What a circuit diagram doesn't tell us

[Andreas Thaler]

Some time ago I sent the circuit diagram for the Nikon MD-4 Motor Drive to an electronics engineer who is particularly well-known in Germany.

I wanted to know whether a practitioner with a lot of experience who also develops circuits commercially could read the electronic function of the motor drive from it.

I was surprised that he couldn't.

He recognized some of the components, but how they all work together as a system was unclear to him. To be fair, I must note that he is not a photo technology specialist,

The IC in particular, that rectangle in the center of the circuit diagram, probably an undocumented Nikon product, makes the circuit a mystery.

I knew that, but it's still astonishing that even the grouping of known electronic components didn't help to understand the electronic processes in the motor drive.


Tribute to Larry Lyells

That's why the circuit descriptions in the manufacturers' repair manuals or the articles by Larry Lyells in The Camera Craftsman and SPT Journal are so important.

Only then can you get a more detailed insight into how the electronics of a photographic device work.


With the cameras of the 1980s

these descriptions became more complex or were no longer included in the repair manuals because the electronics were too highly integrated.

The repairman then no longer replaces individual transistors or diodes, but rather ICs that contain complex circuits.

Here, a detailed description no longer made sense, only the relevant IC, which is responsible for function X or Y, was replaced.

This is something you should know if you dare to repair a Minolta Dynax or Canon EOS.

It is not without reason that there are hardly any repair reports to this electronic cameras be found on the web.

 

[koraks]

I knew that, but it's still astonishing

How's that astonishing? It's a black box. Most electrical engineers I know aren't clairvoyant. It's a bit like pointing a software engineer at a computer that's not even running, demanding they tell you which websites have been visited with it.

That's why the circuit descriptions in the manufacturers' repair manuals or the articles by Larry Lyells in The Camera Craftsman and SPT Journal are so important.

It's also why every electrical engineer will pull up any datasheets of IC's (and often monolithic components as well) in a circuit if they want to understand what's going on. To make another parallel: if you drop a taxi driver in the middle of a country they've never visited, you don't expect them to magically find their way. They'll fire up Google Maps.

The repairman then no longer replaces individual transistors or diodes, but rather ICs that contain complex circuits.

In bigger devices with a lot of 7400-logic and whatnot, you would replace a decade counter etc. if it happened to be shot. This would be in the 1970s up to early 1980s or so. However, I suspect in the camera repair domain, this has never been very common at all. Documentation on ICs would primarily serve as means to understand the functioning of the system as a whole for fault diagnosis - and estimating whether a repair was economically feasible.

It is not without reason that there are hardly any repair reports to this electronic cameras be found on the web.

Indeed. They weren't being repaired much.

 

[Chan Tran]

Yes if there is an IC he needs the data sheet for the IC to know what it does. The diagram for the IC itself is a whole lot more complicated than the entire diagram for the unit. But of course he doesn't need to know the internal diagram of the IC just the functional description from the data sheet. Many of these IC's are ASIC which is designed to be use only for that product and thus there is no data sheet outside of Nikon.

 

[Andreas Thaler]

How's that astonishing? It's a black box. Most electrical engineers I know aren't clairvoyant. It's a bit like pointing a software engineer at a computer that's not even running, demanding they tell you which websites have been visited with it.



It's also why every electrical engineer will pull up any datasheets of IC's (and often monolithic components as well) in a circuit if they want to understand what's going on. To make another parallel: if you drop a taxi driver in the middle of a country they've never visited, you don't expect them to magically find their way. They'll fire up Google Maps.



In bigger devices with a lot of 7400-logic and whatnot, you would replace a decade counter etc. if it happened to be shot. This would be in the 1970s up to early 1980s or so. However, I suspect in the camera repair domain, this has never been very common at all. Documentation on ICs would primarily serve as means to understand the functioning of the system as a whole for fault diagnosis - and estimating whether a repair was economically feasible.



Indeed. They weren't being repaired much.

Very good, then I don't have to be sad when I try again to understand some of the circuits of my favorites.

However, the details become clear when you understand the corresponding functional area in the camera.

For example, I am currently looking for the function of the ratchet gear that is used in the Nikon F4 to control the aperture.

More precisely, the function of the coupled perforated disk that runs through a photointerrupter and generates pulses.

You can research this endlessly.

 

[ic-racer]

I had already made my own schematic for my Durst power supply before the original schematic arrived from Durst. I had the same experience, as all the ICs were left as blank boxes. Whereas my schematic included the functionality of the ICs. Fortunately the 1980s design used all 'off the shelf' components and nothing custom made for Durst. So I was able to download the tech sheets for all the components.

 

[4season]

Documenting the inner workings of an IC makes sense if you are supporting outside developers who want to use your product. But in this case, all MD-4 development was done in-house.

 

[armadsen]

It's also possible that the physical IC has markings on it that identify it even though the schematic doesn't. But as others have said, if it's a custom design that Nikon did, that won't help you.

 

[koraks]

For example, I am currently looking for the function of the ratchet gear that is used in the Nikon F4 to control the aperture.

Yes, and I also recognize that this is where your expectations from this electrical engineer arose from. With mechanical devices and solid-state/discrete component electronics, the function is apparent from the physical layout - there's a direct and visible relationship between the physical realm and the functional realm. This is exactly one of the things many people struggled with when computers (in a very general sense) made their way into society. All of a sudden, there was a class of devices whose behavior could not be understood by simply dissecting them, or at least not by means available to regular consumers or even repairmen.

Moreover, the behavior of these devices was not linear, but complex - while "back in the old days", button A on a machine would always do function A'. Again, there was a direct and linear relationship between a physical artefact and functional behavior. "The computer way" is context-dependent: button A might do A' while the machine is doing one thing, or B' if it's doing something else. Again, many people struggled with this - and bright engineers did not always recognize or acknowledge this difficulty. It took endless pounding by Steve Jobs-like people to get engineers to at least feign simplicity at the surface.

If the above sounds a little abstract: have you ever helped ma/grandma program her VCR back in the early 1990s? There you go.

 

[Andreas Thaler]

Yes, and I also recognize that this is where your expectations from this electrical engineer arose from. With mechanical devices and solid-state/discrete component electronics, the function is apparent from the physical layout - there's a direct and visible relationship between the physical realm and the functional realm. This is exactly one of the things many people struggled with when computers (in a very general sense) made their way into society. All of a sudden, there was a class of devices whose behavior could not be understood by simply dissecting them, or at least not by means available to regular consumers or even repairmen.

Moreover, the behavior of these devices was not linear, but complex - while "back in the old days", button A on a machine would always do function A'. Again, there was a direct and linear relationship between a physical artefact and functional behavior. "The computer way" is context-dependent: button A might do A' while the machine is doing one thing, or B' if it's doing something else. Again, many people struggled with this - and bright engineers did not always recognize or acknowledge this difficulty. It took endless pounding by Steve Jobs-like people to get engineers to at least feign simplicity at the surface.

If the above sounds a little abstract: have you ever helped ma/grandma program her VCR back in the early 1990s? There you go.

But there is also a generation that knows nothing about the electronic basics of computers. They know nothing about transistors and how they switch, or about the currents that flow through them and why they do it.

Recently I was sitting at the table working on a repair project when a young service technician from our electricity supplier rang the doorbell. He had come for an appointment to replace our old electricity meter with a „smart“ one.

This was my opportunity to have a technical discussion, I thought.

When he saw the open camera on the table and its circuit board, he was speechless. I would deal with „the basics“. He couldn't classify what he saw. Resistors, a transistor, diodes ...

So I went to the new electricity meter and asked him how it works, how it switches. The technician couldn't tell me anything about that, but he knew how to assemble the meter.

He is a trained technician for system electronics.

Another young man, a graduate of an electronics engineering school, told me that they had learned the analogue basics of electronics during their training, but that it was only a marginal topic.

On the other hand, he was well versed in digital electronics.

But if I don't know the basics, how am I supposed to understand what is based on them?

---

When I try to repair a camera, it is clear that I want to understand how it works.

This is no longer possible with the often complex mechanics, and there is still something to see in contrast to the integrated electronics.

So I look at how it could all work in principle and of course I want to know more.

Ultimately, the basic assumption remains that a computer processes data with programs, and that this data comes partly from analog signals that are digitized.

This is where knowledge of analog electronics helps: wipers, resistance tracks, optoelectronic switches, voltage dividers, potentiometers, etc.

Digital data is then output again via analog actuators; piezoelectric elements, motors, relays.

From all of this, I can make a basic assumption about how the electronics of a camera work.

And it is completely clear that I start with the simple analog component and not with the CPU.

 

[koraks]

But there is also a generation that knows nothing about the electronic basics of computers.

Always has been. Again, the issue is not so much the people around you, but your expectations of them. Why expect from a field technician that he has the knowledge of an electrical engineer? What good does it do to this guy or his job if he can explain how a transistor switches? It doesn't make him more or less capable of performing the job at hand. Ask a car mechanic about entropy and you'll get a hazy look. So what?

So I went to the new electricity meter and asked him how it works, how it switches.

What do you mean, 'how it switches'? It's a meter; as a rule, it doesn't switch. It monitors current across one or three phases and communicates this to the outside world in usually three ways (for a smart meter). The question you asked was unanswerable to begin with.

But if I don't know the basics, how am I supposed to understand what is based on them?

The same way you don't need to understand the biochemistry behind enzymes in order to bake a decent loaf of bread, or that you don't have to understand the physics of the Higgs boson in order to machine a perfectly usable set of gears. There's always a line to draw and your understanding is always limited. The question is what you do and don't need to know to perform a job. Doing a job well involves doing it efficiently, and this involves not getting lost in details that don't matter. This is what the service engineer did who replaced your meter; he used the knowledge needed to do his job. The same for the EE who specialized in digital circuits; he doesn't care or need to care about analog and RF, simply because that's someone else's job. And that someone else only has a basic understanding of digital processing. Specialization is everywhere in society, it has arguably been with us for the last 12000 years or so when we started settling and farming and it's one of the key principles behind society functioning to begin with. It's not a bad thing - it's a vital necessity.

 

[BrianShaw]

There seems a knowledge gap regarding levels of maintenance, and levels of the associated technicians, skills, and tools/equipment: Field, depot, and factory maintenance are familiar military terms for very different maintenance activities. In different domains there could be different terms but the basic concept should be in place. Recognizing which is which could help better set those expectations.

 

[Andreas Thaler]

Always has been. Again, the issue is not so much the people around you, but your expectations of them. Why expect from a field technician that he has the knowledge of an electrical engineer? What good does it do to this guy or his job if he can explain how a transistor switches? It doesn't make him more or less capable of performing the job at hand. Ask a car mechanic about entropy and you'll get a hazy look. So what?



What do you mean, 'how it switches'? It's a meter; as a rule, it doesn't switch. It monitors current across one or three phases and communicates this to the outside world in usually three ways (for a smart meter). The question you asked was unanswerable to begin with.



The same way you don't need to understand the biochemistry behind enzymes in order to bake a decent loaf of bread, or that you don't have to understand the physics of the Higgs boson in order to machine a perfectly usable set of gears. There's always a line to draw and your understanding is always limited. The question is what you do and don't need to know to perform a job. Doing a job well involves doing it efficiently, and this involves not getting lost in details that don't matter. This is what the service engineer did who replaced your meter; he used the knowledge needed to do his job. The same for the EE who specialized in digital circuits; he doesn't care or need to care about analog and RF, simply because that's someone else's job. And that someone else only has a basic understanding of digital processing. Specialization is everywhere in society, it has arguably been with us for the last 12000 years or so when we started settling and farming and it's one of the key principles behind society functioning to begin with. It's not a bad thing - it's a vital necessity.

We have a fundamental difference of opinion.

This is because, as a lone fighter, I am confronted with everything that is built into photo equipment. From the screw to the IC. And so I have to deal with everything if I want to understand it. There is no division of knowledge and work because no one else is sitting at the table with me.

Someone who has completed an apprenticeship, regardless of the subject, should know the basics, otherwise they won't know where they are going.

The examples I have given are not semi-skilled fitters, but rather specialists or engineers.

I cannot know and be able to do everything, but I want to try to learn the basics when repairing, so that I can explain to others what I am doing, why and to what level.

If I don't understand all of this, my readers won't understand it either.

 

[Chan Tran]

We have a fundamental difference of opinion.

This is because, as a lone fighter, I am confronted with everything that is built into photo equipment. From the screw to the IC. And so I have to deal with everything if I want to understand it. There is no division of knowledge and work because no one else is sitting at the table with me.

Someone who has completed an apprenticeship, regardless of the subject, should know the basics, otherwise they won't know where they are going.

The examples I have given are not semi-skilled fitters, but rather specialists or engineers.

I cannot know and be able to do everything, but I want to try to learn the basics when repairing, so that I can explain to others what I am doing, why and to what level.

If I don't understand all of this, my readers won't understand it either.

for the electronic parts a tech would follow the service manual closely. Checking all the things the service manual tells you to do. You wouldn't understand how it does it but you can indeed diagnose the problem and fix it.

 

[ic-racer]

My only sucessfull experience with 'unnusual' semiconductors for photography equipment came over ten years ago. On a whim I called defunkt Omega (then called Omega-Satter, or something like that). The person answering the phone, to my amazement, indicated that the semiconductor was actually avilable. Yes, it turned out that in the failed power supplies, that device was the failure point. BWT it was part of a motor driving circuit for the filter's servo motors. I suspect it fails when the "White LIGHT" button is presse and the filters don't zero, so the servo motors keep running, and runnig, and running...

D5500 Power Supply--Complete Rebuild From Scratch

ewww, while the bifurcated sockets are OK, machine pins are more reliable...

www.photrio.com

 

[Andreas Thaler]

for the electronic parts a tech would follow the service manual closely. Checking all the things the service manual tells you to do. You wouldn't understand how it does it but you can indeed diagnose the problem and fix it.

That makes sense, but I don't learn much from it.

Since the manufacturers don't document their ICs, as we've seen, all that's left is studying the principles.

That is, how a digital circuit works in principle, how analog is converted to digital and back to analog, and how the peripherals might work in this regard.

What actually happens in the camera circuit in question remains hidden.

That's unsatisfactory, but I understand the background.

 

[Alan Edward Klein]

When I worked for Sperry Rand Univac computers in the late 1960s, everything was discrete components, no IC's. So you had to understand and follow all the logic. Getting the problem down to a bad transistor and replacing it to fix a circuit card was standard practice.

 

[Andreas Thaler]

When I worked for Sperry Rand Univac computers in the late 1960s, everything was discrete components, no IC's. So you had to understand and follow all the logic. Getting the problem down to a bad transistor and replacing it to fix a circuit card was standard practice.

When did this end?

 

[BCM]

I spent years (as a EE) reverse engineering ICs that had no markings and can tell you that it is very common for larger companies to either purchase lots of custom ICs with very minor changes or use commercially available ICs and simply not have them marked (or remove markings). Developing your own motor drive as some expect that Nikon did is somewhat inefficient due to the relatively small market. If you really need to know, you could go through thousands of data sheets on-line and start by finding an identical pin count IC then looking at the support components to see if the pin numbers match with what is being done in your schematic. Lots of work for a limited return IMHO.

 

[Chan Tran]

Even when I have the data sheet of an IC, I only study the functionality of the IC and not how the internal of the IC works. It's too complicated for me. I do not think you need to go to that level of learn exaclty how the internal circuit inside an IC works.

 

[Chan Tran]

When did this end?

It ended when they invented the IC. If you build a computer with discrete component it would be too hard and it won't run very fast because the components placed so far apart.

 

[BrianShaw]

In the 21st century, do repair shops really “fix circuit cards” versus replacing when something goes bad on them? Referring to products with an active support chain still in place.

In the present quest I understand the necessity to do such because replacements aren’t available. Yet most of the repair documentation seems focused on the lower-level repair scheme, hinted at earlier, of find the fault and replace the “line-replaceable unit”, which would be a card rather than any discrete device on the card. The greatest challenge seems that the type of repair being attempted is not easily done with the type of documentation available to consumers or repair shops because those documentation were never intended to support that level of repair.

 

[4season]

Today, no individual can possess even a basic understanding of all aspects of human knowledge - there's simply too much to know! This is what makes civilization's "hive-mind" such a powerful thing. This is also why journals such as Nature and Science contain much which is written at a layman's level: Even the most brilliant reader will be completely unfamiliar with some of the work being done by others.

 

[koraks]

When did this end?

Around 1970.

In fact, on those Univac machines, it was also unusual for the user to replace discrete components. It's possible that the US Military had its own ways, but in the commercial world, this absolutely wasn't the way things were done back then. A service tech would come in, yank a module from the machine and put a new one in. These modules may then have been refurbished back at the manufacturer's facilities.

In the 21st century, do repair shops really “fix circuit cards” versus replacing when something goes bad on them?

Absolutely not, apart from perhaps very exceptional cases which will be hard to find. It's also likely that in 3d world countries component-level repairs are more usual given the low cost of labor in relation to parts cost. You'd see stuff like USB connectors being replaced etc.

Developing your own motor drive as some expect that Nikon did is somewhat inefficient

It certainly is, and it only happens/happened in situations where there were very specific requirements. In the case of cameras, a driving force behind this was the need for extreme miniaturization (relative to what was common in the market at the time) and a high degree of complexity, resulting in highly integrated and specialized solutions. This was enabled by the manufacturing volumes and the high degree of vertical integration of the camera makers. This is one reason why companies like Nikon and Canon for some time actually kept up in semiconductor manufacturing (up until ca. 25 years ago).

We have a fundamental difference of opinion.

As far as I'm concerned the main disagreement is that I try not to have expectations of other people that are unrealistic.

 

[MattKing]

I wouldn't expect anyone to be able to diagnose and repair individual components unless they were in a position that included expectations about being able to build things from individual components.
Someone, for example, who specialized in restoring historic equipment.
A circuit diagram is as useful for understanding how a piece of equipment functions is as useful as a street map (with nothing on it but street names) is when trying to understand a city.

 

[Andreas Thaler]

As far as I'm concerned the main disagreement is that I try not to have expectations of other people that are unrealistic.

Of course, I don't expect anyone to see things that aren't supposed to be seen.

An IC remains unknown if it is undocumented, unless you can analyze it, as said in this thread.

But the circuit diagram for the MD-4 that I posted here doesn't just consist of an IC, but of other electronic components and assemblies. Here the expert recognized a motor brake, but he didn't know what the whole thing was supposed to be.

If I look at a circuit diagram for an espresso machine and don't know what an espresso machine is, the circuit diagram won't help me either.

So: A circuit diagram is only of limited use.

If you know the device, you can assign functional units.

Even if I don't have an IC in the circuit, five discrete transistors connected together are enough to make things confusing, unless I'm familiar with this sub-circuit.

Nevertheless, I was surprised that the professional didn't get any further. And it relieves me, because then I don't have to crack such nuts.