EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: SArepairman on July 08, 2013, 08:07:34 pm
-
I bought a broken spectrum analyzer. The RF section on it is broke according to the self test and it does not display anything on screen.
Does anyone here have a service manual for this unit? I called up IFR and they were unhelpful, I searched google for hours and found only a users manual with no mention of any kind of service.
I am way out of my league in repairing this device. I checked the voltage inputs into the main RF can, and they seem ok. I found a sharpie X inside the RF shielded compartment where the main input leads into.
The circuit is very complicated and I feel that I will get absolutely nowhere without some kind of manual. I checked some tantalum capacitors and saw no signs of obvious damage.
target practice?
-
Hard to make a guess when you don't know what happened to the unit, but a failing SA front end is mostly caused by too much RF on the input killing the attenuator and/or the first mixer.
Just assuming that when you say there's nothing on the screen you mean no signal but still a straight line?
-
No, there is no line. It is just a blank display.
I proved some oscillators with my oscilloscope and the 10 and 100 mhz are working. Probing the input can output leads to noise on the oscope, IDK if it is too high a frequency (I have only a 100 MHZ scope)
Like, the text and attenuation settings and sweep width and all the controls work, the grid works, but there is no display of the signal and the self test just says RF section failed. With a big F.
When I spin the frequency knob I get some dots on the screen as it spins if the intensity is set to high.
The reference output in the front 100MHZ works fine too.
-
I found that the extension board that all the cards plug into is filthy with a black oxide. I am resoldering it all.
-
resoldering that board lead to the light on and inside of the YIG controller to light up. However the synthesizer one and two modules are not lighting up.
-
when I turn the frequency knob the dots appear all over the display, does that mean maybe there is some kind of way over loaded signal going into something ?
by dots I mean very faint dots that appear all over the screen like fairy dust or some thing.
is there any reason why there might be 16 ohms between Vcc and ground in a shielded can filled with 74 series logic?
Seems bad to me.
But then again I am not a expert at gigahertz range equipment.
there are 74LS73
74S00
74F191
and some other IC's
all in a shielded block with ~16 ohms between vdd and gnd. Is the current In my 34401 resistance tester turning on some kind of magical RF transistor/circuit element ruining the readings?
and a SN74LS390 is reading 18 ohms in the adjacent block between VDD and ground,
ah ha.
the 16 ohm "short" is not occuring in the shielded can but something that leads to the shielded can.
I would if it is just a low value bleeder resistor. seems unlikey though.
It all leads back to the original can, there is a 25 ohm resistance between a power rail and a ground.
Hmm. I don't think the OXO is located there. Time to unscrew some hard line.
so the question is does a 25 ohm resistance belong between a power rail and ground in the RF input section (where the hard line from the input goes)?
-
I disconnected the suspicious RF section and I found that the voltage on the rail which goes into the 35 ohm to ground area remains at 5.15 volts, meaning that it is not a short causing a voltage rail to drop.
The 5.15 volt rails carries 100mA on it.
I don't know where to proceed from here.
-
I've noticed that the red light actually turns on when things are plugged in incorrectly, so YIG controller assy is a suspect now. The YIG assy also has a weird bodge job on it, there are a few resistors and capacitors on a little board covered in electrical tape connected by long wires leading into it.. I don't know what to make of that. They could have fit in the original casing, what the hell?
-
Sometimes you just need the service manual to fix stuff... From what you describe it seems that the log detector is screwed up somehow, but it could be a problem with several other sections. I am not familiar with this model, and I also could not find a lot of information on the net. However, I think you misspelled the model name and it should be an A8000?
If you are lucky there is some printing or silkscreen inside so you can at least figure out some kind of block diagram.
Otherwise all you can do is check all the supply voltages and see if there is anything strange happening there.
If the YIG is not working you'd expect a flat line, but who knows what they did in the software...
-
Well, I hooked up an oscilloscope to the "analog video" output on the back of the unit.
The oscilloscope picks up a noise signal , and the DC offset of the noise signal is changed by the attenuation setting and the thickness of the noise signal is changed by adjusting the resolution bandwidth.
I found the manual for a similar unit, the A7750 (sorry it was the A8000 and not the 8000A), and it says that the analog video output is tapped off right before the sweep digitizer. Could this oscilloscope reading indicate that the ADC is broken?
I also noticed that pressing the Log/Lin button changes the amplitude and noise of the analog video output..
What the hell is this analog video output for? Are you supposed to hook it up to a TV?
Or is it supposed to show the spectrum analyzer display on a scope?
-
My ono sokki has an analog video output mirroring the internal display
-
do you just set the oscope time base to match the time base on the spectrum analyzer?
-
great, something I did or did not do caused the cal to display fail too. :scared:
-
http://www.repeater-builder.com/test-equipment/aeroflex/pdfs/ifr-a7550-spectrum-analyzer-sm.pdf (http://www.repeater-builder.com/test-equipment/aeroflex/pdfs/ifr-a7550-spectrum-analyzer-sm.pdf)
This manual has several schematics used in my unit (for instance the log amplifier board) but it is very confusing and it does not feature nice trouble shooting directions as I had hoped :(
Maybe for starters I will replace as many IC's as possible, for instance the CD4053BE (there are two different versions in the unit and a tiny bit of flux residue, I am thinking it has been replaced before) and some op-amps. I sure hope that one of these RF transistors has not broken :'(
Is the procedure for these things similar to what the president does to his cabinet after he takes office?
-
I think a systematic approach would be better than replacing stuff randomly and hoping that helps, so start at the beginning. (Or the end...)
What is missing from the display, is it just the Rf trace? Is the text and the graticule there?
-
I'd be checking the analogue output on a scope with a signal applied to the input, you should be able to see the peak. If it's there then the most likely issue if the graticule is on the display is the ADC and it's associated circuitry.
-
I doubt it's the log amp.
I suspect that you have multiple issues. The spreading, rather than a spike, and it not being on frequency shouts that you have a local oscillator that isn't locked. If that hump increases when you increase the amplitude of the input then the log amp is probably ok.
If you're not getting that shape you see on the scope on your SA screen then you've also got an issue with the circuitry that goes into the ADC.
So you either have multiple circuits that aren't working as they should, or a power supply issue. The other possibility is that you have a processor contol issue, so that the local oscillators and the ADCs aren't being controlled correctly.
Can you put a function generator into the ADC instead of feeding the log amp into it? That should show as something sensible on the screen.
-
The MC12009 is a prescaler, which is nothing more than a divider and this one can divide by 5 or 6. If you wonder why it can do both 5 and 6 read a bit about about PLL circuits, it is not relevant here.
The IC is a bit special because there is a ECL-TTL level converter in there, that is why it has so many power connections.
A VCC of 4.7 volts seems a bit on the low side, did you measure the different voltages out of the power supply? It should be closer to 5 volts I think. Is the VCC of the 74 series also 4.7 volts?
If you want to know if it is still working probe the input and output with a scope and see what is going on.
If the LED is labelled LOCK I would think that is supposed to lit when things are normal, so when the PLL is locked. Not sure what they were thinking when they designed the thing of course.
-
When I connected a supply 5v to it (together with the 74 series which run on 4.7) I draw .25 amps from the psu but no results come of this
-
I think a 250 ma current draw over .7 volt increase is odd....
Hmmm
-
I replaced a busted 2n2907 transistor with a 3906 that supplied neg 12 to input attenuator. It uses a transistor BC it cuts power during mem test.. I don't think that was the hole problem but after mucking around with the iron I saw it catch a output... and then it crashed. I feel that the problem may be
in that area
-
Why cut the power to the attenuator during memory test? Makes no sense to me really.
-
Why cut the power to the attenuator during memory test? Makes no sense to me really.
Well I thought it was odd to have a 74 series chip controlling the -12 V rail so I shorted the transistor. It failed the self check. I unshorted the transistor (the device has -12V applied to that terminal while its running) and it passed the self check, meaning that it must cut power temporarily during the check...
Any other ideas? I went after a LM380 that I thought might be bad but its not the problem. Something on the ADC side has a 20 ohm resistance to ground, but I can't find it. I did a resistance map and thought the LM380 by the input from the RF shit was the problem but the "short" is still there after the IC was removed.
When I do a resistance test across a capacitor then one way it gives me 60ohm resistance and the other way gives me a 0 resistance that eventually rises to 20 ohms.... I Checked everything in the area and it seems to be the point of lowest resistance.
The thing is its a bitch to trouble shook because the short only apears when I plug the unit into the extension board, but I unplugged every thing that I could unplug from the unit and it seems that they decided to route some kind of ground loop in a insane way, I don't know if its some star ground or what.
I also noticed that the board has a screw terminal connector that touches the ground plane by about half a milimeter. When the board is removed the screw terminal does not make a connection to ground, but when the board is inserted and the tension from the housing causes it to buckle a bit the screw terminal grounds that section. I don't know if the engineers wanted to route the ground through the housing at that point... because the ground plane was 95% removed from under the screw terminals, just the corners overlap.... but only make electrical contact while the board is under stress.
FUCK :--
I tried putting a insulator between the screw terminal and the chasis ground incase its not supposed to make contact there but it did not matter.
-
I also read the datasheet and tested that division chip, it works fine in a divide by 5 configuration, turning a 100mhz oscillation into 20mhz outputs, as per the datasheet.
I have located suspicious transistors on one of the synthesizer boards. I left a ohm meter running to measure the resistance between power rails as I plugged in various cards.
I noticed that there was a significant drop in resistance when I inserted the synthesizer #2 board. I probed around this unit and I found that there is a J310 transistor which when tested with a resistance in circuit test shows a short and when I do the diode test it shows up as a short.
http://www.fairchildsemi.com/ds/J3/J310.pdf (http://www.fairchildsemi.com/ds/J3/J310.pdf)
I am not familiar with this type of circuit element but since it is a short circuit I think it is suspicious? Or is it supposed to be SC when there is no voltage applied? The diode test works both ways on them aswell, showing 0.03V.
Am I onto something here?
The fact that all four J310's behave in the same way makes me think that this is their natural behavior but it seems odd to me. There may inductors there I suppose.... I can't get to it without desoldering shield cans on the back.
-
I suspect that if you check the circuitry round that, there will be an inductor pulling the gate down to the source as it's designed to be run with Vgs = 0V to simplify biasing.
-
This is how they are connected in the A7750, and they are not even on the +5 rail as you can see.
-
Damnit every lead I tried is bad so far. I took apart the YIG filter can now, and tested some transistors, I got excited because I noticed one of the 4 wire RF transistors was shorted. I desoldered the boards and I found that they grounded 3/4 legs of a transistor. :palm:
Now I need to clean with acetone resolder and reassemble and look for trouble elsewhere.
Do you think that replacing all the op-amp on this board would be problematic as a part of operation clean sweep?
I wanted to replace every single IC component which I can get on digikey cheap and is not programmed.
-
I checked a whole shit load of transistors, of the 66382 type.
Everyone I checked has a Vbe of around 800mV +- 4mV. There is one, in the YIG sweeper, that has a Vbe of 750mV, which is very off compared to all the other ones (I checked like 30 in the log amp, and all the other ones on the YIG assy are all almost 800mV.
Is this type of drift something to suspect? I would have to order this part from those over priced ware house places.
I will desolder and measure to make sure the circuit is not causing a error in the measurement then I will swap it with something that looks unimporant in the log amp lol
can anyone think of a transistor that won't be missed in the log amp schematic?
FUCK THAT GOD DAMN TRANSISTOR WORKS FINE GOD DAMNIT NO NO NO NO |O
-
all my dedicated followers :-DD (ok, PA0PBZ gets credit) should be pleased to hear that instead of going on witchhunts I followed the signal and I found a sweep generator that is not working.
There is a overheating HEF4040BP (42C+ with no outputs changing) connected to a circuit with ADC's and DAC.s that are stuck dead in the water.
Considering that many of these parts are legacy and would require ordering from odd vendors at exorbant prices (10 bit ADC for 45$!!!), ad573jn, I began to desolder the parts in the non working circuit block and test them on breadboard.
Hopefully the damage is confined to the HEF4040BP ripple counter, but I need to check near by circuits for operation.
There is also a MK4501N that may be causing trouble.
Thankfully the Mp7531 DAC and the expensive ADC work fine.
Parallel devices are a pleasure to test as you can do it with a simple pulse generator and oscilloscope with no need for any kind of SPI communication or micro controllers.
Maybe some of you have those fancy serial decoders..... hooray old equipment
-
Good to see that you are finally getting somewhere. I already thought about telling you that instead of poking around randomly in the poor thing maybe it was time to attack it in a more ehm... structural way, like following the imaginary block diagram and see which blocks are suspicious. Anyway, let's hope that you are on to something and it is not just one of many problems. Time will tell...
-
Well I found one problem with my random approach but it was getting old.
Also I previously wrote:
I also noticed that the board has a screw terminal connector that touches the ground plane by about half a milimeter. When the board is removed the screw terminal does not make a connection to ground, but when the board is inserted and the tension from the housing causes it to buckle a bit the screw terminal grounds that section. I don't know if the engineers wanted to route the ground through the housing at that point... because the ground plane was 95% removed from under the screw terminals, just the corners overlap.... but only make electrical contact while the board is under stress.
Any ideas on this?
(http://s24.postimg.org/rxx9s7bro/jewbacka2.jpg)
the fault appears to have been present since day one.
It seems that it bypasses a star ground. The continuity test shows a connection but of higher resistance .
I just don't understand why the engineers decided to build it so that it only contacts the ground plane when its in the housing with stress applied.
The picture is bad but the corner towards the middle of the board makes contact, not the outer corner as it would seem from the "under tension" part of the diagram. The ground plane was removed from under the suspicious screw terminal in a non square shape, which to me indicates that the layout guy specifically wanted it not to make contact.
The housing for this unit is very robust, I don't think it was damaged.
-
The picture is bad
You have no idea... ;)
-
I can't figure out how the attenuator works in this unit.
The 7550 model uses a relay system for the attenuator, I cracked this can open there are no relays, only some transistors. (I Am waiting for parts and curious).
The 7550 model uses voltage rails to activate these relays, while this model has no inputs (it combined a few of the RF elements which used to be in seperate cans into one big one)....
WTF is going on? Is the YIG assy somehow handling attenuation? The 7550 schematics shows relays which switch in resistor (and RF crap). This thing has no inputs to the attenuator at all. not one wire.
-
This thing has no inputs to the attenuator at all. not one wire.
What makes you so sure it is the attenuator?
-
This thing has no inputs to the attenuator at all. not one wire.
What makes you so sure it is the attenuator?
Because it is the first thing that the input goes to.
It looks like some of those PCB capactors/inductors, three 4 legged transistors and some SMD passives.
I fixed my camera so I will take some pictures soon.
Unless the block diagram for the 8000A is different and the attenuator is not the first thing the RF input hard line connects to.
-
That does sound like a curious arrangement. I'll wait until I see the pictures before attempting to guess what's there.
-
ah curious, I noticed when I put the center frequency high (~1ghz) the synth lock #1 comes up, only if the bandwidth is narrow, I need to reread the theory of operation I think thats a symptom of something i read..
-
This thing has no inputs to the attenuator at all. not one wire.
It looks like some of those PCB capactors/inductors, three 4 legged transistors and some SMD passives.
So if there are no wires, how are you going to power the transistors? My guess is it is some kind of overload protection device.
-
This thing has no inputs to the attenuator at all. not one wire.
It looks like some of those PCB capactors/inductors, three 4 legged transistors and some SMD passives.
So if there are no wires, how are you going to power the transistors? My guess is it is some kind of overload protection device.
I didn't read that description too carefully before, but it could well be a limiter. The alternative is that it is powered from a supply on the RF line.
-
This thing has no inputs to the attenuator at all. not one wire.
It looks like some of those PCB capactors/inductors, three 4 legged transistors and some SMD passives.
So if there are no wires, how are you going to power the transistors? My guess is it is some kind of overload protection device.
I didn't read that description too carefully before, but it could well be a limiter. The alternative is that it is powered from a supply on the RF line.
I put the oscilloscope to the power supply rails and I did not see any kind of signals being transmitted when I changed the attenuation settings.
Here is some RF photography:
http://s9.postimg.org/lde3d20jx/DSC_0104.jpg (http://s9.postimg.org/lde3d20jx/DSC_0104.jpg)
here is the unit that makes various frequencies and some other stuff
http://s15.postimg.org/3s46ehell/DSC_0106.jpg (http://s15.postimg.org/3s46ehell/DSC_0106.jpg)
the attenuator is on the back side of where the sharpie X is (I need to unscrew it and take pictures of the actual attenuator section but it has the same little transistors you see in the top left quadrant)
here is the interface board:
http://s12.postimg.org/ql5axlun0/DSC_0108.jpg (http://s12.postimg.org/ql5axlun0/DSC_0108.jpg)
-
http://s15.postimg.org/3s46ehell/DSC_0106.jpg (http://s15.postimg.org/3s46ehell/DSC_0106.jpg)
the attenuator is on the back side of where the sharpie X is (I need to unscrew it and take pictures of the actual attenuator section but it has the same little transistors you see in the top left quadrant)
Those are almost certainly FETs used to amplifier the RF signal
The printed filter in the top left section is a low pass with an added bandstop.
It would really help to have a photo of the other side and it should then be possible to put together a good guess at the block diagram.
-
http://s14.postimg.org/ta2e0wej4/DSC_0113haha.jpg (http://s14.postimg.org/ta2e0wej4/DSC_0113haha.jpg)
the hard line input is from the left.
My description was a bit off I was thinking of the previous picture but its pretty similar.
the orange wire is 5.15 volts and it does nothing (signal wise!!).
here is the input attenuator from the 7550, this one makes sense to me.
http://s8.postimg.org/9i61x4bnn/Untitled.png (http://s8.postimg.org/9i61x4bnn/Untitled.png)
Voodoo Bullshit (https://www.youtube.com/watch?v=5pL9gxu5LV8#)
and the top coaxial connector is for a tracking generator or something
-
I'd say this is the first mixer: input LPF, mixer, LO buffer amp (on top, the input is the LO input), BPF and output.
So either someone has stolen your input attenuator, it never had one or you have to find the source of the clicking sound when you adjust the attenuator...
-
I'd say this is the first mixer: input LPF, mixer, LO buffer amp (on top, the input is the LO input), BPF and output.
So either someone has stolen your input attenuator, it never had one or you have to find the source of the clicking sound when you adjust the attenuator...
So the scope has no input attenuator but it has an attenuation function?
wtf??
the scope has no real room for an attenuator, no screw holes for one. I thought it was there to protect the rest of the circuitry? :scared: :scared: :scared: :scared:
There is no click when I attenuate.
-
That is only 1 of the 3 options :P
-
That is only 1 of the 3 options :P
? can attenuation be done afterwords somehow? But it does not click.
-
That is only 1 of the 3 options :P
?
I said: "either someone has stolen your input attenuator, it never had one or you have to find the source of the clicking sound when you adjust the attenuator..."
Ah, stop editing your posts after I reply please :)
I think the input attenuator should be the first thing, but who knows what they were thinking. Is there nothing in the users manual about this?
-
That is only 1 of the 3 options :P
?
I said: "either someone has stolen your input attenuator, it never had one or you have to find the source of the clicking sound when you adjust the attenuator..."
but it has a attenuation button, which responds in software, and it has no screw holes or anything where a stolen attenuator would go
the manual even says its not an additional feature (the a8000 user manual!!)
-
What are the attenuator settings according to the manual? Is it something like 0-60dB in 10 dB steps?
-
What are the attenuator settings according to the manual? Is it something like 0-60dB in 10 dB steps?
yes its 0-60db in 10db steps from the front panel
-
Found a picture of the unit in another A8000, and it indeed looks like the input goes directly to that module, and the thing I called the LO input is not used there too. But it looks like it is missing a module so who knows...
(http://img.auctiva.com/imgdata/9/1/5/9/9/6/webimg/681000198_o.jpg)
Oh wait.... Is this the unit you bought? ;)
-
yes
that guy has to be a real crook if he bent the rf input hardline and stole a module :clap:
OH man, it better not be that silver can on the outside housing of the unit that i threw under the bed and forgot about.
but that still would not really make sense. I think its something custom made for the antenna, pics soon
-
Well, there is clearly something missing and the input rigid coax seems to have had some bending.
Looks like there was some creative maintenance on this one.
Oh well, looks like you have at least $ 105.39 worth of parts there.
And... you could have told us about the missing stuff.
The silver can on the outside (bottom unit) looks interesting :-/O
(http://img.auctiva.com/imgdata/9/1/5/9/9/6/webimg/681000279_o.jpg)
-
I don't really see missing parts. The one missing slot is for some option, and the external thing is just some thing.
There are no screw holes to hold any additional modules besides the 12.3 MHz thing and a 75dB attenuator. And the battery is missing too.
If there was a missing modules there would be screw holes or standoffs for it. there are none.
actually there is one stand off with no screw, but there cannot be a module there because there is a coaxial connector sticking out right next to it. It would have to be a very oddly shaped module.
and of course the tracking generator is absent.
-
Yeah, but what about the 2 flatcables and the mounting hole (?) just below the rigid input cable on the bottom unit?
(and the missing attenuator, and probably the LO or at least the LO cable)
-
Yeah, but what about the 2 flatcables and the mounting hole (?) just below the rigid input cable on the bottom unit?
(and the missing attenuator, and probably the LO or at least the LO cable)
One is for the external thingamabob and the other one might be for the tracking generator.
http://s24.postimg.org/rtp7nbimc/DSC_0116.jpg (http://s24.postimg.org/rtp7nbimc/DSC_0116.jpg)
-
One is for the external thingamabob and the other one might be for the tracking generator.
I admire your optimism but I don't share the feeling. There's just a too nice place for another module in the top left corner on that picture.
Maybe it's for the TG but as I said before, you are missing the input attenuator and as far as I can see also the LO.
-
http://s24.postimg.org/rtp7nbimc/DSC_0116.jpg (http://s24.postimg.org/rtp7nbimc/DSC_0116.jpg)
-
the diagram for the 0-75DB attenuator in the 7550 manual would be a perfect fit for the space under it.
http://s16.postimg.org/gertm0mzo/Untitled.jpg (http://s16.postimg.org/gertm0mzo/Untitled.jpg)
-
unless of course you replace the 0-60 with the 75! :--
-
the diagram for the 0-75DB attenuator in the 7550 manual would be a perfect fit for the space under it.
http://s16.postimg.org/gertm0mzo/Untitled.jpg (http://s16.postimg.org/gertm0mzo/Untitled.jpg)
Yes, that is about the place where I'd put it when designing a thing like this, and the rigid cable still looks like it was washed too hot. I'm thinking that this cable could have been the one from the LO to the mixer module, the connector you marked as a TG connection.
I assume that the last picture is the inside of the module you retrieved from under your bed? Let me see if I can say something stupid about it.
-
is it possible there is no LO? I don't see the 5 pin connector that is shown in the picture and I don't see where it plugs in.
the whole RF mother board looks very different, alot of the ribbon cables are not installed.
and the box is the external can
-
I agree that someone may have stolen the attenuator but the big free spot has no external screw marks. I removed the screws for the oven osc in the back and they were screwed on extremely tight and it left a mark in the back. The spot for the screw holes for the "missing" thing is pristine.
-
the systems engineering of my unit is completely different.
http://s9.postimg.org/nica6witq/lo_l.jpg (http://s9.postimg.org/nica6witq/lo_l.jpg)
I am pretty confident there was nothing where the receive IF box is, unless it was screwed in poorly at the factory.
but there were screw marks where the VCO should be, but there is only one free port in the unit, a coaxial screw on the yig assy.
-
There's definitely something very very odd going on.
I rather suspect that it is indeed missing things, including the input attenuator.
-
does the VCO have a good reason to connect to a yig filter assy?
-
I don't think its remotely possible to source modules for this thing.
:palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm: :palm :palm: :palm:
-
but then again where the hell is the control port for the input attenuator? it does not use a 10 pin connector.
-
Gentlemen, we can rebuild him.
-
Hey I noticed that the little can on the outside is one transistor short of a happy meal.
http://s24.postimg.org/iynl320wz/Untitled.png (http://s24.postimg.org/iynl320wz/Untitled.png)
http://s15.postimg.org/l4sowthm1/Untitled2.png (http://s15.postimg.org/l4sowthm1/Untitled2.png)
the VCO is missing in this thing, but the housing is even kind of eerily reminiscent of the actual VCO.
any idea what that machine on the side of the box is?
-
I think that I can hack in a aftermarket attenuator and a soon as I can figure out exactly what that external module is I can maybe make the missing part of the VCO circuit. I can maybe get something milled out.
-
As far as I can see the external module is the VCO. Now to figure out how to connect the stuff and get the cables. The attenuator is not your biggest problem, you could even put an external one on the input, or if you can find out what the control signals are you can fit something else. I even have a few laying around that probably will do the job.
-
As far as I can see the external module is the VCO. Now to figure out how to connect the stuff and get the cables. The attenuator is not your biggest problem, you could even put an external one on the input, or if you can find out what the control signals are you can fit something else. I even have a few laying around that probably will do the job.
but there are several missing transistors, the VCO schematic has more components then that.
and i want to get this thing working sooo bad i have a nice long bent antenna i found in the trash that I want to fix
-
The schematic you have is not for this model, so why worry about one missing transistor?
-
Well, because the thing looks custom made. I dont think its the same box that fits into the empty VCO slot.
The area where the VCO is missing from has three screw marks, and the external "vco' module measures 52x54mm.
If the external VCO module was placed internally then it could at most use one screw, due to the spacing of the holes and the dimensions of the box, leaving two screw marks unexplained.
-
An image of the screw holes, it just does not add up with the module, it was not meant to be here I don't think.
http://s7.postimg.org/3y18z259l/DSC_0119.jpg (http://s7.postimg.org/3y18z259l/DSC_0119.jpg)
also the adjust hole is unused.
-
Hard to tell without seeing the SA, so I have to accept your conclusions. What you can do is look at the block diagram of the A7550 and figure out what connections (and modules?) you are missing. Still looks like its not going to be easy though.
-
http://s22.postimg.org/4nw0mqsf3/DSC_0121.jpg (http://s22.postimg.org/4nw0mqsf3/DSC_0121.jpg)
a picture showing the screw holes in comparison to the module
At most it can use two screws, I noticed some things look different then the 7550, but the marks make me nervous.
-
also, any idea what the screw coaxial cables are called? I need to buy two of them.
From what I see it seems that the little box on the outside contains 6 transistors of the same type and one odd ball out. I don't know if the odd ball out is the YO84 circle with a squigly line in it or the HYIA3101 transistor.
it seems to be neither because all three units connect to three different prongs of it.
what kind of component would be used here?
-
also, any idea what the screw coaxial cables are called? I need to buy two of them.
The cables are called semi-rigid coax, the connectors are probably SMA.
From what I see it seems that the little box on the outside contains 6 transistors of the same type and one odd ball out. I don't know if the odd ball out is the YO84 circle with a squigly line in it or the HYIA3101 transistor.
it seems to be neither because all three units connect to three different prongs of it.
what kind of component would be used here?
The odd ball is the mixer, like the YO84. The amplifier that is behind the mixer on the other model is probably in another compartment in your A8000.
It looks perfectly fit to do the job of the missing VCO.
-
These guys do a good range of semi-rigid cables at a sensible price
http://www.gigatronix.co.uk/default.php?_minc=home (http://www.gigatronix.co.uk/default.php?_minc=home)
-
I found another problem on the board, a z8340B counter/timer is not outputting clocks to the ADC and dac on the sweep digitizer board, I have a suspicion that the chip I removed (hef4040) was actually getting hot because the reset pin was stuck high and it was receiving a low for the clock as the z8340B was linked directly to the HEF4040, which was over heating., unfortunately during the extraction one leg of the HEF4040 got ripped off (now I learned to use the hot air gun), fortunately it only cost 1$, so I have a replacement coming. I am in the process of studying the 8430b datasheet.
I think I might have to desolder the counter/timer chip and test it on a breadboard, trying to make sense of the commands its receiving is difficult without a logic analyzer since its a parallel chip, but perhaps the documentation will clue me in on some kind of test.
I ordered RG-316 male-male coaxial cables on ebay, I hope they work at least to an extent. I am not sure if that could be considered semi rigid coaxial cable. I bought 2 for 20$, and I figure even if they are not ideal for the unit it should at least allow testing and maybe come in handy for something else.
Does anyone think that the HEF4040 could be damaged by having static inputs on the clock and reset pins and or cause it to experience over heating? The datasheet mentions nothing.
http://www.nxp.com/documents/data_sheet/HEF4040B.pdf (http://www.nxp.com/documents/data_sheet/HEF4040B.pdf)
I think that this chip should be functioning even if the analog part of the board is missing the VCO because I did get it to display a trace on the CRT once.
-
Does anyone think that the HEF4040 could be damaged by having static inputs on the clock and reset pins and or cause it to experience over heating? The datasheet mentions nothing.
If by static you mean 'not changing' then no, it is perfectly alright to have static inputs.
-
Does anyone think that the HEF4040 could be damaged by having static inputs on the clock and reset pins and or cause it to experience over heating? The datasheet mentions nothing.
If by static you mean 'not changing' then no, it is perfectly alright to have static inputs.
Well then I am pretty sure it was part of the problem.
I am replacing parts down the line of some clock signals that are not going through.
Hopefully it is the 74HCT138, but I am starting to suspect the z80 cpu, which I will replace, and I will also back up the PROMS.
-
So I got my cables and hooked in that external module. I found that the ribbon cables cause power supply resetable fuse to blow. I hooked up the external VCO suspect in a manner so that the "slow tune" input came from the YIG filter assy and the REC IF output went to the mixer that the input goes to. I assume the YIG has the tuner voltage because there is nothing else on the board that has a coaxial SMA screw connector.
HOWEVER, there is one of those non-screw SMA connectors that is unplugged on the PC interface board, I don't know what it does but it is near some DACs, so maybe I need to connect the TUNE input using a SMA SNAP to SMA screw cable. I will study the schematic and post my analysis.
I hooked it up to a 12V bench power supply instead of using the ribbon cable and monitored the analog output, I saw nothing, however the ADC and Sweep dac are still disabled (waiting on z80 cpu and z8430 counter/timer and 74 series logic), so I don't know if I should expect any change.
I was surprised by the fact that the unhooked ribbon cables caused power supply problems, I also found it strange that the lines on the ribbon cable that are to power the VCO are only 9V and have a high ripple, I must track down these ribbon cables and perhaps replace some bad capacitors.
And of course the YIG filter has the ribbon cable thats not used either, and causes the resettable fuse to trigger. It would be nice if both ribbon cables were used (one for the external module and one for the YIG filter). I need to track them down. Maybe they are blowing for a reason.
-
I wish I was not made a fool of, that "7th" part was actually the god damn coax connector. |O :-DD
So the mixer is not present on the external board.
-
So what you have in the 'external' module is this. Now where is the rest?
-
God only knows!.
-
Here is the picture of the only unused connector in any of the RF stuff (it is a SMA screw connector in red, and the unknown bead looking thing is circled along with the connector.
I figured that maybe this is a mixer but I don't know.
The picture is of the YIG assy.
http://s23.postimg.org/qtlrlzkqx/yig.jpg (http://s23.postimg.org/qtlrlzkqx/yig.jpg)
the circled unknown device looks kind of like a marble
-
Well, considering that the mixer should have been plugged into what that weird lever mechanism on the side is I am wondering if this spectrum analyzer is a machine off the original IFR-A8000 assembly line, modified to test the mixers. A mixer would be inserted into the thing and the machines operation would be tested?
I am suspicious of the YIG filter thing because there is no transistor after whatever that glass bead is. :palm:
What is the purpose of the mixer in that diagram anyway, I thought it was something for mixing two high frequency signals, but it is just connected to a DC bias (2n2222) and a single input (course adjust).
I thought a mixer was a X inside of a circle anyway..... ???
I'm guessing that the ~ in a circle is actually some kind of VCO module symbol. I would expect this to look like a silver can?
I hope in some way the YIG thing works as a VCO...
Does anyone have any ideas about something I can maybe take out of a cell phone and use in this circuit? Or some other radio device?
I can probably cast a RF can in the backyard out of aluminum. I can't find any kind of mention of a YO-92 or a Y092 VCO.
I am able to find YO92 YIG tuned oscillators, for sale on ebay though, not of the 2222 type.
-
Is it normal for 6 of these RF transistors to draw 250mA when they are doing nothing?
There is only 5 ohms between the power rail and the transistor (I checked all of them, they all get equally hot), so its not just one shorting out the whole party.
I am guessing that they should probably not be getting 12 volts. Any idea how to determine the voltage for this little module?
It is possible that every single transistor in it is blown, I need to compare one to the others on the board.
god damnit, I fried one of the transistors in the external module, but its OK because I can probobly remove one from the other pathways (I don't have enough open connections to use all three outputs) and I don't have the generator.
I should probobly put a 470 ohm resistor or even a 520 ohm resistor to limit the current, so it matches up with the schematic.
the path that the transistor failed on had a cap screwed on tight to the smb connector, so it was definatly unused. :phew:
this repair is exciting I keep breaking stuff :D
-
It wouldn't surprise me that those FETs are taking 40mA each.
-
It wouldn't surprise me that those FETs are taking 40mA each.
yea but one broke.. maybe it broke because I reversed the current for a second (I hooked up the bench supply backwords :palm:)
When I probed them with a thermocouple I found they got to like 55C after 20 seconds of operation.
The little ribbon cable that was used for the weird lever outside the casing had a 10 ohm resistor in series with the supply on a little PCB, since I removed the module I soldered on my own chinese 10 ohm metal film resistor, which is slightly smaller and since its Chinese it does not really dissipate heat well, so it started smoking.
I know the resistor broke because I noticed suddenly it started drawing 500mA when I plugged it in, so I opened the housing an did a diode and resistance test, I found that one of the transistors on the middle path changed characteristics compared to the others, so I removed its connection to VCC and then I did a short test at 12V, the supply draw dropped to 200mA for 5 transistors.
Do you think I should put more current limiting or was it just the brief reverse voltage test that broke it?
-
If you look at page 2-42 (page 64 in the pdf file) of the service manual for the A7550 you can see the explanation of the VCO module, the left part (the red line, see attachment) seems to be what is inside that module.
The transistors in there have a maximum Ic of 70 mA, and the normal operating range is 15-30 mA. But I can't see if the transistors in your module are the same.
And yes, you probably killed one by reversing the supply.
The circle with the ~ in it is an oscillator, in the A7550 it is the 1350-2350 VCO and it could be a yig oscillator, I don't know
One of the 3 outputs is for the tracking generator, so it is not strange that it is not connected.
Still we don't know where the oscillator is, do you have a picture of the test jig/fixture that was on the outside of the case together with that module? Does it have any connections?
-
I read the explanation for the VCO module but I am a bit confused because there are only 2 free connections in my spectrum analyzer in the RF section, and as far as I know the VCO does not hook up to the PC control board (which has a free SMB connector on it), but I really don't think the board is rated to handle any high frequencies. It would seem to support the missing module theory.... I think that the reference oscillator module and the LO summing module were combined in my unit, but there are no free coax connectors on it. Reason for it is that one of my modules has 12 coaxes, and the LO summer has 4 and the reference oscillator has 8 in the pictures for the 7550, so 8+4 = 12, but I need to double check the schematics on it.
I know that the VCO has something to do with YIG because when I search YO92 you get some ebay results for yig voltage controlled oscillators.
30*6 = 180, 15*6 = 90, not even close to the 250 it was getting....
Here is a picture of the fixture:
Top:
http://s14.postimg.org/n9g9vfvjz/DSC_0130.jpg (http://s14.postimg.org/n9g9vfvjz/DSC_0130.jpg)
Bottom:
http://s18.postimg.org/620noubdj/DSC_0126.jpg (http://s18.postimg.org/620noubdj/DSC_0126.jpg)
The top (with the lever and the large coaxial screw) is connected to the module by a short piece of hard line into the input section (where the fork in the road is). The coaxial cap (and blown transistor) is in the middle fork of the road, which, if you go according to the diagrams (despite the three paths being identical) is connected to the optional generator module (J5).
The orange wire is connected to the middle two pins of the 10 pin jumper. If I measure the two free ten pin ribbon cables I find that one cable has 0V on those pins with respect to ground (floating) and the other cable has a 5.1 and a floating pin with respect to ground. However the cable which has the 5.1V is kinked in a way that makes me believe it connects to the YIG assy (I found that when the cable is inserted backwards into the YIG it allows the SA to start and does not appear to cause any problems). If I insert the other cable into the YIG (with floats on both middle pins) one way it makes a short and prevents the unit from starting and the other way it makes a sickly noise and appears to hurt the SA so I disconnect power immediately. This causes a IRF540 mosfet to over heat but not get damaged (at least as far as the diode test tells me), It is a candidate for replacement. I smelled the PCB and its defiantly the IRF540 that gave the hot plastic smell.
Considering the kinks in the cable with 5.1V I will assume that it does indeed gointo the YIG assy and that the other cable might be malfunctioning on one end, so I will now attempt to track this cable down.
-
HMMMMMMMMM this is odd.
The middle two pins on my ribbon cables might be shorting.
I measured 5.1 volts on one of the middle pins on the ribbon cable, but when I trace it back I see that there are absolutely no connections to that pin.
What the heck! I checked the cables for continuity and they seem to be fine, and I checked the cables for shorts or traces at the base (under the plastic) and they also seem to be fine.
:-//
It would appear both of the ribbon cables are disconnected where the 12V connection to the external module would be.
This leads me to believe there is a missing ribbon cable, that perhaps goes from the YIG assy directly to the outside. Once I reassemble this thing I will probe the middle pins on the YIG for voltages.
Hmm, the ribbon cable has 5.1 volts and its connected to a switching IC.
the yig assy also has 5.1 volts, while the other cable is disconnected on the middle pins.
this mean that there is no 12V ribbon cable source for the external transitor module, meaning it must run off 5.1 volts?
-
I replaced the 74 series chips (2x 74LS74 and 2x 74LS138 and the HEF4040 and a 74LS00.
These chips were all in line to the sweep ramp DAC (which is still undergoing shipment).
Unfortunately despite the replacement of the 74LS00 (which functions as a clock for the HEF4040) and the 74LS138's (which act as the START AND STOP signals for the clock) the sweep generator is not being activated.
However, the new HEF4040 chip is not over heating.
I am waiting in the mail for the DAC (which probobly wont fix anything), and the Z8430 and the new Z80 cpu.
An opamp which triggers the convert pulse from the video in signal is under suspicion.
I also replaced a filter capacitor on the 5V rail which was allowing too much ripple for the board, but it had no effect.
-
oH MY FUCKING GOD I GOT MOVEMENT.
I REPLACED THE EVIL Z80 CPU AND NOW, DESPITE THE LACK OF A SWEEP DIGITIZER DAC, STUFF SHOWS UP ON THE SCREEN WHEN I BOOT IT.
:o
It shows a huge trace for a second and fades
if i replace the dac and that 7400 chip I think it will be time for hard core logic analyzer use to determine whats going on.
-
I resocketed a CRT controller and now it is over heating and everything broke.
(http://s22.postimg.org/3lvrc4fxr/ahab.png) >:( >:( >:( >:( >:( >:( >:( >:( >:( >:( >:( >:( >:( >:( >:(
why would resocketing this chip cause it to short out. so weird.
from now on I will just desolder the larger dips, I guess the mechanical stress must have cracked the die or something. :palm:
-
Well I looked at the memory of various ROM chips on this board, I found that the larger chips had what seems to be data on them, but without checksums what can I do...
There are also a few 74 series PROM's (used as lookup tables) that I need to desolder and look at. But It looks like I need a custom device to read them.
-
Do you have make some more progress on the repair?
-
:palm: golden rule of old test equipment, always backup roms before repair, still more and more a dead CPU is sounding like a problem in my own 68K based unit so i thank you for going through everything you tried, :-+
-
im in the process of replacing every component on a board that experienced 12V on the 5V rail (thankfully no memory or anything) due to my negligence
i took a break for a while, I am just replacing all the jellybean parts on all the modules that are shady.
Still waiting on a DAC from china though. :--
If I get it to the point where it was before I will hook up a nice logic analyzer to things to try to get a feel for what is going on.
maybe replacing 50$ of this 7400 series crap will fix it.
-
You should have a little bit more care, you keep over volting and reversing the voltage everywhere :--
-
You should have a little bit more care, you keep over volting and reversing the voltage everywhere :--
yea I spend too much time on it in one sitting and get frustrated, I have trouble taking breaks
eventually my self imposed economic warfare will translate to carefullness though
-
I can't figure why the thing won't display anything, power rails are ok...
how do these old machines typically boot up? I can see the clocks are working
i am getting very sick of this thing it was wayyy beyond my ability :--
-
in my ones case, it is held idle by a reset signal while it waits for the 5V rail to become stable, then after a time constant releases it, the CPU then scans through its address lines until it finds an instruction and goes from there
-
Well it looks like this unit has two inverters connected to the reset line of the z80, I am guessing thats acting as a delay.
I will probe a bunch of stuff with the logic analyzer to see if there is any activity on startup, hopefully its not hanging when it tries to read the roms :(
-
Hmmm I wonder why the reset line has a square wave on it, I don't think hex inverters without feedback should be generating square waves....
The reset line goes to multiple boards but as far as I know it should be be sunk by anything its connected to (they all appear to be inputs)...
hMMMMMMMM
-
if its a square wave then it could be the watchdog timer, something making it unhappy or a general timeout for a status update,
-
if its a square wave then it could be the watchdog timer, something making it unhappy or a general timeout for a status update,
But as far as I know none of those chips has the ability to pull to ground, they are all inputs.
-
what about on the CPU board, something feeding in to the inverters,
for reference here is the schematic for my Motorola 68000 based ones reset and watchdog,
unless it pulls low "TIMCLR" in the upper right LOW in under 64ms it will trigger a reset, (also worth noting the INIT line on mine is an open collector output so until i probed it it was possible for other boards to pull it down,
mine ended up with a reset roughly 4 times a second thanks to the big time constant, and I'm now tracking down what may be causing the timeout,
hope it helps,
-
what about on the CPU board, something feeding in to the inverters,
for reference here is the schematic for my Motorola 68000 based ones reset and watchdog,
unless it pulls low "TIMCLR" in the upper right LOW in under 64ms it will trigger a reset, (also worth noting the INIT line on mine is an open collector output so until i probed it it was possible for other boards to pull it down,
mine ended up with a reset roughly 4 times a second thanks to the big time constant, and I'm now tracking down what may be causing the timeout,
hope it helps,
I need to study that more but the inverter is just hooked up to the 5V line.. I figure the delay caused by two inverters is enough to let the thing stabilize?
5v -> inverter -> inverter -> z80 reset pin
-
unless they are also open collector outputs, e.g. 74LS05 with a RC time constant on one or both, i would very much say not, the time delay of 2 unloaded inverters is approaching nanoseconds, equally that would mean it would be releasing the reset line at ~3V which makes near no sense,
unless you have a schematic to share it really sounds like you haven't traced it out right...
-
Oh no, its from the schematic.
There is a RC delay in the first inverter (which I forgot to mention)., 47uF-33kohm.
-
ok, I found A problem.
when i desocketed all the chips I mentioned the unit stopped working, it has been siting on a shelf for a long time. I went back to it and had a look, the custom character ROM chip cracked.. i guess it took a while for the strain to actually break the chip visually.
:'(
nightmare scenario
bloody hell de solder those chips or cut them out of circuit with a dremel never ever touch those fucking 30 year old sockets!!!
-
really you can't use a normal rom and program on it the program?
-
really you can't use a normal rom and program on it the program?
I don't have the program. All I have a half opened chip that says A8000 CHAR on it, on a little piece of paper. I just wanted to resocket the chips, as this is a common trouble shooting method.
IFR was utterly useless when I contacted them about the A8000, my only hope is that some other hobbyist will one day risk taking it out of their machine and giving me a image of it... then I just need to hope there are no revision changes....
unobtanium.
i think I am going to take it out into a field some where and shoot it while its not looking at me...
-
Put them on ebay at least you will earn something.
In the forum there will be someone that can help you.
I wish it because this posts are very interesting.
-
Put them on ebay at least you will earn something.
In the forum there will be someone that can help you.
I wish it because this posts are very interesting.
I am going to look at the chip under a microscope, maybe only bond wires broke
re-reading this makes me feel rather silly (at my witch hunt tactics and short temper) but I am still proud of even attempting such a task... david vs goliath ;)
nothing a bit of white fuming nitric acid cannot decapsulate. its quite easy to make a few ml of 90%+ using drain cleaner and food preservatives with ghetto distillation techniques, or so it seems. I think you only need a drop or three. so long your starting reactants are devoid of water then the nitric acid produced is of much higher purity then the commonly available 70% NA.
I am defiantly NOT Paying for commercially made WFNA though. those prices are scandalous. not for this "corpse"
all those people buying expensive nitric acid for decapsulating chips are probably being ripped off. you can re-distill it over dry sulfuric acid several times to increase the purity to anhydrous, especially for the drop or two required for chip die extraction,
all yoou need is a test tube and a candle. you mix the ingredients and heat to start the reaction, then you insert a clean but cold nitric acid resistant object into the test tube (such as a narrower test tube or glass rod that was stored in a freezer), but not as to touch the liquid. the pure nitric acid vapors will condense on the cooled object, which you then pull out of the test tube and swab onto your chip.
also this should be done outside with a fan blowing over your work bench so you do not get any nitric acid fumes in your lungs. from what I understand it will choke yoou, burn your lungs and give you cancer. treat it like plutonium.
-
Don't know that "procedure".
Let's update on the situation as soon as you have make progress.
-
Better late than never!
Here are the ROM images from my IFR A-8000.
I hope they will help to bring your king back to live.
-
Better late than never!
Here are the ROM images from my IFR A-8000.
I hope they will help to bring your king back to live.
my god ;D
-
Better late than never!
Here are the ROM images from my IFR A-8000...
Just to correct the record (5 years late):
There is a slight error in the CP_U11_3271-5409-340_v3.4_F32E.BIN file, U11 is a 27128 (not a 27256 as U9)
The file should be half the size at 16384 bytes and the checksum 7997.
Inspecting the contents of the file shows that the first 16384 bytes are repeated in the second half (ie 0000-1FFF is repeated from 2000-3FFF)
I have imaged the EPROMS in my own analyser and can confirm that, with the exception of the above, all of the checksums agree.
Jim