Grundig Satellit 300 whinge

The Satellit 300 was a digital PLL synthesis radio receiver from the

80s, and I bought it new, so I must have had it for getting on for thirty years.

formatting link

I suppose I can't complain that much given that it still worked until I dropped a wooden curtain rail onto it. The outside was unmarked, but it would no longer pick up MW (though VHF was OK, don't know about the other bands), the tuning wheel didn't work, and the illumination for the LCD panel was absent.

The first problem suggested an issue with the ferrite rod antenna (not used for VHF of course), and a minute or so after I'd opened the thing up I spotted a broken connection. The second problem was also a wire that had broken off the tuning wheel mechanism. The third was a broken filament bulb.

The first two faults would probably not have occurred if the wiring had any slack in it at all.

Also, this was an expensive radio - hadn't they heard of headers?

formatting link
formatting link

Several more things broke while I was trying to put it back together, though it's all working now - I think.

Sylvia.

Reply to
Sylvia Else
Loading thread data ...

Gosh, what a hairball. Hadn't these people heard of ribbon cables? Hadn't they heard of electronic packaging in general?

Packaging is half the job in this business. As my old mentor Melvin Goldstein used to say, when you've got the schematic, you're 10% done.

John

Reply to
John Larkin

Half the job? I seemed to spend more time sorting out connectors than I ever did on picking the right active components, and I used a great many different active components

If the schematic was drawn with a clear idea of how the active devices were going to be laid out and connected to the outside world, this might be right. In real life the packaging has a way of getting parts of the schematic redrawn.

When you start shipping your timing signal between boards on coax cable, routed through mixed signal DIN-41312 connecters, and find youself stuck with using triple-extended Eurocards to make room for enough connectors, the schematic gets to be a finally minor part of the documentation.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

The old Grundig Satellites from the late 60's and 70's were much worse. OTOH, as a kid I got to repair some and that provided a very nice boost for the hobby budget. Those things broke all the time, especially the really big ones that almost had the size of a briefcase. They were ok if you left them in one spot and just turned the dial. But woe to him who thought the handle meant they could be transported frequently.

The way it usually went was the owner didn't trust a kid like me with that because these radios had cost them a fortune when new. So they gave them to a radio & TV repair shop. Those guys eventually threw their hands up in the air, gave up. With that, the radio had reached basket case status and I got it. It was nice to see the jaws drop when they worked again.

Some defects were outright mean. Such as super-thin enameled wires that corroded off inside the IF filter cans. It seems they must have used some aggressive flux and under there it didn't get cleaned out too well. So when I got one of those surrendered cases the first order of biz was to ohm out all those windings.

He is right.

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

A schematic should certainly include all the connectors, and their part numbers and pin numbers, and all the mounting holes' LEDs, and test points. How else could you lay out the PC board?

In real life the packaging has a way of getting parts

Not in my world.

John

Reply to
John Larkin

What struck me as well is that the thing must have been a nightmare to manufacture. Presumably all those connections were soldered by hand, and somehow they have to address the problem I had - how to close the case without breaking one of them.

I wonder what proportion of the completed units had to be rejected because they didn't work properly.

Sylvia.

Reply to
Sylvia Else

Which is why the Japanese in that era rose to the top in the electronics market.

In modern times, the Chinese are now slowly but surely working their way there.

Reply to
kreed

because=20

It looks like they spent most of their effort making it look good on the =

outside. I have some old German meters my father got from a = Messerschmitt=20 factory in Germany during WW2, and the wiring was very neat. But of = course=20 it was simpler than a SW radio. Still, I would expect much better = attention=20 to detail in a German product.

I recently bought a DSP shortwave radio on eBay for $15 plus $10 = shipping.=20 It was made in China and all the labels and instructions were in = Chinese. By=20 the time I contacted them and got a translation, I had figured out most = of=20 it. I have not used it much, but it seems pretty nice:

formatting link
p3984.m1497.l2649

Paul=20

Reply to
P E Schoen

I

it

the

t
g
n

ad

How indeed? But the realities of printed circuit layout - like things sticking out of other boards - have a way of adjusting the original magnificent simplicities.

We already knew that you didn't live in the real world. Presumably these things happen behind your back - what the boss doesn't know doesn't upset him ...

-- Bill Sloman,

Reply to
Bill Sloman

Hey, Guys -

She said it's about 30 years old. What were you doing back then? Searching for connectors or getting the product out?

Reply to
John S

[...]

My world is the same as John's is. Maybe you live in a different one?

Da boss usually sits in design reviews so da boss gets to know if this should happen :-)

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

Joerg, Joerg, Joerg! All you contribute here is feeding trolls. Bye... again. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Jim, Jim, Jim! When do you learn about modern newsreaders that allow to selectively pick what one wants to have displayed ...

Tsk .. tsk .. tsk :-)

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

Please, Joerg, don't bait the old prejudicial fart.

Reply to
John S

I draw schematics, fill out the form to create new library parts, define the outline/mounting/connectors/heatsinks/LEDs, do the stackups and trace impedance rules, plan and often do the parts placement, oversee layout, and check/tweak the PCB before it's Gerbered. I often lay out critical high-speed or low-level sections myself. Once in a while I do an entire board, just to keep in practice.

We Gerbered this today:

ftp://jjlarkin.lmi.net/ESM_pcb.gif

The thing in the middle is an LTC2242 250 MHz, 12-bit ADC.

The only new thing, of late, is that I don't assign FPGA or uP BGA pins on the initial schematic. That would almost always make routing impossible. We have a net name convention like

NAME_F FPGA i/o

NAME_P_FL lvds pair + NAME_N_FL lvds pair -

NAME_FD FPGA dedicated function

NAME_FC FPGA clock

NAME_U ARM uP i/o

and so on. Sometimes we specify a bank, too. That lets The Brat pick the pins for best routing. We rip the final netlist and feed that into the Quartus thing, which tells us if her pin assignments are OK. In some cases, nothing but the Altera software can decide for sure if it's OK.

John

Reply to
John Larkin

I generally use a convention sorta like:

SECTION_NAME_TYPE where: SECTION = controlling subsection NAME = signal function TYPE = I2C (SDA, SCL), SPI (SCLK, MOSI, etc.), TxD, Clock, etc

/ (prefix) = negative active. Gets converted to "_n" suffix in the VHDL wrapper (negative active signals are fairly rare in the logic).

  • (suffix) = positive differential

- (suffix) = negative differential

+V.V = positive voltages

-V.V = negative voltages

I don't much care if it's an FPGA or UC I/O. I care about its function and which section of the logic owns it.

Do you make a good cut at the FPGA design before layout or just define the pins and their functions?

Reply to
krw

We usually finish the board layout before we finish the FPGA design; often we don't start the FPGA until after the board is gerbered. I do a pin count and a sketch-level suggestion of parts placement/routing/bank usage, and Liz makes it actually happen. A lot of negotiation follows. I have a lot of sympathy for people who have to use outside layout services, or throw it over the wall to another department. I can see how they could wind up with two or three board spins before it works.

The rules for pin usage on the Altera chips aren't all in writing, so that's why we load the pinout into the Quartus software before we gerber the board. We had one case recently where a cmos logic level ball was next to an LVDS pair, and Quartus told us there would be too much crosstalk. Fixing the layout would be almost impossible, so we declared the signal rate on the cmos pin to be 0 Hz. That's the cheat for that particular problem. We've also discovered some power supply issues on certain banks for certain kinds of signals, again from the software and not from the manuals.

John

Reply to
John Larkin

the

I've done it both ways. It helps to have your own layout people but without that luxury it just means working out of a hotel room while you sit over the layouter's shoulder. ;-) Without direct interaction between the designer and layout person, agreed, anything more than a trivial design is hopeless. I've seen the results. :-(

Got that part. Some don't cut the Gerber until the FPGA is first simulated and fit. I've never done it that way, though.

Reply to
krw

This product's printed wiring was manually-cut tape on a transparency.

The schematic art was drawn by hand on (possibly pre-formatted) paper or vellum, using a T-square, pen, ink and possibly some art transfer materials.

The point to point connections, and schematic correspondence were checked visually, by humans, sometimes with the aid of a photocopier and coloured pencils/pens/crayons.

The bill of material was, at the very most, typed on paper, before being (eventually, if ever), manually entered into some kind of database.

RL

Reply to
legg

Consumer-grade connectors in the 80's had their own issues, never mind the cost. Looks like the display used a printed edge connector - that would have been considered quite progressive for the time.

I've an Emerso/Sangean/Olympia ATSxxxx clone from the ~ same era. First thing to go was normally the 'digital pot' tuning dial stuck out the side. It has electromechanical wipers that can misalign on impact.....

I've resoldered an internal solid-conductor shielded cable termination three times (first time after buying it as 'dead' scrap). You don't see any solid conductors inside the Satellite. Count your blessings.

Mine's still in daily use, in spite of having a melted front face, aquired when positioned on a test shelf, in too close proximity (12 inches) to a multi-KW passive load, sometime in the early 90s. It was a first order EMC policeman, and a useful indicator of the functioning status in the nearby switch-mode DUTs.

RL

Reply to
legg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.