suppose that the board vendor might have made some bad ones. Making sure that there is a meniscus on both sides for all joints is a good idea. (probably not easy on the sockets). If there is no "top side" trace and just a pad, it doesn't matter. Only pads/through holes with top side traces will need this extra touch up. There are a number of PTH without component leads, but they have smaller drill holes. I really hope that those are not bad, that could be a real problem to fix. (scraping solder mask and soldering a small gauge wire, etc.)

Greg H.

[miraculon]

I finally succeeded in getting the B channel on my amplifier to come to life a couple of days ago.  This was after over a week of frustration, going over the board multiple times with a soldering iron, and even trying to obtain a new amplifier from Egon.  He won't be available until next week, so I persisted.  Even after going over the board multiple times I still found a couple of pins on the first DIP channel B socket that didn't look quite right.  I touched up those two along with a couple of other possibly cold solder joints and Voila!  My channel B came alive.  It doesn't see the same number of strokes as channel A when configured the same, but at least it's giving me some data now.

We had a huge thunderstorm train overhead yesterday evening with tornadoes a few miles SE of here, and my Blitzortung kept going into interference mode just because of the number of strikes it was seeing.  My Boltek got up to over 500/min. but with the Blitzortung's interference mode kicking in at 30/sec I spent the evening 'tuning' my Blitzortung to keep it from going offline due to 'interference'.  At the peak of the storm I had all amps set to gain of 4 with a threshold of 220 mV each.  Overnight while I slept my detection became much less as the storms moved a little further away.  I'm currently set at gain on A 4 & 5 with a 240 mV threshold and gain on B 5 & 5 with a 190 mV threshold.

My overall impression is that babysitting a lightning detector during a thunderstorm isn't much fun.  I certainly hope that Egon, Tobi, et al get the 'Automatic' mode working real soon now.  Interference mode could also use a little fine tuning for those of us who see severe thunderstorms on occasion.

I installed MyBlitzortung and it works fine standalone, but I can't make it play nice at all in a template.  Better coding skills than mine are needed for that job, but right now I'm out of ideas!

In summary, don't give up.  Even after you have all of the components installed correctly there may still be cold joints hidden here and there.  Once you reheat and/or resolder them your board should come alive.  I thought I had a broken trace for the longest time due to backwards electrolytic capacitor removals, but not so.  I didn't ruin my board like I feared through going over and over it with a soldering iron.  BTW, I went back to my small-wedge-tipped iron from 25 yrs. ago after I gave up trying to get good joints with my new off-brand digital temperature-controlled iron.  Just like 25 yrs. ago, the old iron didn't destroy anything.  It's higher temperature allowed me to solder much much more quickly and effectively.  Two to fours hours a day of bending over a circuit board seems about all my old back and eyes can tolerate any more.

Congrats on getting everything up and running. Your participation and waveforms look great. 
Your Nebraska station fills an important and needed location in the network.

Greg H.

[W7DRM (old call was W3DRM)]

I finally got a loaner scope (B&K 1474 30MHz) and some really nice new Tektronix P6100 probes I bought on eBay. Now, I am ready to begin troubleshooting my RED Amplifier and see if I can find out why the B channel is dead. The A channel shows a signal but nothing but low-level noise on the B side.

I am going to try to use a PC-based signal generator that is part of the PC scope software but first, I need to know if anyone can tell me what are the typical input voltages that are seen at the antenna inputs to the RED amplifier board. I don't want to blow-out my op-amps by overdriving them!

Also, given the circuit description in the RED documentation it looks like the circuitry is designed with a 1KHz hi-pass filter and a following 50KHz low-pass filter. So, I'm going to start off with a test frequency of around 5KHz. I'm very limited to my frequency range since this is an output from the sound card in my pc but, my main purpose is to find where I am losing my signal as I trace it through the various stages on the Amp board.

[Cutty Sark Sailor]

I finally got a loaner scope (B&K 1474 30MHz) and some really nice new Tektronix P6100 probes I bought on eBay. Now, I am ready to begin troubleshooting my RED Amplifier and see if I can find out why the B channel is dead. The A channel shows a signal but nothing but low-level noise on the B side.

I am going to try to use a PC-based signal generator that is part of the PC scope software but first, I need to know if anyone can tell me what are the typical input voltages that are seen at the antenna inputs to the RED amplifier board. I don't want to blow-out my op-amps by overdriving them!

Also, given the circuit description in the RED documentation it looks like the circuitry is designed with a 1KHz hi-pass filter and a following 50KHz low-pass filter. So, I'm going to start off with a test frequency of around 5KHz. I'm very limited to my frequency range since this is an output from the sound card in my pc but, my main purpose is to find where I am losing my signal as I trace it through the various stages on the Amp board.

Morning... need more coffee... but:
Half split method:
I wouldn't inject anything just yet. Antennas are microvolts. But 25KHz'd be fast enough... it'd probably inject just by wrapping the sound out wire around the antenna... (haven't tried it).
but
Let's do it quickly:
 Aim the B antenna same direction as A.  Now, start in middle and work backwards or forwards, half-split... comparing sigs at various points each channel. Prob locate loss in a jiffy.
Also about a .1uf NP cap an' a wire could jump A channel sig into corresponding point of B channel... .
Now..
coffee....

Mike

[miraculon]

5KHz is probably a good generator frequency. The bandpass filter is from 1KHz to 50KHz so anything in the middle of that range is fine. Personally, I would choose about 10KHz.

If you can adjust your generator to a few milli-Volts and start at the lower gain settings, you can inject the signals. The worst thing that can happen is that you overdrive the inputs and/or clip the amplifier outputs.

The good thing about a generator is that you can put in a known input and gain and see if you get a pure sine output with the correct amplitude. That way you can eliminate any distortion or gain anomalies in the amplifiers. After that, antenna placement in a low-noise field is critical.

The only problem I see with the PC generator would be that the noise floor of the PC is probably higher than normal antenna signals. You also might run into a large ground loop problem. A "real" generator would be better.

Given that you don't have an actual signal generator, Cutty's method will probably be a better bet. Just make sure the antennas themselves are OK. (no short/open)

Greg

[W7DRM (old call was W3DRM)]

I finally got a loaner scope (B&K 1474 30MHz) and some really nice new Tektronix P6100 probes I bought on eBay. Now, I am ready to begin troubleshooting my RED Amplifier and see if I can find out why the B channel is dead. The A channel shows a signal but nothing but low-level noise on the B side.

I am going to try to use a PC-based signal generator that is part of the PC scope software but first, I need to know if anyone can tell me what are the typical input voltages that are seen at the antenna inputs to the RED amplifier board. I don't want to blow-out my op-amps by overdriving them!

Also, given the circuit description in the RED documentation it looks like the circuitry is designed with a 1KHz hi-pass filter and a following 50KHz low-pass filter. So, I'm going to start off with a test frequency of around 5KHz. I'm very limited to my frequency range since this is an output from the sound card in my pc but, my main purpose is to find where I am losing my signal as I trace it through the various stages on the Amp board.

Morning... need more coffee... but:
Half split method:
I wouldn't inject anything just yet. Antennas are microvolts. But 25KHz'd be fast enough... it'd probably inject just by wrapping the sound out wire around the antenna... (haven't tried it).
but
Let's do it quickly:
 Aim the B antenna same direction as A.  Now, start in middle and work backwards or forwards, half-split... comparing sigs at various points each channel. Prob locate loss in a jiffy.
Also about a .1uf NP cap an' a wire could jump A channel sig into corresponding point of B channel... .
Now..
coffee....

Mike

Well, as I mentioned in my post, I am using a V2 Loop Antenna that is already fully assembled. Thus, I don't have the ability to independently move each loop as they are physically attached and more or less held in place by the wire loops themselves. I'm not so worried about external noises being picked up by the antennas at this point in my troubleshooting. I just want to be able to insert a low-level sine wave into the amplifier inputs (both inputs at same frequency and in phase with each other) and then begin walking through the circuits with the scope to see where things either don't match or the signal disappears all together.

As for technique used, we are all different. I will probably start by checking the A & B outputs of the amplifier board just to be certain one or the other really isn't working. I don't know that for a fact except for what I see on the displayed blitzortung user data. The red channel always shows a signal and the green channel is almost always a straight line. Both of them have some ripple frequency on them which may just be stray ac being picked-up. I can't explain why I occasionally see a matching signal on the green channel though.

In the attached images you will see that one of them has both red and green traces but the other only shows a red trace. I can't figure out why this is happening. Sometimes the green trace is there but has no amplitude.

[W7DRM (old call was W3DRM)]

Okay, not sure why the second image didn't post. Anyway, here it is...

[W7DRM (old call was W3DRM)]

5KHz is probably a good generator frequency. The bandpass filter is from 1KHz to 50KHz so anything in the middle of that range is fine. Personally, I would choose about 10KHz.

If you can adjust your generator to a few milli-Volts and start at the lower gain settings, you can inject the signals. The worst thing that can happen is that you overdrive the inputs and/or clip the amplifier outputs.

The good thing about a generator is that you can put in a known input and gain and see if you get a pure sine output with the correct amplitude. That way you can eliminate any distortion or gain anomalies in the amplifiers. After that, antenna placement in a low-noise field is critical.

The only problem I see with the PC generator would be that the noise floor of the PC is probably higher than normal antenna signals. You also might run into a large ground loop problem. A "real" generator would be better.

Given that you don't have an actual signal generator, Cutty's method will probably be a better bet. Just make sure the antennas themselves are OK. (no short/open)

Greg

Guess I'll try injecting the signals directly into the antenna inputs first with as low a signal as I can and go from there. I see the input has a 2.2K parallel resistor and a series 10K resistor plus a blocking/coupling 100nF capacitor so I don't have to worry about any DC component at the input. Based on the scope signals, the PC sound card output looks very clean.

If this doesn't work then I may try to devise an inductive coupling on the inputs.

[miraculon]

If this doesn't work then I may try to devise an inductive coupling on the inputs.

When I worked for "Famous Auto Maker" we had a residual DSP noise spur at 610KHz that was coupling from the rear deck speakers to the backlight AM antenna. It actually was radiating from the voice coil into the antenna.

If you have a spare speaker laying around it might work for your inductive drive. Your speaker-level audio outputs should be OK, if your ears can take the noise. I had a spare ferrite rod of the type that I used for my sniffer box that I was able to drive right off the generator. No audible noise with this!

Greg

[W7DRM (old call was W3DRM)]

Here is another sample of my signal:

[W7DRM (old call was W3DRM)]

Yippee, I found the cause of my dead channel B. It turned out to be a bad solder joint on pin 3 of IC8. Once I got a working o'scope it took about 30 seconds to find the missing signal and then another few minutes to heat-up the ole soldering iron and touch-up the solder on pin 3 of IC8. Once I did that, I had good signal traces throughout and all looks just fine now.

Now, it's off to trying to find a quiet place to put the loop antenna and get everything ready for this summer's lightning storms.

I used the signal generator feature of the Soundcard Oscilloscope V1.41 software on my laptop along with a dual-trace o'scope to trace the signal through the circuitry. I set the output of the signal (both channels) to a Sine wave at 5Khz and 0.500 amplitude for the test. Once I found the bad connection, it was a simple matter to just reheat the pin connection and all was well. This was a connection that looked perfect even with a 20x loop - so, you can't just go by looks of the solder joints.

And lastly, the new "show noise" feature on the V6.0 firmware really helps to show visually the signals coming through the system.

[miraculon]

Yippee, I found the cause of my dead channel B. It turned out to be a bad solder joint on pin 3 of IC8. Once I got a working o'scope it took about 30 seconds to find the missing signal and then another few minutes to heat-up the ole soldering iron and touch-up the solder on pin 3 of IC8. Once I did that, I had good signal traces throughout and all looks just fine now.

Now, it's off to trying to find a quiet place to put the loop antenna and get everything ready for this summer's lightning storms.

I used the signal generator feature of the Soundcard Oscilloscope V1.41 software on my laptop along with a dual-trace o'scope to trace the signal through the circuitry. I set the output of the signal (both channels) to a Sine wave at 5Khz and 0.500 amplitude for the test. Once I found the bad connection, it was a simple matter to just reheat the pin connection and all was well. This was a connection that looked perfect even with a 20x loop - so, you can't just go by looks of the solder joints.

And lastly, the new "show noise" feature on the V6.0 firmware really helps to show visually the signals coming through the system.

That is great news. I am glad that you found the culprit.

Greg

[W7DRM (old call was W3DRM)]

I am going to try to build my own plastic housing for both the amp and the controller boards. I've been watching some of the TapPlastics (http://www.tapplastics.com/) videos and it looks like it will be a pretty easy job to build a box using their techniques. If I'm successul, I'll post some photos of my creations.