Have designed and built a rack mount hybrid that essentially integrates two different devices into one homogeneous system without deconstructing components, some are relocated but not reassigned. One is a 3M Peltor DECT Comm 2815 handset internal components with additional modules to formalise use, connecting to a further piece of hardware to equalise the duplex signals for introduction to existing hardwired duplex comms systems used in broadcast and entertainment. The Peltor DECT system is now discontinued and 3M will not assist. The 2815 unit is rare, I own 3, 2 have now ‘failed’ in development after working; first was component damage second remains mystery 3rd not going to use until diagnosis is known leading towards repair of other two. Can anyone help/recommend/assist in getting this resolved please as I can go no further without assistance from knowledgeable electronics engineering- I am an inventor/ electrician not electronics hence my plea for assistance/ guidance please…

 

Welcome to AAC.

This is going to require the ordinary steps for repairing an electronic device. Success will depend on a few things:

• Is there documentation—a service manual, schematics, &c.
  
  
• How the device is constructed—is it solidly built with connectors for board-to-board signals, does it have well laid out and marked PCB(s), is it heavily dependent on SMD parts of the smallest type, &c.
  
  
• Does it rely on ASICs or other proprietary components that are unavailable—are the chips scrubbed of identification or marked with only house numbers, &c.

I suspect there are no manuals or schematics and that there are some proprietary components, but that’s not a show stopper. You need to do two things:

1. Start looking on eBay and elsewhere for used units that you can buy—working is great, of course, but inoperable might be even better. I am guessing any working examples will sell for quite. high price, but if you can get inoperable units cheap, they can be used for repair parts and possibly be combined into fewer, but working devices. The only thing that would make this not very useful is if there is a very common failure mode that involves unobtainable parts.

2. Document the device yourself—

• Take well lit, sharply focused photos of each board and component. Make sure that any part numbers and color codes are clearly visible. For laser etched parts whose designations are hard or impossible to read, you can use correction fluid to make them legible: very lightly brush or otherwise apply the fluid to the IC then wipe it off completely before it can dry. It will fill the etching and make it miraculously more readable. Obviously be careful doing it, and use the smallest amount—literally a drop. You can also use a brightly colored crayon.
  
  
• Collect whatever documentation you can find. It is often the case that clues to the repair can come from unexpected sources contained in whatever the manufacturer provides in the way of documents.

Once you have the photos (including ones of your application, with the bits in place, and all wiring), you can post them here along with a description of the problem in detail. What doesn’t work? What do you expect it to do? What have you tried? What changed before it stopped working? (&c)

Good luck, I hope you can find an answer and, with documentation in hand, you can surely get some help here.

 

Welcome to AAC.

This is going to require the ordinary steps for repairing an electronic device. Success will depend on a few things:

• Is there documentation—a service manual, schematics, &c.
  
  
• How the device is constructed—is it solidly built with connectors for board-to-board signals, does it have well laid out and marked PCB(s), is it heavily dependent on SMD parts of the smallest type, &c.
  
  
• Does it rely on ASICs or other proprietary components that are unavailable—are the chips scrubbed of identification or marked with only house numbers, &c.

I suspect there are no manuals or schematics and that there are some proprietary components, but that’s not a show stopper. You need to do two things:

1. Start looking on eBay and elsewhere for used units that you can buy—working is great, of course, but inoperable might be even better. I am guessing any working examples will sell for quite. high price, but if you can get inoperable units cheap, they can be used for repair parts and possibly be combined into fewer, but working devices. The only thing that would make this not very useful is if there is a very common failure mode that involves unobtainable parts.

2. Document the device yourself—

• Take well lit, sharply focused photos of each board and component. Make sure that any part numbers and color codes are clearly visible. For laser etched parts whose designations are hard or impossible to read, you can use correction fluid to make them legible: very lightly brush or otherwise apply the fluid to the IC then wipe it off completely before it can dry. It will fill the etching and make it miraculously more readable. Obviously be careful doing it, and use the smallest amount—literally a drop. You can also use a brightly colored crayon.
  
  
• Collect whatever documentation you can find. It is often the case that clues to the repair can come from unexpected sources contained in whatever the manufacturer provides in the way of documents.

Once you have the photos (including ones of your application, with the bits in place, and all wiring), you can post them here along with a description of the problem in detail. What doesn’t work? What do you expect it to do? What have you tried? What changed before it stopped working? (&c)

Good luck, I hope you can find an answer and, with documentation in hand, you can surely get some help here.

Awesome, thank you so much for such a concise reply, most of which you state are indeed already available, and indeed, the crucial such as circuit diagrams are not to be found anywhere, and boy have I tried! I only have one working unit left, out of the three I have, and am obviously reluctant to do a thing with it, other than have ready for someone that knows how to interrogate it to understand what it is up to and how it does it - I sure do not! Other than it does what is expected to do. User manuals I have, but 3M didn’t make accessible the all important circuit diags., partly because, we think , that it was also part of the NATO comms that were very similar to look and design to these… (?) Standby for documentation etc. Again, many thanks for taking the time.

 

In addition to the excellent comments from "Y", for anyone tasked with diagnosing the failures, an understanding of how the system is actually supposed to function will provide a lot of help.

 

Folks, I have managed to start the compilation of documents relevant, and attach them here for your perusal. If this is insufficient please tell me what else I need to realise and it shall be so if I can. I would again like to thank you all so far for being so welcoming and assistive. Here we go then….

 

That’s great, and useful, but photos are going to be the most valuable initial thing and while the third document contains many there are too small and too out of focus to use for troubleshooting.

We will really need high resolution, well lit, sharp photos of the Peltor’s PCBs (both sides) and connections. Since we can’t see what is happening there, it is the only way to gather enough infornation to help.

A concise problem statement is also very important. It needs to answer several questions:

• What is supposed to happen?
  
  
• What doesn’t happen?
  
  
• What happens instead?
  
  
• If things were OK at first but then not—
  
  what changed? (e.g.: moved device to new location)
  
  how did it fail? (e.g.:became noisier and nosier and then stopped)
  
  
• What have you done to fix it?
  
  
• What were the results (even if it didn’t fix it, if anything changed)

This is general framework and should be adapted to your circumstances. The materials you provided is likely to be very helpful if it can be given a context by these two things.

 

Standby please! I am involved in some particularly absorbing work for the next week or so.. but will return with appropriate Information for y’all.

 

Folks I am back! I truly hope you are still interested in helping me out? Been knocked down with this viral cough thing; anyhoo, I shall be taking the HD photos as requested over the weekend, and will re-read through all requests etc., and see what else it was you all so kindly suggested would help too. Thank you for your patience! I am still in need of much help, nothing has changed since we last conversed, with this project or life for that matter !

 

That’s great, and useful, but photos are going to be the most valuable initial thing and while the third document contains many there are too small and too out of focus to use for troubleshooting.

We will really need high resolution, well lit, sharp photos of the Peltor’s PCBs (both sides) and connections. Since we can’t see what is happening there, it is the only way to gather enough infornation to help.

A concise problem statement is also very important. It needs to answer several questions:

• What is supposed to happen?
  
  
• What doesn’t happen?
  
  
• What happens instead?
  
  
• If things were OK at first but then not—
  
  what changed? (e.g.: moved device to new location)
  
  how did it fail? (e.g.:became noisier and nosier and then stopped)
  
  
• What have you done to fix it?
  
  
• What were the results (even if it didn’t fix it, if anything changed)

This is general framework and should be adapted to your circumstances. The materials you provided is likely to be very helpful if it can be given a context by these two things.

YA’AKOV, finally I have managed to get these off to you, file 3/3 has failed due to size, I must get an hour of shut-eye, so will re-send that later, Apols. Do hope that this makes greater sense, and that progress can now be made… fingers crossed! Kind regards Michael + Attachments 3 of 4 one more to re-send later…

 

Well, hello.

That’s quite a rabbit hole you’ve handed me. Very nicely documented but with so many moving parts it will take time to absorb. I am going to try to invest more time in a bit, but here’s what I have so far:

FIRST: Please consider the idea that there is NOTHING wrong with your adapter. If I read correctly, you have exactly one failure that might be attributed to it. Smashing the board with a hammer might indicate you need better armor on your rack case, but it surely doesn’t explain the only failure you’ve seen any better than “it died of old age” or, “I really should have been wearing an ESD bracelet when I handled one of the three last examples in the world…”. Boards do die, so…

1. There is a TI chip (tape covering the identifying marks) which is pretty crucial to understanding what is up with things because it appears that the power circuit is connected directly to it—so what is under that tape is important.

2. The board was manufactured by Kirk Telecom, now part of Spectrum, who make DECT products. However, Kirk manufactured DECT base stations and so I am very curious about whether any of those have an application for you. I don’t know what the 2815 firmware is like, and though I expect it is at least partly compatible with the Kirk base stations since:

• It is DECT compliant, and;
• It is from the same OEM.

But, I don’t know—still it seems worth investigating.

3. Your brute force approach to this problem has left you with a bit of a mess. I certainly commend you on the thoroughness of your work. I am confident that you have been accounting for everything as you go along, but treating everything as a black box is certainly showing some scaling challenges.

I understand it is beyond your (current) skill set, but some reverse engineering and designing of your own parts to accomplish, for example, the power supply, certainly seems in order.

Barring that, a more orderly interconnection using things like flex PCB (which you can design surprisingly easily and have produced surprisingly cheaply), PCB with connectors rather than jumpers, and things of that sort could reduce the chaos.

Don‘t misunderstand, I think you’ve done an admirable job of keeping the chaos to the pre-disaster level with the routing and dressing visible in the photo, but it is still a tracing nightmare and would be very hard to repair by someone other than you.

In any case, while not directly related to your current complaint, it is foundational and makes working out the problem that much harder.

4. My first pass surmise, given that I don’t have the device(s) to handle is a power related problem that is damaging the boards. Since the large chip—which might be an ASIC, or possibly an FPGA—seems from the surface likely to be mediating the power on sequence the trouble could be there.

If this is the case it is very bad news since an ASIC is custom and an FPGA has custom cores (firmware) and so unless you use the donor board approach, they would be unrepairable. The problem could be voltage spikes, ESD, both, neither—it’s a crapshoot at this point, but with worse odds. Like playing craps with a pair of D100s…

I don’t even know what the rules of such a game would be…
​

5. You still have the most basic troubleshooting to do: follow the power signal until it stops. You (or your local expert assistant) need to trace the B+ from where it enters the board, and each step along the way. On the lower left of the PCB, there appears to be a power supply circuit, with some luck, you might find a part of it that can be replaced needs to be.

The large square component is an inductor, and it looks like it is part of a DC-DC convertor circuit. The button on the PCB edge, I am guessing, is the power button. That button is just a signal, but if you check the high side of it (the leg not connected to “ground”) and there is no power, then it can’t turn the device on.

If you trace the B+ through that power supply section and it makes it all the way to a pin (or pins) on the big TI chip, it will be sad—because that makes the probably unobtainable part the likely suspect.

6. On root causes… there are so many moving parts, which makes this much harder. But since your only real modification here is on the power supply side, it seems likely related. Just what taking things out of cases and wiring them directly would do isn’t clear.

But, the fact they run for an extended time during burn in, then die, points me in the direction of a noisy power supply that might be sending spikes big enough to unalive key component(s).

7. I think you need to seriously consider the idea of engineering some of your own hardware (with the help of a qualified designer, and with you learning what you can). This is a losing game, I fear. The special parts that make this at all feasible are rarer than polite and respectful answers in a C language programming forum—not a sustainable business model.

I know it seems impossible to do, but I think it really could be done. It might involve adapting other models to work with the system, or replacing the old radios with newer, obtainable ones, or something—just not this. Please don’t take this as criticism intended to put your efforts in a bad light, it surely isn’t.

While it is critical, as far as I am concerned you have done an excellent job with what you have to work with. It just seems that the scale of the adaptations has reached the limit where things will now start to break off a brittle arrangement of things regularly and keep you chasing “problems” when the solution itself is the root cause.

Some things I mentioned in 3, above, might add a bit of life to it. But I think it is inevitable that this will stop being possible as you run out of parts, as they mysteriously die.